This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Schiller, J. H. Search for Related Content PubMed PubMed Citation Articles by Schiller, J. H.

Clinical Trial Design Issues in the Era of Targeted Therapies

University of Wisconsin. Madison, Wisconsin

	ABSTRACT

Top ABSTRACT DEFINING CLINICAL END POINTS DEFINING THE TARGET TRIAL DESIGN OPEN DISCUSSION REFERENCES

 
Although our lack of effective therapies mandates the development of new, molecularly targeted therapies, a number of issues in the design of clinical trials and their end points remain to be considered.

	DEFINING CLINICAL END POINTS

Top ABSTRACT DEFINING CLINICAL END POINTS DEFINING THE TARGET TRIAL DESIGN OPEN DISCUSSION REFERENCES

 
Designing ethical clinical trials has never been easy, but in this age of increased public and political scrutiny and of the development of "cytostatic" drugs, it has become even more problematic. The classical end point for drug approval has been survival, demonstrated equivalency or noninferiority, or improved time to progression (TTP) plus improvement in symptom benefit or quality of life (1) . There are problems with these, however. Even the end point of survival, which would appear to be most unambiguous, has been potentially confounded by the advent of more effective second- and even third-line therapies. TTP has the advantage that it is not confounded by second-line therapy or crossover designs. However, it is difficult to measure and complicated by the fact that it tends to occur between scheduled observations and thus will vary depending on the frequency of scheduled evaluations. Response rates have the advantage that any evidence of tumor shrinkage is almost certain to be due to the drug in question; however, like TTP, response rate is also dependent on the frequency of patient assessment and, furthermore, does not necessarily correlate with survival. The clinical benefit of both TTP and response rate can be questioned if the patient is asymptomatic to begin with or if treatment neither prolongs life nor reduces symptoms (2 , 3) . Biomarkers as end points need to be validated before being used as a surrogate for activity or survival; as discussed below, this is often difficult to do.

	DEFINING THE TARGET

Top ABSTRACT DEFINING CLINICAL END POINTS DEFINING THE TARGET TRIAL DESIGN OPEN DISCUSSION REFERENCES

 
Another major issue in trial design, particularly given the "targeted therapies," which, by definition, inhibit a specific "target," concerns patient eligibility and whether enrollment should be restricted to patients whose tumor expresses the target of interest. Although again this would appear to be relatively straightforward, numerous problems exist. These center around identifying the target in the tissue of interest (surprisingly easier said than done!) and the frequency with which the target is expressed in the patient population of interest. In the former case, even the most straightforward of targeted inhibitors sometimes surprises us. For example, with the epidermal growth factor receptor (EGFR) inhibitors, effectiveness does not correlate with overexpression of the "target," the EGFR (4, 5, 6) . In addition, obtaining tissue can be difficult, particularly for lung cancer, and surrogate tissues, such as skin or buccal mucosa, have not been validated. In the latter case, if the marker of interest is expressed in only a minority of patients, sample size will need to be prohibitively large to determine a benefit. This problem occurs in attempting to correlate antitumor activity of trastuzumab, the monoclonal antibody against the Her2/neu receptor, with expression of the receptor. In both lung cancer and breast cancer, response rates and survival increase with overexpression of the receptor; however, unlike breast cancer, the overexpression of Her2/neu in lung cancer is low, making randomized Phase III trials nearly impossible (7) .

	TRIAL DESIGN

Top ABSTRACT DEFINING CLINICAL END POINTS DEFINING THE TARGET TRIAL DESIGN OPEN DISCUSSION REFERENCES

 
Until many of the biological issues are worked out, further consideration should be given to developing innovative trial designs that will allow us to identify which drugs should be the focus of further investigation. Statisticians have developed models that should allow for predicting which Phase II drugs are likely to yield positive Phase III results (8) , identifying early stopping rules (9 , 10) , expanding a randomized Phase II trial to Phase III (11 , 12) , and designing two-stage Phase II studies (13) . Another novel trial design is the "randomized discontinuation" design, in which all patients receive a cytostatic drug for one to two cycles. Those patients who have stable disease at the end of this period of time are randomized to placebo or to continue the drug. The purpose of this trial design is to try to control for patients with disease that is growing so rapidly that the patient never has time to take enough therapeutic doses of the drug before progressing. The advantage of such a design is that it enriches the patient population for those with slowly growing disease, e.g., those most likely to derive a benefit.

In the meantime, what do we do? Continue to depend on preclinical models to identify which drugs/patients/sequences to test? Again, these are notoriously unreliable. Despite preclinical data suggesting that the efficacy of cytotoxic agents against human tumor xenografts is markedly enhanced by coadministration of the EGFR tyrosine kinase inhibitor gefitinib (14) , randomized clinical trials involving chemotherapy have failed to show any benefit to the combination over chemotherapy alone (15 , 16) . Preclinical models would have suggested that the EGFR inhibitors were more likely to work in squamous cell carcinomas, which tend to express higher levels of EGFR; however, clinical studies have shown that patients with bronchioloalveolar carcinoma are more likely to respond (17 , 18) . Until we have better preclinical and clinical trial designs that would allow us to predict clinical benefit to new molecularly targeted agents, perhaps we should continue to do what we have always done—build on our clinical observations. For example, clinical observations that should be pursued include those suggesting that response to EGFR inhibitors is improved among nonsmokers, patients with bronchioalveolar carcinoma, women, and those who develop skin rash or improved symptomatology while on the agent (18, 19, 20, 21) . After all, ultimately the best "experimental animal" for human cancer is the human patient.

	OPEN DISCUSSION

Top ABSTRACT DEFINING CLINICAL END POINTS DEFINING THE TARGET TRIAL DESIGN OPEN DISCUSSION REFERENCES

 
Dr. Joan H. Schiller: In the IDEAL trials, about a third of the patients had symptom improvement. Although far fewer actually had objective responses, all responding patients did have symptom response. So, one question that one could ask is whether patients with symptom improvement are more likely to respond or to have improved TTP than those without symptom improvement. In another study that we are in the process of developing in the Eastern Cooperative Oncology Group, all the patients at the start will get single-agent gefitinib and be monitored for symptom improvement. Those that do show symptom response will get the combination carboplatin/paclitaxel/gefitinib, whereas nonresponders will go on to carboplatin/paclitaxel. So the idea is to try to identify whether symptom improvement really can predict for improved outcome.

Dr. Ramaswamy Govindan: That design does not answer the question. One approach would be to have those with symptom improvement go on to chemotherapy with or without gefitinib.

Dr. Schiller: It does not answer the question directly, but we are hypothesizing that the response rate to carboplatin/paclitaxel here is going to be the typical 20% and that with the combination in responders, we should see higher than the 10% response rate that we have seen typically in the past.

Dr. Thomas Lynch: Unless it turns out that the same patients are responding to either therapy, and then you may see a lower response rate. That is one of the explanations for why INTACT1 and 2 were negative: you are not really adding anything with the combination because the patients who respond to gefitinib are the ones who are going to respond to chemotherapy.

Dr. Roy Herbst: What is your window? How long do you look for symptom response?

Dr. Schiller: It is not long, 3 weeks, I think. Another popular way of selecting a population group that is likely to respond is by enriching for patients who have slowly progressive disease, so they have time to respond to therapy. I think we might need to start getting more statisticians at meetings like this to discuss trial design issues; in particular, novel trial designs that would allow us to identify more precisely from the Phase I or II studies which agents are most likely to succeed in Phase III studies. We need statistical modeling to help here.

	FOOTNOTES

 
Presented at the First International Conference on Novel Agents in the Treatment of Lung Cancer, October 17–18, 2003, Cambridge, Massachusetts.

Requests for reprints: Joan H. Schiller, University of Wisconsin Medical School, 600 Highland Avenue, Room K4/548, Madison, WI 53792-0001. Phone: (608) 263-5343; Fax: (608) 263-8613; E-mail: jhschill{at}wisc.edu

	REFERENCES

Top ABSTRACT DEFINING CLINICAL END POINTS DEFINING THE TARGET TRIAL DESIGN OPEN DISCUSSION REFERENCES

 

1. Schilsky R End points in cancer clinical trials and the drug approval process. Clin Cancer Res, 8: 935-8, 2002.[Abstract/Free Full Text]
2. Scher HI, Heller G Picking the winners in a sea of plenty. Clin Cancer Res, 8: 400-4, 2002.[Abstract/Free Full Text]
3. Von Hoff DD There are no bad anticancer agents, only bad clinical trial designs: twenty-first Richard and Hinda Rosenthal Foundation award lecture. Clin Cancer Res, 4: 1079-86, 1998.[Abstract]
4. Saltz L, Rubin M, Hochster H, et al Cetuximab (IMC-C225) plus irinotecan (CPT-11) is active in CPT-11-refractory colorectal cancer (CRC) that expresses epidermal growth factor receptor (EGFR). Proc Am Soc Clin Oncol, 20: 3a 2001.
5. Baselga J Current status of anti-EGFR tyrosine kinase receptor therapies. Ann Oncol, 13(Suppl 5): 6 2002.
6. Arteaga CL, Baselga J Clinical trial design and end points for epidermal growth factor receptor-targeted therapies: implications for drug development and practice. Clin Cancer Res, 9: 1579-89, 2003.[Free Full Text]
7. Langer C, Adak S, Thor A, Vangel M, Johnson D Phase II Eastern Cooperative Oncology Group (ECOG) pilot study of paclitaxel (p), carboplatin (c), and trastuzumab (t) in her-2/neu (+) advanced non-small cell lung cancer (NSCLC): early analysis of E2598. Proc Am Soc Clin Oncol, 20: 315a 2001.
8. Freidlin B, Breathnach OS, Johnson BE A model to select regimens for Phase III trials for patients with advanced-stage non-small cell lung cancer. Clin Cancer Res, 9: 917-22, 2003.[Abstract/Free Full Text]
9. Wieand S, Schroeder G, O’Fallon JR Stopping when the experimental regimen does not appear to help. Stat Med, 13: 1453-8, 1994.[Medline]
10. Baum M, Houghton J, Abrams K Early stopping rules–clinical perspectives and ethical considerations. Stat Med, 13: 1459-69, 1994.[Medline]
11. Inoue LY, Thall PF, Berry DA Seamlessly expanding a randomized Phase II trial to Phase III. Biometrics, 58: 823-31, 2002.[CrossRef][Medline]
12. Steinberg SM, Venzon DJ Early selection in a randomized Phase II clinical trial. Stat Med, 21: 1711-26, 2002.[CrossRef][Medline]
13. Herndon J A design alternative for two-stage, Phase II, multi-center cancer clinical trials. Control Clin Trials, 19: 440-50, 1998.[CrossRef][Medline]
14. Sirotnak F, Zakowiski M, Miller V, Scher H, Kris M Efficacy of cytotoxic agents against human tumor xenografts is markedly enhanced by coadministration of ZD1839 (Iressa), an inhibitor of EGFR tyrosine kinase. Clin Cancer Res, 6: 4885-92, 2000.[Abstract/Free Full Text]
15. Johnson D, Herbst R, Giaconne G, et al ZD1839 ("Iressa") in combination with paclitaxel and carboplatin in chemotherapy-naive patients with advanced non-small-cell lung cancer: results from a Phase III clinical trial (INTACT2). Ann Oncol, 13(Suppl 5): 127-8, 2002.
16. Giaccone G, Johnson DH, Manegold C, et al A Phase III clinical trial of ZD1839 ("Iressa") in combination with gemcitabine and cisplatin in chemotherapy-naive patients with advanced non-small-cell lung cancer (INTACT 1). Ann Oncol, 13(Suppl 5): 2 2002.
17. Modjtahedi H, Eccles S, Sandle J, et al Differentiation or immune destruction: two pathways for therapy of squamous cell carcinomas with antibodies to the epidermal growth factor receptor. Cancer Res, 54: 1695-701, 1994.[Abstract/Free Full Text]
18. Kris M, Natale R, Herbst R A Phase II trial of ZD1839 ("Iressa") in advanced non-small cell lung cancer patients who had failed platinum- and docetaxel-based regimens. Proc Am Soc Clin Oncol, 21: 292a 2002.
19. Shah NT, Miller VA, Kris MG, et al Bronchioalveolar histology and smoking history predict response to gefitinib. Proc Am Soc Clin Oncol, 22: 628 2003.
20. Perez-Soler R, Chachoua A, Huberman M, et al A Phase II trial of the epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor, OSI-774, following platinum-based chemotherapy, in patients (pts) with advanced, EGFR-expressing, non-small cell lung cancer (NSCLC). Proc Am Soc Clin Oncol, 20: 310a 2001.
21. Fukuoka M, Yano S, Giaccone G, et al Multi-institutional randomized Phase II trial of gefitinib for previously treated patients with advanced non-small-cell lung cancer. J Clin Oncol, 21: 2237-46, 2003.[Abstract/Free Full Text]

This article has been cited by other articles:

				W. Slichenmyer Clinical Trial Strategies to Facilitate the Transition to Registration-Directed Trials Am. Assoc. Cancer Res. Educ. Book, April 1, 2005; 2005(1): 297 - 302. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Schiller, J. H. Search for Related Content PubMed PubMed Citation Articles by Schiller, J. H.