Since January 2020 Elsevier has created a COVID-19 resource centre with

free information in English and Mandarin on the novel coronavirus COVID-

19. The COVID-19 resource centre is hosted on Elsevier Connect, the
company's public news and information website.

Elsevier hereby grants permission to make all its COVID-19-related

research that is available on the COVID-19 resource centre - including this
research content - immediately available in PubMed Central and other
publicly funded repositories, such as the WHO COVID database with rights
for unrestricted research re-use and analyses in any form or by any means
with acknowledgement of the original source. These permissions are
granted for free by Elsevier for as long as the COVID-19 resource centre
remains active.
 Regulatory Toxicology and Pharmacology 118 (2020) 104813

Contents lists available at ScienceDirect

Regulatory Toxicology and Pharmacology

journal homepage: www.elsevier.com/locate/yrtph

Contemporary Review

Considerations for setting occupational exposure limits for novel

pharmaceutical modalities
Jessica C. Graham *, Jedd Hillegass, Gene Schulze
Bristol Myers Squibb, 1 Squibb Drive, New Brunswick, NJ, 08903, USA

A R T I C L E I N F O A B S T R A C T

Keywords: In order to develop new and effective medicines, pharmaceutical companies must be modality agnostic. As
Occupational exposure limit science reveals an enhanced understanding of biological processes, new therapeutic modalities are becoming
Worker safety important in developing breakthrough therapies to treat both rare and common diseases. As these new modalities
Occupational health
progress, concern and uncertainty arise regarding their safe handling by the researchers developing them, em­
Exposure control band
ployees manufacturing them and nurses administering them. This manuscript reviews the available literature for
Pharmaceutical modalities
Vaccine safety emerging modalities (including oligonucleotides, monoclonal antibodies, fusion proteins and bispecific anti­
Protein therapeutic bodies, antibody-drug conjugates, peptides, vaccines, genetically modified organisms, and several others) and
PET Tracer provides considerations for occupational health and safety-oriented hazard identification and risk assessments to
Oncolytic virus enable timely, consistent and well-informed hazard identification, hazard communication and risk-management
Engineered bacteria decisions. This manuscript also points out instances where historical exposure control banding systems may not
be applicable (e.g. oncolytic viruses, biologics) and where other occupational exposure limit systems are more
applicable (e.g. Biosafety Levels, Biologic Control Categories).

Occupational Safety and Health (NIOSH) hazardous drug list (Nations U,

2019; NIOSH, 2016). However, guidance is also needed for those pre­
1. Introduction paring the SDSs and assigning hazard classifications.
As the bulk of novel modality pharmaceuticals being evaluated/
Until recently, small molecule drugs were the primary focus of the developed are in the discovery phase of drug development, they tend to
pharmaceutical industry. As the scientific field advances through an lack appreciable nonclinical and clinical data, including pharmacolog­
enhanced understanding of biological processes, the role of genetics and ical potency and toxicity information. Compounds with limited data,
the interplay among peptides/DNA/RNA, and how these interactions such as those in early discovery research and development, are placed
relate to both the cause and cure of disease, many new therapeutic into occupational exposure control bands (ECBs; also commonly
modalities are becoming important in developing breakthrough thera­ referred to as occupational exposure bands [OEBs] or occupational
pies to treat both rare and common diseases. In addition to novel mo­ health categories [OHCs]). Occupational banding/categorization sys­
dalities, increasingly potent and persistent medicines are being designed tems essentially pair a hazard determination with an acceptable occu­
to enable lower doses and less frequent dosing. As compounds become pational exposure concentration range along with appropriate exposure
more potent, even seemingly small amounts of dermal or inhalation controls and handling practices. These bands are assigned based on
exposure can pose significant health hazards to the employee who is historical experience and information, read-across strategies, in silico
synthesizing and manufacturing the drug, to the health care worker who evaluations, in vitro screening data, and in vivo data (where available)
is administering the drug, to the patient performing self-administration conducted to elucidate a compound’s pharmacological and/or toxico­
(e.g. subcutaneous injection of protein therapeutics at home) and/or to logical characteristics and subsequent hazard assumptions and
others in the shared facilities, clinics, or homes where these activities are classifications.
taking place. There have been several attempts to identify and What follows is an overview of occupational hazards and risks
communicate the hazards associated with drugs via the safety data sheet associated with several of the most broadly utilized pharmaceutical
(SDS), Globally Harmonized System of Classification and Labelling of modalities. Literature searches were conducted to identify key
Chemicals (GHS) categorizations, and the National Institute for

* Corresponding author.
E-mail address: [email protected] (J.C. Graham).

https://doi.org/10.1016/j.yrtph.2020.104813
Received 7 May 2020; Received in revised form 13 August 2020; Accepted 26 October 2020
Available online 2 November 2020
0273-2300/© 2020 Elsevier Inc. All rights reserved.
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

Abbreviations: HSV-1 herpes simplex type 1

IC50 half maximal inhibitory concentration
aa amino acid ICH M7 International Council on Harmonization M7 guidance
AAV adeno-associated virus Ig immunoglobulin
ADA Anti-drug antibody i.t. intratracheal
ADC antibody drug conjugate IV intraveous
ADCC Antibody dependent cellular cytotoxicity kDa kilodalton
ADME adsorption, distribution, metabolism and elimination kg kilogram
ALL acute lymphoblastic leukemia L late
ATC adoptive T-cell therapy LAB lactic acid bacteria
ALARA as low as reasonably achievable LAG-3 lymphocyte-activation gene-3
APC Antigen presenting cell LBP live bio-therapeutic product
API active pharmaceutical ingredient LNA locked nucleic acid
ASO antisense oligonucleotide mAbs monoclonal antibodies
BA bioavailability MHC major histocompatibility complex
BCC biologic control category MW molecular weight
BCMA B-cell maturation antigen MMAD mass median aerodynamic diameter
BMS Bristol-Myers Squibb mg milligram
bsAb bispecific antibody mRNA messenger ribonucleic acid
BSL biosafety level miRNA micro ribonucleic acid
CAR chimeric antigen receptor NET neuroendocrine tumor
Cas9 CRISPR associated protein 9 NIOSH National Institute for Occupational Safety and Health
CBI cyclopropabenzidole OEB occupation exposure band
CD cluster of differentiation OEL occupational exposure limit
CDC Centers for Disease Control ON oligonucleotide
CDC complement dependent cytotoxicity OSHA Occupational Safety and Health Administration
CNS central nervous system PBD pyrrolobenzodiazepines
COVID coronavirus disease PCR polymerase chain reaction
CRISPR clustered regularly interspaced short palindromic repeats PD pharmacodynamics
CRS cytokine release syndrome PD-1 programmed death-1 cell surface membrane receptor
CT computed tomography PD-L1 programmed death-ligand 1
CTLA-4 cytotoxic T lymphocyte-associated antigen 4 PEG polyethylene glycol
CYP450 cytochrome P450 PET positron-emission tomography
Da Dalton PK pharmacokinetic
DAMP damage-associated molecular pattern molecules ppm parts per million
DAR drug-to-antibody ratio pRb retinoblastoma protein
DM1 mertansine PSA prostate specific antigen
DNA deoxyribonucleic acid QC quality control
DSEN dermal sensitizer notation QSAR quantitative structure-activity relationships
E− early RNA ribonucleic acid
E2F E2 factor RNAi RNA interference
ECB exposure control band Rnase H ribonuclease H
EG exposure guideline SDS Safety data sheet
EGFR epidermal growth factor receptor SGN-35 auristatin
EpCAM epithelial cell adhesion molecule siRNA small interfering
EpoFc erythropoietin-Fc fusion molecule 5 sc exposure control band 5 special case
Fab antigen binding fragment TCR T cell receptor
Fc crystallizable fragment T-DM1 maytansine
FcRn neonatal Fc receptor TIL tumor-infiltrating lymphocyte
FDA Food and Drug Administration TIM-3 T-cell immunoglobulin and mucin-domain containing-3
FPF fine particle fraction Tmax the time after administration of a drug when the maximum
FIXa factor IXa plasma concentration is reached
FX factor X TNF tumor necrosis factor
GHS Globally Harmonized System of Classification and TMDD target-mediated drug disposition
Labelling of Chemicals TTC threshold of toxicological concern
GM genetically modified T-VEC Talimogene laherparepvec
GM-CFS granulocyte-macrophage colony-stimulating factor μg microgram
GSD geometric standard deviation UNA unlocked nucleic acid
h hour VLP virus-like particle
HBEL health based exposure limit WHO World Health Organization
HER2 human epidermal growth factor receptor 2

2
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

toxicology and pharmacology information on each of the modalities deeply embedded in occupational health and safety practices, particu­
with a focus on occupationally relevant data including occupational larly in the pharmaceutical industry (Naumann et al., 1996; Zalk and
exposure case studies and inhalation studies for the modalities described Nelson, 2008; NIOSH, 2019). Additionally, such systems are elements of
(See Supplementary Table 1). Information discussed for each modality well-developed, current hazard communication programs (e.g., United
includes: Nations 2019 Globally Harmonized System of Classification and Label­
ling of Chemicals) (Nations U, 2019). Occupational health categoriza­
(1) Background: a brief background on the modality; tion and compound handling practice systems are considered standard
(2) How they work: an introduction to how the drugs within this practice throughout the pharmaceutical industry in both research and
modality work; manufacturing operations.
(3) Marketed drugs: examples of marketed drugs; The occupational categorization system was designed to give guid­
(4) ADME: the documented absorption, distribution and elimination ance, based on historical experience, on safe handling practices for
(ADME) properties; compounds with limited data as a stopgap until additional relevant data
(5) Health hazards associated with therapeutic use: health hazards could be generated. For pharmaceuticals with robust data sets,
observed or expected after therapeutic administration as well as compound-specific occupational exposure limits (OELs) are established
those observed in relevant nonclinical studies; to protect employees. However, when there is limited data for a com­
(6) Occupational hazard and exposure considerations: a summary of pound, occupational exposure banding is often employed to establish
the occupational exposure risk considerations and occupationally occupational exposure constraints. While an OEL is a specific airborne
relevant hazards; and concentration limit usually presented in units of μg/m3 or parts per
(7) Occupational exposure banding guidance: a recommendation for million (ppm), an occupational ECB is a range of airborne concentra­
an occupational exposure control band based on the occupational tions to which exposure to a compound should be controlled to ensure
health hazards and risks. worker safety (See Table 1).

This work provides guidance in regards to characterizing the occu­

pational hazards of new and emerging modalities to enable timely, 2.2. Application of occupational exposure control banding
consistent and well-informed hazard identification, hazard communi­
cation and risk-management decisions. Occupational exposure banding (also known as hazard banding or
health hazard banding) is a systematic evaluation process utilized to
2. Occupational exposure control banding assign chemicals/compounds to “bands” based on selected health effect
endpoints (e.g. inherent toxicity, pharmacological effects, etc.). The
2.1. Background of occupational exposure control banding basic premise of the ECB classification system is to place chemicals into
categories based on their inherent toxicity and potency, which offers a
The concept of using hazard-based categories to communicate po­ simplified solution for controlling worker exposures to compounds in
tential occupational health concerns, signal workers and employers to the workplace. Briefly, an initial hazard assessment is conducted in an
the need for risk management, and inform exposure control re­ effort to identify potential exposure ranges expected to represent
quirements has been utilized for decades. The original occupational negligible risk for the physical (e.g. corrosivity), toxicological, and/or
health categorization practices were developed in the pharmaceutical pharmacological effect(s) of concern. The mechanism of pharmacolog­
industry and such hazard classification and category-based systems are ical action, in vitro/in vivo potency, preclinical dose-response related
effects, bioavailability (inhalation, oral and dermal), therapeutic dose,

Table 1
Example of an exposure control band (ECB) system.
ECB Range Relevant Compounds Rationale Examples
(μg/m3)a

1 ≥1000 Compounds of very Caffeine

low toxicity/potency

2 100 - < Compounds of low Permitted exposure of >1000 μg/day (>1 mg/day) for compounds Antibiotics of tetracycline, aminoglycoside and
1000 toxicity/potency of low toxicity which are not potent. Compounds that may cause fluoroquinolones class; some cardiovascular, antiviral,
mild, reversible acute affects (e.g. skin/eye irritation). and central nervous system (CNS) drugs

3 10 - < Compounds of A TTC of 1000 μg/day is recommended for relatively unstudied Some cardiovascular drugs, statins
100 intermediate toxicity/ compounds that may be intermediately potent or toxic.
potency

4 1 - < 10 Potent/Toxic A TTC of 100 μg/day is recommended for relatively unstudied Some potent cardiovascular, metabolic, antiviral and
compounds compounds that are not likely to be highly potent, highly toxic, or CNS drugs, early discovery APIs, some chemically
carcinogenic, have no a priori evidence of unusual potency or synthesized peptides
toxicity and are not considered mutagenic (Dolan et al., 2005;
Kroes et al., 2004; Cramer et al., 1978; Bercu and Dolan, 2013).

5 0.1 - <1 Highly toxic/potent A TTC of 10 μg/dayb is recommended for relatively unstudied Toxic oncology drugs, potent compounds, chemically
compounds compounds that may be highly potent or highly toxic with limited synthesized peptides, antibody drug conjugates, steroids
data to indicate they may produce pharmacologic or toxic effects at
very low doses (Dolan et al., 2005; Kroes et al., 2004; Cramer et al.,
1978).

5 special <0.1 Especially potent/toxic A TTC of 1 μg/dayb is recommended in the absence of sufficient Especially potent/toxic compounds, protein nucleic
case compounds data for anti-cancer drugs, which are developmental toxicants, acids
mutagenic, or may be carcinogenic (Dolan et al., 2005; Bercu and
Dolan, 2013; Stanard et al., 2015).
a
The banding recommendations presented reflect the assumption that an employee will inhale 10 m3 of air daily during his/her 8-h shift (Derelanko, 2017).
b
A threshold of 1.5 μg/day is recommended for relatively unstudied compounds which may be mutagenic or carcinogenic (Guideline, 2018).

3
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

and the spectrum and severity of clinically observed adverse effects of a BMS and is presented in Table 2. Essentially, active biologic materials
specific drug substance, all provide the basis for the hazard assessment. that are made through biological processing (i.e. cell culture), can be
Preclinical data such as QSAR (in silico predictive systems) and animal categorized into one of two Biologic Control Categories (BCC) (BCC A or
data is also considered in the hazard assessment, and one or more BCC B). BCC A or BCC B is assigned based on the hazards and potency of
compound characteristic may be responsible for placing a compound in the biologic of concern (See Fig. 1). For potent or toxic therapeutic
a specific ECB. It is generally prudent to assign compounds to more proteins or other biologic compounds, BCC B exposure controls (and
protective bands earlier in their development, and as new data emerges, corresponding handling practices) should be utilized. Note that while
subsequently adjust their occupational exposure limits/bands to less Table 2 illustrates the banding system that BMS implements for bi­
restrictive bands and corresponding handling practices (with the goal ologics, alternative schemes are also utilized in the pharmaceutical in­
being to ensure workers in the early development space are adequately dustry which are equally effective in controlling exposures.
protected). Additionally, for compounds expected to pose special hazards, a
The banding system is integrated into the organization’s engineering compound-specific risk assessment can be completed to confirm
controls and thus may be different across organizations. For illustrative whether additional exposure controls are needed (e.g. BCC B controls
purposes, Table 1 presents a typical pharmaceutical compound banding with additional personal protective equipment). Also, note that once a
system and is the one employed by Bristol Myers Squibb (BMS). The sufficient data package is available to establish an OEL (also referred to
cutoffs for the bands/categories presented are based on several factors, as a health-based exposure limit [HBEL] or an exposure guideline [EG]),
including approaches based on the threshold of toxicological concern the BCC is of limited applicability.
(TTC) (NIOSH, 2019; Dolan et al., 2005; Kroes et al., 2004; Cramer et al.,
1978; Gould et al., 2016; Guideline, 2018). While the TTC may not be 2.4. Banding decision tree
originally derived for occupational purposes, the principles have been
applied successfully in the field of occupational health and safety for the In order to assist in the selection of the appropriate ECB or BCC for
establishment of safe occupational exposure limits (Chebekoue and pharmaceutical modalities discussed in subsequent sections, a decision
Krishnan, 2017, 2019; Carthew et al., 2009; Hoersch et al., 2018). TTC tree was generated as shown in Fig. 1. The primary consideration for
limits can be applied using an assumed breathing volume of 10 m3 and band selection is pharmacological potency in vivo, and details regarding
100% inhalation bioavailability. The banding recommendations derivation of the pharmacological potency levels which can differentiate
included in this manuscript for compounds including small molecules, between bands, bioavailability considerations, and additional toxicities
antibody drug conjugates, oligonucleotides and biologic material made which may warrant an additional safety factor are described in more
through chemical synthesis, are based on those presented in Table 1. For detail within this article (see Sections 3.1.7 and 3.4.7). This decision tree
more information on occupational exposure banding systems and ap­ focuses specifically on ECB 4, 5, and 5 special case (for small molecules)
plications including suggested exposure controls and handling practices since early in drug development when the banding approach is most
see the following references (Naumann et al., 1996; Zalk and Nelson, applicable, limited data are available such that less restrictive bands
2008; NIOSH, 2019; Ader et al., 2005; Garrod and would not be considered, and active pharmaceutical ingredients (APIs)
Rajan-Sithamparanadarajah, 2003). would generally default to one of these three bands. It should be noted
that this decision tree, and the doses included therein, should be
considered as a rough guide for initial band selection. Ultimate selection
2.3. Banding considerations for biologics
of the band should come from a qualified occupational toxicologist, and
rely on the consideration of a number of additional criteria as described
Due to the general instability of biologic therapeutics, differences in
in this article, including the innate hazards of the therapeutic as well as
their manufacturing (generally closed processes to protect sterility) and
its pharmacokinetic and pharmacodynamic profiles among others.
their potential limited bioavailability (BA) via the inhalation, oral and
dermal routes (Gould et al., 2018; Pfister et al., 2014a; Bos and Mei­
3. Occupational exposure control banding considerations for
nardi, 2000; Krause and Sahin, 2019), it can be argued that a different
pharmaceutical modalities
set of exposure controls can be utilized to protect employees when
working with biologics as compared to small molecules. Based on this
3.1. Small molecules
information as well as extensive industrial hygiene monitoring con­
ducted by BMS showing that airborne concentrations of biologics are
3.1.1. Background
most often <1 μg/m3 (data not shown), a simplified two-band system
Small molecule drugs (<900 Da [Da]) are generally designed to
was developed for biologics for use when there is insufficient informa­
freely enter cells (Dougherty and Pucci, 2011). Once inside a cell, small
tion available to calculate an OEL. Therefore, a two-category system for
molecule drugs can interact with proteins, receptors and deoxy­
banding mid-to high-molecular weight (MW) biologics made through
ribonucleic acid (DNA). This is different from drugs that have a large
biological processes (i.e., mammalian cell culture) is implemented at
MW, such as monoclonal antibodies, which are not able to penetrate
cells very easily even once they are systemically bioavailable. While
Table 2 there are exceptions, generally oral bioavailability significantly de­
Example of a biologic-specific banding system.
creases when the molecular size exceeds 900 Da.
Biologic Range Relevant Examples
Control (μg/m3) Compounds
3.1.2. How they work
Category
(BCC)
Small molecules exert pharmacologic effects through various
mechanisms of action, including but not limited to: 1) agonism/antag­
A Biologics with low Mid- to high- MW biological
onism of specific receptors (e.g. tamoxifen), 2) enzyme inhibition (e.g.
≥1
to moderate compounds, therapeutic
toxicity/potency proteins, PEGylated proteins, apixaban), 3) hormonal interaction (e.g. levonorgestrel), 4) alkylation
antibodies, adnectins (e.g. lomustine), and 5) inhibition of transporters (e.g. dapagliflozin).
B <1 Especially toxic/ Potent proteins, bispecific Their small size allows for the possibility of rapid diffusion across cell
potent biologics antibodies or other large membranes so that they can reach intracellular sites of action (Dough­
molecule biologics as erty and Pucci, 2011; Veber et al., 2002).
determined by a case-by-case Approximately 20% of all small molecule drugs approved during the
assessment
period of 2000–2008 were prodrugs (Huttunen et al., 2011) and they

4
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

Fig. 1. Occupational Exposure Control Banding Decision Tree for Therapeutic Modalities. Use of this decision tree, and the potencies included herein, should be
considered as a rough guide for initial band selection. Ultimate selection of the band should come from a qualified occupational toxicologist, and rely on the
consideration of a number of additional criteria such as the innate hazard of the therapeutic, its pharmacokinetics, and pharmacodynamics, among others. For
example, an additional safety factor may need to be applied for compounds which are mutagenic or teratogenic, resulting in a more restrictive band.
Footnotes: a ADC banding is generally driven by the warhead potency; b Follow guidance for radiolabeled compounds, banding is based on the API; c Biologic
materials made through biological processes (i.e. cell culture); d This biologically significant effect should also be clinically relevant; e Note that for live viruses,
enhanced control measures may be required. Abbreviations: aa = amino acids; sc = special case.

remain a significant portion of drugs being developed (Rautio et al., derivatives that are only one or two chemical or enzymatic steps away
2017). Prodrugs are inactive derivatives of active drug molecules that from the active parent drug. Prodrugs are typically designed to improve
must undergo an enzymatic and/or chemical transformation in vivo to the pharmacokinetic (PK) profile of the active drug and can also facili­
release/become the active drug, which can then elicit its desired phar­ tate the intracellular localization of drugs. The liver is generally assumed
macological effect in the body (Huttunen et al., 2011). to be the major site of first-pass metabolism of a small molecule drug
administered orally.
3.1.3. Marketed drugs There are many different routes of excretion including via the urine,
Small molecules make up the majority of marketed pharmaceuticals. bile, sweat, saliva, tears, milk, and stool. The majority of drugs are
Examples of small molecule pharmaceutical drugs include many car­ eliminated via pathways that involve the kidneys and/or the liver. A
diovascular drugs (e.g. Eliquis® or Apixaban), antivirals (e.g. Dacla­ major characteristic of compounds excreted in urine is that they are
tasvir or Daklinza®), and diabetes therapeutics (e.g. Farxiga® or polarized (i.e., charged) and water-soluble. Drugs that are lipid soluble
Dapagliflozin), to name a few. One example of a prodrug is Vyvanse® (a are not readily removed by the kidneys and require hepatic metabolism
form of amphetamine used to treat attention-deficit/hyperactivity dis­ (e.g. phase I and phase II biotransformation reactions) to increase their
order) which is designed to have less abuse potential than other am­ water solubility for possible urinary excretion (Kapusta et al., 2007).
phetamines due to the slower release of the active drug following Drugs with a MW exceeding 300 Da and with polar and lipophilic groups
bioconversion in the intestine/liver. Another example of a prodrug is are more likely to be excreted in bile (Lu et al., 2019).
omeprazole (a proton pump inhibitor used to treat acid reflux and ul­
cers) which is bioactivated site-selectively in the acidic conditions of the 3.1.5. Health hazards associated with therapeutic use
stomach. Due to their small size, the oral and inhalation BA of small molecules
is often quite high (>50%). In general, the hazards associated with small
3.1.4. ADME molecule drugs are due to their pharmacological effects or exaggerated
The absorption of a small molecule drug is highly variable and pharmacology. Hazards can also be due to off-target effects.
dependent on multiple factors including its hydrophobicity/hydrophi­ In general, the hazards of prodrugs are the same as for small mole­
licity, size, molecular charge/ionization state, and plasma/protein cules. Prodrugs can be metabolized into more active or less active forms,
binding potential. Small molecule drugs are generally administered which can contribute to their pharmacology and/or toxicity.
orally and are designed to have high oral BA, which can give them an
ease of use advantage over high MW drugs that require parenteral 3.1.6. Occupational hazard and exposure considerations
administration. Small molecule pharmaceuticals are designed to be Small molecules can elicit effects via the oral, dermal, inhalation and
stable alone and/or in formulation enabling a long shelf-life and the systemic routes (in addition to others) therefore exposure via each of
ability to reach cellular targets intact with consistent pharmacological these occupational exposure routes is of concern. Regarding inhalation
potency (Chen et al., 2018). In most cases, prodrugs are simple chemical exposure, it is important to consider the molecule’s potential to exert

5
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

direct effects on the lung (e.g. target receptor present in the lung, pul­ ‘unlocked’ (UNA – unlocked nucleic acid) with LNAs generally having
monary vasodilator) as well as the compound’s systemic BA via the higher stability resulting in a longer half-life and stronger base-pairing
inhalation route. In the absence of inhalation BA data, it is acceptable to potential than their unlocked counterparts. A typical ON is a short
conservatively assume 100% of the inhaled dose reaches the systemic chain DNA or RNA molecule, with a MW > 7000 Da, manufactured by
circulation. There is also the potential for dermal absorption for this an isolated/enclosed process such as solid phase synthesis followed by
class of compounds and therefore the potential for dermal toxicity due to preparative chromatographic purification and downstream processing.
dermal absorption and subsequent systemic exposure. This biosynthetic process is closer to a chemical process than a biological
process.
3.1.7. Occupational exposure banding guidance recommendation
When assigning a small molecule to an ECB (Fig. 1), the goal is to 3.2.2. How they work
ensure that employees are adequately protected throughout their tenure ONs are currently being investigated for the treatment of a variety of
working with the compound of concern. Since the data to support the diseases, and are primarily being administered by parenteral injection.
calculation of the OEL is not available in the early development space, ON drugs target mRNA and are generally synthesized to match a specific
assumptions need to be made to enable the determination of the nucleotide sequence of interest. Pharmacology is dependent on Watson-
appropriate ECB. These assumptions include information and confi­ Crick base pairing between the drug and an mRNA target molecule, a
dence in the dose where biologically significant (and clinically relevant) scenario that provides for both high affinity and exquisite specificity.
pharmacology is expected. Adhering to the suggested doses where bio­ ONs such as RNAi compounds are important tools for therapeutic use
logically significant pharmacology is observed in Fig. 1 provides a 100- because of the roles they play in controlling gene expression. ONs can
fold safety factor from the midpoint of the band for a 50 kg individual alter gene expression through a variety of mechanisms by targeting
(bodyweight recommended when establishing permissible daily expo­ mRNA for degradation by cellular RNase H activation or blocking
sure limits) (Agency, 2014). It is important to note that cytotoxic com­ ribosome initiation of protein translation (Templin et al., 2000).
pounds, mutagens, teratogens and hormones may require a greater
safety factor due to their potential for severely toxic effects. It is BMS’s 3.2.3. Marketed drugs
practice that the default band for pharmacologically-active, small Though there are many clinical trials ongoing (>100 ONs are in
molecule APIs can be considered to be ECB 4, however this default band clinical trials), there are only a few ON-based drugs approved to date,
should be reconsidered if a molecule is expected to be extremely which include Vitravene® (no longer marketed) and Macugen® (Stein
potent/toxic or based on professional judgment. and Castanotto, 2017).
Another helpful piece of information when establishing the ECB is
the projected (or actual) lowest human therapeutic dose. Compounds 3.2.4. ADME
which may have a therapeutic dose of <1 mg/day may need to be placed Once in the systemic circulation, ONs are generally taken up endo­
into ECB 5 or 5 special case in order to allow for an acceptable margin of cytically and distributed to the lysosomes or target mRNA. ONs cleared
safety, in this instance defined as the exposure margin between the from the circulation are taken up by the liver, kidneys, spleen, and bone
therapeutic dose and the midpoint of the occupational exposure band. marrow. In a study in which rats received an ON dosed as a single
Information on the PK of the compound such as half-life and oral BA, intratracheal (i.t.) instillation, ONs rapidly left the lung via absorption
along with pharmacodynamics (PD), should also be considered if into the systemic circulation and were renally excreted within minutes
available. For example, a compound with a long half-life (on the order of (Moschos et al., 2011).
days) may bioaccumulate, resulting in a lower acceptable occupational
exposure than for a compound with a half-life of hours. For compounds 3.2.5. Health hazards as a class/modality associated with therapeutic use
that are prodrugs, the ECB should be consistent with that of the more Adverse effects associated with an ON are generally related to the
active form, taking into account the rate of formation of the active and sequence’s ability to interfere with normal cellular function. Health
the route of exposure. hazards associated with ONs include target-mediated effects, effects
resulting from the binding of the ON to off-target DNA/RNA, and effects
3.1.8. Establishing default bands due to tissue accumulation (liver, kidney and lung [inhalation
In the situation where very limited data is available, the API can be exposure]).
assigned to a default band, based on the in vitro/in vivo potency, Oligonucleotides are generally not associated with genetic toxicity,
knowledge and/or experience with the therapeutic target and the pro­ developmental or fertility effects and have been well tolerated in the
jected human efficacious dose (if available). As mentioned previously, clinic. The primary effect of ON administration in rodents appears to be
the default band employed by BMS for pharmacologically-active APIs is a result of pro-inflammatory effects. Additionally, preclinical studies
ECB 4, however this default band should be reconsidered if a molecule is with ONs have demonstrated histopathologic alterations in the liver and
exquisitely potent or toxic or based on professional judgment resulting kidneys, which were dependent upon dose level, dosing frequency, and
from knowledge or experience with the therapeutic target and modality. duration of treatment (Alton et al., 2012). Effects on tissues are generally
For example, some modalities are known to have potent toxicity (e.g. related to deposition and accumulation. The most common effect in the
ADCs) and therefore can be assigned to a default ECB based on prior clinic is transient, blood-level-dependent effects stemming from inter­
experience or information on pharmacologically, toxicologically or action with plasma proteins. Theoretical causes for concern with ONs
structurally similar compounds. documented in the literature include the potential incorporation of
degradation products of phosphorothioate nucleotides into newly syn­
3.2. Oligonucleotides thesized DNA, as well as the binding of ONs to DNA resulting in triplex
formation that could ultimately induce site-specific mutations (Vasquez
3.2.1. Background et al., 2000). While a genotoxicity hazard due to DNA incorporation is
Oligonucleotides (ONs) are a novel class of therapeutic agents unlikely based on available data, it has been recommended that the
comprised of the nucleotides adenine, guanine, cytosine, thymine, and potential for triplex formation should be assessed if appropriate (Berman
uracil. ONs are usually made up of 13–25 nucleotides and are designed et al., 2016).
to hybridize specifically to DNA or RNA sequences. ONs include anti­
sense oligonucleotides (ASO), DNA duplexes, Aptamers, Spiegelmers, 3.2.6. Occupational hazard and exposure considerations
and RNA interference compounds (RNAi) such as small interfering RNA In the workplace, potential routes of exposure to ONs include inha­
(siRNAs) and microRNAs (miRNAs). ONs can be ‘locked’ (LNA) or lation, ingestion (via mucocilliary transport from the lung, hand to

6
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

mouth contact, etc.) and dermal contact. Overall, the inhalation BA is 3.3.3. Marketed drugs
expected to be low (Guimond et al., 2008), oral BA is generally unknown Approximately 140 peptide drugs were in clinical trials in 2015 with
and dermal BA will depend on such properties as the MW of the ON. >500 in preclinical development (Fosgerau and Hoffmann, 2015). Ex­
Regarding occupational exposures to ONs, the primary concern is amples of peptide-based medicines include Lupron™ for the treatment
immunogenicity due to their foreign nucleic acid structure. Based on a of prostate cancer and Byetta™ (exenatide) for Type 2 Diabetes. With
favorable PK profile, inhalation delivery is a therapeutic option for ONs regards to route of administration, most peptide therapeutics are in­
and pharmacologically relevant doses for the treatment of localized lung jectables (e.g. administered subcutaneously or intravenously) (Fosgerau
diseases will likely be in the range of 1 mg/kg or less (Templin et al., and Hoffmann, 2015).
2000). Regarding inhalation exposure, adverse effects associated with
inhaled ONs such as lung inflammation are: 1) typically dose related, 2) 3.3.4. ADME
primarily occur at high toxicological doses and 3) have been observed to Peptides are cleared by the same catabolic pathways used to elimi­
be reversible upon termination of exposure, suggesting regression of nate endogenous and dietary proteins, and are generally regarded as
lung inflammation in humans after exposure is terminated (Alton et al., having a predictable metabolism (Taft et al., 2009). They tend to be
2012). After inhalation exposure to ONs, reversible lung inflammation chemically and physically unstable in the general environment, are
has been observed in preclinical studies; however, no toxicity or prone to hydrolysis and oxidation, have a tendency for aggregation,
increased lung inflammation has been reported in patients or healthy exhibit a relatively short circulating plasma half-life and fast elimina­
individuals (Alton et al., 2012). A pro-inflammatory response observed tion, and have low membrane permeability (Fosgerau and Hoffmann,
in studies in higher species such as primates is expected to be of minimal 2015). Several drug delivery strategies employ binding the peptide of
consequence since the mixed monocellular infiltrate characteristic of the interest to the circulating protein albumin as a means of obtaining an
response in rodents is absent even after long-term exposure in monkeys. extended half-life, leading to less frequent dosing, in some cases only
However, the effect of prolonged inhalation exposure to ONs remains once weekly (Fosgerau and Hoffmann, 2015).
unknown as inhaled concentrations have been low and clinical trials
have been short in duration. 3.3.5. Health hazards as a class/modality associated with therapeutic use
It is important to note that each ON may have different intrinsic Inhalation of therapeutic peptides has been examined in the clinic
toxicological and PK properties following typical occupational routes of and in preclinical studies; however, in these studies the therapeutic
exposure. Therefore, as with small molecules, the specific chemistries of target(s) was present in the lung (Hartmann et al., 2015; Fellner et al.,
ONs warrant individual risk assessments. LNAs are expected to be more 2016; Kuehl et al., 2016; Onoue et al., 2011; Walker et al., 2017).
of a concern than naturally occurring DNA/RNA in the workplace. Inhalation studies with peptides have shown that they have the potential
to produce local immunogenicity and irritation upon chronic exposure
3.2.7. Occupational exposure banding guidance (Fellner et al., 2016). Given that therapeutic doses generally need to be
Despite their complexity in formulations and large molecular sizes, administered with a delivery device to ensure sufficient lung exposure,
ONs are considered more similar to small molecules than biologics in occupationally relevant airborne exposures to therapeutic peptides are
that they are manufactured by chemical synthesis processes and as such, expected to be lower than the levels which would cause adverse effects
are generally expected to follow the small molecule quality guidelines in the lung. The potential for direct lung effects (resulting from the
issued by regulatory agencies and the ICH. ONs can be evaluated in a therapeutic target being present in the lung), however, would need to be
similar manner to small molecules, and assigned to an ECB according to considered. For peptides where there is no target in the lung, the inha­
the system outlined in Fig. 1. The recommended default ECB for ONs lation route would be less of a concern, although inhalation BA would
with insufficient data is similar to that recommended for small mole­ still need to be considered. An evaluation of inhalation studies with
cules: ECB 4 (1 - <10 μg/m3; <100 μg/day). ONs which are not expected peptides conducted by Pfister et al. (2014) reported BAs of up to 100%,
to be highly toxic or carcinogenic can be assigned to ECB 4 (1 - <10 μg/ with most being <50% (Pfister et al., 2014a).
m3; <100 μg/day). ONs which may be highly toxic, mutagenic or
carcinogenic can be assigned to ECB 5 (0.1 - <1 μg/m3; <10 μg/day). 3.3.6. Occupational hazard and exposure considerations
Peptides are generally unstable in the environment (e.g. degrade
3.3. Peptides upon prolonged exposure to light, temperature-sensitive), therefore any
pharmacological activity may be lost over time should the material be
3.3.1. Background present on a work surface (Krause and Sahin, 2019). Although there are
Peptides are short chains of amino acid monomers linked by amide a few marketed oral peptides, in general peptides are not orally
bonds, the covalent chemical bonds formed when the carboxyl group of bioavailable due to their instability in the GI tract (Fosgerau and Hoff­
one amino acid reacts with the amino group of another. Peptide thera­ mann, 2015). Dermal exposure would also be of limited concern due to
peutics are distinguished from proteins on the basis of size. Generally, a the size of these compounds, the fact that they are naturally occurring
peptide is comprised of approximately 50 amino acids or less (Baldo, proteins and their relative instability under environmental conditions.
2015). Exposure via inhalation is a concern, especially when there are potential
targets present in the lung (e.g. compounds that target glucocorticoid
3.3.2. How they work receptors). Additionally, enzyme proteins have been reported to cause
Naturally occurring peptides function in crucial physiological roles occupational allergies such as asthma (Basketter et al., 2015). Needle
such as hormones, neurotransmitters, growth factors, ion channel li­ sticks and sharps exposures are also a concern as there is a potential for
gands, or anti-infectives (Fosgerau and Hoffmann, 2015). Peptides as immunogenicity upon systemic exposure (Fosgerau and Hoffmann,
therapeutics are recognized for being highly selective and efficacious as 2015).
well as relatively safe and well tolerated. They also are generally asso­
ciated with lower production complexity compared with protein ther­ 3.3.7. Occupational exposure banding guidance recommendation
apeutics and small molecules. Current development efforts involve Based on the chemical synthesis used, small size and higher BA of
peptide targets with emerging peptide technologies inclusive of multi­ these compounds compared with large MW biologics, peptides can be
functional and cell penetrating peptides, as well as peptide drug banded according to the small-molecule ECB system (See Fig. 1). Even if
conjugates. made via a biologic process, their small size precludes them from the
BCC system. Similar to small molecules, peptides can be placed in ECB 4
as a default band. A more conservative ECB can be selected if the

7
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

compound is an extremely potent peptide (i.e. Calcitonin), if it can cause Based on the types of biological targets and modes of action, bispe­
direct lung effects or exhibit pharmacology at occupationally relevant cific immunotherapies can be divided into three main categories:
exposures and/or if the compound is a respiratory sensitizer. In the case
of enzyme proteins, suggested industry exposure and handling guide­ (1) T-cell redirectors: Cytotoxic effector cell redirector bsAbs target a
lines are presented in Basketter et al., 2015 (Basketter et al., 2015). tumor-associated antigen (e.g. CD19, CD20, epithelial cell
adhesion molecule [EpCAM], B-cell maturation antigen [BCMA])
3.4. Biologics and the T-cell receptor/CD3 complex, which activates cytotoxic T
lymphocytes, thereby redirecting T-cell cytotoxicity to malignant
3.4.1. Background cells.
Biologics are considered therapeutic proteins, antibodies, enzymes, (2) Tumor-targeted immunomodulators: These bispecific immuno­
adnectins or other active biologic materials that are made through therapies bind to a tumor-associated antigen and an immuno­
biological processing (i.e. cell culture). modulating receptor, such as CD40. Such compounds are usually
Monoclonal antibodies: Monoclonal antibodies are typically large designed to be inactive until binding the tumor antigen, thereby
molecules which have a MW > 140 kDa and are designed to target localizing immune stimulation to the tumor environment, while
specific proteins. minimizing immune activation elsewhere. This is expected to
Bispecific antibodies: Bispecific antibodies (bsAbs) are emerging as induce powerful activation of tumor-specific T cells with reduced
the next generation of antibodies. Bispecific antibodies generally have a risk of immune-related adverse events.
MW ranging from 50 to 60 kDa, and have the potential to improve (3) Tumor-targeted dual immunomodulators: Bispecific compounds
clinical efficacy as well as safety by targeting two distinct immunoreg­ that bind two distinct immunomodulating targets, often
ulatory pathways (Dahlen et al., 2018; Brinkmann and Kontermann, combining targeting of PD-1 or PD-L1 with that of lymphocyte-
2017). activation gene 3 (LAG-3) or T-cell immunoglobulin and mucin-
Probodies: Probodies are similar to prodrugs but are antibodies domain containing-3 (TIM-3). The rationale is to induce supe­
engineered to remain inert until activated proteolytically in diseased rior tumor immunity compared to monospecific antibodies to the
tissue. In principle, any therapeutic antibody can be converted into same targets (Dahlen et al., 2018).
probody form. There are two types of probodies, conventional IgG-based
probodies and probody-drug conjugates (Polu and Lowman, 2014). Probodies: Probodies leverage the upregulation of protease activity
Fusion proteins: Fusion proteins generally consist of a peptide (or in diseased tissue (e.g. cancer, inflammatory diseases) to achieve
other short-lived effector domain) coupled to a ‘carrier’, which is usually disease-tissue-specific therapeutic activity. The probody remains intact
a protein or peptide that contributes to the functional properties of the and is blocked from binding to the antigen target, however, once the
resultant fusion protein. As peptides have a short half-life owing to linker peptide is cleaved by proteases that are selectively activated in the
proteolytic degradation and are usually rapidly cleared (within minutes) diseased-tissue, the masking peptide is released, allowing the active
via the kidneys, the peptide can be linked to a protein/fusion partner to antibody to bind to its target, resulting in tissue-specific activity (Polu
enable a more stable molecule with an extended half-life. The crystal­ and Lowman, 2014). Preclinical studies have demonstrated that a
lizable Fc region of human IgG1 antibody is commonly utilized as a therapeutic antibody with known on-target toxicity can be reengineered
fusion partner for the effector molecule(s) because it extends the fusion as a probody retaining potent in vivo efficacy, but with greatly reduced
protein half-life by recycling via the salvage neonatal Fc receptor (FcRn) side effects (Desnoyers et al., 2013).
receptor, and protects the molecule from lysosomal degradation. The Fusion proteins: Fusion proteins have been used in the biophar­
linked effector peptide may have widely varying properties contributing maceutical industry for over 25 years to improve the PK properties of
to recognition, binding, and toxicity, while its fused partner may aid in otherwise short half-life biologics (Strohl, 2015). Fusion proteins offer
stability and targeting of the chimeric polypeptide (Baldo, 2015). the potential to target a therapy to the location of disease by combining a
targeting component such as a mAb with a therapeutic peptide or pro­
3.4.2. How they work tein such as a cytokine. This avoids short half-lives, dose-limiting tox­
Monoclonal antibodies: Antibodies, or immunoglobulins (Igs), are icities, and sub-optimal localization of the therapy, which may be
large Y-shaped molecules with Fab (fragment, antigen binding) regions encountered if the unbound therapeutic peptide or protein was sys­
and an Fc (fragment crystallizable) region. Fab regions consist of two temically administered. Many approved fusion proteins work as agonists
variable domains that are designed to recognize and bind to specific (e.g. alefacept) or antagonists (e.g. belatacept, etanercept) of receptor
antigens. Binding triggers the endocytotic internalization of the mAb function, or by a direct targeted cytotoxic killing effect (e.g.
and subsequent lysosomal degradation (Ryman and Meibohm, 2017). denileukin-diftitox). Antibody-cytokine fusion proteins, often referred
The Fc region interacts with cell surface receptors, and allows the mAb to as immunocytokines, are being utilized to employ the tumor-targeting
to activate the immune system (Kennedy et al., 2018). Antibody thera­ ability of mAbs to guide the cytokines specifically to tumor sites where
peutics can have activity via antibody dependent cellular cytotoxicity they can stimulate anti-tumor immune responses while avoiding
(ADCC) or complement dependent cytotoxicity (CDC). In ADCC medi­ dose-limiting systemic toxicity (Young et al., 2014; Hutmacher and Neri,
ated effector activity, the Fc portion of the mAb binds to an FccR on an 2019).
effector cell, such as a monocyte, macrophage, or natural killer cell, and
the Fab domains bind to cell surface receptors on the target cell. This 3.4.3. Marketed drugs
leads to the destruction of the target cell via phagocytosis by the immune Monoclonal antibodies: Currently, most mAbs developed are hu­
cell or release of cytokines leading to cell death (Ryman and Meibohm, manized or fully human. The use of mAbs has expanded exponentially
2017). Monoclonal antibodies are potent and highly selective and they during the last decade and currently covers several therapeutic areas,
can be agonists, antagonists or neutralizing. such as oncology, respiratory diseases, hematology, immunology, car­
Bispecific antibodies: Bispecific antibodies can target two antigens diovascular diseases, and inflammatory diseases (Singh et al., 2018). In
at once. Companies are developing bispecific antibodies with the 1992, the FDA approved the first therapeutic mAb Muromonab-CD3
objective of creating drugs which encompass the function of two (trade name Orthoclone OKT3) to reduce acute rejection in patients
monospecific drugs or which have properties that cannot be achieved with organ transplants (Cai, 2018). As of May 10, 2018, the FDA had
with a mixture of monospecific compounds. However, the development approved 80 therapeutic mAbs with indications includinging
of bsAbs is considerably more challenging than development of con­ immune-mediated disorders (e.g. brodalumab, dupilumab) and cancer
ventional mAbs. indications (e.g. avelumab, ocrelizumab) (Cai, 2018; Kaplon and

8
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

Reichert, 2018). (Trivedi et al., 2017).

Cancer immunotherapies have been established as a highly effective Protein therapeutics are cleared via the same catabolic pathways
therapeutic option. Ipilimumab was a first-in-class T-cell potentiator utilized to eliminate endogenous and dietary proteins (Taft et al., 2009).
that works by blocking cytotoxic T-lymphocyte antigen-4 (CTLA-4), a Due to the high binding specificity and affinity of the mAb for its target,
critical protein receptor that downregulates the immune system (Tarhini target-mediated drug disposition (TMDD) is a major route of elimination
et al., 2010). Other approved mAbs include programmed death 1 cell for many mAbs with a membrane-standing target, especially at low
surface receptor (PD-1) and programmed death-ligand 1 (PD-L1) in­ doses and concentrations. TMDD consists of receptor-mediated endo­
hibitors such as nivolumab which binds to the PD-1 receptor and blocks cytotic internalization of the IgG molecule and subsequent lysosomal
its interaction with PD-L1 and PD-L2, releasing PD-1 pathway-mediated degradation. The rate of elimination of a protein therapeutic is depen­
inhibition of the immune response, including the anti-tumor immune dent on the expression of the target receptor, the affinity of the mAb for
response (Singh et al., 2018). the receptor, the dose of the mAb, the rate of receptor therapeutic
Bispecific antibodies: T-cell redirecting therapies are the most protein internalization, and the rate of catabolism within the target cell.
established class of bsAbs, with two approved products, blinatumomab It is important to note that the antibodies cleared primarily by TMDD
and catumaxomab (later withdrawn), and several others in clinical and will have dose-dependent nonlinear elimination. Additionally, the for­
preclinical development (Dahlen et al., 2018). In July 2017, the FDA mation of anti-drug antibodies (ADAs) can neutralize mAbs and cause
approved blinatumomab (aka Blincyto, CD3XCD19), a bsAb for acute them to be cleared faster (Vande Casteele and Gils, 2015).
lymphoblastic leukemia (ALL) in adults and children. Blinatumanb tar­ The clearance of therapeutic monoclonal antibodies (mAbs) typically
gets CD19, a protein on the surface of normal and B-cell malignancies, does not involve cytochrome P450 (CYP450)-mediated metabolism or
and the CD3 receptor on the surface of cytotoxic T lymphocytes. Addi­ interaction with cell membrane transporters (Ferri et al., 2016).
tionally, the first FDA-approved non-oncology bsAb was emicizumab Although both the kidney and liver can metabolize proteins by hydro­
(bsAb recognizing coagulation factor IXa and the substrate factor X) for lysis, there is minimal clearance of protein therapeutics via conventional
the treatment of hemophilia A in 2017 (Labrijn et al., 2019; Lenting renal and biliary excretion mechanisms (Taft et al., 2009). Unlike small
et al., 2017). molecules, mAbs are too large to be filtered by the kidneys and are not
Probodies: Currently a pipeline of probody candidates in oncology eliminated in the urine in healthy individuals (Ryman and Meibohm,
are in preclinical development and clinical trials with the potential to 2017). Biliary excretion accounts for a very small amount of the elimi­
reduce side effects and broaden the range of effective doses. However, nation of IgG antibodies. Thus, IgG elimination occurs mostly through
there is no approved probody for therapeutic use (Autio et al., 2020). intracellular catabolism by lysosomal degradation to amino acids after
Fusion proteins: More than a dozen fusion proteins have received uptake by either pinocytosis, an unspecific fluid phase endocytosis, or by
regulatory approval for human therapy including etanercept, belatacept a receptor-mediated endocytosis (Ryman and Meibohm, 2017).
and denileukin-diftitox (Strohl, 2015). Most approved fusion protein
therapies are Fc-fusion proteins, which utilize the Fc fragment of human 3.4.5. Health hazards as a class/modality associated with therapeutic use
IgG1 to interact with Fc receptors on immune cells. Importantly, the As biologics are specific for their target(s) and are generally
neonatal Fc receptor (or FcRn) normally functions to transport Igs across administered systemically, they can be especially potent. Bispecific an­
cells and to protect circulating Ig from degradation. tibodies, for example, can be quite potent, with in vitro potency in the
Etanercept (Enbrel®) was the first chimeric fusion protein to gain low picomolar range and in vivo potency at doses <0.1 mg/kg. In an FIH
regulatory approval in 1988. Etanercept is a dimeric fusion protein (MW study where catumaxomab was administered IV, fatal acute liver failure
150 kDa) consisting of a tumor necrosis factor α (TNF-α) receptor ligand- and cytokine release-associated systemic toxicity were observed at doses
binding region linked to the Fc portion of human IgG1 which acts by as low as 10 μg (Mau-Sorensen et al., 2015).
binding tumor necrosis factor (TNF) thereby inhibiting the interaction of Often, mAbs are not tested for their genotoxicity or carcinogenicity
this cytokine with cell surface TNF receptors and ultimately reducing the since they are not expected to interact directly with DNA or other
ensuing inflammatory response. Etanercept is indicated for the treat­ chromosomal material. Concerns for effects on fertility and develop­
ment of rheumatoid and other forms of arthritis and is one of the most mental toxicity generally depend on the pharmacology of the biologic.
commercially successful fusion protein therapies (Baldo, 2015). For example, compounds which activate the immune system have the
Belatacept (Nulojix®) is a fusion protein of the Fc fragment of human potential to cause developmental toxicity, evidenced by the fact that
IgG1 and the extracellular domain of CTLA-4, indicated as prophylaxis spontaneous abortions have been observed in ePPND monkey studies
against organ rejection in adult patients receiving a kidney transplant. with immune-modulators, therefore developmental toxicity is a concern
Belatacept blocks CD28-mediated T-cell activation and the production for biologics with these modes of action (Brennan et al., 2010). Another
of cytokines by binding CD80/CD86 on antigen presenting cells. example involves biologics which target EGFR, a receptor involved in
fetal development, which is likely essential for normal organogenesis,
3.4.4. ADME cell proliferation, and cell differentiation in the developing embryo. The
Due to their large size and potential to be broken down in the GI mAbs bevacizumab, cetuximab, panitumumab, and trastuzumab bind
tract, most “large molecule” biologics (e.g. mAbs, bsAbs, fusion pro­ with the EGFR and inhibit the function of the receptor, resulting in
teins) are primarily administered by intravenous (IV) infusion or injec­ developmental toxicity (Halsen and Kramer, 2011). Additionally, mAbs
tion rather than via the oral route. They are designed to be stable in the IgG2 subclass are able to cross the placenta (Halsen and Kramer,
molecules and to have long half-lives typically in the range of days or 2011).
weeks supporting intermittent administration (Trivedi et al., 2017). Immunogenicity is a concern with administration of biologic mate­
Additionally, the conjugation or pegylation of proteins can extend the rials. The immunogenic potential of a biologic increases with the pro­
half-life further (Baldo, 2015). Fusion proteins have been noted to have portion of foreign protein, and therefore humanized mAbs have less
significantly shorter half-lives then general mAbs (Baldo, 2015). sensitizing potential than murine mAbs (Halsen and Kramer, 2011).
Notably, large molecules have different PK profiles than conven­ Although Fc fusion proteins are generally safe, adverse events observed
tional small molecule drugs. The PK and serum half-life of bsAbs may or with fusion protein therapies include Type I, II, III and IV hypersensi­
may not be linear depending on the presence of an Fc domain, the tivities such as anaphylaxis, cutaneous manifestations, infusion, and
absence of which can lead to linear PK. For example, blinatumomab is a injection site reactions (~20–50% of patients), and cytokine release
recombinant non-glycosylated protein that does not have an Fc domain, syndrome (CRS) (Baldo, 2015). The mechanism of action should also be
thus it does not undergo FcRN-mediated recycling and has a short taken into account when determining the health hazards associated with
elimination half-life of roughly 2 h and negligible renal clearance the fusion protein being assessed. Many fusion proteins effect the

9
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

immune system, so increased risk of infection, reduced immune system target is present in the lung since adverse effects would not be dependent
function and autoimmune responses are a concern. For example, the on systemic BA in this case.
most common adverse effects of etanercept are relatively mild and Due to the high potency and generally low clinical doses associated
consist of fever, headache, injection-site reactions, mild allergic re­ with bsAbs, the default BCC for bsAbs is considered to be BCC B (<1 μg/
actions, and pruritus (Baldo, 2015). Importantly, some fusion m3), even after taking into account the low inhalation BA and high MW.
protein-induced skin responses may represent direct targeting events
that are not genuine hypersensitivities and are similar to, for example, 3.5. Antibody-drug conjugates
agents that bind EGFR causing non-immune-mediated adverse cuta­
neous events (Baldo, 2015). Proteins conjugated with a polyethylene 3.5.1. Background
glycol (PEG) carry additional safety concerns around the lack of Antibody drug conjugates (ADCs) are a novel class of therapeutic
biodegradability of the PEG component. agents typically developed as treatments for cancer that are character­
It has been shown that FcRn is expressed in both the upper and ized by an antibody scaffold covalently modified with a variable number
central airways in non-human primates as well as in humans. After of small-molecule payloads. The payloads are generally highly potent
deposition of the aerosolized protein in the lung, transport of the Fc- cytotoxic chemicals (known as warheads) and are bound to the antibody
fusion proteins occurs through binding to FcRn in the epithelial cells scaffold via synthetic linkers. Such immuno-conjugates combine the
followed by transport to the underlying tissue and ultimately into the anti-tumor potency of highly cytotoxic small-molecule drugs (300–1000
systemic circulation. The bioavailabilities for Fc-fusion molecules in Da, with subnanomolar IC50 values) with the high selectivity, stability,
non-human primates have been in the range of 20–50% when given by and favorable pharmacokinetic profile of monoclonal antibodies (Drake
inhalation (Dumont et al., 2005). Monomeric fusion proteins are ex­ and Rabuka, 2017). These drugs are hybrid entities combining both
pected to have higher bioavailabilities than dimeric due to their biologic and small-molecule characteristics, where the antibody serves
decreased size among other factors (Dumont et al., 2005). For example, to target the small molecule specifically to the intended cell type.
the BA of the therapeutic protein erythropoietin-Fc fusion molecule
(~112 kDa) in monkeys was ~6% with the dimer and ~35% with the 3.5.2. How they work
monomer (Dumont et al., 2005). Additionally, many biologics are ADCs are currently being investigated for the treatment of a variety
designed to have long half-lives and may bioaccumulate upon repeated of cancers. The mechanistic basis for the ADC activity includes its spe­
dosing. cific binding to the cellular target, triggering ADC cellular internaliza­
tion by pinocytosis, followed by the intracellular release of the payload
3.4.6. Occupational hazard and exposure considerations (Roberts et al., 2013). Therefore, after administration into the systemic
As biologics are generally specific for their target(s) and can be quite circulation, the antibody fragment guides the complex to the targeted
potent, they should be evaluated on a case-by-case basis. The potential tumor antigen where it becomes internalized into the cancer cell (Hasan
for accumulation due to long half-lives, the pharmacodynamic profile et al., 2018). The intracellular release of the payload in its active form
and the potential for developmental toxicity should be considered in the results in cell death. The three components of ADCs, namely the anti­
hazard assessment for biologics. body, the linker and the cytotoxic drug, have their own inherent char­
Regarding exposure risks, exposures via the dermal, oral and inha­ acteristics and limitations which influence each other, thus finding the
lation routes are generally expected to be low. Compounds >500 Da are best combination of these components to design an ideal ADC is a
not expected to be bioavailable via dermal exposure (Bos and Meinardi, complicated task (Hasan et al., 2018). Although this targeted approach
2000). Low systemic exposure potential is generally expected via the appears straightforward, its translation to clinical practice has been
inhalation route of exposure (e.g. BA<1%) in the occupational setting problematic with initial attempts failing due to inappropriate linker
for biologics with a MW > 10 kDa (Gould et al., 2018; Pfister et al., systems or insufficiently potent cytotoxins, resulting in unfavorable
2014b). Limited studies conducted on the inhalation BA of Fc-fusion therapeutic indices. Modern ADCs tend to use highly potent cytotoxic
proteins, have shown BAs of 20–50%, which is substantially higher molecules, such as derivatives of calicheamicin, maytansine and auri­
than the BA observed with mAbs, emphasizing the importance of statins (Senter, 2009; Ducry and Stump, 2010). A number of these
considering the binding component when estimating the inhalation BA molecules have demonstrated substantial in vivo antitumor activity. To
of fusion proteins (Gould et al., 2018; Dumont et al., 2005; Pfister et al., translate the fundamental advantage of ADCs in targeting
2014b). Low BA may not be the case when there are drug targets present tumor-selective or tumor-specific antigens in clinical practice, it is
in the lung/skin. essential that linkers be of sufficient stability to minimize systemic
Regarding sensitization potential, there are currently no validated in exposure to the cytotoxin and the resulting toxicities, while still
vitro or in vivo models for respiratory sensitization potential. Available providing sufficient targeted intracellular release of the cytotoxic agent
screening tools are generally based on historical data for known respi­ (Roberts et al., 2013).
ratory sensitizers; therefore, the structure of the biologic should be
examined for potential similarities to known respiratory sensitizers. 3.5.3. Marketed drugs
Occupational exposures to therapeutic proteins (like other drugs) There are currently four ADCs marketed in the US, although there are
should be kept to a minimum (de Lemos et al., 2018). While exposure over 60 ADCs in clinical trials (Garcia-Alonso et al., 2018). Gemtuzumab
can never be truly proven to be zero, occupational exposure limits, en­ ozogamicin (Mylotarg™) was the first marketed ADC, originally
gineering controls, administrative controls, and PPE are necessary to approved in 2000 for treatment of CD33-positive acute myeloid leuke­
protect employees. Dermal exposure to drugs is never recommended and mia (AML). However, it was withdrawn from the market in 2010 due to
this remains the recommendation for biologics. treatment-related toxicity concerns. Mylotarg™ was reapproved in 2017
with a lower recommended dose and altered dosing schedule for the
3.4.7. Occupational exposure banding guidance recommendation treatment of adults with newly diagnosed CD33-positive acute myeloid
The default BCC for mAbs, probodies and fusion proteins which are leukemia (AML) and for patients ages 2 years and older with relapsed or
expected to exhibit pharmacology at doses of ≥1 μg/kg is considered to refractory CD33-positive AML. In 2011, brentuximab vedotin (Adcet­
be BCC A (≥1 μg/m3), after accounting for the low systemic BA ris©), an anti-CD30 monomethyl auristatin E (MMAE) conjugate gained
following inhalation (e.g. less than 1%) for compounds with a MW > 10 market approval for the treatment of relapsed/refractory Hodgkin
kDa. For biologics where a pharmacological effect is expected at <1 μg/ lymphoma and systemic anaplastic large cell lymphoma. In 2013, tras­
kg, BCC B is generally appropriate. It is also important to take into tuzumab emtansine (Kadcyla©), an anti- HER2 directed ADC conjugated
consideration the potential for direct lung effects and when the drug to DM1 (also called mertansine, a potent antimicrotubule agent) was

10
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

approved for the treatment of HER2+ metastatic breast cancer. Addi­ payload. The toxicity of other types of linkers would need to be evalu­
tionally, in 2017, another ADC was approved by the FDA, which ated on a case-by-case basis. Potential issues include the possibility of
included a cytotoxic agent from the class of calicheamicins. Inotuzumab linking to endogenous proteins or other cellular macromolecules, or
ozogamicin (Besponsa©), which targets CD22 was approved for relapsed altered immune responses.
or refractory B-cell precursor acute lymphoblastic leukemia (Garcia-A­ For the ADC as a whole, it is important to note that even though the
lonso et al., 2018). payload is connected to an antibody, it does not mean that the toxicity to
the payload will be reduced. There is a strong possibility that the
3.5.4. ADME payload could be released despite not having a (tumor) binding site.
ADCs are large molecules (~150 kDa) typically administered sys­ Therefore, it is critical to account for the hazards associated with the
temically by intravenous (IV) infusion (Malik et al., 2017). After sys­ payload and not to assume the linker to the antibody is systemically
temic administration, the distribution of the intact ADC is generally stable.
confined to the plasma, interstitial fluid, and lymph (Han and Zhao,
2014). They often possess long plasma residence times because of their 3.5.6. Occupational hazard and exposure considerations
dynamic interactions with the FcRn (Garg and Balthasar, 2007) and are In the workplace, potential routes of exposure to the intact ADC or its
therefore administered intermittently, typically once every 1–4 weeks respective components include inhalation, ingestion (via mucocilliary
(Malik et al., 2017). In most, but not all cases, ADCs exhibit nonlinear PK transport from the lung, hand to mouth contact, etc.) and dermal contact
and dose-dependent clearance due to saturable target binding and (depending on MW, etc.). The overlying concern with occupational
elimination (Levy, 1994). Plasma stability and the extent of deconju­ exposure to these types of molecules is the highly potent and toxic na­
gation in circulation can be estimated with direct measurements of the ture of the payload component. A critical attribute of any ADC is the
conjugated fraction (Kraynov et al., 2016). The free payload concen­ amount of drug loading, or the average ratio of conjugated payload to
trations can be used to evaluate payload dependent toxicity, although antibody. This is referred to as the drug-to-antibody ratio (DAR). Since
this toxicity is difficult to distinguish from toxicity that occurs when the the DAR dictates the amount of payload delivered per internalized
conjugated antibody is taken up into normal tissues (Poon et al., 2013). antibody, it strongly influences both efficacy and toxicity. In addition,
ADCs are designed to bind membrane-bound target proteins that can depending on the conjugation and linker technologies used, high-DAR
facilitate internalization into tumor cells by receptor-mediated endo­ ADCs can have poor biophysical characteristics (e.g., hydrophobicity,
cytosis (Ritchie et al., 2013). Because of the high affinity of the initial aggregation) that reduce efficacy and increase toxicity (Drake and
binding interaction, ADCs undergo target-mediated drug disposition Rabuka, 2017). Another important factor to consider when assessing the
(TMDD), meaning that the properties of the target influence the PK of toxicity of the ADC and individual components is the relative size of the
the drug (Mager and Jusko, 2001). Target expression, internalization, component molecules. While the payload and linker may have molar
turnover, accessibility and binding affinity all impact the PK of ADCs. masses in the several hundreds of Da, the antibody will typically have a
After administration, ADCs distribute to occupy target sites that are molar mass of around 150,000 Da, though antibody fragments may be
present in both normal and diseased tissue. significantly smaller. As such, the payload element of the ADC may only
constitute less than 0.5% by weight of the total MW; a potentially
3.5.5. Health hazards as a class/modality associated with therapeutic use important fact when safe exposure levels are typically quoted in mass
The health hazards of ADCs may be dissected into the relative parts terms. Of course, with smaller antibody fragment utilization, this effect
or building blocks of these hybrid molecules and can also be evaluated will be reduced and the toxic payload will constitute a greater propor­
for the ADC as a whole. There are the obvious hazards associated with tion of the compound mass.
the highly potent cytotoxic payload, hazards associated with the During each step to the manufacturing process, the hazards may
monoclonal antibody scaffold, and hazards associated with the linker change based on the nature of the components being utilized in that step
molecules, which can then vary depending if the linker is bound with the of the manufacturing process. The relatively low toxicity of antibody
payload or antibody. proteins, and the common processing of such macromolecular materials
The potent cytotoxic payloads typically utilized in ADCs include in enclosed solution or suspension forms, results in a relatively low risk
DNA double strand breakers such as Calicheamicin (Mylotarg©) or of intolerable exposure by main traditional routes (inhalation, ingestion
esperamycin; DNA alkylating agents such as duocarmycin, cyclo­ and skin absorption). In addition, the BA of such large biologically
propabenzidole (CBI), or pyrrolobenzodiazepines (PBD); and inhibitors derived macromolecules by traditional exposure routes (ingestion and
of tubulin polymerization such as maytansine (T-DM1), auristatin (SGN- airborne inhalation) is low due to instability of proteins in the GI tract as
35), or tubulysin (Beck et al., 2017). These agents typically induce cell well as low inhalation BA of such large molecules from the respiratory
death in rapidly dividing tissues and induce hematologic and GI toxicity. tract. The processes used for the manufacture of payload and subsequent
Genotoxicity, and reproductive and developmental toxicities are also conjugation to the antibody scaffold are typical “wet chemistry”
associated with exposure to these highly potent cytotoxic drugs, and covering all normal synthetic chemistry and pharmaceutical processing
neurotoxicity is a known adverse effect of the tubulin disruptors steps for small molecules, which includes the synthesis of the payload
(Remesh, 2012). Radio-nucleotides were also historically used as ADC and linker and associated purification and isolation steps including
payloads, but have been replaced in modern ADCs by the use of these chemical reaction, chromatography distillation, filtration, crystalliza­
highly potent cytotoxic drugs (Hasan et al., 2018). tion, drying and lyophilization, with solvent recovery and emission
The antibody scaffolds may have potential toxicity resulting from the control. Given the known toxicities of payload materials, activities such
intended and unintended pharmacological action as well as immuno­ as payload manufacture and handling of pure payload material prior to
genic effects associated with exposure to proteins. Proteins are thought conjugation, and especially any processes involving open handling of
to have lower potential to cause harm, and most of these are humanized such materials, especially in a dry powder form, are considered a high
antibodies with a lower potential for off target immunological events. risk of unacceptable exposure. Similarly, ancillary activities such as
They are designed to preferentially target cancer cells rather than Quality Control (QC) testing and cleaning, where exposure to the
normal tissues, which can lower the hazard of off target pharmacology. payload as either a trace residue or component of a sample may occur,
The linkers are typically small molecules with hazards that may vary should also be considered.
and need to be assessed on a case-by-case basis. Linking strategies that Major concerns during synthesis and conjugation will include all
take advantage of the properties of endogenous amino acids (such as activities where manual intervention is required, transfer of materials
engineered antibodies or peptides) are unlikely to be toxic on their own, between processes except in sealed transit routes, and in recovery and
or to significantly contribute to the toxicity of either the antibody or the storage of the high toxicity material in a form which may present

11
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

enhanced emission risks by certain routes; for example as a dry friable transported rapidly into cells by amino acid transporters (uptake time
solid for airborne transfer, or as a solution in an organic solvent for 3–5 min) (Clarke, 2018).
transdermal transfer. The high toxicity of the payload and of the ADC NETs have receptors for somatostatin, a hormone that regulates the
requires a more rigorous consideration of the routes of exposure than endocrine system. 68Ga-Dotatate (AAA’s NETSPOT®) is a positron-
might be typical for small molecule applications where typically only emitting analogue of somatostatin, which works by binding to somato­
airborne and (occasionally) surface transfer routes are considered. It is statin receptors enabling imaging of NETs. 68Ga-DOTATATE is
important to note that a payload could be released from an ADC despite comprised of DOTA-TATE, an amino acid peptide, with a covalently
not having a binding site. Therefore, it is critical to protect directly for bonded DOTA bifunctional chelator, bound to radionuclide gallium-68
the contribution of the payload and not to assume the linker to the (Clarke, 2018).
antibody is systemically stable. The extreme toxicity of the payload
warrants consideration of all routes, with control of hand contamination 3.6.4. ADME
in particular being a concern as this is a major transfer route into ocular Distribution of a PET agent is going to depend on its receptor spec­
and ingestion routes of exposure. The high toxicity and uncertainty of ificity and intended area(s) of localization. Based on clinical studies, the
exposure uptake efficiencies means that other activities that may lead to organs with the highest radiation from 18F radiotracers are the urinary
exposure to trace levels of residues, such as might occur during manual bladder and gallbladder followed by the kidney, liver, pancreas and
cleaning of contaminated equipment, may be significant, and the po­ lungs (Zanotti-Fregonara et al., 2013). The elimination half-lives are
tential and mechanisms for equipment and containers to become generally in the order of hours, for example with those of Axumin® and
contaminated on exposed external surfaces should also be assessed. 68Ga-DOTATATE being 110 min and 68 min, respectively (Clarke,
2018).
3.5.7. Occupational exposure banding guidance recommendation
Occupational exposure banding recommendations are based on the 3.6.5. Health hazards as a class/modality associated with therapeutic use
inherent hazards associated with exposures to each component of the The pairing of a radioactive element with a targeted drug/compound
ADC manufacturing process. Therefore, those components containing results in a dual hazard, one from the toxicity of the compound or drug
the intact or generally active payload, which would include the payload substance and one from the radioactive effects. Non-ionizing radiation is
itself and the payload plus the linker, which are all small molecules essential to life, but excessive exposures will cause tissue damage. All
components, are assigned to ECB 5 special case (<0.1 μg/m3) due to forms of ionizing radiation have sufficient energy to ionize atoms that
their highly potent cytotoxic hazards. The intact ADC, while retaining may destabilize molecules within cells and lead to tissue damage
the cytotoxic hazards, is assigned to ECB 5 (0.1 - < 1 μg/m3) due to its (Occupational Safety, 2020). The hazards attributable to the small
mixed biological/small molecule dilution effect. The intact antibody molecule or target-oriented compound are aligned with those described
components are assigned to Biologic Control Category A (≥1 μg/m3). herein under the appropriate modality.
The linker molecules are typically evaluated on a case-by-case basis, but
generally fall within ECB 3 (10 - <100 μg/m3; <1000 μg/day) or ECB 4 3.6.6. Occupational hazard and exposure considerations
(1 - <10 μg/m3; <100 μg/day). With regards to the occupational hazard of radiation exposure, the
guiding principle of radiation safety is that exposure be controlled to “as
3.6. Positron-emission tomography (PET) tracers low as reasonably achievable” (“ALARA”) (Control CfRadiatio, 2015).
Radioactive hazards are addressed using the techniques detailed by
3.6.1. Background several organizations (Occupational Safety, 2020; EPA, 2020); and are
Positron-emission tomography (PET) is a nuclear medicine func­ focused on limiting the time spent near a radiation source, maximizing
tional imaging technique that is used to observe biochemical processes the distance from the radiation source and shielding from a radiation
in the body and for the detection of many diseases. PET imaging tracers source (Directive C, 2013).
enable molecular imaging that relies on derangement of physiological
and biochemical processes for the detection of many diseases (Hicks 3.6.7. Occupational exposure banding guidance recommendation
et al., 2006). The compounds and/or drug substances which are “attached” to the
tracer should be addressed as if there were no radioactive tracers present
3.6.2. How they work and would maintain their ECB or BCC that existed prior to the
The PET imaging system detects pairs of gamma rays emitted indi­ radiolabeling.
rectly by a positron-emitting radionuclide tracer, which is introduced
into the body on a biologically active molecule. This tracer can be a 3.7. Chimeric antigen receptor (CAR) T cell therapies
small or large molecule that has been labeled with a radioactive element
(commonly fluorodeoxyglucose, 11C, 18F, 64Cu, 89Zr) (Hicks et al., 3.7.1. Background
2006). Adoptive T-cell therapy (ATC) refers to the use of ex vivo culture and
cellular engineering to modify a patient’s own lymphocytes in order to
3.6.3. Marketed drugs elicit anti-viral, anti-inflammatory, or anti-tumor effects. There are
While the first radiopharmaceutical approved by the FDA was so­ currently three forms of ACT being developed for cancer therapy spe­
dium Fluoride F-18 in 1972, the majority of PET imaging agents have cifically, including tumor-infiltrating lymphocytes (TILs), T cell receptor
been approved in the past 10 years (Clarke, 2018). Examples of (TCR) T cells, and chimeric antigen receptor (CAR) T cells (June et al.,
approved PET agents include Axumin® and 68Ga-DOTATATE, indicated 2018). CAR T-cell therapy involves the isolation of a patient’s autolo­
for imaging prostate tumors and neuroendocrine tumors (NETs), gous T cells via leukapheresis, genetic modification of said cells ex vivo
respectively. using viral and non-viral transfection methods (e.g. viral transduction,
The PET Tracer Axumin® is indicated for Positron Emission DNA-based transposons, CRISPR/Cas9), modified cell expansion in
Tomography—Computed Tomography (PET/CT) imaging in men with culture, and finally, re-infusion of the CAR T-cells back into the patient.
suspected prostate cancer for men with elevated blood prostate specific This genetic modification involves transgenic expression of a CAR for a
antigen (PSA) levels following prior treatment. Axumin® contains flu­ cell surface receptor(s) or antibodies (e.g. CD19, CD22, BCMA) specific
ciclovine, a fluorine 18 (F 18) labeled synthetic amino acid. Since amino to cancer targets, essentially re-directing a patient’s T cells to specif­
acids are key nutrients for tumor growth, fluciclovine is readily incor­ ically target and destroy tumor cells (Miliotou and Papadopoulou,
porated into dividing tumor cells. It is administered as an injection, is 2018).

12
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

3.7.2. How they work Table 3

As opposed to normal TCRs, CARs are able to recognize a number of Adverse effects associated with CAR T therapies (modified from Miliotou and
different elements expressed on the surface of tumor cells including Papadopoulou, 2018) (Miliotou and Papadopoulou, 2018).
unprocessed antigens, carbohydrates, and glycolipids, all without Type of Toxicity Cause
requiring antigen presentation by the major histocompatibility complex “On-target on-tumor” Rapid oncolysis of large tumor
(MHC) (Schmidt-Wolf et al., 1991). The fact that CAR recognition is Significant release of tumor cell components into the
independent from MHC class I and II restriction means that CAR T-cells systemic circulation
of both CD8+ and CD4+ subsets can be redirected to recognize the tumor “On-target off-tumor” Engagement of a related antigen on healthy tissues
cell directly, thereby offering a fundamental antitumor advantage since
“Off-target off-tumor” Inflammatory response outside of the targeted tumor
loss of MHC-associated antigen presentation by tumor cells is a primary tissue
mechanism of cancer immunoevasion (Garrido et al., 2016).
Cytokine Release Release of pro-inflammatory cytokines (IFN-γ, IL-6,
CAR-mediated tumor elimination by redirected CD8+ and CD4+ T-cells
Syndrome (CRS) TNF-α) by CAR T-cells, resulting in supra-physiological
occurs primarily via cytolysis mediated by either perforin and granzyme serum levels
exocytosis, or death receptor signaling through Fas/Fas-ligand or tumor
Neurotoxicity Systemic cytokines trafficking to the cerebrospinal fluid,
necrosis factor (TNF)/TNF-receptor (Miliotou and Papadopoulou, 2018; thereby causing diffuse encephalopathy
Chmielewski et al., 2013).

3.7.3. Marketed drugs cardiac dysfunction, renal impairment, hepatic failure, and dissemi­
CAR T cells are the first form of gene transfer therapy to gain com­ nated intravascular coagulation appearing 1–2 h after the first infusion
mercial approval by the U.S. FDA, and there now more than 250 clinical (Lee et al., 2014). The incidence rate of CRS in patients receiving CD19
trials evaluating the effectiveness of this therapy (June et al., 2018). The CAR T-cell therapy ranged from 54 to 91%, including severe CRS in
majority are directed against hematological malignancies, although 8.3–43% (Hay et al., 2017).
there are ongoing efforts to apply this therapeutic approach more Neurotoxicity is often observed concurrently with CRS, and is the
broadly to solid tumors such as glioblastoma (Brown et al., 2015) and result of increased circulating levels of cytokines crossing the blood
head and neck cancer (van Schalkwyk et al., 2013). To date, there are brain barrier into to the cerebrospinal fluid (Prudent and Breitbart,
two FDA-approved CD19-directed autologous CAR T-cell therapies: (1) 2017). This neurotoxicity presents as seizures, delirium, aphasia, and
tisagenlecleucel (Kymriah™), approved for adults with relapsed or re­ hallucinations, and is largely reversible (June et al., 2018). In rare in­
fractory diffuse large B-cell lymphoma, and you adult patients up to age stances, CRS can evolve into CAR T-cell related encephalopathy syn­
25 with relapsed or refractory acute lymphoblastic leukemia, and (2) drome (CRES) or fulminant hemophagocytic lymphohistiocytosis (HLH;
axicabtagene ciloleucel (Yescarta™), approved for various types of also known as macrophage activation syndrome) (Miliotou and Papa­
B-cell lymphoma that have either not responded to, or have relapsed dopoulou, 2018). Patients receiving tisangenlecleucel and axicabtagene
following two or more lines of systemic therapy. Given the high cost of ciloleucel during clinical trials reported neurotoxicity in percentages of
autologous CAR T-cell therapy, there are efforts to develop allogenic 15% and 28%, respectively (Buechner et al., 2017; Neelapu et al., 2017).
CAR-T-cells sourced from healthy donors and cryopreserved for future Less common hazard considerations include “on-target off-tumor”
use as an “off-the-shelf” product. Cellectis, for example, is developing toxicity resulting from target engagement of target antigen on non-
allogenic CAR T-cell therapies targeting a number of protein markers pathogenic tissues (Morgan et al., 2010), anaphylaxis resulting from
including CD19, CD22, CD38 and CS1 (Yip and Webster, 2018). host recognition of infused foreign components (Maus et al., 2013), and
the risk of insertional oncogenesis as observed in gene therapy of he­
3.7.4. ADME matopoietic stem cells for X-linked severe combined immunodeficiency
The conventional ADME criteria that apply to modalities such as and chronic granulomatous disease (Hacein-Bey-Abina et al., 2008).
small molecules and biologics, are not necessarily applicable to CAR T- From an occupational perspective, common hazards include acci­
cells. Because a CAR transgene is permanently integrated in the T-cell dental punctures with needles or contaminated sharps, spills and
genome, the equivalence to ADME for this modality is cell infusion, splashes on the skin and mucus membranes, and inhalation exposure to
trafficking, proliferation, persistence, and apoptosis (Mueller et al., infectious aerosols. Since workers tasked with manufacturing CAR T
2017; Wang, 2017). In consideration of approved CAR T-cell therapies cells will be handling cells isolated from individual patients, the nature
specifically, both Kymriah™ and Yescarta™ exhibited an initial rapid of the hazard will be dependent on the patients themselves (e.g.,
expansion immediately following infusion (peak levels occurred within potentially infected with viruses pathogenic for humans such as HIV,
the first 7–14 days for Yescarta™), followed by a decline to near baseline adenovirus, etc.).
levels (by 3 months for Yescarta™). For Kymriah™, the transgene was
present in the blood and bone marrow and was measurable beyond 2 3.7.6. Occupational hazard and exposure considerations
years, with high distribution to bone marrow (44%, 67%, and 69% of Occupational hazards associated with manufacturing of CAR T cells
that present in the blood at Day 28, Month 3, and Month 6, respectively) are going to be similar to those encountered in any cell culture labora­
(Novartis, 2017; Food and Administration D, 2018). tory or medical laboratory where the handling of blood and body fluids
occurs (Vormittag et al., 2018). Strict adherence to sterile technique and
3.7.5. Health hazards as a class/modality associated with therapeutic use standard microbiological practices will ensure that occupational expo­
Toxicities associated with CAR T-cell therapy are summarized in sures are limited.
Table 3. The most common adverse effect following CAR T-cell infusion
is immune activation known as CRS, similar to what has been observed 3.7.7. Occupational exposure banding guidance recommendation
following infusion of therapeutic mAbs, bispecific antibodies, and sys­ A BSL is a biocontainment designation system with requirements
temically administered IL-2 (Bonifant et al., 2016). CRS is the result of intended to protect personnel from potentially harmful pathogenic
elevated inflammatory cytokines (interferon gamma, granulocyte exposure in a research or manufacturing environment (Table 4) (Control
macrophage colony-stimulating factor, IL-10, IL-6) being released once CfD and Prevention., 2009). Assigning an exposure control band is not
the CAR T-cell engages surrogate antigens. Other innate immune system appropriate for this modality. Rather, CAR T cell should only be handled
cells also become activated and subsequently release additional soluble by individuals trained in proper BSL procedures as recommended by the
mediators. CRS presents clinically as high fever, malaise, fatigue, CDC for any human or other animal sourced material.
myalgia, nausea, anorexia, tachycardia/hypotension, capillary leak,

13
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

Table 4 including entry of the virus into the host cell and entry of the virus
Biosafety levels and examples. genome into the nucleus. The proteins encoded by E1A specifically, are
Biosafety Applicable to Example: Microbes Example: produced immediately after infection to modulate the cell cycle, recruit
Levela Occupational Exposure Cells cellular proteins, and regulate the expression of cellular and viral genes
Scenarios (Wechman et al., 2016). The E1A protein binds the retinoblastoma
1 Low-risk microbes that Nonpathogenic protein (pRb) resulting in the release of E2F (E2 factor) and cell cycle
pose little to no threat of Escherichia coli arrest. The release of E2F subsequently triggers activation of viral genes
infection in healthy that eventually lead to the generation of new virions, lysis of the infected
adults
cell, and spread of the viral progeny to adjacent cells (Garcia-Moure
2 Moderate-risk microbes Staphylococcus aureus; Human and et al., 2017). Cancer cell lysis also releases tumor antigens, and path­
that pose moderate Shiga toxin producing primate cells
ogen- and damage-associated molecular pattern molecules (DAMPs)
hazards to laboratory E. Coli; Adenoviruses (CAR T cells)
personnel and the capable of stimulating tumor-infiltrating antigen presenting cells that
environment activate innate and adaptive immune responses (Keller and Bell, 2016).
3 Microbes (indigenous or Mycobacterium
exotic) that can cause tuberculosis 3.8.3. Marketed drugs
serious or potentially There exists and extensive number of completed or ongoing clinical
lethal disease(s) through trials utilizing adenoviruses. A 2018 search of clinicaltrials.gov found
respiratory transmission
more than 180 clinical trials utilizing some form of adenovirus as cancer
4 High-risk microbes Ebola virus therapeutics (Baker et al., 2018). These trials have focused on a plethora
which pose a high risk of
of tumor types ranging from prostate carcinoma to glioblastoma multi­
aerosol-transmitted
infections and infections
forme, and have been examined as monotherapies or in combination
caused by these with previously approved chemotherapeutics (Rosewell Shaw and
microbes are frequently Suzuki, 2016). While none have been approved to date by any Western
fatal and without country health authority, there is one oncolytic adenovirus that was
treatment or vaccines
approved in 2005 by the Chinese State Food and Drug Administration
a
Note that enhanced control measures may be required depending on the for the treatment of squamous cell cancer of the head and neck (HNC),
strain and pathogenicity of the micro-organism being handled. and is marketed under the brand name Oncorine (Garcia-Moure et al.,
2017; Garber, 2006). Oncorine (H101) is an E1B-55K/E3B-deleted
3.8. Oncolytic adenoviruses oncolytic adenovirus that demonstrated efficacy and safety in HNC in
a phase III trial, obtaining an overall response rate of nearly 80% in
3.8.1. Background combination with cisplatin, with only mild flu-like symptoms as side
Oncolytic virotherapy is an emerging approach in the treatment of effects (Xia et al., 2004).
human cancer. Recently, the oncolytic virus Talimogene laherparepvec
(T-VEC; Imlygic™) was approved for the treatment of advanced mela­ 3.8.4. ADME
noma. T-VEC, an attenuated herpes simplex type 1 (HSV-1) encoding Clinical studies on oncolytic adenoviruses have demonstrated varied
Granulocyte-Macrophage Colony-Stimulating Factor (GM-CSF), pro­ PK depending on the route of administration, adenovirus serotype and
vided evidence of the potential of locally delivered genetically modified associated genetic modifications, such that generalizations regarding
replicating viruses as oncology agents, and opened the door for the pharmacokinetic (PK) parameters cannot be made. For example, in a
development of viruses capable of tumor oncolysis via systemic Phase 1 study of enadenotucirev (a group B Ad11p/Ad3 chimeric
administration (Rehman et al., 2016). Within this arena, development oncolytic adenovirus) administered intravenously to patients with
has focused primarily on the oncolytic adenoviruses given the fact that epithelial solid tumors, the half-life was short (16.7 min) and indepen­
these non-enveloped, double-stranded DNA viruses possess genomes dent of dose and administration frequency (Machiels et al., 2019).
that are easily engineered, and because they exhibit low levels of Post-infusion viral shedding (i.e., release of adenovirus from injection
pathogenicity following administration (Baker et al., 2018). While sites and patient secretions) manifested as buccal shedding and rectal
dozens of different adenoviruses have been described, they all share a shedding (both of which were related to dose level), and to a lesser
common architecture consisting of an icosahedral capsid composed of extent, urine shedding. In a second example, a Phase 1 dose escalation
up to seven different structural proteins. One of these, the fiber, is a study examined the safety and host immune response of telomelysin
trimeric protein located on each of the twelve vertices of the virion, (OBP-301) following a single intratumoral injection in patients with
which protrudes from the capsid like an antennae and directly influences advanced solid tumors (Phadke, 2008). Telomelysin is a
adenovirus tropism (Russell, 2009). telomerase-specific replication-selective adenovirus in which the human
telomerase reverse transcriptase promotor element drives expression of
3.8.2. How they work E1A/E1B genes linked with an internal ribosome entry site. This study
Oncolytic adenoviruses function by selectively infecting and repli­ indicated that despite intratumoral dosing, viral DNA was detected in 5
cating in cancer cells while sparing non-cancerous cells. This selectively of 9 patient plasma samples, as early as 30 min and as late as 14 days
is accomplished via a number of subtle modifications in early adenoviral post-treatment. Viral DNA was also detected in the sputum of one pa­
genes, and is summarized in the review by Baker et al., 2018 (Baker tient, indicating systemic dissemination following intratumoral injec­
et al., 2018). The general viral cycle involves adenoviral recognition of tion (Phadke, 2008). Systemic dissemination of oncolytic adenoviral
its specific receptor on the cell surface thus triggering its internalization. DNA via vascular transduction has also been demonstrated in a study
Following this internalization, the adenovirus migrates through the examining post-mortem tissues of cancer patients treated with this
microtubules and subsequently introduces the viral genome inside the modality either intratumorally or intravenously, and whose deaths were
nucleus. The adenovirus genome is composed of linear, double-stranded caused by disease progression rather than the treatment vector (Koski
DNA of approximately 36 kilobases, which can be divided into early (E) et al., 2015). Oncolytic adenoviral DNA was recovered in a wide array of
and late (L) genes based upon their expression across the infection cycle tissues, including both injected and non-injected tumors and various
(Seth and Higginbotham, 2000). While the L genes encode structural normal tissues, including the brain.
proteins, which package the viral DNA into the adenovirus virion during
the final stages of replication, the E genes regulate viral replication itself,

14
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

3.8.5. Health hazards as a class/modality associated with therapeutic use 3.8.7. Occupational exposure banding guidance recommendation
Generally, the use of oncolytic adenoviruses appears to be reason­ Adenoviruses should be handled as dictated by the applicable Bio­
ably safe following local administration (i.e., intratumoral) and lower logical Safety Level (BSL) (See Table 4) (Control CfD and Prevention.,
systemic doses, although the development of newer generations of ad­ 2009). For other virus types, the appropriate BSL as outlined in the
enoviruses expressing transgenes and possessing altered capsids or Biosafety in Microbiological and Biomedical Laboratories guidance
different promoters could alter the safety profile (Buijs et al., 2015). prepared by the Centers for Disease Control and the National Institutes
Viral shedding of engineered adenovirus vectors could theoretically of Health should be followed (Control CfD and Prevention., 2009).
result in homologous recombination between wildtype adenoviruses of
the same subgroup, leading to new wildtype adenoviruses that possess 3.9. Engineered bacteria
transgenes or have expanded tissue tropism due to retargeting strategies
(Buijs et al., 2015; Singh et al., 2013). To date, no such recombination 3.9.1. Background
events have been observed in clinical trials. Bacteria can be genetically modified (GM) to produce a continuous
Given the mechanism of action of oncolytic adenoviruses, one might and inexpensive supply of proteins/molecules such as human hormones,
consider CRS an outcome of concern. However, in the thousands of interleukins and antibodies within specific organs or tissues (Pine­
patients that have received adenoviruses therapeutically, only one pa­ ro-Lambea et al., 2015). An enhanced understanding of the role of the
tient (who consequently, had an underlying metabolic disorder) has human gut microbiome in health and diseases has increased interest in
died as a result of CRS (Sibbald, 2001). Based on this prior extensive the use of live bio-therapeutic products (LBPs), such as GM bacteria, for
clinical experience, the risk of severe CRS and toxicity is generally the treatment or prevention of disease. Also important, especially in the
deemed to be minimal, especially considering the neoantigens expressed context of developing countries, is that GM bacteria can be administered
by the virus are not typically present in normal tissue and therefore, orally, which makes the requirement for hygienic syringes and needles
cross-reactive T-cell responses are unlikely to occur (Larson et al., 2018). unnecessary. The use of this technology for vaccines is in the spotlight
since, in addition to the ease of oral administration, GM bacteria are
3.8.6. Occupational hazard and exposure considerations relatively inexpensive to propagate/manufacture and transport (resis­
Occupational exposure to an adenovirus is generally through the tant to poor refrigeration), and thus the ongoing revolution in GM
upper respiratory tract via aerosol exposure, although other potential bacteria for vaccination fits well with the World Health Organization
routes of exposure include via mucus membranes (splash of virus to eye, (WHO) agenda and recommendations.
nose, mouth), parenteral (needle stick or sharp object injury) and con­
tact with non-intact skin (Larson et al., 2018). In general, adenovirus 3.9.2. How they work
infection most commonly causes respiratory illness, although, depend­ GM bacteria serve as drug delivery systems, which, upon oral
ing on the infecting serotype, they may cause other illnesses such as administration, can carry the drug and deliver it to a local environment
gastroenteritis, conjunctivitis, cystitis, and rash (Stanford. Stanford En­ such as the gut or mucosa, without the need for systemic exposure. GM
viro, 2020). Adenovirus infection of the respiratory systems typically bacteria are being regarded as a technologically viable, economically-
can present as a spectrum of ailments ranging from the common cold feasible, safe delivery modality for drugs/proteins to localized areas in
syndrome to pneumonia, croup, and bronchitis. In terms of exposure the body (Ferreira et al., 2017). The insertion of plasmid vectors, which
scenarios specifically, special consideration must be given to the po­ encode proteins such as enzymes, antibodies, antigens and cytokines
tential for viral shedding and human-to-human transmission. In the into living bacteria, enables these GM bacteria to produce or deliver the
clinical setting, viral shedding can occur from injection sites and patient therapeutic proteins/molecules to the target site. GM bacteria have also
excretions, and prevalence has been demonstrated to increase with been designed to convert pro-drugs to their active form at the localized
higher doses and systemic administration (Kimball et al., 2010; Keedy disease site (Pinero-Lambea et al., 2015). GM bacteria can bypass
et al., 2008; Tian et al., 2009). Preclinically, viral shedding could result problems associated with conventional cancer chemotherapies, such as
in exposure via contact with animal secreta, excreta and bedding. poor selectivity and limited tumor penetrability, and can be finely
Pregnant individuals and those who are severely immunocompro­ engineered to sense and respond to the tumor microenvironment (For­
mised are particularly susceptible to adenovirus exposure and should bes, 2010). Local delivery of recombinant proteins via GM bacteria in
exercise extreme caution when handling adenoviruses, as infection affected organs also has many advantages as systemic side effects are
could lead to systemic disease (e.g. hepatitis). There is a paucity of data generally avoided.
in terms of occupational exposures to oncolytic adenoviruses. However,
a study in which health-care workers involved in the aerosolized 3.9.3. Marketed drugs
administration of adeno-associated virus (AAV) during a Phase II cystic Currently, there are no approved GM bacterial therapeutics. Many
fibrosis study determined that individuals were exposed to an estimated have been studied in clinical trials and been shown to be promising
0.0006% of the administered doses based on airborne vector particle components of therapies for numerous cancer types and immune-related
concentration. At this level of exposure, the prevalence of symptoms diseases, while exhibiting a low incidence of adverse events (Pine­
(typically associated with adenovirus exposure) was very low, the ro-Lambea et al., 2015; Ferreira et al., 2017). Many of the GM bacteria
spectrum of symptoms was similar in both active and control health care utilized in clinical trials are similar to those sold over the counter as
workers, and there were no reported negative health effects (Croteau probiotics (Dreher-Lesnick et al., 2017) and natural isolates obtained
et al., 2004). It is assumed that oncolytic adenoviruses and AAVs would from the microbiota of healthy individuals. They are most often lactic
behave similarly in this exposure scenario. Therefore, if proper handling acid bacteria (LAB) and to a lesser extent Escherichia coli strains (Pine­
precautions and engineering controls are implemented as described in ro-Lambea et al., 2015).
the proceeding section, occupational exposure is expected to be limited.
While the focus here has been placed on adenoviruses specifically, 3.9.4. ADME
there are a number of other oncolytic virus platforms in development for Microbes are commonly ingested via a variety of fermented foods
the treatment of cancer including herpes simplex virus 1, measles virus, and drinks. They then go on to live in the intestinal milieu. Once a new
coxsackievirus, and poliovirus, among others (Buijs et al., 2015). host has ingested the bacteria, the bacteria must colonize otherwise they
Occupational exposure considerations as outlined for adenoviruses will rapidly transit through the gastrointestinal system and be elimi­
previously would be applicable to these other virus types as well. nated in fecal matter. Additionally, bacteria which have successfully
colonized the intestinal environment by establishing a niche and
reaching replication levels that ensure stability and survival can shed

15
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

from the host into fecal matter (Browne et al., 2017). Once it has entered 3.10.3. Marketed drugs
the external environment, there is a potential that the GM bacteria can A few examples of marketed vaccines include Flumist® Quadriva­
enter a new susceptible host. lent, Gardasil® and BioThrax®. Flumist® Quadrivalent, a vaccine
against influenza, contains a live attenuated influenza virus that con­
3.9.5. Health hazards as a class/modality associated with therapeutic use tains four vaccine virus strains for administration via an intranasal spray
The main health hazards associated with GM bacteria are the risk of (MedImmune. FluMist Quadri, 2013). Gardasil®, a vaccine targeted
infection and the risk of transient or stable transfer of identified genes of against human papillomavirus (HPV), is a recombinant quadrivalent
concern (e.g. antibiotic resistance genes) in the LBP to another species vaccine prepared from the purified virus-like particles (VLPs) of the
present in host sites exposed to product colonization. major capsid (L1) protein of HPV Types 6, 11, 16, and 18 and is
administered with Alum adjuvant (Siddiqui and Perry, 2006). Bio­
3.9.6. Occupational hazard and exposure considerations Thrax®, a subunit vaccine for Anthrax (Bacillus anthraxis), consists of
Bacteria are key elements for human health as evinced by the gut an avirulent, nonencapsulated strain of Bacillus anthracis and proteins
microbiota and the numerous health disorders in axenic animals and is administered with Alum adjuvant (BioSolutions, 2013).
(Pinero-Lambea et al., 2015). The occupational safety assessment of a Several vaccine platforms along with the relevant occupational
novel bacterial species should take into account information on the hazards and control guidance as well as examples of currently marketed
bacterial strain(s) and source(s) of the strain (e.g. original donor), the vaccines are presented in Table 5. Coronavirus disease or COVID-19
proposed subject population, mechanism of action, the intended route of vaccine candidates exist for each of the vaccine platforms and several
administration and the delivery method. If there are genetic modifica­ of these are noted (Le et al., 2020). The WHO’s authoritative and
tions to the strain, the stability of those genetic modifications should be continually updated list of COVID-19 vaccine candidates is located at:
taken into account and assessed. As it is important to ensure https://www.who.int/publications/m/item/draft-landscape-of-covi
product-related infections can be treated should they arise, strain d-19-candidate-vaccines.
characterization information including details about the presence of
virulence factors or toxins and the strain’s antibiotic resistance profile 3.10.4. ADME
(inclusive of knowledge of any antibiotic resistance genes present in the PK studies are generally not required or conducted for vaccines
GM bacteria) should be assessed/considered. because the kinetic properties of antigens do not provide useful infor­
mation for determining dose recommendations. Differences in distri­
3.9.7. Occupational exposure banding guidance recommendation bution, persistence and elimination of the components of a vaccine can
Given that GM bacteria are microbes, they should be handled as be attributed to several factors such as the route of administration, the
dictated by the applicable BSL (Control CfD and Prevention., 2009). number of injections, the quantity of antigen and the species under study
(Faurez et al., 2010). Whatever the route of administration, the site of
injection is the area where the vaccine components tend to remain the
3.10. Vaccines most concentrated and persist the longest, however vaccine components
have been observed in organs throughout the body including immune
3.10.1. Background cells and the blood (Faurez et al., 2010). Vaccinated persons have been
Vaccines provide the main method of prophylaxis against pandemic shown to shed vaccine viruses, however the shed live attenuated viruses
viruses and other infectious diseases. The introduction of human vac­ are generally stable and do not revert to virulent strains (Fiore et al.,
cines has had a tremendous impact on global health by dramatically 2009).
reducing the mortality and morbidity caused by infectious diseases.
Vaccination is considered one of the most cost effective and successful 3.10.5. Health hazards as a class/modality associated with therapeutic use
medical interventions ever introduced. Vaccines have inevitably pre­ There are safety concerns associated with various vaccine platforms.
vented diseases, complications, and the death of millions of people by Although attenuated vaccines elicit long-lasting immunity, safety issues
protecting against many deadly and debilitating infectious diseases including the administration of live viral or bacterial components and
including smallpox and polio (Kallerup and Foged, 2015). The World their capacity to replicate, infect immunocompromised individuals, and
Health Organization (WHO) currently recommends routine immuniza­ revert to pathogenic forms are of concern. For example, whole virus
tion against 22 different diseases (WHO, 2019). vaccines are considered too reactive, particularly in young children
(Wood and Robertson, 2004). In the case of plasmid DNA, there is a
3.10.2. How they work concern that the DNA vaccine could possibly integrate into host ge­
The ideal vaccine elicits the exact immune response that occurs nomes, increasing the risk of malignancy (via insertional mutagenesis)
during natural infection, safely inducing a strong humoral and cellular (Kim and Jacob, 2009). Adjuvanated vaccines have only recently been
immune response. Vaccines work by mimicking disease agents and licensed so there remains some uncertainty regarding the safety of some
stimulating the immune system to build up defenses against them. adjuvants in humans (Keitel and Atmar, 2009). Additionally, adjuvanted
Vaccines are generally created to express one or more antigens present formulations typically elicit a higher frequency of injection site, and
on a pathogen to prime an individual’s immune system, so that if the occasionally systemic, reactions when compared with non-adjuvanted
individual is exposed to the pathogen in the future, the immune system formulations (Keitel and Atmar, 2009). The possibility of adverse re­
will respond swiftly with a specific defense. Briefly, antigen-presenting actions increase when multiple doses are required to provide immunity.
cells (APCs; e.g., dendritic cells, macrophages) take up and process the Both aerosol-administered and intradermally administered vaccines
pathogen-specific antigen(s) and present them to CD4+ T cells via the have been well tolerated and immunogenic. Respiratory adverse events
MHC class II pathway and then to B cells, ultimately resulting in the are rare and mild. Intradermal vaccines have been associated with ex­
production of antibodies. For most infectious diseases, primary protec­ pected mild local injection-site reactions (Satti et al., 2014). Common
tion is mediated by existing antibodies, whereas for intracellular path­ adverse effects in vaccine recipients include fatigue and headache. In
ogens (e.g. mycobacterium tuberculosis), protection is mediated by studies with FluMist®, adverse effects included runny nose or nasal
MHC class I-restricted CD8 T cell responses. The goal of vaccination is to congestion, headache, sore throat, tiredness/weakness, muscle aches,
produce a memory of the vaccine antigen, so that if an individual is cough or chills and rarely wheezing that required bronchodilator ther­
exposed to the pathogen in the future, the memory cells will recognize it apy or that was associated with significant respiratory symptoms (Fiore
and the immune system will be able to respond more effectively than if it et al., 2009).
had never encountered the pathogen.

16
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

Table 5
Vaccine platforms, associated occupational hazard and exposure control guidance, and banding paradigmsa.
Vaccine Platform Potential Vaccine Example(s)b Occupational Hazards and Exposure Control Banding Paradigmc
Components Guidance

Live Attenuated Deoptimized live Marketed: Flumist® Quadrivalent (Influenza) COVID- See Section 3.8 (Oncolytic adenoviruses), 3.9 BSLs (Table 4)
Virus or attenuated virus or 19 Candidates in Development: Attenuated Influenza (Engineered Bacteria) and Section 3.10
Bacteria bacteria expressing an antigenic portion of the Spike protein (Vaccines)
Inactivated Virus Adenovirus, measles, Marketed: IPOL® (Polio) COVID-19 Candidates in See Section 3.8 (Oncolytic adenoviruses) and BSLs (Table 4)
influenza or other viral Development: Whole-Virion Inactivated Section 3.10 (Vaccines)
vector
Protein Subunit Peptide, S protein, Marketed: BioThrax® (Anthrax) See Sections 3.1 (Small molecules), 3.3 Small molecule banding
peptide antigens COVID-19 Candidates in Development: Recombinant (Peptides), 3.4 (Biologics), 3.9 (Engineered paradigm (Table 1) or
protein (RBD-Dimer) + Adjuvant Bacteria) and Section 3.10 (Vaccines) BCCs (Table 2)
DNA DNA, Plasmid DNA Marketed: None currently approved for use in See Section 3.2 (Oligonucleotides) and Small molecule banding
humans; West Nile-Innovator DNA vaccine for equines Section 3.10 (Vaccines) paradigm (Table 1)
(West Nile Virus)
COVID-19 Candidates in Development Include: Spike
DNA vaccine + Adjuvant
RNA mRNA, small activating Marketed: None currently approved COVID-19 See Section 3.2 (Oligonucleotides) and Small molecule banding
RNA, encapsulated Candidates in Development: LNP-encapsulated mRNA Section 3.10 (Vaccines) paradigm (Table 1)
mRNA
Virus-like Proteins Marketed: Gardasil® (human papillomavirus) COVID- See Section 3.1 (Small Molecules) and Section Small molecule banding
particle (VLP) 19 Candidates in Development: Plant-derived VLP + 3.10 (Vaccines) paradigm (Table 1) or
Adjuvant BCCs (Table 2)
a
Not an exhaustive list.
b
COVID-19 Candidate information obtained from the WHO’s authoritative and continually updated list of COVID-19 vaccine candidates located at: https://www.
who.int/publications/m/item/draft-landscape-of-covid-19-candidate-vaccines (accessed 07August 2020).
c
Note that BCCs are utilized at BMS, however alternative banding paradigms for biologics are also used in the pharmaceutical industry which are equally effective in
controlling exposures.

3.10.6. Occupational hazard and exposure considerations national regulatory bodies). Stringent vaccine biosafety control mea­
Many vaccines are expected to be avirulent in humans. Inactivated sures with specific enhancements, defined as BSL2 enhanced (BSL2-e;
pandemic influenza vaccines present no biosafety risks provided that the pandemic influenza vaccine) and BSL3 enhanced (BSL3-e; pandemic
results of the inactivation steps show complete virus inactivation, as the influenza vaccine) are designed to manage the risk from vaccine pro­
viral vaccine is rendered incapable of replication (Organization WH, duction and quality control using such viruses during the pre-pandemic
2005). Concerns arise where workers are immunologically naive to a or inter-pandemic period (Organization WH, 2005). In the United
virus and potentially susceptible to infection. In the instances where Kingdom, it is necessary to use BSL4 facilities, but in other parts of the
viruses are utilized, there are concerns around human infection with the world (e.g. United States) working with highly pathogenic avian influ­
vaccine virus during manufacturing and of a subsequent outbreak enza (HPAI) viruses is acceptable under BSL3-e (Wood and Robertson,
originating from the manufacturing plant (Organization WH, 2005). 2004). Note that while this guidance exists, a risk assessment should be
Inhalation exposure to live viruses or bacteria are of concern. Upon performed that will depend on the nature of the strain and the pandemic
inhalation, live attenuated viruses have the potential to replicate in the period declared by WHO (Organization WH, 2005).
mucosa of the respiratory tract (Chen and Subbarao, 2009).
Human-to-human transmission is expected to be unlikely since replica­ 4. Conclusion
tion is attenuated and the virus titers would be below those considered
necessary for human infection (Organization WH, 2005). Potential Novel therapeutic drugs and technologies of increasing complexity
adverse effects may be at worst, similar to those observed upon intra­ and potency are being developed and handled throughout pharmaceu­
nasal administration of live attenuated influenza, which consist of some tical research, development and manufacturing organizations and
effects of local viral replication (e.g., nasal congestion). With the clinics. As biological processes become elucidated and novel drug plat­
exception of surrogate viruses, the use of wild type pandemic-like vi­ forms continue to be discovered, classical small-molecule approaches
ruses to develop pandemic vaccine strains presents considerable are not always appropriate or adequate and so other types of modalities
biosafety risks to personnel in vaccine development and manufacturing are required to address these targets (e.g. protein–protein and pro­
facilities (Organization WH, 2005). Additionally, viruses may be ex­ tein–nucleic acid interactions). From an occupational safety perspective,
pected to survive for at least a short time (hours) on surfaces and thus this means staying current with the emerging literature on the hazards
provide a potential means of infection for workers (Organization WH, observed and expected with therapeutics spanning multiple modalities.
2005). Therefore, the wild-type pandemic-like viruses present the Occupational safety professionals must be proactive and aware of the
greatest occupational health concern. pharmacological, toxicological and physicochemical characteristics of
the new modalities residing within their respective drug development
3.10.7. Occupational exposure banding guidance recommendation pipelines to ensure occupational exposures are appropriately controlled
Vaccine production facilities are designed to maintain the sterility of and the risk of adverse effects, both pharmacological and toxicological,
the product. Therefore, although there is opportunity to be exposed to in the workplace is minimized. Occupational exposure control banding
the virus and/or antigens, occupational exposures are expected to be systems represent an acceptable approach to ensuring the safe handling
limited. The WHO biosafety risk assessment and guidelines for the of many new modalities, however it is important to be aware that certain
production and quality control of human influenza pandemic vaccines modalities do not fit historical banding paradigms and require other
presents guidance to vaccine manufacturers on the safe production of means of classification, such as BSLs or BCCs.
influenza vaccines (Organization WH, 2005). According to the WHO
guideline, all laboratory procedures involving live highly pathogenic Funding sources
influenza viruses should take place at a high level of biological
containment (e.g., BSL3 and above, as recommended by WHO and This research was funded by Bristol Myers Squibb.

17
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

Declaration of competing interest Browne, H.P., Neville, B.A., Forster, S.C., Lawley, T.D., 2017. Transmission of the gut
microbiota: spreading of health. Nat. Rev. Microbiol. 15 (9), 531–543. https://doi.
org/10.1038/nrmicro.2017.50. Epub 2017/06/12, PubMed PMID: 28603278.
The authors declare that they have no known competing financial Buechner, J., Grupp, S., Maude, S., Boyer, M., Bittencourt, H., Laetsch, T., et al., 2017.
interests or personal relationships that could have appeared to influence Global registration trial of efficacy and safety of CTL019 in pediatric and young adult
the work reported in this paper. patients with relapsed/refractory (R/R) acute lymphoblastic leukemia (ALL): update
to the interim analysis. Clin. Lymphoma, Myeloma & Leukemia 17, S263–S264.
https://doi.org/10.1016/j.clml.2017.07.030.
Acknowledgements Buijs, P.R., Verhagen, J.H., van Eijck, C.H., van den Hoogen, B.G., 2015. Oncolytic
viruses: from bench to bedside with a focus on safety. Hum. Vaccines Immunother.
11 (7), 1573–1584. https://doi.org/10.1080/21645515.2015.1037058. Epub 2015/
The authors would like to acknowledge David Brandwene (IES En­ 05/23, PubMed PMID: 25996182; PubMed Central PMCID: PMCPMC4514197.
gineers) for his help with literature searches, Dr. Kate Sokolowski for Cai, H., 2018. Therapeutic monoclonal antibodies approved by FDA in 2017. MOJ
creating the excellent graphical abstract, Dr. Wendy Luo for initial Immunology 6. https://doi.org/10.15406/moji.2018.06.00198.
Carthew, P., Clapp, C., Gutsell, S., 2009. Exposure based waiving: the application of the
contributions to the manuscript and Dr. Michael Graziano for his sup­ toxicological threshold of concern (TTC) to inhalation exposure for aerosol
port of this work. ingredients in consumer products. Food Chem. Toxicol. 47 (6), 1287–1295.
Chebekoue, S.F., Krishnan, K., 2017. Derivation of occupational thresholds of
toxicological concern for systemically acting noncarcinogenic organic chemicals.
Appendix A. Supplementary data Toxicol. Sci. 160 (1), 47–56.
Chebekoue, S.F., Krishnan, K., 2019. Derivation of internal dose-based thresholds of
Supplementary data to this article can be found online at https://doi. toxicological concern for occupational inhalation exposure to systemically acting
organic chemicals. J. Occup. Environ. Hyg. 16 (4), 308–319.
org/10.1016/j.yrtph.2020.104813.
Chen, G.L., Subbarao, K., 2009. Live Attenuated Vaccines for Pandemic Influenza.
Vaccines for Pandemic Influenza. Springer, pp. 109–132.
References Chen, C., Wang, Z., Zhang, Z., Liu, X., Kang, S.S., Zhang, Y., et al., 2018. The prodrug of
7,8-dihydroxyflavone development and therapeutic efficacy for treating Alzheimer’s
Ader, A.W., Farris, J.P., Ku, R.H., 2005. Occupational health categorization and disease, 3. In: Proceedings of the National Academy of Sciences of the United States
compound handling practice systems—roots, application and future. J. Chem. of America, vol. 115, pp. 578–583. https://doi.org/10.1073/pnas.1718683115.
Health Saf. 12 (4), 20–26. https://doi.org/10.1021/acs.chas.8b12407. Epub 2018/01/04, PubMed PMID: 29295929; PubMed Central PMCID:
Agency, E.M., 2014. Guideline on Setting Health Based Exposure Limits for Use in Risk PMCPMC5777001.
Identification in the Manufacture of Different Medicinal Products in Shared Chmielewski, M., Hombach, A.A., Abken, H., 2013. Antigen-specific T-cell activation
Facilities. independently of the MHC: chimeric antigen receptor-redirected T cells. Front.
Alton, E.W., Boushey, H.A., Garn, H., Green, F.H., Hodges, M., Martin, R.J., et al., 2012. Immunol. 4, 371. https://doi.org/10.3389/fimmu.2013.00371. Epub 2013/11/26,
Clinical expert panel on monitoring potential lung toxicity of inhaled PubMed PMID: 24273543; PubMed Central PMCID: PMCPMC3822734.
oligonucleotides: consensus points and recommendations. Nucleic Acid Therapeut. Clarke, B.N., 2018. PET Radiopharmaceuticals: what’s new, what’s reimbursed, and
22 (4), 246–254. https://doi.org/10.1089/nat.2012.0345. Epub 2012/07/20, what’s next? J. Nucl. Med. Technol. https://doi.org/10.2967/jnmt.117.205021.
PubMed PMID: 22809313; PubMed Central PMCID: PMCPMC3426204. Epub 2018/02/14, PubMed PMID: 29438008.
Autio, K.A., Boni, V., Humphrey, R.W., Naing, A., 2020. Probody therapeutics: an Control CfD, Prevention. Biosafety in Microbiological and Biomedical Laboratories
emerging class of therapies designed to enhance on-target effects with reduced off- (BMBL), 2009. US Government Printing Office, St Louis, MO.
tumor toxicity for use in immuno-oncology. Clin. Canc. Res. 26 (5), 984–989. Control CfD, 2015. CDC Radiation Safety.
Baker, A.T., Aguirre-Hernandez, C., Hallden, G., Parker, A.L., 2018. Designer oncolytic Cramer, G.M., Ford, R.A., Hall, R.L., 1978. Estimation of toxic hazard–a decision tree
adenovirus: coming of age. Cancers 10 (6). https://doi.org/10.3390/ approach. Food Chem. Toxicol. 16 (3), 255–276. https://doi.org/10.1016/s0015-
cancers10060201. Epub 2018/06/16, PubMed PMID: 29904022; PubMed Central 6264(76)80522-6. Epub 1978/06/01, PubMed PMID: 357272.
PMCID: PMCPMC6025169. Croteau, G.A., Martin, D.B., Camp, J., Yost, M., Conrad, C., Zeitlin, P.L., et al., 2004.
Baldo, B.A., 2015. Chimeric fusion proteins used for therapy: indications, mechanisms, Evaluation of exposure and health care worker response to nebulized administration
and safety. Drug Saf. 38 (5), 455–479. https://doi.org/10.1007/s40264-015-0285-9. of tgAAVCF to patients with cystic fibrosis. Ann. Occup. Hyg. 48 (8), 673–681.
Epub 2015/04/03, PubMed PMID: 25832756. https://doi.org/10.1093/annhyg/meh066. Epub 2004/10/28, PubMed PMID:
Basketter, D.A., Kruszewski, F.H., Mathieu, S., Kirchner, D.B., Panepinto, A., 15507460.
Fieldsend, M., et al., 2015. Managing the risk of occupational allergy in the enzyme Dahlen, E., Veitonmaki, N., Norlen, P., 2018. Bispecific antibodies in cancer
detergent industry. J. Occup. Environ. Hyg. 12 (7), 431–437. immunotherapy. Ther Adv Vaccines Immunother 6 (1), 3–17. https://doi.org/
Beck, A., Goetsch, L., Dumontet, C., Corvaia, N., 2017. Strategies and challenges for the 10.1177/2515135518763280. Epub 2018/07/13, PubMed PMID: 29998217;
next generation of antibody-drug conjugates. Nat. Rev. Drug Discov. 16 (5), PubMed Central PMCID: PMCPMC5933537.
315–337. https://doi.org/10.1038/nrd.2016.268. Epub 2017/03/18, PubMed de Lemos, M.L., Badry, N., Kletas, V., Fabbro, J., Tew, A., 2018. Safe handling of
PMID: 28303026. monoclonal antibodies: too large to be hazardous? J. Oncol. Pharm. Pract. 24 (3),
Bercu, J.P., Dolan, D.G., 2013. Application of the threshold of toxicological concern 218–220. https://doi.org/10.1177/1078155217698846. Epub 2017/12/30,
concept when applied to pharmaceutical manufacturing operations intended for PubMed PMID: 29284346.
short-term clinical trials. Regul. Toxicol. Pharmacol. 65 (1), 162–167. https://doi. Derelanko, M.J., 2017. The Toxicologist’s Pocket Handbook. CRC Press.
org/10.1016/j.yrtph.2012.06.012. Epub 2012/06/27, PubMed PMID: 22732128. Desnoyers, L.R., Vasiljeva, O., Richardson, J.H., Yang, A., Menendez, E.E., Liang, T.W.,
Berman, C.L., Barros, S.A., Galloway, S.M., Kasper, P., Oleson, F.B., Priestley, C.C., et al., et al., 2013. Tumor-specific activation of an EGFR-targeting probody enhances
2016. OSWG recommendations for genotoxicity testing of novel oligonucleotide- therapeutic index. Sci. Transl. Med. 5 (207) https://doi.org/10.1126/
based therapeutics. Nucleic Acid Therapeut. 26 (2), 73–85. https://doi.org/ scitranslmed.3006682, 207ra144. Epub 2013/10/18, PubMed PMID: 24132639.
10.1089/nat.2015.0534. Epub 2016/03/16, PubMed PMID: 26978711. Directive C, 2013. 59/Euratom of 5 December 2013 Laying Down Basic Safety Standards
BioSolutions, E., 2013. Biothrax Package Insert. Emergent BioSolutions Rockville, MD. for Protection against the Dangers Arising from Exposure to Ionising Radiation, and
Bonifant, C.L., Jackson, H.J., Brentjens, R.J., Curran, K.J., 2016. Toxicity and Repealing Directives 89/618/Euratom, 90/641/Euratom, 96/29/Euratom, 97/43/
management in CAR T-cell therapy. Mol Ther Oncolytics 3, 16011. https://doi.org/ Euratom and 2003/122/Euratom. Euratom.
10.1038/mto.2016.11. Epub 2016/09/15, PubMed PMID: 27626062; PubMed Dolan, D.G., Naumann, B.D., Sargent, E.V., Maier, A., Dourson, M., 2005. Application of
Central PMCID: PMCPMC5008265. the threshold of toxicological concern concept to pharmaceutical manufacturing
Bos, J.D., Meinardi, M.M., 2000. The 500 Dalton rule for the skin penetration of chemical operations. Regul. Toxicol. Pharmacol. 43 (1), 1–9. https://doi.org/10.1016/j.
compounds and drugs. Exp. Dermatol.: Viewpoint 9 (3), 165–169. yrtph.2005.06.010. Epub 2005/08/16, PubMed PMID: 16099564.
Brennan, F.R., Morton, L.D., Spindeldreher, S., Kiessling, A., Allenspach, R., Hey, A., Dougherty, T.J., Pucci, M.J., 2011. Antibiotic Discovery and Development. Springer US.
et al., 2010. Safety and immunotoxicity assessment of immunomodulatory Drake, P.M., Rabuka, D., 2017. Recent developments in ADC technology: preclinical
monoclonal antibodies. mAbs 2 (3), 233–255. https://doi.org/10.4161/ studies signal future clinical trends. BioDrugs 31 (6), 521–531. https://doi.org/
mabs.2.3.11782. Epub 2010/04/28, PubMed PMID: 20421713; PubMed Central 10.1007/s40259-017-0254-1. Epub 2017/11/10, PubMed PMID: 29119409;
PMCID: PMCPMC2881251. PubMed Central PMCID: PMCPMC5696438.
Brinkmann, U., Kontermann, R.E., 2017. The making of bispecific antibodies. mAbs 9 Dreher-Lesnick, S.M., Stibitz, S., Carlson Jr., P.E., 2017. Regulatory considerations for
(2), 182–212. https://doi.org/10.1080/19420862.2016.1268307. Epub 2017/01/ development of live biotherapeutic products as drugs. Microbiol. Spectr. 5 (5)
11, PubMed PMID: 28071970; PubMed Central PMCID: PMCPMC5297537. https://doi.org/10.1128/microbiolspec. Epub 2017/10/05, BAD-0017-2017.
Brown, C.E., Badie, B., Barish, M.E., Weng, L., Ostberg, J.R., Chang, W.C., et al., 2015. PubMed PMID: 28975881.
Bioactivity and safety of IL13Ralpha2-redirected chimeric antigen receptor CD8+ T Ducry, L., Stump, B., 2010. Antibody-drug conjugates: linking cytotoxic payloads to
cells in patients with recurrent glioblastoma. Clin. Canc. Res. : Off. J. Am. Assoc. monoclonal antibodies. Bioconjugate Chem. 21 (1), 5–13. https://doi.org/10.1021/
Canc. Res 21 (18), 4062–4072. https://doi.org/10.1158/1078-0432.CCR-15-0428. bc9002019. Epub 2009/09/23, PubMed PMID: 19769391.
Epub 2015/06/11, PubMed PMID: 26059190; PubMed Central PMCID: Dumont, J.A., Bitonti, A.J., Clark, D., Evans, S., Pickford, M., Newman, S.P., 2005.
PMCPMC4632968. Delivery of an erythropoietin-Fc fusion protein by inhalation in humans through an
immunoglobulin transport pathway. J. Aerosol Med. 18 (3), 294–303. https://doi.
org/10.1089/jam.2005.18.294. Epub 2005/09/27, PubMed PMID: 16181004.

18
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

EPA, 2020. United States Environmental Protection Agency: Federal Guidance for receptor-modified T-cell therapy. Blood 130 (21), 2295–2306. https://doi.org/
Radiation Protection. 10.1182/blood-2017-06-793141. Epub 2017/09/20, PubMed PMID: 28924019;
Faurez, F., Dory, D., Le Moigne, V., Gravier, R., Jestin, A., 2010. Biosafety of DNA PubMed Central PMCID: PMCPMC5701525.
vaccines: new generation of DNA vectors and current knowledge on the fate of Hicks, R.J., Dorow, D., Roselt, P., 2006. PET tracer development–a tale of mice and men.
plasmids after injection. Vaccine 28 (23), 3888–3895. Canc. Imag. 6, S102–S106. https://doi.org/10.1102/1470-7330.2006. Epub 2006/
Fellner, R.C., Terryah, S.T., Tarran, R., 2016. Inhaled protein/peptide-based therapies for 11/23, 9098. PubMed PMID: 17114061; PubMed Central PMCID:
respiratory disease. Mol Cell Pediatr 3 (1), 16. https://doi.org/10.1186/s40348-016- PMCPMC1805077.
0044-8. Epub 2016/04/22, PubMed PMID: 27098663; PubMed Central PMCID: Hoersch, J., Hoffmann-Doerr, S., Keller, D., 2018. Derivation of an inhalation TTC for the
PMCPMC4839019. workplace based on DNEL values reported under REACH. Toxicol. Lett. 290,
Ferreira, A.K., Mambelli, L.I., Pillai, S.Y., 2017. Intervening in disease through 110–115.
genetically-modified bacteria. Best Pract. Res. Clin. Gastroenterol. 31 (6), 693–697. Hutmacher, C., Neri, D., 2019. Antibody-cytokine fusion proteins: biopharmaceuticals
Ferri, N., Bellosta, S., Baldessin, L., Boccia, D., Racagni, G., Corsini, A., 2016. with immunomodulatory properties for cancer therapy. Adv. Drug Deliv. Rev. 141,
Pharmacokinetics interactions of monoclonal antibodies. Pharmacol. Res. 111, 67–91. https://doi.org/10.1016/j.addr.2018.09.002. Epub 2018/09/12, PubMed
592–599. https://doi.org/10.1016/j.phrs.2016.07.015. Epub 2016/07/21, PubMed PMID: 30201522.
PMID: 27438459. Huttunen, K.M., Raunio, H., Rautio, J., 2011. Prodrugs–from serendipity to rational
Fiore, A.E., Bridges, C.B., Cox, N.J., 2009. Seasonal Influenza Vaccines. Vaccines for design. Pharmacol. Rev. 63 (3), 750–771. https://doi.org/10.1124/pr.110.003459.
Pandemic Influenza. Springer, pp. 43–82. Epub 2011/07/09, PubMed PMID: 21737530.
Food, Administration D, 2018. Yescarta Package Insert. June, C.H., O’Connor, R.S., Kawalekar, O.U., Ghassemi, S., Milone, M.C., 2018. CAR T
Forbes, N.S., 2010. Engineering the perfect (bacterial) cancer therapy. Nat. Rev. Canc. 10 cell immunotherapy for human cancer. Science 359 (6382), 1361–1365. https://doi.
(11), 785–794. org/10.1126/science.aar6711. Epub 2018/03/24, PubMed PMID: 29567707.
Fosgerau, K., Hoffmann, T., 2015. Peptide therapeutics: current status and future Kallerup, R.S., Foged, C., 2015. Classification of Vaccines. Subunit Vaccine Delivery.
directions. Drug Discov. Today 20 (1), 122–128. https://doi.org/10.1016/j. Springer, pp. 15–29.
drudis.2014.10.003. Epub 2014/12/03, PubMed PMID: 25450771. Kaplon, H., Reichert, J.M., 2018. Antibodies to watch in 2018. mAbs 10 (2), 183–203.
Garber, K., 2006. China approves world’s first oncolytic virus therapy for cancer https://doi.org/10.1080/19420862.2018.1415671. Epub 2018/01/05, PubMed
treatment. J. Natl. Cancer Inst. 98 (5), 298–300. https://doi.org/10.1093/jnci/ PMID: 29300693; PubMed Central PMCID: PMCPMC5825203.
djj111. Epub 2006/03/02, PubMed PMID: 16507823. Kapusta, D., 2007. Drug excretion. In: Enna, S.J., Bylund, D.B. (Eds.), xPharm: the
Garcia-Alonso, S., Ocana, A., Pandiella, A., 2018. Resistance to antibody-drug Comprehensive Pharmacology Reference. Elsevier, New York, pp. 1–2.
conjugates. Canc. Res. 78 (9), 2159–2165. https://doi.org/10.1158/0008-5472. Keedy, V., Wang, W., Schiller, J., Chada, S., Slovis, B., Coffee, K., et al., 2008. Phase I
CAN-17-3671. Epub 2018/04/15, PubMed PMID: 29653942. study of adenovirus p53 administered by bronchoalveolar lavage in patients with
Garcia-Moure, M., Martinez-Velez, N., Patino-Garcia, A., Alonso, M.M., 2017. Oncolytic bronchioloalveolar cell lung carcinoma: ECOG 6597. J. Clin. Oncol. 26 (25),
adenoviruses as a therapeutic approach for osteosarcoma: a new hope. J Bone Oncol 4166–4171. https://doi.org/10.1200/JCO.2007.15.6927. Epub 2008/09/02,
9, 41–47. https://doi.org/10.1016/j.jbo.2016.12.001. Epub 2017/12/12, PubMed PubMed PMID: 18757331; PubMed Central PMCID: PMCPMC2654378.
PMID: 29226089; PubMed Central PMCID: PMCPMC5715440. Keitel, W.A., Atmar, R.L., 2009. Vaccines for Pandemic Influenza: Summary of Recent
Garg, A., Balthasar, J.P., 2007. Physiologically-based pharmacokinetic (PBPK) model to Clinical Trials. Vaccines for Pandemic Influenza. Springer, pp. 431–451.
predict IgG tissue kinetics in wild-type and FcRn-knockout mice. J. Pharmacokinet. Keller, B.A., Bell, J.C., 2016. Oncolytic viruses-immunotherapeutics on the rise. J. Mol.
Pharmacodyn. 34 (5), 687–709. https://doi.org/10.1007/s10928-007-9065-1. Epub Med. (Berl.) 94 (9), 979–991. https://doi.org/10.1007/s00109-016-1453-9. Epub
2007/07/20, PubMed PMID: 17636457. 2016/08/06, PubMed PMID: 27492706.
Garrido, F., Aptsiauri, N., Doorduijn, E.M., Garcia Lora, A.M., van Hall, T., 2016. The Kennedy, P.J., Oliveira, C., Granja, P.L., Sarmento, B., 2018. Monoclonal antibodies:
urgent need to recover MHC class I in cancers for effective immunotherapy. Curr. technologies for early discovery and engineering. Crit. Rev. Biotechnol. 38 (3),
Opin. Immunol. 39, 44–51. https://doi.org/10.1016/j.coi.2015.12.007. Epub 2016/ 394–408. https://doi.org/10.1080/07388551.2017.1357002. Epub 2017/08/10,
01/23, PubMed PMID: 26796069; PubMed Central PMCID: PMCPMC5138279. PubMed PMID: 28789584.
Garrod, A.N., Rajan-Sithamparanadarajah, R., 2003. Developing COSHH Essentials: Kim, J.H., Jacob, J., 2009. DNA Vaccines against Influenza Viruses. Vaccines for
dermal exposure, personal protective equipment and first aid. Ann. Occup. Hyg. 47 Pandemic Influenza. Springer, pp. 197–210.
(7), 577–588. https://doi.org/10.1093/annhyg/meg089. Epub 2003/10/08, Kimball, K.J., Preuss, M.A., Barnes, M.N., Wang, M., Siegal, G.P., Wan, W., et al., 2010.
PubMed PMID: 14530184. A phase I study of a tropism-modified conditionally replicative adenovirus for
Gould, J., Callis, C.M., Dolan, D.G., Stanard, B., Weideman, P.A., 2016. Special endpoint recurrent malignant gynecologic diseases. Clin. Canc. Res. : Off. J. Am. Assoc. Canc.
and product specific considerations in pharmaceutical acceptable daily exposure Res 16 (21), 5277–5287. https://doi.org/10.1158/1078-0432.CCR-10-0791. Epub
derivation. Regul. Toxicol. Pharmacol. 79 (Suppl. 1), S79–S93. https://doi.org/ 2010/10/28, PubMed PMID: 20978148; PubMed Central PMCID:
10.1016/j.yrtph.2016.05.022. Epub 2016/05/29, PubMed PMID: 27233924. PMCPMC2970766.
Gould, J.C., Carvajal, I., Davidson, T., Graham, J., Hillegass, J., Julien, S., et al., 2018. Koski, A., Bramante, S., Kipar, A., Oksanen, M., Juhila, J., Vassilev, L., et al., 2015.
Bioavailability of protein therapeutics in rats following inhalation exposure: Biodistribution analysis of oncolytic adenoviruses in patient Autopsy samples reveals
relevance to occupational exposure limit calculations. Regul. Toxicol. Pharmacol. vascular transduction of noninjected tumors and tissues. Mol. Ther. 23 (10),
100, 35–44. https://doi.org/10.1016/j.yrtph.2018.10.003. Epub 2018/10/07, 1641–1652. https://doi.org/10.1038/mt.2015.125. Epub 2015/07/15, PubMed
PubMed PMID: 30291877. PMID: 26156245; PubMed Central PMCID: PMCPMC4817914.
Guideline, I.H.T. (Ed.), 2018. Assessment and Control of Dna Reactive (Mutagenic) Krause, M.E., Sahin, E., 2019. Chemical and physical instabilities in manufacturing and
Impurities in Pharmaceuticals to Limit Potential Carcinogenic RISK M7. storage of therapeutic proteins. Curr. Opin. Biotechnol. 60, 159–167. https://doi.
International Conference on Harmonization of Technical Requirements for org/10.1016/j.copbio.2019.01.014. Epub 2019/03/13, PubMed PMID: 30861476.
Registration of Pharmaceuticals for Human Use. ICH), Geneva. Kraynov, E., Kamath, A.V., Walles, M., Tarcsa, E., Deslandes, A., Iyer, R.A., et al., 2016.
Guimond, A., Viau, E., Aube, P., Renzi, P.M., Paquet, L., Ferrari, N., 2008. Advantageous Current approaches for absorption, distribution, metabolism, and excretion
toxicity profile of inhaled antisense oligonucleotides following chronic dosing in characterization of antibody-drug conjugates: an industry white paper. Drug Metab.
non-human primates. Pulm. Pharmacol. Therapeut. 21 (6), 845–854. https://doi. Dispos. 44 (5), 617–623. https://doi.org/10.1124/dmd.115.068049. Epub 2015/
org/10.1016/j.pupt.2008.08.001. Epub 2008/09/02, PubMed PMID: 18761414. 12/17, PubMed PMID: 26669328.
Hacein-Bey-Abina, S., Garrigue, A., Wang, G.P., Soulier, J., Lim, A., Morillon, E., et al., Kroes, R., Renwick, A.G., Cheeseman, M., Kleiner, J., Mangelsdorf, I., Piersma, A., et al.,
2008. Insertional oncogenesis in 4 patients after retrovirus-mediated gene therapy of 2004. Structure-based thresholds of toxicological concern (TTC): guidance for
SCID-X1. J. Clin. Invest. 118 (9), 3132–3142. https://doi.org/10.1172/JCI35700. application to substances present at low levels in the diet. Food Chem. Toxicol. 42
Epub 2008/08/09, PubMed PMID: 18688285; PubMed Central PMCID: (1), 65–83. https://doi.org/10.1016/j.fct.2003.08.006. Epub 2003/11/25, PubMed
PMCPMC2496963. PMID: 14630131.
Halsen, G., Kramer, I., 2011. Assessing the risk to health care staff from long-term Kuehl, P.J., Boyden, T., Dobry, D.E., Doyle-Eisele, M., Friesen, D.T., McDonald, J.D.,
exposure to anticancer drugs–the case of monoclonal antibodies. J. Oncol. Pharm. et al., 2016. Inhaled PYY(3-36) dry-powder formulation for appetite suppression.
Pract. 17 (1), 68–80. https://doi.org/10.1177/1078155210376847. Epub 2010/07/ Drug Dev. Ind. Pharm. 42 (1), 150–156. https://doi.org/10.3109/
30, PubMed PMID: 20667850. 03639045.2015.1036067. Epub 2015/05/27, PubMed PMID: 26006332.
Han, T.H., Zhao, B., 2014. Absorption, distribution, metabolism, and excretion Labrijn, A.F., Janmaat, M.L., Reichert, J.M., Parren, P.W., 2019. Bispecific antibodies: a
considerations for the development of antibody-drug conjugates. Drug Metab. mechanistic review of the pipeline. Nat. Rev. Drug Discov. 18 (8), 585–608.
Dispos. 42 (11), 1914–1920. https://doi.org/10.1124/dmd.114.058586. Epub Larson, C., Oronsky, B., Varner, G., Caroen, S., Burbano, E., Insel, E., et al., 2018.
2014/07/23, PubMed PMID: 25048520. A practical guide to the handling and administration of personalized
Hartmann, E.K., Ziebart, A., Thomas, R., Liu, T., Schad, A., Tews, M., et al., 2015. transcriptionally attenuated oncolytic adenoviruses (PTAVs). OncoImmunology 7
Inhalation therapy with the synthetic TIP-like peptide AP318 attenuates pulmonary (9), e1478648. https://doi.org/10.1080/2162402X.2018.1478648. Epub 2018/09/
inflammation in a porcine sepsis model. BMC Pulm. Med. 15, 7. https://doi.org/ 20, PubMed PMID: 30228948; PubMed Central PMCID: PMCPMC6140583.
10.1186/s12890-015-0002-6. Epub 2015/04/17, PubMed PMID: 25879802; Le, T.T., Andreadakis, Z., Kumar, A., Roman, R.G., Tollefsen, S., Saville, M., et al., 2020.
PubMed Central PMCID: PMCPMC4346123. The COVID-19 vaccine development landscape. Nat. Rev. Drug Discov. 19 (5),
Hasan, M., Alam, S., Poddar, S.K., 2018. Antibody-drug conjugates: a review on the 305–306.
epitome of targeted anti- cancer therapy. Curr. Clin. Pharmacol. 13 (4), 236–251. Lee, D.W., Gardner, R., Porter, D.L., Louis, C.U., Ahmed, N., Jensen, M., et al., 2014.
https://doi.org/10.2174/1574884712666180802095521. Epub 2018/08/04, Current concepts in the diagnosis and management of cytokine release syndrome.
PubMed PMID: 30073930. Blood 124 (2), 188–195. https://doi.org/10.1182/blood-2014-05-552729. Epub
Hay, K.A., Hanafi, L.A., Li, D., Gust, J., Liles, W.C., Wurfel, M.M., et al., 2017. Kinetics 2014/05/31, PubMed PMID: 24876563; PubMed Central PMCID:
and biomarkers of severe cytokine release syndrome after CD19 chimeric antigen PMCPMC4093680.

19
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

Lenting, P.J., Denis, C.V., Christophe, O.D., 2017. Emicizumab, a bispecific antibody Pfister, T., Dolan, D., Bercu, J., Gould, J., Wang, B., Bechter, R., et al., 2014b.
recognizing coagulation factors IX and X: how does it actually compare to factor Bioavailability of therapeutic proteins by inhalation–worker safety aspects. Ann.
VIII? Blood 130 (23), 2463–2468. Occup. Hyg. 58 (7), 899–911. https://doi.org/10.1093/annhyg/meu038. Epub
Levy, G., 1994. Pharmacologic target-mediated drug disposition. Clin. Pharmacol. Ther. 2014/06/25, PubMed PMID: 24958792.
56 (3), 248–252. https://doi.org/10.1038/clpt.1994.134. Epub 1994/09/01, Phadke, A., 2008. 442. Viral pharmacokinetics and host immune response following
PubMed PMID: 7924119. intratumoral injection of a conditional, replicative, oncolytic adenovirus telomelysin
Lu, J.-D., Xue, J., 2019. Chapter 101 - poisoning: kinetics to therapeutics. In: Ronco, C., (OBP-301). Mol. Ther. 16, S168. https://doi.org/10.1016/S1525-0016(16)39845-8.
Bellomo, R., Kellum, J.A., Ricci, Z. (Eds.), Critical Care Nephrology, third ed. Pinero-Lambea, C., Ruano-Gallego, D., Fernandez, L.A., 2015. Engineered bacteria as
Content Repository Only!, Philadelphia. 600-29.e7. therapeutic agents. Curr. Opin. Biotechnol. 35, 94–102. https://doi.org/10.1016/j.
Machiels, J.P., Salazar, R., Rottey, S., Duran, I., Dirix, L., Geboes, K., et al., 2019. A phase copbio.2015.05.004. Epub 2015/06/13, PubMed PMID: 26070111.
1 dose escalation study of the oncolytic adenovirus enadenotucirev, administered Polu, K.R., Lowman, H.B., 2014. Probody therapeutics for targeting antibodies to
intravenously to patients with epithelial solid tumors (EVOLVE). J Immunother diseased tissue. Expet Opin. Biol. Ther. 14 (8), 1049–1053. https://doi.org/10.1517/
Cancer 7 (1), 20. https://doi.org/10.1186/s40425-019-0510-7. Epub 2019/01/30, 14712598.2014.920814. Epub 2014/05/23, PubMed PMID: 24845630.
PubMed PMID: 30691536; PubMed Central PMCID: PMCPMC6348630. Poon, K.A., Flagella, K., Beyer, J., Tibbitts, J., Kaur, S., Saad, O., et al., 2013. Preclinical
Mager, D.E., Jusko, W.J., 2001. General pharmacokinetic model for drugs exhibiting safety profile of trastuzumab emtansine (T-DM1): mechanism of action of its
target-mediated drug disposition. J. Pharmacokinet. Pharmacodyn. 28 (6), 507–532. cytotoxic component retained with improved tolerability. Toxicol. Appl. Pharmacol.
https://doi.org/10.1023/a:1014414520282. Epub 2002/05/10, PubMed PMID: 273 (2), 298–313. https://doi.org/10.1016/j.taap.2013.09.003. Epub 2013/09/17,
11999290. PubMed PMID: 24035823.
Malik, P., Phipps, C., Edginton, A., Blay, J., 2017. Pharmacokinetic considerations for Prudent, V., Breitbart, W.S., 2017. Chimeric antigen receptor T-cell neuropsychiatric
antibody-drug conjugates against cancer. Pharm. Res. (N. Y.) 34 (12), 2579–2595. toxicity in acute lymphoblastic leukemia. Palliat. Support Care 15 (4), 499–503.
https://doi.org/10.1007/s11095-017-2259-3. Epub 2017/09/20, PubMed PMID: https://doi.org/10.1017/S147895151600095X. Epub 2017/01/05, PubMed PMID:
28924691. 28049548; PubMed Central PMCID: PMCPMC5496800.
Mau-Sorensen, M., Dittrich, C., Dienstmann, R., Lassen, U., Buchler, W., Martinius, H., Rautio, J., Karkkainen, J., Sloan, K.B., 2017. Prodrugs - recent approvals and a glimpse of
et al., 2015. A phase I trial of intravenous catumaxomab: a bispecific monoclonal the pipeline. Eur. J. Pharmaceut. Sci. 109, 146–161. https://doi.org/10.1016/j.
antibody targeting EpCAM and the T cell coreceptor CD3. Canc. Chemother. ejps.2017.08.002. Epub 2017/08/08, PubMed PMID: 28782609.
Pharmacol. 75 (5), 1065–1073. https://doi.org/10.1007/s00280-015-2728-5. Epub Rehman, H., Silk, A.W., Kane, M.P., Kaufman, H.L., 2016. Into the clinic: Talimogene
2015/03/31, PubMed PMID: 25814216. laherparepvec (T-VEC), a first-in-class intratumoral oncolytic viral therapy.
Maus, M.V., Haas, A.R., Beatty, G.L., Albelda, S.M., Levine, B.L., Liu, X., et al., 2013. J Immunother Cancer 4, 53. https://doi.org/10.1186/s40425-016-0158-5. Epub
T cells expressing chimeric antigen receptors can cause anaphylaxis in humans. 2016/09/24, PubMed PMID: 27660707; PubMed Central PMCID:
Cancer Immunol Res 1, 26–31. Epub 2014/01/17. PubMed PMID: 24432303; PMCPMC5029010.
PubMed Central PMCID: PMCPMC3888798. Remesh, A., 2012. Toxicities of anticancer drugs and its management. ijbcp 1. https://
MedImmune. FluMist Quadrivalent [Package Insert], 2013. MedImmune, Gaithersburg, doi.org/10.5455/2319-2003.ijbcp000812.
MD. Ritchie, M., Tchistiakova, L., Scott, N., 2013. Implications of receptor-mediated
Miliotou, A.N., Papadopoulou, L.C., 2018. CAR T-cell therapy: a new era in cancer endocytosis and intracellular trafficking dynamics in the development of antibody
immunotherapy. Curr. Pharmaceut. Biotechnol. 19 (1), 5–18. https://doi.org/ drug conjugates. mAbs 5 (1), 13–21. https://doi.org/10.4161/mabs.22854. Epub
10.2174/1389201019666180418095526. Epub 2018/04/19, PubMed PMID: 2012/12/12, PubMed PMID: 23221464; PubMed Central PMCID:
29667553. PMCPMC3564878.
Morgan, R.A., Yang, J.C., Kitano, M., Dudley, M.E., Laurencot, C.M., Rosenberg, S.A., Roberts, S.A., Andrews, P.A., Blanset, D., Flagella, K.M., Gorovits, B., Lynch, C.M., et al.,
2010. Case report of a serious adverse event following the administration of T cells 2013. Considerations for the nonclinical safety evaluation of antibody drug
transduced with a chimeric antigen receptor recognizing ERBB2. Mol. Ther. 18 (4), conjugates for oncology. Regul. Toxicol. Pharmacol. 67 (3), 382–391. https://doi.
843–851. https://doi.org/10.1038/mt.2010.24. Epub 2010/02/25, PubMed PMID: org/10.1016/j.yrtph.2013.08.017. Epub 2013/09/10, PubMed PMID: 24012707.
20179677; PubMed Central PMCID: PMCPMC2862534. Rosewell Shaw, A., Suzuki, M., 2016. Recent advances in oncolytic adenovirus therapies
Moschos, S.A., Frick, M., Taylor, B., Turnpenny, P., Graves, H., Spink, K.G., et al., 2011. for cancer. Curr. Opin. Virol 21, 9–15. https://doi.org/10.1016/j.
Uptake, efficacy, and systemic distribution of naked, inhaled short interfering RNA coviro.2016.06.009. Epub 2016/07/06, PubMed PMID: 27379906; PubMed Central
(siRNA) and locked nucleic acid (LNA) antisense. Mol. Ther. 19 (12), 2163–2168. PMCID: PMCPMC5138135.
https://doi.org/10.1038/mt.2011.206. Epub 2011/10/06, PubMed PMID: Russell, W.C., 2009. Adenoviruses: update on structure and function. J. Gen. Virol. 90 (Pt
21971426; PubMed Central PMCID: PMCPMC3242665. 1), 1–20. https://doi.org/10.1099/vir.0.003087-0. Epub 2008/12/18, PubMed
Mueller, K.T., Maude, S.L., Porter, D.L., Frey, N., Wood, P., Han, X., et al., 2017. Cellular PMID: 19088268.
kinetics of CTL019 in relapsed/refractory B-cell acute lymphoblastic leukemia and Ryman, J.T., Meibohm, B., 2017. Pharmacokinetics of monoclonal antibodies. CPT
chronic lymphocytic leukemia. Blood 130 (21), 2317–2325. https://doi.org/ Pharmacometrics Syst. Pharmacol. 6 (9), 576–588. https://doi.org/10.1002/
10.1182/blood-2017-06-786129. Epub 2017/09/25, PubMed PMID: 28935694; psp4.12224. Epub 2017/06/28, PubMed PMID: 28653357; PubMed Central PMCID:
PubMed Central PMCID: PMCPMC5731220. PMCPMC5613179.
Nations U. Globally Harmonized System of Classification and Labelling of Chemicals Satti, I., Meyer, J., Harris, S.A., Thomas, Z.-R.M., Griffiths, K., Antrobus, R.D., et al.,
(GHS)2019. 2014. Safety and immunogenicity of a candidate tuberculosis vaccine MVA85A
Naumann, B.D., Sargent, E.V., Starkman, B.S., Fraser, W.J., Becker, G.T., Kirk, G.D., delivered by aerosol in BCG-vaccinated healthy adults: a phase 1, double-blind,
1996. Performance-based exposure control limits for pharmaceutical active randomised controlled trial. Lancet Infect. Dis. 14 (10), 939–946.
ingredients. Am. Ind. Hyg. Assoc. J. 57 (1), 33–42. https://doi.org/10.1080/ Schmidt-Wolf, I.G., Negrin, R.S., Kiem, H.P., Blume, K.G., Weissman, I.L., 1991. Use of a
15428119691015197. Epub 1996/01/01, PubMed PMID: 8588551. SCID mouse/human lymphoma model to evaluate cytokine-induced killer cells with
Neelapu, S.S., Locke, F.L., Bartlett, N.L., Lekakis, L.J., Miklos, D.B., Jacobson, C.A., et al., potent antitumor cell activity. J. Exp. Med. 174 (1), 139–149. https://doi.org/
2017. Axicabtagene ciloleucel CAR T-cell therapy in refractory large B-cell 10.1084/jem.174.1.139. Epub 1991/07/01, PubMed PMID: 1711560; PubMed
lymphoma. N. Engl. J. Med. 377 (26), 2531–2544. https://doi.org/10.1056/ Central PMCID: PMCPMC2118875.
NEJMoa1707447. Epub 2017/12/12, PubMed PMID: 29226797; PubMed Central Senter, P.D., 2009. Potent antibody drug conjugates for cancer therapy. Curr. Opin.
PMCID: PMCPMC5882485. Chem. Biol. 13 (3), 235–244. https://doi.org/10.1016/j.cbpa.2009.03.023. Epub
NIOSH, 2016. NIOSH List of Antineoplastic and Other Hazardous Drugs in Healthcare 2009/05/06, PubMed PMID: 19414278.
Settings. US Department of Health and Human Services, Centers for Disease Control Seth, P., Higginbotham, J., 2000. Advantages and disadvantages of multiple different
and Prevention, National Institute for Occupational Safety and Health, DHHS methods of adenoviral vector construction. Methods Mol. Med. 45, 189–198.
(NIOSH) Publication. Number 2016-161 (Supersedes 2014-138). https://doi.org/10.1385/1-59259-079-9:189. Epub 2000/01/01, PubMed PMID:
NIOSH. reportTechnical Report: the NIOSH Occupational Exposure Banding Process for 21341057.
Chemical Risk Management. US Department of Health and Human Services, Centers Sibbald, B., 2001. Death but one unintended consequence of gene-therapy trial. CMAJ
for Disease Control and Prevention, National Institute for Occupational Safety and (Can. Med. Assoc. J.) 164 (11), 1612. Epub 2001/06/14. PubMed PMID: 11402803;
Health, DHHS (NIOSH) Publication No 2019-132. 2019. PubMed Central PMCID: PMCPMC81135.
Novartis, A., 2017. Kymriah Package Insert. Siddiqui, M.A.A., Perry, C.M., 2006. Human papillomavirus quadrivalent (types 6, 11,
OSHA. Occupational Safety and Health Administration: Safety and Health Topics - 16, 18) recombinant vaccine (Gardasil®). Drugs 66 (9), 1263–1271.
Radiation, 2020. Singh, G., Robinson, C.M., Dehghan, S., Jones, M.S., Dyer, D.W., Seto, D., et al., 2013.
Onoue, S., Aoki, Y., Matsui, T., Kojo, Y., Misaka, S., Mizumoto, T., et al., 2011. Homologous recombination in E3 genes of human adenovirus species D. J. Virol. 87
Formulation design and in vivo evaluation of dry powder inhalation system of new (22), 12481–12488. https://doi.org/10.1128/JVI.01927-13. Epub 2013/09/13,
vasoactive intestinal peptide derivative ([R(15, 20, 21), L(17), A(24,25), des-N(28)]- PubMed PMID: 24027303; PubMed Central PMCID: PMCPMC3807923.
VIP-GRR) in experimental asthma/COPD model rats. Int. J. Pharm. 410 (1–2), Singh, S., Kumar, N.K., Dwiwedi, P., Charan, J., Kaur, R., Sidhu, P., et al., 2018.
54–60. https://doi.org/10.1016/j.ijpharm.2011.03.021. Epub 2011/03/23, PubMed Monoclonal antibodies: a review. Curr. Clin. Pharmacol. 13 (2), 85–99. https://doi.
PMID: 21419198. org/10.2174/1574884712666170809124728. Epub 2017/08/12, PubMed PMID:
Organization WH, 2005. WHO Biosafety Risk Assessment and Guidelines for the 28799485.
Production and Quality Control of Human Influenza Pandemic Vaccines, 2007. Stanard, B., Dolan, D.G., Hanneman, W., Legare, M., Bercu, J.P., 2015. Threshold of
Pfister, T., Dolan, D., Bercu, J., Gould, J., Wang, B., Bechter, R., et al., 2014a. toxicological concern (TTC) for developmental and reproductive toxicity of
Bioavailability of therapeutic proteins by inhalation—worker safety aspects. Ann. anticancer compounds. Regul. Toxicol. Pharmacol. 72 (3), 602–609.
Occup. Hyg. 58 (7), 899–911. Stanford. Stanford Environmental Health and Safety: Adenovirus Fact Sheet, 2020.

20
J.C. Graham et al. Regulatory Toxicology and Pharmacology 118 (2020) 104813

Stein, C.A., Castanotto, D., 2017. FDA-approved oligonucleotide therapies in 2017. Mol. Vormittag, P., Gunn, R., Ghorashian, S., Veraitch, F.S., 2018. A guide to manufacturing
Ther. 25 (5), 1069–1075. CAR T cell therapies. Curr. Opin. Biotechnol. 53, 164–181. https://doi.org/10.1016/
Strohl, W.R., 2015. Fusion proteins for half-life extension of biologics as a strategy to j.copbio.2018.01.025. Epub 2018/02/21, PubMed PMID: 29462761.
make biobetters. BioDrugs 29 (4), 215–239. https://doi.org/10.1007/s40259-015- Walker, M.P., Cowlen, M., Christensen, D., Miyamoto, M., Barley, P., Crowder, T., 2017.
0133-6. Epub 2015/07/17, PubMed PMID: 26177629; PubMed Central PMCID: Nonclinical safety assessment of SPX-101, a novel peptide promoter of epithelial
PMCPMC4562006. sodium channel internalization for the treatment of cystic fibrosis. Inhal. Toxicol. 29
Taft, D.R., 2009. Chapter 9 - drug excretion. In: Hacker, M., Messer, W., Bachmann, K. (8), 356–365. https://doi.org/10.1080/08958378.2017.1366602. Epub 2017/10/
(Eds.), Pharmacology. Academic Press, San Diego, pp. 175–199. 07, PubMed PMID: 28984146.
Tarhini, A., Lo, E., Minor, D.R., 2010. Releasing the brake on the immune system: Wang, Z., 2017. Clinical pharmacological considerations on CAR-T cell therapy for
ipilimumab in melanoma and other tumors. Cancer Biother. Radiopharm. 25 (6), cancer. J. Pharmacol. Clin. Res. 3 https://doi.org/10.19080/JPCR.2017.03.555619.
601–613. https://doi.org/10.1089/cbr.2010.0865. Epub 2011/01/06, PubMed Wechman, S.L., Rao, X.M., McMasters, K.M., Zhou, H.S., 2016. Adenovirus with DNA
PMID: 21204754; PubMed Central PMCID: PMCPMC3011989. packaging gene mutations increased. Virus Release. Viruses. 8 (12) https://doi.org/
Templin, M.V., Levin, A.A., Graham, M.J., Aberg, P.M., Axelsson, B.I., Butler, M., et al., 10.3390/v8120333. Epub 2016/12/22, PubMed PMID: 27999391; PubMed Central
2000. Pharmacokinetic and toxicity profile of a phosphorothioate oligonucleotide PMCID: PMCPMC5192394.
following inhalation delivery to lung in mice. Antisense Nucleic Acid Drug Dev. 10 WHO, 2019. WHO Recommendations for Routine Immunization Summary Tables [cited
(5), 359–368. https://doi.org/10.1089/oli.1.2000.10.359. Epub 2000/11/18, 2020 August 7]. Available from: https://www.who.int/immunization/policy/immu
PubMed PMID: 11079575. nization_tables/en/.
Tian, G., Liu, J., Zhou, J.S., Chen, W., 2009. Multiple hepatic arterial injections of Wood, J.M., Robertson, J.S., 2004. From lethal virus to life-saving vaccine: developing
recombinant adenovirus p53 and 5-fluorouracil after transcatheter arterial inactivated vaccines for pandemic influenza. Nat. Rev. Microbiol. 2 (10), 842–847.
chemoembolization for unresectable hepatocellular carcinoma: a pilot phase II trial. Xia, Z.J., Chang, J.H., Zhang, L., Jiang, W.Q., Guan, Z.Z., Liu, J.W., et al., 2004. [Phase III
Anti Canc. Drugs 20 (5), 389–395. https://doi.org/10.1097/ randomized clinical trial of intratumoral injection of E1B gene-deleted adenovirus
CAD.0b013e32832a2df9. Epub 2009/03/17, PubMed PMID: 19287305. (H101) combined with cisplatin-based chemotherapy in treating squamous cell
Trivedi, A., Stienen, S., Zhu, M., Li, H., Yuraszeck, T., Gibbs, J., et al., 2017. Clinical cancer of head and neck or esophagus]. Ai Zheng 23 (12), 1666–1670. Epub 2004/
pharmacology and translational aspects of bispecific antibodies. Clin. Transl. Sci. 10 12/17. PubMed PMID: 15601557.
(3), 147–162. https://doi.org/10.1111/cts.12459. Epub 2017/03/16, PubMed Yip, A., Webster, R.M., 2018. The market for chimeric antigen receptor T cell therapies.
PMID: 28297195; PubMed Central PMCID: PMCPMC5421745. Nat. Rev. Drug Discov. 17 (3), 161–162. https://doi.org/10.1038/nrd.2017.266.
van Schalkwyk, M.C., Papa, S.E., Jeannon, J.P., Guerrero Urbano, T., Spicer, J.F., Epub 2018/01/30, PubMed PMID: 29375140.
Maher, J., 2013. Design of a phase I clinical trial to evaluate intratumoral delivery of Young, P.A., Morrison, S.L., Timmerman, J.M., 2014. Antibody-cytokine fusion proteins
ErbB-targeted chimeric antigen receptor T-cells in locally advanced or recurrent for treatment of cancer: engineering cytokines for improved efficacy and safety.
head and neck cancer. Hum Gene Ther. Clin. Dev. 24 (3), 134–142. https://doi.org/ Semin. Oncol. 41 (5), 623–636. https://doi.org/10.1053/j.seminoncol.2014.08.002.
10.1089/humc.2013.144. Epub 2013/10/09, PubMed PMID: 24099518. Epub 2014/12/03, PubMed PMID: 25440607; PubMed Central PMCID:
Vande Casteele, N., Gils, A., 2015. Pharmacokinetics of anti-TNF monoclonal antibodies PMCPMC4354941.
in inflammatory bowel disease: adding value to current practice. J. Clin. Pharmacol. Zalk, D.M., Nelson, D.I., 2008. History and evolution of control banding: a review.
55 (S3), S39–S50. J. Occup. Environ. Hyg. 5 (5), 330–346. https://doi.org/10.1080/
Vasquez, K.M., Narayanan, L., Glazer, P.M., 2000. Specific mutations induced by triplex- 15459620801997916. Epub 2008/03/20, PubMed PMID: 18350442.
forming oligonucleotides in mice. Science 290 (5491), 530–533. Zanotti-Fregonara, P., Lammertsma, A.A., Innis, R.B., 2013. Suggested pathway to assess
Veber, D.F., Johnson, S.R., Cheng, H.Y., Smith, B.R., Ward, K.W., Kopple, K.D., 2002. radiation safety of (1)(8)F-labeled PET tracers for first-in-human studies. Eur. J.
Molecular properties that influence the oral bioavailability of drug candidates. Nucl. Med. Mol. Imag. 40 (11), 1781–1783. https://doi.org/10.1007/s00259-013-
J. Med. Chem. 45 (12), 2615–2623. https://doi.org/10.1021/jm020017n. Epub 2512-x. Epub 2013/07/23, PubMed PMID: 23868334; PubMed Central PMCID:
2002/05/31, PubMed PMID: 12036371. PMCPMC3844925.

21