Response to bortezomib in refractory type I cryoglobulinemia

ESAs not the culprit: More studies requiredTo the Editor: We read with interest the recently published study by Katodritou et al. [1], which addresses the important issue of whether Erythropoiesis-stimulating agents (ESAs) impact survival in multiple myeloma patients.However, it is impossible to draw meaningful conclusions from this retrospective study because of important limitations in both its methodology and reporting.In general, the well-known drawbacks inherent in nonrandomized studies rarely allow such studies to provide a valid estimate for the efficacy of a treatment [2], which is especially true in a setting where the indications for treatment are also strong predictors of outcome.The ESA and none ESA treated populations differed in almost all important prognostic factors.Patients started on ESAs presented with higher age; higher International staging system (ISS) stage; lower platelet count, hemoglobin levels, and serum albumin; and higher serum creatinine, Lactatdehydrogenase (LDH), and b2microglobin.They had more frequent light chain paraproteinemia and achieved a lower response rate to cancer therapy.These discrepancies illustrate that any attempts to compare outcomes between the ESA-treated group and the non-ESA treated group are prone to misleading results and interpretations.Further, the survival curves for both patient groups depart from each other immediately after start of the comparison and continue to depart from each other over the next 50 months of follow up, which is years after initiation and termination of ESA therapy.This clearly demonstrates that the differences observed in this study between ESA-treated and non-ESA-treated patients are the result of the selection of two patient populations with very different prognostic features and not the consequence of treatment with ESAs.To control the differences between the ESA-treated and non-ESA treated populations it would have required that all possible confounding factors were identified and measured accurately.Although a multivariate analysis was used in this article to control the marked differences between both groups for several prognostic factors, it is likely that marked differences also existed for unmeasured factors.Furthermore, only limited information was given regarding the details of the multivariate analysis.The article reported neither the results for the full model nor the process used to select variables for building the model, which makes it difficult to fully evaluate the quality of the multivariate analysis results.As it is highly likely that patients with the worst prognoses were selected to receive ESAs, the significance of ESA in the multivariate model may only indicate that ESA use is a good surrogate for poorer performing patients.Additional shortcomings are lack of specifications about the initiation of ESA treatment.As ESAs were started in patients with a baseline hemoglobin value ranging from 7.3 g/dl to 11.9 g/dl, obviously, the decision for beginning ESA therapy was left to the discretion of individual physicians, not standardizing for Hb levels, pretreatment, type of myeloma treatment, or age.Further, it is likely that a mix of patients with different phases of disease were analyzed because it is not made clear whether patients analyzed received their ESA treatment during the first or later line of chemotherapy.Information is also missing regarding time to the first dose of ESA, whether ESAs were given concomitantly with chemotherapy and whether the proportions of patients with or without chemotherapy were comparable between both groups.In this study, the median duration of ESA therapy was only 6 weeks (range 4-10 weeks), which is strikingly shorter than that used in any other meaningful study evaluating the effects of ESAs.In spite of this short treatment duration, the authors report an ESA response rate of 68% but do not mention how the response was defined.It is also unclear why the start time for progression-free survival and overall survival differed.Progressionfree survival was defined from start of treatment (assumed to be chemotherapy), although overall survival was defined from diagnosis.Finally, the generalization of the findings from this study is also limited because the data were collected from only one center and were generated over a period of 19 years, during which the patterns of care for multiple myeloma have changed significantly.Finally, if the authors would really believe in their data they would have refrained from arguing that, ''ESAs still maintain their therapeutic value.''Presently, an impact of ESA use on survival, either positive or negative, cannot be determined on the basis of this article and on a previous report claiming the opposite, namely, a prolongation of survival in ESA-treated patients [3].A prospective, randomized trial would be the optimal scenario for answering the pertinent question of how ESA use impacts survival in cancer patients.The second-best approach would be a meta-analysis of all previously conducted studies that randomized patients prospectively to an ESA-treated group or to a control group (best supportive care or placebo control).An international group of investigators is currently aiming to accomplish this task.