|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2003.05.100 on June 13 2003 © 2003 American Society for Clinical Oncology
Functional Imaging in Phase I Studies: Decorations or Decision Making?Laboratory of Clinical Pharmacology, United States Food and Drug Administration, Rockville, MD UNTIL RECENTLY, phase I clinical studies were narrowly focused and empiric. As drug targeting has become more sophisticated, there has been a trend toward addition of correlative scientific studies to clinical trials. Maximizing the information obtained from phase I studies is desirable, but an assessment of the goals of such work is essential. Additional layers of laboratory and clinical tests are expensive, logistically difficult, and consume scarce technological resources. Thus, we should expect results that contribute to additional development, rather than those that serve as decorations. In our attempts to find early measures of drug activity and proof-of-concept for mechanism of action, many biomarkers have emerged, including noninvasive functional imaging probes. Magnetic resonance imaging (MRI) and positron emission tomography (PET) have become indispensable in certain situations for diagnosis and staging. In such settings, the key element for measuring the value of these imaging tools is whether there are changes in management of patient care or outcome. A similar standard could be used to assess the relevance of the emerging use of these tools in early drug development, but how have these techniques changed our decision making? CASE STUDY: COMBRETASTATIN A4 PHOSPHATE (CA4P) Three articles published in this issue of Journal of Clinical Oncology report on the same phase I trial of weekly doses of CA4P.1–3 CA4P is one of more than a dozen antivascular compounds in development, including both small molecules and antibodies.4
The maximum-tolerated dose was established on the basis of ataxia, syncope, and motor neuropathy, with one fatality caused by ischemia.1 Tumor pain and cardiovascular effects were also observed. The phase I trial design maximized the number of patients who received a bioactive dose by using small cohort sizes, dose doubling, and intrapatient dose escalation. Because of the efficiency of these approaches, 27 of 34 patients received
The PET data for oxygen-15–labeled water showed a noticeable decrease in tumor perfusion 30 minutes after dosing of CA4P.2 However, the two-fold decrease (mean 49% at For the MRI results, there is great emphasis on technical details; the authors caution that "the extent of reduction is difficult to relate to preclinical models and to data from other sites." Overall, we are left with a strong qualitative sense that tumor perfusion was decreased in some patients. Taken together with MRI data from other CA4P trials,4,6,7 it appears that the MRI and PET results are harmonious with the view that some patients had decreased perfusion in their tumors, but the changes were not as dramatic as those observed in rodent tumors. The purpose of phase I trial designs is to guide more rigorous activity assessment in phase II. Traditionally, the main design parameters have been dose and schedule. A series of questions can help us to assess whether imaging added value to the transition from phase I to phase II. DID IMAGING HELP TO ASSESS WHETHER THE MECHANISTIC GOALS WERE ACHIEVED? Yes. Without the results of the imaging studies, it would not have been possible to determine if tumor perfusion was reduced. DID IMAGING ASSIST DOSE SELECTION FOR PHASE II? Marginally. There was little difference between the dose selected by conventional toxicity criteria and the dose recommended from imaging of tumor perfusion. These investigators conclude that 68 mg/m2 would be the recommended phase II dose on the basis of toxicity criteria, and 52 to 68 mg/m2 would be the recommended phase II dose on the basis of imaging results. The inability of imaging to identify a lower dose that reduces tumor perfusion is not a failure of the technology, but a failure of animal models to predict quantitatively the impact of CA4P on human tumors. The authors have cited rodent data that raised expectations for a "wide therapeutic index" for CA4P and activity at "nontoxic" doses. Such terms are not supported by the data from this work or any of the other human studies. In humans, CA4P has less impact on tumor perfusion and more toxicity than in rodents. Readers are familiar with optimistic projections from animal models that are not attained in humans. Although the rhetoric about a wide therapeutic index for CA4P needs to be discarded, this is not necessarily a cause to discard immediately either this drug or the whole concept of targeting tumor vasculature. Our everyday experience is that a drug with a narrow therapeutic index can be used successfully. DID IMAGING PROVIDE ASSISTANCE FOR SCHEDULE SELECTION FOR PHASE II? Yes. For a drug with a novel mechanism of action, it can be reasonable to test three schedules in phase I, but ordinarily only one schedule will move ahead to phase II. How do we use phase I results to select a schedule for additional evaluation? The once-a-week schedule for CA4P used in this trial is intermediate in dose frequency between previous reports for once every 3 weeks6 and five times every 3 weeks.7 The tolerable dose range for a single infusion and the clinical toxicity patterns appear to be similar for all three schedules. Empirically, if the toxicity profile provides the opportunity to give a dose of 52 mg/m2 once every 3 weeks, three times in 3 weeks (weekly), or five times in 3 weeks, we would assume that the higher dose-intensity would be preferred in early studies to maximize potential benefit.
These authors did not make an explicit recommendation for a delivery schedule. In the discussion of PET results,2 there is one sentence that raises the possibility that "there might be a benefit. . . from exploring a dose schedule involving greater than once-a-week dosing." Although the authors were focusing on normal tissues, the data from the 24-hour observations showed that the mean tumor perfusion at doses Why not consider even more frequent dosing for CA4P—perhaps daily? Radiation therapy is almost always given for 5 to 7 days over 4 to 6 weeks. Although daily intravenous infusions are demanding, the endostatin phase I studies8–10 were conducted on a daily schedule for median periods of 2 or 3 months, with a reported range for individual patients up to 426 daily doses.9 Of course, there is always a balance between the practical aspects of maintaining long-term intravenous dosing compared with the penalty for improperly rejecting a potentially active agent in favor of an easier schedule. As with endostatin, daily infusions do not ensure substantial activity, but do provide a method for minimizing false-negative results. If the proof-of-concept via the intravenous route shows activity, the development of orally available analogs might readily be stimulated. CAN IMAGING SELECT SUBPOPULATIONS ENRICHED FOR RESPONSE? Possibly. Conventional phase I or II trial designs provide no guidance regarding patient selection. When overall activity is high, the need for selecting subpopulations using biomarkers is not highly compelling. No efficacy was found in this trial, despite the use of phase II dose level in 27 patients, so CA4P is not broadly active. However, responses were reported in other trials5,7 and the therapeutic approach is novel, so perhaps it would be worthwhile to search for a sensitive subpopulation of patients. If a patient’s tumor does not have a substantial reduction in perfusion as assessed by imaging, why should further doses be administered? In other words, the phase II population could be enriched if the only patients who are treated are those whose tumor vasculature responds. An early decision can be reached to switch to another treatment strategy, rather than waiting for overt failure and more resistant disease. In breast cancer, the use of human epidermal growth factor receptor 2 and hormone receptors to guide treatment are relevant precedents for testing enriched populations. The selection criteria for an enriched subpopulation are not obvious from the current results. Both MRI and PET revealed substantial interpatient variability in baseline tumor perfusion; thus, either a threshold for absolute tumor perfusion rate or relative decreases might be used, and both should be observed further in subsequent studies. ARE IMAGING TOOLS WIDELY AVAILABLE FOR A BROAD RANGE OF PROBES? No. However, both MRI and PET tools for measuring blood flow are available at major centers. MRI is obviously well established as a superb anatomic imaging tool, but is now moving toward the domain of functional imaging. The extensive discussion by the authors of the technical limitations of dynamic contrast enhanced MRI could have the unintended effect of driving readers to conclude that the technology is impractical. In the years since this study was initiated, dynamic contrast enhanced MRI has become more widely available and more readily interpretable, but not every clinical MRI facility can conduct blood perfusion studies. Similarly, although the use of oxygen-15–labeled water is a well-established procedure for quantitative monitoring of human blood flow, not every institution can make labeled water and interpret its images. Despite the hard work toward improving success with new drugs, it is safe to assume that there will continue to be failures, and we must be able to learn from these disappointments. We need tools to determine more specifically why failure occurred. Did the drug fail to influence its target? Did the tumor grow despite adequate modulation of the target? Was the animal model poorly predictive for human disease? Was the clinical trial design flawed? Is there any strategy that can reverse the apparent failure? The CA4P reports in this issue provide us with relevant information for many of these questions, but not all of the answers. Thus, for these studies, the report card for the added value from imaging is excellent. The prognosis for the drug itself is uncertain. There remains much work for both CA4P and the general class of antiangiogenic drugs. Attack of the tumor’s blood supply, either the existing blood vessels or the growth of new blood vessels, is a strategy that has substantial appeal, even though the first round of clinical studies have not met expectations. The ability to generalize about the prospects for other novel therapeutics under development to benefit from imaging depends heavily on specific efforts to develop appropriate probes. It is not likely that, as in the case of CA4P, probes will already be available at the time phase I studies begin. Perhaps the reason that we do not have more examples of success (or, at least, the ability to dissect to the root cause of failure) is not an inherent lack of appreciation for the technology, but rather the lack of appropriate tools during early clinical studies. The value of having diagnostic imaging tools ready when the therapies enter the clinic is demonstrated by the experience with CA4P. Conversely, production of the tools too late in the development process creates the worst possible decoration: one that missed its chance to make a difference in decision making. REFERENCES
1. Rustin GSJ, Galbraith SM, Anderson H, et al: Phase I clinical trial of weekly combretastatin A4 phosphate (CA4P): Clinical and pharmacokinetic results. J Clin Oncol 21:2815–2822, 2003
2. Anderson HJ, Yap JT, Miller MP, et al: Assessment of pharmacodynamic vascular response in a phase I trial of combretastatin A4 phosphate. J Clin Oncol 21:2823–2830, 2003
3. Galbraith SM, Maxwell RJ, Lodge MA, et al: Combretastatin A4 phosphate has tumor antivascular activity in rat and man as demonstrated by dynamic magnetic resonance imaging. J Clin Oncol 21:2831–2842, 2003
4. Thorpe PE, Chaplin DJ, Blakey DC: The first international conference on vascular targeting: Meeting overview. Cancer Res 63:1144–1147, 2003
5. Tozer GM, Prise VE, Wilson J, et al: Combretastatin A4 phosphate as a tumor vascular targeting agent: Early effects in tumors and normal tissues. Cancer Res 59:1626–1634, 1999
6. Dowlati A, Robertson K, Cooney M, et al: A phase I pharmacokinetic and translational study of the novel vascular targeting agent combretastatin A-4 phosphate on a single-dose intravenous schedule in patients with advanced cancer. Cancer Res 62:3408–3416, 2002 7. Stephenson JP, Gallagher M, Sun W, et al: Phase 1/pharmacokinetic trial of the endothelial toxin combretastatin A4P (CA4P) administered as an iv bolus on a daily x 5 schedule every 21 days. Proc Am Assoc Cancer Res 41:544, 2000 (abstr 3469)
8. Eder JP Jr, Supko JG, Clark JW, et al: Phase I clinical trial of recombinant human endostatin administered as a short intravenous infusion repeated daily. J Clin Oncol 20:3772–3784, 2002
9. Herbst RS, Hess KR, Tran HT, et al: Phase I study of recombinant human endostatin in patients with advanced solid tumors. J Clin Oncol 20:3792–3803, 2002
10. Thomas JP, Arzoomanian RZ, Alberti D, et al: Phase I pharmacokinetic and pharmacodynamic study of recombinant human endostatin in patients with advanced solid tumors. J Clin Oncol 21:223–231, 2003
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
52 mg/m2. Despite an adequate, bioactive dose in 80% of patients, there were no complete or partial responses. 
