A collection of brief reviews of selected issues follows that can pose as controversies or extra challenges in clinical development (Fig. 8.35). Some topics, such as investigator-initiated INDs and compassionate use, have been perennially discussed within the pharmaceutical industry. Others, such as quality of life end points and pharmacogenomics, represent disciplines and concepts that are relatively new in their application to drug development. The purpose of this section is to introduce these issues and some of the key concerns surrounding them to the reader.
In an investigator-initiated IND, an investigator, typically an academician, who normally also serves as principal investigator on the studies under the IND/CTA, submits the IND under his or her own name (Fig. 8.36). Thus, the investigator is literally the sponsor of the IND. The role of a pharmaceutical company in this process generally may involve providing permission to reference existing drug master files and other regulatory submissions, including other INDs, and providing supplies of study drug. It is generally held that the advantages of allowing investigators to study new drugs under an investigator-initiated
IND consist of conservation of resources and a lower level of regulatory scrutiny. IND creation by a sponsor company is a significant undertaking in terms of time and expense due to the high standards that exist for the quantity and quality of data in industry-sponsored INDs. Investigator INDs typically are more rudimentary in nature and resources that are expended by the investigator rather than the company. Similarly, it is generally held that investigator-initiated INDs are subjected to a lesser degree of regulatory scrutiny and that potential issues such as clinical holds pending submission of additional data or changes in protocol design are less likely to occur.
On the negative side, studies carried out under the investigator-initiated IND are done with a total loss of sponsor control of key study elements, data analysis and interpretation, and handling of safety data. Whereas there are certainly instances where mutual good faith has allowed projects to proceed under these conditions, the risk of essentially giving a company asset to an investigator for some period of time is very real and must be clearly understood and balanced against the perceived advantages.
It is a generally held view among laymen that "patients with serious diseases are always looking for clinical trials to participate in" (Fig. 8.36). Seemingly reasonable corollaries to this would be that most of these patients do, in fact, participate in clinical trials and that, assuming a reasonably promising drug under study, recruitment of patients should not be difficult. The reality is quite different. For example, one study estimated that less than 2% of white adult cancer patients in the United States participated in a National Cancer Institute clinical trial in the common tumors from 2000 to 2002. Even lower participation was found among racial and ethnic minorities . On a more parochial level, subject recruitment issues and challenges are a day-to-day reality for anyone involved in conducting clinical trials. Enrollment of patients into clinical trials of new drugs is affected by a variety of factors. Scientifically meaningful and medically safe evaluation of new drugs usually requires a highly defined study population, particularly in early patient trials. Thus, complex inclusion and exclusion criteria in r Investigator Initiated vs. Company INDs:
c Controls in the design e Quality c Resources c Regulatory scrutiny r Access to new drugs / Access to willing patients -Myth vs Reality: c. Reasons c Outside help to find patients c Enrollment aids ( advertisements, networks, databases, internet )
study protocols markedly reduce the population available for enrollment.
Community physicians may be less eager to refer patients to clinical trials because either loss of patients or loss of control is a concern. Notably, in the one area of medicine, pedi-atric oncology, where participation in clinical trials is the rule rather than an exception, most clinical care is initiated and overseen through regional medical centers rather than by community-based physicians. Geographic access to centers is limited in many parts of the country. Study requirements (frequency of visits, testing, etc.) are frequently burdensome and, while appropriate in the context of the study, are far in excess of standard care. Concerns about random treatment assignment, especially to placebo, may exist.
Aside from traditional recruitment aids such as newspaper and radio advertising, other actions may aid recruitment. Working with sites within a health maintenance organization (HMO) or other organized network may maximize referrals. Making sure that there is awareness of a clinical development program by relevant patient advocacy groups and ensuring that studies are reflected in relevant databases, most notably the "Clinical Trials Data Bank," may also be helpful. Note that a sponsor of a clinical trial of a drug intended for treatment of a serious or life-threatening disease is required by federal regulations to submit information to this database, but that any sponsor may do so .
Patient expectations are heterogeneous and depend on a multiplicity of factors (Fig. 8.37). For healthy volunteer subjects ("professional subjects") participating in clinical pharmacology studies, the primary if not sole motivation is financial compensation, at times joined by some elements of altruism, curiosity, and desire to be with one's friends. On the other hand, for patients with life-threatening diseases, participation in a study of a new and promising drug may represent a last chance at receiving an effective medical therapy. Between these two extremes are an almost unlimited number of permutations of expected medical benefit, once again curiosity and altruism, economic benefit (free exams, free medication, stipends), and social benefits (meet people, get attention).
There are strongly held notions in academia, not without basis in fact, that suppressive publication practices are widespread in industry (Fig. 8.37). This has led to a requirement by several leading international medical journals that trials r Patient expectations: c Heterogeneous c Spectrum from altruism to money r Publications: c Industry vs. Academic perceptions c Some practical concerns c Publication policy for sponsors will need to be registered in a public clinical trial registry in order to be considered for publication . There are equally strongly held notions in industry, one again with some merit, that less than rigorous publication practices (e.g., based on unreliable data or questionable analyses) are widespread in academia.
Attempts to publish based on incomplete or unreliable data (e.g., premature single-center publication from multicenter studies, publication prior to data lock, selective publication of data) or unsound analyses (post hoc statistics, invalid assumptions, etc.) clearly do not serve the interest of a sponsor or of any author. Efforts at preventing these occurrences are, if carefully presented, not likely to be construed as suppressive and more likely to be seen as facilitating timely dissemination of reliable data and analyses to the peer-review community. Review periods allowing for thorough consideration of intellectual property protection, if not unreasonable, are similarly accepted as standard practice and not ill perceived.
Sponsors should speak very clearly to their unequivocal support of the dissemination of scientific information, but build intellectual property protection (required review period by sponsor) and injunctions against inappropriate data use into study contract and protocol. The language should be carefully considered but once agreed should be adhered to without exception. In agreements with investigators especially at academic institutions, a company will require an opportunity to review and comment on a possible publication.
Randomized, double-blinded, placebo-controlled clinical trials, while generally regarded as the gold standard for scientific proof of the efficacy and safety of most new drugs, are limited in their application by ethical, scientific, and practical considerations (Fig. 8.38). In certain instances, while generally accepted effective standard treatment does exist, withholding it and using placebo may be acceptable, as in, for example, antihistamines in allergic rhinitis. In other clinical settings as, for example, in virtually all serious infections, the sequelae of withholding treatment would be medically unacceptable, thus mandating the use of a positive control of currently available approved therapy.
Use of an approved product as a positive control then poses the question of whether an approved treatment really works? This is less of a problem when it is possible to show that a r Issues in use of placebo: o Nature of disease o Available treatment o Approval creep with active controls r Quality of Life Endpoints: o QOL as basis for approval vs.
QOL as "phamacoeconomic endpoint"
new treatment is more effective than standard care. In that case, even if the assumption is that standard care is no better than placebo, demonstration of superiority for the test treatment is a compelling demonstration of efficacy. In the instance where equivalence to the standard treatment is shown, however, demonstration of efficacy is critically dependent on the robustness of the original demonstration of efficacy of the approved comparator. Note the concept of approval creep, i.e., showing equivalence to an approved product of dubious efficacy, when your product is approved, the next one down the line uses it as a comparator to gain approval and so on down the line .
Placebos will probably be scientifically appropriate and a regulatory necessity for a long time to come in many disease states. Uncritical condemnation of placebo-controlled studies as unethical leads one to the unpalatable alternatives of either approving drugs on the basis of inadequately controlled studies or not approving new drugs, neither of which serve the interests of either ethics or society.
It is to be hoped that new medicinal treatments will, by virtue of their being effective, have a measurable impact on health-related QOL (Fig. 8.38). Historically, the basis for new drug approval has accepted this hope as a reality by implicitly assuming that when biological evidence of efficacy exists, there will be health-related benefits to the patient. QOL end points, either as add-ons to a traditional efficacy study or as free-standing studies, have become increasingly important in supporting comparative product claims and drug pricing and reimbursement. QOL is not, however, necessarily different from efficacy. QOL may serve as basis for approval. The best examples of this are in the area of oncology. Traditional "biological" oncology end points, such as tumor response and time to event variables (especially survival), have historically served as bases for new drug approval. Although FDA Oncology Division guidances over the past 20 years have invoked QOL end points as an acceptable bases for approval, until relatively recently, sponsors have not sought approval based on these end points. Recent examples of approvals based on QOL end points include mitox-antrone in prostate cancer and porfimer sodium in esophageal and non-small cell lung cancer .
Compassionate use, also called treatment IND, is provision of an unapproved drug to a patient outside of a formal clinical trial (Fig. 8.39). In the usual course of events, the patient should be enrolled in a clinical trial if eligible and if an appropriate study exists. In general, the drug should have been shown to have provided direct benefit to the patient on an earlier study, or there should be extremely compelling medical reasons for believing that it will benefit the patient during compassionate use. The patient's disease should either be serious and progressive or highly symptomatic, and approved therapies should have been tried and found to be either ineffective or not well tolerated. Compassionate use requires notification to and approval by the applicable regulatory authority.
For the right drug and patient, compassionate use can meet an important humanitarian need. However, valid instances are r Compassionate use: c Definition c Uncommon o Ethical considerations c Practical impact k Pharmacogenomics: c Metabolism c Safety c Efficacy
Fig. 8.39. Selected Issues & Controversies uncommon. If many valid compassionate use situations arise, the sponsor might want to consider whether a drug might be suitable for accelerated approval. Poorly managed compassionate use programs can quickly get out of control, require inordinate amounts of time to administer, endanger patients, create legal issues, generate large amounts of data of uncertain reliability, and cast doubts on conclusions drawn from reliable data. Using promises of ongoing compassionate use to attract patients into short clinical trials is associated with all of the problems above, as well as with potential ethical issues.
Variation in response to drugs among individuals has long been recognized. One of the mechanisms for such variation is genomic variation in the population (Fig. 8.39) At the most readily observed level, this variation results in "responders" and "nonresponders" in clinical trials and clinical practice and also in the occurrence of drug toxicity in some treated individuals but not in others (called pharmacogenomics). More recently, individual variations in drug metabolism, particularly with respect to cytochrome P450 (CYP450) mediated liver metabolism, began to be systematically explored in an effort to understand the bases for individual differences in drug efficacy and toxicity. Numerous well-studied examples (e.g., dextromethorphan, tri-cyclic antidepressants, beta-blockers, narcotic analgesics) exist of interindividual variation in the expression of a CYP450 drug-metabolizing enzyme (phenotypic expression) resulting in varying systemic drug exposure and consequent clinical sequelae .
Advances in molecular biology, which had their beginnings in the early 1970s supported by powerful computer technologies, have made the identification of specific genetic sequences in an extracted sample of human DNA possible and allowed this technique to be carried out efficiently and economically on large numbers of samples. Thus, phenotypic variation may now be related to variations in genotype. Pharmacogenomics attempts to relate variation in individual drug response to genotypic variation. To date, much of the practical application of this discipline to clinical drug development has been in the area of drug metabolism. Genotyping may be used to efficiently identify and either include or exclude subjects with a given phenotype for a specific CYP450 enzyme. This approach may enhance safety or create a more homogeneous population for efficacy analysis.
Alternatively, DNA may simply be collected from any one of a number of tissue samples and stored for possible retrospective retrieval and genotyping should clinical findings from the study suggest that genetically determined variation in drug metabolism may have been operative. Storage of DNA samples for possible subsequent genotyping and linking to efficacy and safety end points, rather than purely metabolic end points, is an area of active investigation at this time. In concept, many facets of drug response could be genetically determined, although, in distinction to drug-metabolizing enzyme expression, these response phenotypes would be more likely to be based on multiple genes.
Was this article helpful?