VIENNA, AustriaWhen cure or prolongation of survival is no
longer a reasonable possibility, enhancing the cancer patients
quality of life becomes the preeminent objective of treatment, Ian
Tannock, MD, of the Princess Margaret Hospital, Toronto, said at the 10th
European Cancer Conference (ECCO 10). Quality of life is not a
soft endpoint and, when measured appropriately, is every bit as
reproducible as so-called objective measures such as tumor
response, he said.
Dr. Tannock challenged the tendency of clinical trials to concentrate
on the endpoints of tumor response and survival, even in advanced
metastatic disease where improvement in survival is unlikely. If
the goal of treatment is palliation, we should try to measure
palliation directly, he said.
Tumor response is not an endpoint of benefit to the patient, Dr.
Tannock argued, noting that response may reflect patient selection,
performance status, selection of response criteria, errors in
measurement, and other methodologic issues. If we shrink the
patient as rapidly as we shrink the tumor, we havent gained
much, he said.
There are certain principles about measuring quality of life, he
said, and my own bias is that, in most clinical trials,
including some of those that are still ongoing, it is done extremely
badly. He stressed that quality of life must be assessed not by
physicians or family members but, rather, by patients themselves.
Dr. Tannock also criticized trials that compare mean quality of life
at baseline with mean quality of life at a fixed point after
randomization and then compare the mean changes between the two
In contrast, he said, clinical investigators measure tumor size for
each patient at baseline, specify an a priori reduction in size and
duration for which it must be maintained in order for a patient to be
designated a responder, and then calculate the proportion of patients
in each study arm who satisfy these criteria.
Surely the analogy is the same with quality of lifeit is
an individual property of the patient, Dr. Tannock said.
You have to measure a predefined quality of life for each
patient at baseline and decide which quality of life endpoint is
relevant. You specify a priori improvement in quality of life and the
duration for which it must be maintained in order for the patient to
have a palliative response. You measure quality of life repeatedly
and determine the proportion of patients in each arm who satisfy the
criteria and for how long.
Dr. Tannock suggested that this approach could be applied in studies
of hormone-refractory prostate cancer, where chemotherapy may well
decrease PSA or improve bone scans while at the same time worsening
quality of life. He and his colleagues have validated a simple,
disease-specific questionnaire called PROSQOLI (Health-Related
Quality of Life Instrument for Prostate Cancer), which consists of a
six-point present pain intensity scale; an analgesic diary; and
linear analog scales that measure pain, physical activity, fatigue,
walking, appetite, constipation, and urination.
Using this instrument, the investigators found that prednisone plus
mitoxan-trone (Novantrone) was superior to prednisone alone in either
decreasing the pain score by at least 2 points or permitting a
decrease in analgesic dose without increased pain, and that the
combination produced a longer duration of response.
We did not generate a pain improvement at the expense of other
aspects of quality of life, Dr. Tannock said, citing positive
changes in the linear analog scales. Importantly, a substantial
number of patients experienced an improvement in quality of life
despite a rise in PSA levels or, conversely, showed no palliative
response despite a fall in PSA.