Content uploaded by Andrew Walsh
Author content
All content in this area was uploaded by Andrew Walsh on Jan 13, 2018
Content may be subject to copyright.
49
Chimica Oggi - Chemistry Today - vol. 35(6) November/December 2017
presented by the products manufactured. Where the
toxicological evaluation supports a threshold value, this
should be used as an input parameter in risk assessment (11).
Potential absence of thresholds is connected to certain
genotoxic mechanisms and indicates an absence of a dose
under which there would be no effects. Examples of threshold
mechanisms and setting limits for genotoxic substances have
been published by Lovsin Barle et al. (12).
The problem is that, despite their ease of communication and
understanding, quantitative hazard classes alone cannot be
used in the risk assessment and risk mitigation measures
cannot be based on qualitative criteria alone. The
interchangeable usage of these qualitative and quantitative
criteria for describing hazards and potency as we have seen
in this section demonstrates this. For example, for
occupational health purposes processes require specialized
containment to ensure that employees and their environment
are protected from exposure. These processes can limit the
exposure to a certain level, e.g. OEL, which is a quantitative
measure. But designing containment for a qualitative hazard
class is not possible.
HIGH RISKS FOR CROSS-CONTAMINATION
Risk is the probability that some adverse effect (e.g. skin
irritation or cancer) will result from a given exposure to a
chemical. The risk posed by a substance depends both on
the intrinsic properties of the substance (hazard) and of
exposure (1). The intrinsic hazards and the potency of
substances are substance specific. These characteristics are
normally assessed by qualified toxicologists (for definition see
Olson et al., (13)). The outcome of such a toxicological
assessment is a Permitted Daily Exposure (PDE) value, as is
now required by the EU GMP chapter 5. The PDE has been
defined by the ‘Guideline on setting health based exposure
limits for use in risk identification in the manufacture of
different medicinal products in shared facilities” (14) as “a
substance-specific dose that is unlikely to cause an adverse
effect if an individual is exposed at or below this dose every
day for a lifetime”. The PDE and its definition are synonymous
with the Acceptable Daily Exposure (ADE).
For PDE calculations based on toxicological criteria a
critical effect needs to be identified; in this context any
effect is an adverse effect. Identification of the critical
effect dose termed the Point of Departure (POD) which is
used for the PDE calculation (15). Depending on the POD,
several adjustment factors (AF) are used to account for
uncertainties when extrapolating the hazard information to
the target population of human patients and differences in
routes of administration (16).
OELs (Occupational Exposure Limits) are calculated in a very
similar way; the critical difference is in the determination of
the critical effect for the target population of healthy workers
vs. patients, followed by modification of AFs to adjust for this
criterion. Also the OEL is expressed in µg/m3 to extrapolate to
the volume of air inhaled in 8h working time (10 m3).
Similarly to the OEL, PDE values fall into a range that spans
several orders of magnitude. In a study done with over 200
substances in a typical pharma portfolio, there were about
63% of drugs with a PDE ≥100 µg/day, while there were only
2% of substances with PDE < 1 µg/day (figure 1).
Toxic chemicals are defined as those that can produce
injuries or lethal effects on contact with body cells. It is
acknowledged that results of the tests that determine toxicity
are not absolute due to great variability in the target
populations. The term “highly toxic” applies to those
substances for which toxic effects are severe and occur in
small doses. A problem arises in providing adequate
definitions for detailing a list of chemicals of concern,
particularly for regulatory purposes. A comparison of the
doses used in human therapy with doses that produce lethal
doses in animals has been compared by Lovsin Barle et al. (6).
Based on their results, toxic classes based on lethal doses in
animals are not useful for human estimation of acute toxicity.
Potency is determined qualitatively by the dose of the drug
required to produce an effect. An HPAPI(e.g., fentanyl,
estrogen, calcitriol) evokes a desired pharmacological
response in target population of patients at low doses. On the
other hand, these drugs may also evoke unintended, adverse
effects at doses lower than the dose for the intended effects.
These effects may be acceptable in the patient population;
however they are not acceptable in other populations, such
as workers, or in patients who might be exposed to such a
drug as a contaminant.
The definition of an HPAPI varies significantly in the
literature and is not harmonized. The following descriptions
may be found:
1. Based on therapeutic dose (quantitative criteria): A
substance with biological activity at approximately 150
μg/kg of body weight or below in humans (therapeutic
daily dose at or below 10 mg) (7)
2. Based on Occupational Exposure Limit (OEL) (quantitative
criteria): Substance with an OEL at or below 10 μg/m3 of
air as an 8-h time-weighted average (7) or OEL below
1μg/m3 (8)
3. Based on hazards (qualitative and quantitative criteria):
Substance with high selectivity (i.e., ability to bind to
specific receptors or inhibit specific enzymes) and/or with
the potential to cause cancer, mutations, developmental
effects, target organ effects or reproductive toxicity at
low doses
4. Based on risks (quantitative criteria): extreme acute and
chronic toxicity, irreversible effects, strong sensitizer, poor
or no warning properties, quick absorption rate, known
“genic” effects, higher degree of medical intervention
required, affecting sensitive subpopulations (9)
5. Or, by default, a novel compound of unknown potency
and toxicity (7)
The term highly potent is most frequently used as a synonym
to highly active. Due to the perceived risk, certain classes of
medicinal product have previously been required to be
manufactured in dedicated or segregated self-contained
facilities including, “certain antibiotics, certain hormones,
certain cytotoxics and certain highly active drugs” (10). No
definitions for these criteria, including highly actives were
available at that time. This has been since revised; the EU
Guidelines for Good Manufacturing Practice (GMP) for
Medicinal Products for Human and Veterinary Use Chapter 5
now states that a Quality Risk Management process which
includes a potency and toxicological evaluation, should be
used to assess and control the cross-contamination risks
48
Chimica Oggi - Chemistry Today - vol. 35(6) November/December 2017
KEYWORDS: Highly hazardous, high potency, active pharmaceutical ingredient, risk assessment, cross-contamination.
Abstract The terms highly hazardous, highly active and highly potent are often used in mixed or inconsistent contexts
regarding good manufacturing practice and worker safety. This article provides an overview of definitions
and criteria used for classifying active pharmaceutical ingredients (APIs) and the use of these terms for identification of risks in cleaning
validation. While competent toxicological evaluation is required to identify and characterize the health hazards for each drug substance
and then calculate a health based exposure limit, further steps in risk assessment for cross-contamination include batch size, maximal
daily dose of subsequent products, equipment surface areas, cleaning process parameters and process capability. Risks presented by
drug substances cannot be managed by simple classification into highly potent and highly hazardous categories.
Are high potency active
pharmaceutical ingredients (HPAPI)
also high risks for cross-contamination?
INTRODUCTION
Although the terms hazard and potency are thought to be
well understood, they are often mixed or used in inconsistent
contexts. It is acknowledged that the hazards, as qualitative
aspects are much easier to assess, communicate and
perceive by the general public, but hazards alone are
insufficient to assess the risks for populations and define the
risk mitigation strategies that should follow. Scales that include
the qualitative and quantitative aspects of hazards and
potency that can be used in setting priorities of concern have
been proposed for decades. One such example is ranking of
possible carcinogenic hazards by an index that relates to
potency of each carcinogen in rodents to the exposure in
humans (Ames, 1987). In this article we have looked at the
common definitions of human health hazards and potency of
substances, and the possibility for the consistent use of these
terms in risk assessment for cross-contamination of
pharmaceuticals in multipurpose manufacturing facilities.
HIGHLY HAZARDOUS, TOXIC, POTENT AND ACTIVE
Human health hazard assessment is defined as “evaluation of
the information on the intrinsic properties of a substance to
conclude on the hazards to human health” (1). According to
the European Chemicals Agency (ECHA) glossary, a human
hazard is defined as the “intrinsic properties of a substance
that have the potential to cause adverse effects in humans”.
Hazards are therefore qualitative properties of substances
and do not include a quantitative aspect, which involves the
dose of that substance that is needed for these hazards to
appear. Conversely toxicity is defined as degree to which a
substance or mixture of substances can cause adverse
effects to humans. The toxicity of a drug depends on its
dosage and therefore has a quantitative aspect to it’s
definition, which is not an intrinsic property.
While hazards are typically grouped into hazard classes, which
have been harmonized for chemicals (2), there is no consensus
on the term “highly hazardous”. European Medicines Agency
(EMA) has attempted to give guidance on this term in their
Q&A document EMA/CHMP/CVMP/SWP/463311/2016 (3).
While there are some qualitative hazard classes mentioned
there as the basis for definition (eg. genotoxic, carcinogenic,
reprotoxic substances), these terms are combined with the
quantitative aspect, their potency. It was pointed out in the
publication by Lovsin Barle et al. (4), that additional criteria that
require substance specific evaluation of an API may lead to a
low health based exposure limit (HBEL), based on the following:
potential for bioaccumulation, differences in bioavailability for
different routes of exposure, non-daily dosing regime and short
treatment periods of patients.
Similarly to EMA, the National Institute for Occupational Safety
and Health (NIOSH) provides a definition for hazardous drugs
based on qualitative and quantitative criteria
(carcinogenicity, teratogenicity or other developmental
toxicity, reproductive toxicity, organ toxicity at low doses,
genotoxicity and structure and toxicity profiles of new drugs
that mimic existing drugs determined hazardous by the above
criteria). They emphasize the quantitative aspects that
include potency in more detail in the footnote which contains
the term “level of toxicity”. The footnote states “The level of
toxicity reflects a continuum from relatively nontoxic to
production of toxic effects in patients at low doses (for
example, a few milligrams or less).” (5).
PHARMACEUTICAL
CHEMISTRY: APIS, HPAPIS
ESTER LOVSIN BARLE1*, ANDREW WALSH2
*Corresponding Author
1. Global HSE& BCM, Novartis, Basel, Switzerland,
2. President, Center for Pharmaceutical Cleaning Innovation, Hillsborough, USA
Industry perspective
51
50
Chimica Oggi - Chemistry Today - vol. 35(6) November/December 2017 Chimica Oggi - Chemistry Today - vol. 35(6) November/December 2017
In summary, the risks presented by drug substances cannot be
managed by simple classification into highly hazardous and
highly potent categories of drugs. The hazard level (toxicity) of
the compound and the level of exposure (Process Capability)
must be considered together to arrive at a true measure of
the level of risk and categories such as HPAPI are not helpful.
REFERENCES
1. ECHA (2017), ECHA-term versión 2.0.0/20141128 https://echa-
term.echa.europa.eu/home?p_p_id=term_WAR_
termportlet&p_p_lifecycle=0&p_p_state=maximized&p_p_
mode=view&_term_WAR_termportlet_entryId=10760&_term_WAR_
termportlet_srcLang=en&_term_WAR_termportlet_q=hazard&_
term_WAR_termportlet_searchType=define&_term_WAR_
termportlet_curIndex=8&_term_WAR_termportlet_total=19&_term_
WAR_termportlet_cur=1&_term_WAR_termportlet_jspPage=%2Fht
ml%2Fportlet%2Fterm%2Ffull_entry.jsp&_term_WAR_termportlet_
selLang=en
2. European Commission Regulation (EC) No. 1272/2008 of the
European Parliament and of the Council of 16 December 2008 on
Classification, Labelling and Packaging of Substances and
Mixtures, Amending and Repealing Directives 67/548/EEC and
1999/45/EC, and amending Regulation (EC) No. 1907/2006 (2008).
3. EMA/CHMP/CVMP/SWP/169430/2012, (2016). http://www.ema.
europa.eu/docs/en_GB/document_library/Other/2017/01/
WC500219500.pdf (Last checked on Aug 28th 2017)
4. Lovsin Barle E., Jandard C., et al., Pharm. Technol. 41(5): 42-53
(2017)
5. NIOSH List of Antineoplastic and Other Hazardous Drugs in
Healthcare Settings, (2016) https://www.cdc.gov/niosh/topics/
antineoplastic/pdf/hazardous-drugs-list_2016-161.pdf (Last
checked on Aug 28th 2017)
6. Barle E.L., Looser R., et al., Regul Toxicol Pharmacol. 62(3):412-8
(2012).
7. Bornett D., Pharm. Technol., 2008 (4) Supplement (2008). http://
www.pharmtech.com/high-potency-apis-containment-and-
handling-issues (Last checked on Aug 28th 2017)
8. Axon M., Mason-Home J., et al., Chemistry Today 26, 57-61 (2008).
9. Bornett D (2009), An introduction to High-Potent API Classification.
https://www.scribd.com/document/22669477/An-Introduction-to-
High-Potent-API-Classification (Last checked on Aug 28th 2017)
10. EMEA/INS/GMP/14529/2008. Last checked on 24.7.2017 at http://
www.ema.europa.eu/docs/en_GB/document_library/
Other/2009/10/WC500004696.pdf
11. European Commission (EC) Health and Consumers Directorate
General: EU Guidelines for Good Manufacturing Practice for
Medicinal Products for Human and Veterinary Use. Chapter 5:
Production. Adopted August 13, (2014). http://ec.europa.eu/
health/files/eudralex/vol-4/pdfs-en/cap5en.pdf (Last checked on
Aug 28th 2017)
12. Lovsin Barle E., Winkler G.W., et al., Tox Sci, 151: 2–9 (2016).
13. Olson M., Faria E., et al., Regul Toxicol Pharmacol. 79: S19-27
(2016).
14. EMA/CHMP/CVMP/SWP/169430/2012 (2014). http://www.ema.
europa.eu/docs/en_GB/document_library/Scientific_
this may be measured by the amount of residue that
remains after cleaning manufacturing equipment that
might contaminate the next drug product. For a worker, this
may be measured by the amount of drug substance in the
air in the manufacturing area that they might inhale.
For cleaning of manufacturing equipment, the PDE is only
the starting point for determining what levels of drug
residues are safe. A recent article demonstrated how batch
size, the maximal daily dose of the next product, as well as
the equipment surface influence the calculation of safe
residue levels and consequently the level of risks for cross
contamination (18). While these parameters influence the
calculation of limits they provide no input into the level of
Exposure. Other criteria may directly influence the
cleaning capability of a substance such as manufacturing
equipment Materials of Construction (MOC), the cleaning
process parameters (e.g., cleaning agent, temperature,
etc.) and most importantly the physico-chemical
properties of the substances and their product matrices
that affect their cleanability. These factors directly impact
the level of residues and therefore the level of exposure.
The PDE only reveals what the safe levels of these residues
are. The actual level of residues found after cleaning
reveals the level of exposure and this can be used to
measure the level of risk.
In the world of cleaning validation, drug residue levels have
been collected for many years. These drug residue levels
have typically been compared to one or more limits that
are not health-based or arbitrary (19). The drug residue
levels either passed or failed these limits. What has been
missing, in almost all cases, has been an evaluation of the
level of risk that these drug residues actually show. Another
recent article has shown how these residues can be
compared to true health-based limits to determine the
“margin of safety” the removal of these residues provides
which also offers an indication of the level of risk. This
“margin of safety”, or level of risk, can be quantified using
the statistical technique of Process Capability (20).
As with the Toxicity Scale derived from the PDE, it would be
convenient and informative to have a similar scale for Process
Capability. Another Scale has been developed that is
converts the Process Capability value into a value on a
“Process Capability Scale” which ranges from 0 - 10, with 0
being the lowest exposure level and 10 being the highest
exposure level. This scale is also potentially useful as a scale
for Occurrence or Likelihood FMEAs (20). While these scales
have several potential applications, at the most basic level
they can be used to clearly visualize the level of risk, and for
the purposes of this article, why HPAPI does not necessarily
translate to “high risk”.
Table 2 below shows two drugs that have very different
Toxicity Scores and Process Capability Scores and would
clearly present different levels of risk - even though in a
typical FMEA their Risk Priority Numbers (RPNs) would be the
same (10). Drug 1 has a Toxicity Score of 10 (very high
hazard) and would most likely be classified as an HPAPI. But
it has a corresponding Process Capability Score of 1.0
(excellent cleaning) which would constitute a low risk
situation. On the other hand, Drug 2 has a Toxicity Score of
1 (very low hazard) so it would not be classified as an
HPAPI. But Drug 2 has a corresponding Process Capability
Score of 10.0 (very poor cleaning), which indicates that
patients would be exposed to residues exceeding the PDE
and therefore represents a potentially high-risk situation.
• Criteria 1: Genotoxic (specically mutagenic)
compounds that are known to be, or highly likely to be,
carcinogenic to humans
• Criteria 2: Compounds that can produce reproductive
and/or developmental effects at low dosages in animal
studies of ≤1 mg/kg/day
• Criteria 3: Compounds that can produce serious target
organ toxicity or other signicant adverse effects at low
doses in animal studies of ≤1 mg/kg/day
• Criteria 4: Compounds with a high pharmacological
potency i.e. recommended daily dose of <1 mg
• Criteria 5: Compounds with a high sensitizing potential
But it’s important to understand that the PDE is not the sole
parameter that determines the risk of cross-contamination. As
noted in ICH Q9, the level of risk is a function of a hazard and
the level of exposure to that hazard. While the degree of
hazard can be measured or visualized using the Toxicity Scale
mentioned above, this only shows the relative toxicity of drugs
to one another and not the level of risk they pose. It is the
level of exposure to those hazards that reveals the level of risk
to a patient or worker from cross contamination. For a patient,
However, the calculation of the PDE is only the first step in
assessing the risks for cross-contamination in multipurpose
manufacturing facility. The use of the PDE as a tool for
determining risks in cross-contamination has been published
recently (17). In this article, the PDE is converted into a value
on a “Toxicity Scale” which ranges from 1 - 10, with 1 being
the least hazardous and 10 being the most hazardous which
makes it potentially useful as a scale
for Severity in FMEAs (Failure Modes
and Effects Analyses). For example,
using this scale a PDE of 10 µg/day
results in a toxicity score of 5, which
lies in the middle of the scale (Note:
the Scale is a logarithmic scale).
In general there are several groups of
substances that can be considered
HPAPI based on the previously
mentioned criteria, such as low
therapeutic dose (eg., the peptide
hormone oxytocin, guanylate
cyclase stimulant for irritable bowel
syndrome, dopamine agonists for
Parkinson’s disease), drugs that have
adverse effects in low doses (eg.
certain genotoxic antineoplastics,
toxins used in antibody-drug
conjugates (ADCs)) or both of these
criteria (eg. sex steroids and sex
hormone modulators). However there
are some substances that might
have relatively high therapeutic
doses, which are not dosed daily
(eg. bisphosphonates) that could be
considered HPAPI. For the purposes
of PDE and OEL calculations these
doses must be extrapolated to daily
doses, showing that the therapeutic
dose alone is not necessarily a good
indicator of HPAPI (4). In total, there
are only about 14% of substances
in the above study that would fall
into the highly hazardous substances
group based on their PDE. Out of
these 14%, there are four drugs which
would have not been identied as
“highly hazardous” based on the
EMA’s denitions (3), which is also 14%
of substances with a PDE < 10 µg/
day. (see table 1 for details):
Figure 1. Percentage of substances’ PDEs, grouped by order of
magnitude in a study of over 200 drug substances in a typical
pharma portfolio.
Table 1. A table of drugs with PDE <10 µg/day with their therapeutic indication and MoA and their
concordance with EMA’s definition of highly hazardous (3).
Table 2. Risk based on Toxicity Scores and Process Capability Scores.
53
52
Chimica Oggi - Chemistry Today - vol. 35(6) November/December 2017 Chimica Oggi - Chemistry Today - vol. 35(6) November/December 2017
guideline/2014/11/WC500177735.pdf (Last checked on Mar 2nd 2017)
15. Bercu J., Morinello E., et al., Regul Toxicol Pharmacol. 79 Suppl 1:
S48-56 (2016).
16. Sussman R., Naumann B., et al., Regul Toxicol Pharmacol. 79 Suppl
1: S57-66 (2016)
17. Walsh A, Lovsin Barle E, Crevoisier M, Dolan D, Flueckiger A, Ovais
M, Shirokizawa O, Waldron K, An ADE-Derived Scale For Assessing
Product Cross-Contamination Risk In Shared Facilities, May 22, 2017
18. Crevoisier M., Lovsin Barle E., et al., Pharm. Technol., 40(1): 52-56
(2016).
19. Walsh A. Pharmaceutical Engineering, Vol. 31 (2011)
20. Walsh A., Lovsin Barle E., et al., A Process Capability-Derived Scale
For Assessing Product Cross-Contamination Risk In Shared Facilities,
August 2017 (2017) https://www.pharmaceuticalonline.com/doc/
an-ade-derived-scale-for-assessing-product-cross-contamination-
risk-in-shared-facilities-0001 (Last checked on Aug 29th 2017).
Ester Lovsin Barle, PhD, MScTox,
ERT is Corporate Toxicologist
at Lonza since September
2017. Previously she has been
the Head of Health Hazard
Assessment in Novartis Global
HSE & BCM
About the author
