Realizing the Full Value of Clinician Engagement in Pharmaceutical Development and Manufacturing

Introduction

Adoption of ICH Q8-11 has provided a more structured way to define product Critical Quality Attributes (CQAs), the design space, the manufacturing process and the control strategy for establishing acceptance criteria - all linked to the Quality Target Product Profile (QTPP). Utilizing these guidance documents, regulatory agencies have been increasingly effective in promoting a quality culture embedded in an integrated system of continuous improvement. The ICH guidelines have ushered in a paradigm shift from “black-letter” regulatory directives to a more agile risk-assessed control strategy standard. As a result - and in keeping patient safety at the forefront - clinician participation in product lifecycle management must come earlier and at a deeper level of involvement than historically undertaken. While great strides forward have been achieved, many organizations have failed to effectively engage healthcare providers in this process and thus do not realize the ‘full value’ that clinicians can provide.

Product Development and Process Control

Pharmaceutical product and medical device development move forward closely interlocked with safety and efficacy questions assessed with in silico, analytical, toxicological, human studies and ultimately broad patient exposure. The complexity of today’s therapeutic and treatment modalities necessitates an innovative approach to lifecycle management that should include clinician- scientists capable of bridging the science of pharmaceutical and medical device development with the delivery of bedside patient care. Ideally, this starts with the development of the QTPP, defined in ICH Q8R2 as follows: “A prospective summary of the quality characteristics of a drug product that ideally will be achieved to ensure the desired quality, taking into account safety and efficacy of the drug product” [emphasis added].¹ The resulting synopsis would generally include the intended use in clinical settings; dosage strength and route of administration; drug product quality attributes (such as sterility, stability, drug release); container closure system, etc. and should therefore include input from individuals with significant clinical experience.

The underlying objective is to expand the core organizational functions represented on the working groups responsible for the evaluation of manufacturing capability and the control strategy necessary to support the production of a consistently high-quality product. Absent cognitive diversity on these teams, erroneous assumptions arising from intuition or insufficient substantive expertise can undermine the opportunity for success.

Biopharmaceuticals only increase this challenge. These therapeutic modalities are characterized by their heterogeneity, translational, post-translational, chemical and enzymatic modifications producing potential immune-related and on-target effects where toxicological studies may lack high-predictive significance. Clinician input assists in appropriately resourcing the control strategy at points most relevant to product efficacy and patient risk. This feedback is vital to avoid over- resourcing a control strategy for an attribute with little or no impact to the patient or under-resourcing the controls where safe use is critical. By understanding the relevant properties of critical quality attributes (CQA), a better-informed safety monitoring plan can be developed to evaluate specific product attributes and associated outcomes potentially linked to those characteristics. Ultimately, developing and manufacturing a clinically successful product is about much more than demonstrating safety and efficacy. A successful, value-added therapeutic should incorporate a patient-centric design to assure tolerability and a risk management strategy that resources process controls to those points impacting patient-specific risks. Bringing clinicians into these activities provides an important perspective into the challenges that product or device-related attributes might play in either hindering or promoting safe use and compliance.

Subscribe to our e-Newsletters
Stay up to date with the latest news, articles, and events. Plus, get special offers
from American Pharmaceutical Review – all delivered right to your inbox! Sign up now!

Clinical Assessment of Post-Approval Changes

To illustrate an example of clinician involvement and the benefits they can bring, one need only look at post-approval change management. Building on global regulatory initiatives incorporating final guidances Q8 - 11, a draft version of ICH Q12 has been released providing proposed guidelines addressing the commercial phase of the product lifecycle. This important document offers a  framework  to  facilitate the management of post-approval Chemistry, Manufacturing and Control (CMC) changes within an organization’s pharmaceutical quality system. Actively engaging clinicians throughout the product lifecycle facilitates a more informed clinical risk assessment related to these post- approval changes and a foundation upon which to adopt appropriate surveillance programs. Waiting until the change package is ready for submission risks developing clinical work-product lacking the full view of prior optimization efforts and clinical outcome measures.

Since most products are approved before the manufacturing pro- cess is fully optimized, changes in the original manufacturing and quality management design are essential to dependably produce a high-quality product. Once the changes are considered for implementation, a comparability exercise ascertains whether the pre-and post- change products are comparable, to the extent that quality, safety and efficacy are not adversely affected. When comparability has been demonstrated through physicochemical/analytical and biological as- says, then non-clinical or clinical studies with the post-change product are not generally warranted. However, because of the complexity of biological therapeutics even minor changes might ultimately impact the product’s safety or efficacy.² One need only look at the modification of Boeing’s 737 flight control software to recognize a compound failure mode with unanticipated disastrous consequences. While it is highly unlikely that minor variability around a drug substance or product CQA would result in a similar catastrophic event, it is nonetheless important that product quality and safety surveillance programs are in place for on-going assessment of post-change event data.

The clinical post-change assessment begins with a review of the product’s risk management plan (RMP). The RMP serves “to document the risk management system considered necessary to identify, characterize and minimize a medicinal product’s important risks.”³ For biotherapeutic products, the European Medicines Agency’s (EMA) guidelines specifically include requirements to assess the potential impact of a manufacturing change; perform an associated risk analysis; and, update the RMP if warranted in conjunction with the variation or extension submitted with the change.¹ As an element of that assessment and based upon the level of comparability between the pre- and post-change materials a determination can be made concerning the need for additional non-clinical or clinical studies.

Discharging Uncertainty

One of the challenges posed in assessing post-change patient risk is the clinical consequence uncertainty inherent in the attributes potentially impacted.⁴ From a clinical perspective, the proposed changes might theoretically impact the product’s toxicity profile, safety, pharmacokinetic and pharmacodynamic profiles, or immunogenicity. In some instances where an attribute has been associated with a specific characteristic such as drug clearance,⁵ then a narrow-focus study can be conducted.⁶ In the majority of changes, however, the impact to the CQA is limited and falls within the established acceptance limits for that attribute. How then should these changes be assessed and what role might the physician play?

The discovery of drug induced illness is  a  function  of  the  rate  of the suspected  drug-induced  illness  against  the  background  rate  of the associated clinical finding. In other words, the rate of a drug- related illness must equal or exceed the background rate for that illness in order to be linked back to that drug.⁷ This point is critically important to effectively understand the capabilities of any clinical study methodology that might be utilized to evaluate the potential impact of manufacturing-related changes on the safety or efficacy of the product. In considering an approach to assess the possible clinical impact of the changes, it is important to understand the capabilities and limitations of various clinical study and surveillance designs - a point where clinical acumen provides a major contribution.

There are basically four mechanisms to evaluate clinical exposure of post-change batches:

1. Release material in the normal course of distribution and monitor via routine product complaint and pharmacovigilance processes. While spontaneous product complaints and adverse events could be captured, this course of action falls short of good vigilance practices (GVP) for major process changes.
2. Introduce material in the normal course of distribution and follow with an observational study. Observational studies would not be conducive to investigate this question. Incomplete data bases, multiple therapeutics and confounding illnesses, data collection and researcher biases and a lack of batch numbers to link with the drug of interest make this method unworkable.
3. Introduce material in an on-going or conventional clinical trial. While the potential to gather clinical laboratory data would be present, the number of study subjects necessary to identify a new related event would be significant. Absent a narrow therapeutic window and isolated atypical responses linked to a specific attribute with high biologic plausibility, adverse events that might be associated with the change would likely be subsumed by similar events associated to the drug product’s profile.
4. Introduce material in the normal course of distribution and implement a post-approval, batch-specific surveillance study utilizing disproportionality algorithms linked to analytic data.Spontaneously reported event data certainly have shortcomings but also provide a larger dataset that can be evaluated on a per-batch level for signal detection.

Incorporating Surveillance Data

In considering the above options, both clinical trial data and spontaneously reported events have the better opportunity to capture batch identity than observational studies and provide avenues for contemporaneous follow-up for greater event detail. However, considering the large trial population necessary to identify a causal link to a minor attribute variant, a traditional randomized clinical trial is not a plausible option - leaving contemporaneous post-change surveillance as the better alternative. In this scenario, batch specific analyses can be conducted to search for signals that might indicate a post-change issue.

A safety signal is information obtained on a new or known adverse event that may be caused by a medicine and requires further investigation. Further investigation, sometimes tabbed signal clarification, is then conducted to determine if the signal establishes whether or not there is a causal relationship between the drug product and the reported adverse event.^8,9 Post-change events can be aggregated into cohorts and then compared to pre-change batches.

The use of automated datamining algorithms has advanced  pharmacovigilance signal detection and can be used to examine product quality complaints as  well. Although  there  are  a  number of different methodologies, pharmacovigilance maps (PVmaps), proportional reporting ratio  (PRR),  empirical  Bayes  geometric  mean (EBGM), and the lower-bound of the EBGM’s 90% confidence interval (EB05) are among the most commonly cited.¹⁰ These methods utilize disproportionality analysis to identify disparate expression of event frequency comparing one cohort to another. A proportion is established using the number of events of interest (a) for a particular drug batch over all events for that drug batch (a+b), where b is the number of reports not having the event of interest (a/a+b). A similar proportion is found for the comparator set of batches, A/A+B. The ratio of the two proportions is calculated as follows:

(a/a+b)/(A/A+B) = PRR

Utilizing these data outputs, the pharmacovigilance surveillance scientist and clinician can evaluate drug-event combinations expressing disproportionately and determine if the differences might be the result of variances in the pre-and post-products. Although these methods and databases have limitations and datamining is not a tool for establishing a causal relationship between products and AEs, it is useful to identify disproportionate reporting that is then further evaluated for factors such as temporality, dose response, biologic plausibility and strength of  association  to  determine  causality.⁸ Early engagement of the clinician during QTPP development and continued involvement through clinical trials and product and process optimization enables the adoption of focused targets supporting the decision to employ disproportionality-based surveillance to follow post-change product.

Finally, it is worth highlighting that these algorithms might be utilized on data throughout the lifecycle to support on-going manufacturing and development efforts and provide input to the desired control strategy. Advances in datamining algorithms and artificial intelligence¹¹ may lend to an earlier ability to discharge uncertainty from changes associated with product and process optimization.

Author Biography

Dr. John Ayres is a board-certified internist with extensive experience evaluating the human safety risk potentially associated with Critical Quality Attributes (CQAs), manufacturing and environmental excursions, linked to product complaints, or related to counterfeit medication issues including surveillance, risk assessment and management, and regulatory-compliance functions.

References

1. The International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH). Q8(R2) Pharmaceutical Development. Geneva, Switzerland, 2009.
2. P.II.B.1.2.1. Potential impact of a manufacturing change. Guideline on good pharmacovigilance practices (GVP) – P. II EMA/168402/2014
3. Guideline on good pharmacovigilance practices (GVP) – Module V (Rev 2) EMA/838713/2011 Rev 2
4. Rosenberg A.; Verthelyi D.; Cherney B. Managing Uncertainty: A Perspective on Risk Pertaining to Product Quality Attributes as They Bear on Immunogenicity of Therapeutic Proteins. J. Pharm. Sci., 2012; 101 (10):3560-3567.
5. Liu L. Antibody glycosylation and its impact on the pharmacokinetics and pharmacodynamics of monoclonal antibodies and Fc-fusion proteins. J Pharm Sci. 2015; 104(6):1866-1884.
6. Putnam WS, Prabhu S, Zheng Y, Subramanyam M and Wang Y-M C. Pharmacokinetic, pharmacodynamic and immunogenicity comparability assessment strategies for monoclonal antibodies. Trends in Biotechnology. 2010; 28 (10): 509-516.
7. Jick H. The Discovery of Drug Induced Illness. NEJM. 1977;296(9):481-5.
8. FDA Guidance for Industry Good Pharmacovigilance Practices and Pharmacoepidemiologic Assessment, March 2005 Clinical Medical J:\!GUIDANC\6359OCC.doc 03/22/05.
9. CIOMS VIII Practical Aspects of Signal Detection in Pharmacovigilance 2009.
10. Faich G. and Morris J. Adverse reaction signaling and disproportionality analysis: An update. Drug Info J. 2102;46(6):708-714.
11. Zhang S., Zhang C., Yang Q. Data preparation for datamining. Appl Art Intell. 2003;17:375- 381.