The role of quality of life assessment in gastric cancer.

Abstract

To the Editor: Patients with locally advanced gastric cancer (T3/4 with extensive nodal disease) without distant metastases from Europe or the USA generally face a dismal prognosis. At laparoscopy or laparotomy, such patients even without peritoneal deposits or positive cytology are often deemed irresectable and subsequent treatments are palliative in intent. Only those with gastric outlet obstruction, bleeding or an ‘easily’ resectable tumour proceed to simple total or subtotal gastrectomy. Reasons for this practice are based on assumptions that locally advanced disease with lymph node spread is indicative of micrometastatic disease elsewhere and the knowledge that extensive resection will probably be associated with serious morbidity and mortality. My visit to the National Cancer Centre, Tokyo, in March 2000 has impressed on me the very different approach to locally advanced disease in Japan. I watched two D2 total gastrectomies with para-aortic lymph node dissection in patients with extensive nodal disease. These were performed with meticulous surgical techniques and minimal trauma. At present there are only retrospective data to support super-extended surgery (1). Morbidity and mortality are consistent with standard Japanese D2 lymphadenectomy and 5 year survival rates are about 15%. The value of this approach is questioned, however, because many patients still die from recurrent disease, some within 12 months of surgery. Patients will also suffer the deleterious effects of gastrectomy on food intake, appetite, gastrointestinal motility and general well being. Prospective randomized evidence to support super-extended surgery is required. Randomization to D2 and extended lymphadanectomy or palliation alone might seem the logical protocol, although it is understood that this may be considered unethical in a country which boasts negligible morbidity and mortality after D2 gastrectomy. The present JCOG trial, protocol 9501, comparing D2 and D2 plus para-aortic dissection may help resolve this problem if sufficient patients with extensive locally advanced disease are accrued. Clearly survival and standard outcomes are important end-points in such a trial, but a measure of patients’ quality of life (QL) will be critical to interpretation of the data. If length of survival is similar in both groups, then quality of survival becomes paramount. Quality of life may seem a vague philosophical idea, but within a medical context it is defined as a measure of physical, social and emotional well being. Self-completion questionnaires are currently the standard tool used to assess QL and instruments must be reliable and valid. Assessments should be performed in standardized settings, either in the hospital or at the patient’s home, and if a third person is needed to complete the questionnaire this should be recorded. Analyses of QL data are complex because multiple comparisons are made and problems with missing data due to attrition occur. It is therefore essential to choose a QL instrument carefully and to ensure that it is adminstered thoughtfully and that data are analysed correctly (2). A rigourous approach to QL assessment will provide reliable and valid data that can be of clinical use. There are several valid generic measures of QL for patients with cancer that assess functional aspects of QL and commonly occurring symptoms. These may be supplemented by disease-specific modules to improve sensitivity and specificity. The European Organization for Research and Treatment of Cancer Quality of Life Study Group has adopted this ‘modular’ approach to QL assessment and modules are available for breast, lung, head and neck, oesophageal and ovarian cancer (3,4). A module designed to measure QL in patients with gastric cancer has recently been developed (5). This has questions assessing dysphagia, dietary restrictions, upper gastrointestinal discomfort, pain, specific emotional problems and side effects of chemotherapy. Using QL assessment in clinical trials of patients with gastric cancer will help define the role of potentially curative surgery, palliative resection and adjuvant treatment. In trials of locally advanced gastric cancer it may become apparent that the small hope of cure offered by surgery is offset against the likely negative impact on QL, or data may reveal that super-extended surgery can be carried out with minimal impact on QL. This type of information will improve decision making for both the patient and the clinician.