Biology Reference
In-Depth Information
responders are more likely to discontinue and the most anemic patients are
often the most ill and most at risk to die. A more conservative approach uses
the intent-to-treat principal in which all patients who begin therapy remain in
the analysis whether they remain on study or not. Again, the last available
hemoglobin level is imputed for each patient for each time point for the
remainder of the study. This approach probably under-estimates the efficacy of
the agent, because some patients who would have benefited withdraw from the
study or die before that response occurs.
Given these complexities and the importance of developing approaches to
the treatment of anemia in oncology that are superior to the current standard, it
is important that the literature be interpreted critically and that important com-
parisons be done on randomized data sets using consistent and clearly articu-
lated approaches to transfused hemoglobin concentrations and missing data.
Toward optimal patient management
Despite mounting evidence that successful treatment of anemia in patients
with cancer provides benefits in terms of prevented transfusions and enhanced
functional status and quality of life similar to those enjoyed by dialysis
patients, most patients with cancer who have symptomatic anemia are not cur-
rently treated. One reason for the disparity between nephrology and oncology
may be rooted in the different approaches to drug development taken in the
two clinical settings. When rHuEPO was introduced for the treatment of
patients with renal failure undergoing dialysis, patients were treated with
increasing doses until all patients had responded,although individual patients
differed as much as three-fold in terms of dose requirements. The result is an
approach to the treatment of anemia in the dialysis setting in which the physi-
cian can be assured that all patients will ultimately benefit. When rHuEPO was
developed for patients with cancer, the result was that patients were treated
with a fixed dose, sometimes with a single dose increase at least one month
later for non-responding patients; the doses chosen were not those associated
with the highest response rates. Because patients receiving erythropoietic sup-
port during chemotherapy are usually treated for five months or less, this
approach to rHuEPO therapy has resulted in therapeutic algorithms in which
40% or more of patients benefit minimally or not at all. This research has led
to extensive efforts to identify predictors of response [72-74] that might per-
mit non-responders to forgo therapy or discontinue therapy earlier, thereby
conserving resources. Moreover, once the effects of transfusions are eliminat-
ed from the analysis of hemoglobin change data, the response to currently
favored doses of erythropoietic agents (30,000 to 60,000 U weekly of
rHuEPO, 3-5 µg/kg every two weeks of darbepoetin alfa) is slow, with medi-
an times to response of approximately 10 weeks [32]. The failure of current
treatment paradigms to provide rapid relief to all anemic, fatigued patients
with cancer, coupled with the higher dose requirements and per week costs,
