Patients with anemia may develop a chronic anemia that impacts the quality of life, and some of these patients may benefit from administration of recombinant erythropoietin. Ludwig et al identified factors indicating the likelihood of response to erythropoietin administration. The authors are from Vienna and Graz in Austria.
Patient selection: Patients with cancer and chronic anemia. The patients should not have chronic blood loss or nutritional deficiency as the primary cause for the anemia.
Protocol: Human recombinant erythropoietin is administered for 2 weeks.
Laboratory testing:
(1) serum erythropoietin levels
(2) change in hemoglobin
(3) serum ferritin
Rule 1: Based on serum erythropoietin level and change in hemoglobin
|
Serum EPO
|
Change in Hemoglobin
|
Prediction
|
Predictive Power
|
|
>= 100 mU/mL
|
< 0.5 g/dL
|
no response
|
93%
|
|
< 100 mU/mL
|
>= 0.5 g/dL
|
response
|
> 95%
|
|
>= 100 mU/mL
|
>= 0.5 g/dL
|
possible response
|
see note
|
|
< 100 mU/mL
|
< 0.5 g/dL
|
possible response
|
see note
|
where:
• According to Figure 1, page 1059, the probability for response if not in the first group is 80%. However, this includes the patients in the second group, meaning that the predictive value for response in the third and fourth groups is less than 80%.
Rule 2: Based on serum ferritin
If the serum ferritin is < 400 ng/mL then predict response (predictive power 72%).
If the serum ferritin is >= 400 ng/mL, then predict no response (predictive power 88%).
Use of the Rules (Table 5, page 1061):
(1) If the conclusion is that there no response to erythropoietin is evident, then the erythropoietin is discontinued.
(2) If there is evidence of response to erythropoietin, then it is continued.
(3) If there is uncertainty as to response, then the patient should be re-evaluated after a few more weeks of therapy.
