Evidence generated from the randomized study design may be largely ignored by an individual or a community of physicians if the results or strategies being examined do not align with existing beliefs or bias, local practice and referral patterns, and other potentially highly relevant factors unrelated to the trial outcome.
Maurie Markman, MD
Although randomized clinical trials provide what is generally considered gold-standard evidence and play a well-recognized critical role in the development of cancer care delivery models, clinicians do not always implement changes in practice based on these findings. The evidence generated from the randomized study design may be largely ignored by an individual or a community of physicians if the results or strategies being examined do not align with existing beliefs or bias, local practice and referral patterns, and other potentially highly relevant factors (eg, payment for services, resources available to undertake a new therapy) unrelated to the trial outcome.
Examples of this exist in gynecologic oncology, although there is nothing uniquely different about the way data from randomized clinical trials are absorbed in this field compared with other areas of cancer medicine. The published literature in gynecologic oncology simply happens to be particularly robust, which permits the examination that follows.
The question is, how has the gynecologic oncology physician community as an entity dealt with evidence-based randomized trial data that challenge existing therapeutic paradigms and beliefs deeply ingrained in the literature and residency/fellowship training programs?
The answer partly reflects the several-decades-long experience gynecologic oncologists have with the use of cisplatin- based primary intraperitoneal chemotherapy in smallvolume residual advanced ovarian cancer. Three well-conducted phase III randomized trials revealed an overall survival advantage associated with this route of drug delivery compared with intravenous treatment, leading the National Cancer Institute to issue a clinical alert in 2006 to inform the oncology community, patients, and the public of the outcome of these gold-standard, evidence-based studies.1 Yet peer-reviewed, published data revealed that just a minority of potentially eligible patients were treated by the regional approach versus the less effective systemic route of platinum delivery.1 Even considering the potential for added time, effort, and toxicity associated with intraperitoneal therapy, one wonders why more gynecologic oncologists did not employ the approach that the evidence strongly suggests improves survival. A logical extension of that question: Why continue to do so-called evidence-based studies if superior survival outcomes will be ignored by the community of physicians who conducted the research and are principally responsible for the care of patients who might benefit from these findings?
Finally, we turn to an even more controversial topic: the role of surgery in ovarian cancer or, to be specific, the value of a more aggressive versus a less aggressive surgical approach in disease management. For more than 40 years, the gynecologic oncology literature has extensively highlighted case series, retrospective institutional experiences, and sophisticated meta-analyses to support the claim that a critical component to optimizing survival in advanced ovarian cancer is successful surgery that removes as much macroscopic cancer as possible prior to initiating cytotoxic chemotherapy.2 This translates into no visible cancer following completion of the primary surgical procedure. Further, this hypothesis has been carried over into the recurrent disease setting, where it was routine practice to attempt maximal surgical cytoreduction prior to the initiation of second-line chemotherapy and, particularly, where the cancer was thought to remain sensitive to additional platinum-based therapy.
It is not my intention to negate the major role played by surgery, which can be safely performed by expert gynecologic surgeons with acceptable morbidity. However, what is being challenged here is the actual justification for aggressive surgery in ovarian cancer. To date, no published peer-reviewed randomized effort has directly compared primary surgery of advanced ovarian cancer with currently used primary chemotherapy regimens (without surgery). The fundamental issue is the unknown contribution—positive or negative—of an aggressive primary surgical approach in advanced disease versus the possibility that the intervention’s success is largely (although not exclusively) the result of a more or less favorable cancer biology in an individual patient that hypothetically permits or does not permit successful cytoreduction (eg, the presence or absence of diffuse carcinomatosis, extensive nodal involvement).
To date, several phase III randomized trials have examined a closely related question: What is the relative value of primary surgical cytoreduction versus chemotherapy followed by neoadjuvant surgery?3 The results show no difference in overall survival outcome but do indicate a more favorable postoperative morbidity profile for the neoadjuvant approach. Further, recently reported phase III trial data show no benefit for surgical cytoreduction prior to the delivery of platinum-based chemotherapy in recurrent ovarian cancer.4 Neither do the data demonstrate a benefit from lymph node dissection (versus no dissection) following successful cytoreductive surgery in advanced ovarian cancer.5
Finally, that leads to these questions: How will the results of these evidence- based, randomized phase III trials influence the practice of ovarian cancer management in the United States? Will established beliefs and training trump evidence-based phase III clinical trials, as observed with regional cisplatin delivery, or will the physician community follow the evidence?