Durable response to recombinant human erythropoietin in a patient with myelodysplastic syndrome.

Sir, I read with great interest the letter by Aloe Spiriti et al. on a patient with myelodysplastic syndrome (MDS) who had a durable response to recombinant human erythropoietin (rHuEpo). In my opinion, it emphasizes the need for a patient-oriented instead of a disease-oriented approach to the use of this expensive drug. It is well known that only 15 to 20% of MDS patients respond to rHuEpo and that the vast majority of responders are not transfusiondependent. In addition, the doses require to achieve response are high so that treatment is unlikely to be cost-effective. On the other hand, cost-effectiveness represents a problem also for the use of rHuEpo in the setting of uremia. The patient described by Aloe Spiriti et al., however, had inadequate endogenous erythropoietin production and regular need for transfusion, thus meeting two fundamental prerequisites for the use of rHuEpo. In addition, a serum erythropoietin level < 100 mU/mL and female gender (i.e., two features of the reported patient) have been found to be factors predicting response to rHuEpo in patients with myelodysplastic syndrome. Apparently they might also predict a durable response. I agree with Aloe Spiriti et al. that rHuEpo should be employed in any patient with myelodysplastic syndrome who fits into the above criteria, and that, though not cost-effective, treatment should be continued if the need for transfusion is eliminated. Erythropoietin is not approved for treatment of MDS patients, but one of the duties of a physician is to be able to move from public policy implemented for the average patient to one that serves the individual patient’s needs. rHuEpo should be provided to all patients who can medically benefit from it, regardless of whether the basic disorder is an approved condition or not. Using predictive algorithms as those recently developed by us may allow a patient-oriented rather than a disease-oriented approach to rHuEpo treatment of nonrenal anemia. Other examples of individual patients responding to rHuEpo have been previously reported in this journal.