No Easy Answers for Discussing Outcomes With Patients

Maurie Markman, MD

In the discussion about outcomes, which will certainly occur when an individual receives a cancer diagnosis, most likely there will be nothing more important to the patient than the issue of survival.

As a result, it may come as a surprise to many patients, their families, and the public at large that answering this question of survival is difficult for a cancer treatment specialist. It is relevant to acknowledge that these inquiries may concern the anticipated survival of a given patient at a particular stage of disease or that they may be far more specific, such as a request for the survival statistics associated with the care provided by an individual oncologist, clinical practice, or hospital.

Too Many Unknowns

In another branch of medicine, such information might be much easier to provide. Orthopedic surgeons should be able to provide rather detailed data on their own experiences with relevant outcomes, such as the time required before a patient returns to work or the anticipated level of activity following the performance of a particular procedure. Of course, orthopedic prognostic data will be presented based on well-defined pre-intervention factors that include patient age, presence of specific comorbidities, and the extent of joint damage, etc.Why is it so much more difficult in oncology to provide similarly detailed relevant outcome data that would answer how long individual patients with cancer will survive or perhaps be cured of their malignancy? Oncologists have long recognized that 2 major issues, which are completely independent from the quality of care provided or the specific antineoplastic interventions employed, may substantially impact their ability to provide meaningfully accurate information regarding an individual patient’s anticipated survival.

The first factor is the underlying highly complex biology of an individual patient’s cancer, which often remains inadequately understood despite revolutionary changes in our knowledge of the molecular basis of cancer development, growth, progression, and resistance. Therefore, providing prognostic information to an individual patient with a newly diagnosed cancer, as important and relevant as the information may be, is fraught with danger of being terribly wrong. In discussions with patients and their families, oncologists reasonably attempt to deal with this difficult issue by providing patients and their families a wide range of possible outcomes. These physicians also fully appreciate that, to a major extent, survival can be known only by observing the natural history of an individual’s malignancy, which hopefully is favorably influenced by the treatment employed during the course of that illness.

An Example of Uncertainty

The second issue, perhaps even more difficult to clearly define and articulate, is the impact of a number of noncancerrelated clinical factors (eg, physiologic age, presence of comorbidities, severity of experienced toxicities, etc) on the selection and effectiveness of anticancer therapeutics, as well as the potentially substantial independent influence of these factors on an individual patient’s ultimate survival.A particularly striking and well-documented example of the relevance of these issues can be found in an experience in the management of advanced ovarian cancer related to the introduction more than 20 years ago of the novel antineoplastic agent paclitaxel.¹ Today the safety profile of this still widely employed cytotoxic drug is well understood, including the risk of serious-but-manageable acute hypersensitivity reactions, but this was not the case when the agent was first introduced into the clinical trials arena. Despite the remarkable promise of paclitaxel in the treatment of ovarian cancer during this early period in the drug’s clinical development, there was a major legitimate concern for highly unpredictable, sudden, severe, and even fatal treatment-related events, the cause of which was poorly understood.

As a result, during the initial phase III randomized trial examining the utility of a combination of cisplatin plus paclitaxel versus cisplatin plus cyclophosphamide in the primary chemotherapeutic management of advanced ovarian cancer, patients with essentially any known preexisting cardiac concern were excluded from study entry.¹ In addition, because of reports of severe, nonpreventable, and fatal toxicity, it was highly likely that oncologists very carefully screened out otherwise eligible patients whom they perceived as being at a greater risk of dying from a sudden treatment-related cardiac or hypersensitivity event versus death from advanced ovarian cancer. They recognized they could always employ standard-of-care chemotherapy not containing paclitaxel. Patients receiving their first paclitaxel treatment on this protocol required intensive inpatient cardiac monitoring during infusion of the drug.

Fortunately, it was quickly learned that with appropriate prophylactic measures, paclitaxel could be safely administered to the vast majority of patients without serious incident. Further, due to the anticipated delay in release of survival results for this initial phase III frontline advanced ovarian cancer study, a second trial was rapidly initiated,² with one study arm in this follow-up initiative containing the identical cisplatin/paclitaxel schedule with the same study requirements. A standout difference was the critical observation that concerns for serious/fatal cardiac events were now greatly reduced. As a result, patients with ovarian cancer with preexisting comorbidities that likely would have prevented them from participating in the first study were permitted entry into the second trial.

What is the point of providing this rather detailed oncology history lesson? The median overall survival (OS) for patients with advanced ovarian cancer receiving cisplatin/paclitaxel in the initial study was 38 months,¹ and the median survival with the same program, including cancer-staging eligibility, conducted by the same group of physicians in the second study was strikingly only 26.3 months.² Although crosstrial comparisons are always fraught with hazard, it is rational to argue that this observed substantial reduction in OS in this objectively reasonably controlled setting may have been substantially influenced by factors partially or completely unrelated to the cancer or its therapy.

References

1. McGuire WP, Hoskins WJ, Brady MF, et al. Cyclophosphamide and cisplatin compared with paclitaxel and cisplatin in patients with stage III and stage IV ovarian cancer. N Engl J Med. 1996;334(1):1-6. doi: 10.1056/NEJM199601043340101.
2. Muggia FM, Braly PS, Brady MF, et al. Phase III randomized study of cisplatin versus paclitaxel versus cisplatin and paclitaxel in patients with suboptimal stage III or IV ovarian cancer: a gynecologic oncology group study. J Clin Oncol. 2000;18(1):106-115. doi: 10.1200/JCO.2000.18.1.106.