In this issue of Blood, Kuter and colleagues present further information about the incidence and temporal course of bone marrow reticulin fiber formation after treatment with a TPO-mimetic.
Reticulin fiber formation is present to some degree in normal bone marrows of healthy adults,1 whereas collagen fibrosis (trichrome staining) is considered to be abnormal and associated with more severe disease.2 Both reticulin fibrosis and trichrome staining have been associated with myelofibrosis and other hematopoietic disorders. A study by Mufti et al showed that patients with immune thrombocytopenia (ITP) have normal reticulin fiber formation in the bone marrow3 so that excitement over potential benefits of thrombopoietin (TPO)–mimetic drugs in the treatment of ITP has been tempered by emerging reports of bone marrow fibrosis.4,5 These reports were especially concerning because previous animal data suggested that long-term stimulation of bone marrow with TPO (by virally induced overexpression) resulted in a phenotype that resembled primary myelofibrosis.6 In this issue of Blood, Kuter and colleagues present further information about the incidence and temporal course of bone marrow reticulin fiber formation after treatment with the TPO-mimetic romiplostim both in an animal model and in ITP patients.7 Romiplostim is a hybrid protein that contains a TPO receptor–binding region conjugated to the Fc portion of an antibody (peptibody).8
In the first part of this study, the authors describe findings in rats treated with romiplostim at up to 10 times the recommended dose for treatment of patients with chronic ITP. All of the animals showed dose-dependent increases in platelet count with a relatively short treatment course (4 weeks); a subgroup of animals was evaluated 4 weeks after discontinuation of therapy. Within 4 weeks, many of the rats showed dose-dependent changes in bone marrow; both an increase in bone marrow fiber content and hyperostosis, suggestive of collagen fibrosis (although trichrome staining was not done). All of these changes resolved after a 4-week washout period with none of the animals showing residual changes. These findings suggest that, in animals at least, romiplostim actually does alter marrow fiber content, but if treatment is short, the changes induced are reversible. Certainly a longer course of therapy in this animal model is needed to determine whether eventually, these changes become progressive or irreversible.
Next, the authors present data from 2 clinical trials: a retrospective analysis of 271 patients with chronic ITP (treated June 2002 to January 2008) who received different doses of romiplastin and a smaller, prospective study in which bone marrow biopsies were obtained before initiation of treatment as well as at 3, 6, and 9 months after initiation of therapy. Photomicrographs were reviewed by an expert panel to determine level of bone marrow reticulin. In the retrospective study, bone marrow examination was left to the discretion of the investigators and only 11 patients had at least one sample. Of these, 10 patients showed reticulin fiber formation. Given the study design, one can say only that reticulin fiber formation in patients treated with this TPO-mimetic is somewhere between 4% (10 of 271 total patients enrolled) and 91% (10 of 11 who had bone marrow studies). The available data suggest that the real incidence may be closer to 5% to 10%. Of the 10 patients who were positive for reticulin staining, 6 received higher than the currently recommended 1 to 10 μg/kg dosing. Five of these patients had both a pretreatment and on-treatment marrow evaluation. Four of these showed an increase in reticulin formation on treatment, and 3 patients had subsequent bone marrow studies that showed a decrease in reticulin staining with discontinuation of the study drug.
The prospective study involved 10 patients who received doses of romiplostim that were between the recommended 2 to 10 μg/kg. Six patients had both pretreatment and posttreatment bone marrow studies that could be evaluated for reticulin fiber formation. Only one patient had clear increase in reticulin fiber formation upon treatment, which then remained stable for the remaining months. This patient remained on romiplostim and had no change in response to therapy or in other blood counts.
These preliminary studies suggest that bone marrow reticulin fiber formation after TPO-mimetics is real and can occur in some patients receiving the recommended dosage. This fibrosis is reversible after short-term treatment and may be dose related. This is consistent with a previous study in patients with acute myeloid leukemia treated with TPO.9 Preliminary evidence is also presented that perhaps long-term treatment may not cause continuously worsening fibrosis if the dose is within the presently recommended range. Certainly these studies are encouraging, but further studies will not only need to systematically and prospectively evaluate the consequences of short-term administration of these medications, but also address the long-term consequences of treatment with any TPO-mimetic drugs. For this reason, caution is still required when considering long-term management of patients with chronic ITP with TPO-mimetics, and patients should be made aware of this potential complication.
Conflict-of-interest disclosure: The author declares no competing financial interests. ■
