An exploratory study of frequent pain measurement in a cancer clinical trial.

The ideal methodology for quality of life (QOL) measurement in cancer clinical trials matches the evaluation to the anticipated outcomes, thereby increasing the likelihood that clinically relevant changes are captured. The present study explored the importance of such methodological 'tailoring' in a phase II trial of paclitaxel and recombinant human granulocyte-colony stimulating factor (rhG-CSF) for metastatic breast cancer. Prior to the trial, clinical observation suggested that frequent short-lived episodes of pain might occur during this treatment regimen. Twenty-one patients provided longitudinal data for at least three cycles of chemotherapy. To assess transient pain, a routine QOL assessment at baseline and every third cycle was supplemented with pain measurements twice weekly. The interval assessment included a multidimensional QOL instrument (Functional Living Index-Cancer) and measures of psychological state (Rand Mental Health Inventory), symptom distress (Memorial Symptom Assessment Scale), and performance status (Karnofsky Performance Status Score). The frequent pain measurements were acquired using visual analogue and categorical scales for pain intensity (Memorial Pain Assessment Card). From baseline to the end of cycle three, global pain scores declined and the results on other QOL measures were variable. The data obtained using these measures did not reveal the existence of episodic pains. In contrast, the twice weekly pain measurements clearly demonstrated transient severe pains in approximately half the patients. These data highlight the importance of specific measurement of troubling symptoms or other relevant QOL concerns at clinically appropriate intervals during the routine QOL assessment of clinical trials. The additional burden involved in these assessments is warranted if the information derived is highly relevant, would not be adequately captured otherwise and could improve therapy.