Controversies in Management: Chemotherapy for solid tumoursBMJ 1995; 310 doi: https://doi.org/10.1136/bmj.310.6974.246 (Published 28 January 1995) Cite this as: BMJ 1995;310:246
- G M Mead, consultant in medical oncologya
Routine treatment not yet justified
Most patients with advanced or metastatic cancer will at some point be considered for chemotherapy or, occasionally, biological therapy—for example, interferon. These treatments are of limited specificity and commonly result in short term toxicity; they may also be expensive. If cure or increased survival is a realistic possibility these considerations are important but do not alter the treatment approach. However, most cancers are incurable once metastatic and often respond poorly to chemotherapy, which can result in side effects, inconvenience, and financial costs without improvements in symptoms or survival. In practice, partly because of the limited resources available in Britain, chemotherapy is often not discussed with, much less given to, many such patients. Increasingly, however, patients demand access to all available options, and the issue then is should treatment be considered, and if so with single or multiple drugs (with of course variable toxicity and cost) given intensively or non-intensively? Common examples of these diseases include metastatic non-small cell lung cancer, colorectal and upper gastrointestinal cancer, and renal cell cancer. The table shows some of the treatments used.
Though no systematic reviews have been published, anecdotal experience suggests that management policies for these diseases vary widely between surgeons, radiotherapists, and oncologists and also within these groups. Approaches vary considerably between institutions and nations, and in the public and private sector. Desperate patients who seek second opinions may also expose these differences, as will enthusiastic forays into the media by patients or colleagues. The media often highlight new and untested therapies without the tempering effects of peer review or trials. How should we judge treatment efficacy, and is it possible to achieve a consensus with regard to standard therapy for these diseases?
Medical oncologists assess chemotherapy regimens by measuring clinical and radiological response rates together with “time to treatment failure” and survival. Additional and potentially very important end points are control of symptoms and quality of life, as assessed by the patient, and financial cost. In practice all of these variables require interpretation—and naturally patients and their families will have their own views on treatment goals.1
Response rates are perhaps the most commonly used method of assessing efficacy of chemotherapy. Though these give data on the proportion of patients receiving treatment with no benefit, radiological shrinkage of solid tumours should not be overinterpreted, as it often has little or no survival benefit.2 Toxic, expensive drugs such as aldesleukin for renal cell cancer3 and, more recently, paclitaxel for ovarian carcinoma, have surprisingly, been licensed on the basis of response rate without published randomised trial data that allow clinicians to place the efficacy of these drugs in an understandable clinical context. Media and patient pressure for such new and expensive therapies seems in danger of dominating clinical science and sense.
What is meaningful to the patient is an improvement in survival, symptoms, or quality of life. Unfortunately, few studies have compared chemotherapy with supportive care alone. Few, if any of the trials of chemotherapy for metastatic solid tumours, have shown a survival benefit when comparing standard single drugs with more complex combination chemotherapy. A good example of the problems that can occur has been the widespread adoption of potentially toxic cisplatin containing chemotherapy for advanced gastrointestinal malignancy—on the basis of preliminary experience—and where randomised trials have shown no benefit.4 5 In diseases such as non-small cell lung cancer meta-analysis of studies comparing chemotherapy with no chemotherapy are under way—though even here conclusions are open to interpretation and conflict.6 7 8
Measurement of quality of life or control of symptoms in chemotherapy trials seem to be an inherently sensible approach.9 But these data may be difficult to understand and can be marred by denial and the hope given to patients by new or more intensive treatments. Interestingly, in a study in which patients were asked to define what chance of cure, increase in survival, or alleviation of symptoms would make intensive treatment worth while10 quite minimal chances were avidly sought, despite toxicities. This perception was not shared by health care professionals or surrogates, though these people are of course not threatened by loss of life.
It is clearly neither necessary nor possible to formulate an explicit policy whereby all or none of these patients should receive chemotherapy. Many trials do suggest with some regularity certain prognostic features for longer survival, such as younger age, better performance status, and no major organ dysfunction. Thus it should be possible to be selective when recommending treatment, excluding patients for whom the potential benefit is thought to be low. In this regard, oncologists should be no different from any other clinician practising good medicine.
Cancer chemotherapy can be a burdensome approach to treatment, and it has to justify its role in the context of overall health care. Only intelligently designed prospective studies performed in large patient populations can assess the true impact of these treatments when given with palliative intent. New treatments need to be designed and tested in large centres but should not be widely adopted without supporting data from randomised clinical trials comparing them with a standard treatment. We need trials that assess the benefits of non-toxic single drugs compared with intensive combination chemotherapy; these studies should measure not only survival but also more meaningful end points such as control of symptoms and quality of life. With such objective data it will be possible to plan resource allocation and make more intelligent recommendations about treatment to our patients.