Complete remission of refractory immune thrombocytopenia (ITP) with a short course of Romiplostim

To the Editor: Chronic immune thrombocytopenia (ITP) is an autoimmune disorder characterised by antibody-mediated impaired platelet production and increased platelet destruction resulting in decreased platelet counts and variable bleeding (1). The arrival of thrombopoietic (TPO) agents has revolutionised the treatment for chronic, refractory ITP with a high number of patients responding to these treatments even after splenectomy failures (2, 3). However, the use of such agents is complicated by the fact that discontinuation of these drugs may result in rebound thrombocytopenia, while prolonged use has theoretical, but as yet unconfirmed, risks of thrombosis and myelofibrosis (1). In this report, we report the normalisation of platelet counts in a previously refractory ITP patient with a very short course of romiplostim. A 70-year-old male first noticed bruises and low platelet count (5 9 10/L) in October 2007. He was initially treated with steroids, which produced a good response. Drop in platelet counts over the next 3 years prompted the use of cyclosporine (caused renal impairment), low dose rituximab (excellent but short-lived response of 8 months) and mycophenolate (no response). During this period, he developed a bleed in his left eye, which caused partial visual loss and required intravenous immunoglobulins. A second dose of rituximab at a higher dose of 375 mg/m was tried, which produced a ten-month period of remission. Laparoscopic splenectomy was then undertaken in October 2011. However, within a period of 5 months, platelet counts dropped again, and he was referred to our institution. Bone marrow biopsy carried out at this time demonstrated no evidence of myelodysplasia but increased number of megakaryocytes consistent with peripheral consumption. Tests for human immunodeficiency virus, and hepatitis B and C viruses, cytomegalovirus and Helicobacter pylori were negative. Dapsone was next tried that caused methemoglobinemia. Prednisolone was required throughout these years, on and off, for relapses. In October 2012, Romiplostim was commenced at a dose of 2 lg/kg, when the platelet count was 16 9 10/L (see Fig. 1). Because the bruises started getting more prominent, additional prednisolone was given. Platelet count responded well but overshot to 527 and 630. The second dose of Romiplostim at 1 lg/kg was given 3 weeks later, when the count had decreased to 87 9 10/L. Despite this dose, the platelet count continued to drop. Associated bruises and epistaxis, required further rescue steroids. A week later when the platelet count was 14, a third injection of Romiplostim 2 lg/kg was given, when once again, the count shot up to over 400. He then received a further four injections at 2 lg/kg over the next 8 weeks, with intermittent need for steroids for excess bruising and very low platelet counts. However, after the seventh dose in February 2013, his platelet count remained stable with no fluctuations and clinical symptoms. The last few platelet counts until the last check in mid-June remain persistent around 250 9 10/L. Refractory ITP may be defined as a low platelet count persisting after splenectomy, but requiring active treatment to maintain a ‘safe’ platelet count (4). There is no consensus for the appropriate treatment for refractory ITP, although rituximab and TPO-mimetic agents are most commonly considered in these patients. Debate continues about which may be better of these options – rituximab is associated with long-term immunosuppression and consequent clinical implications, while the TPO-agonists have adverse effects, which may include thrombosis, rebound thrombocytopenia and fibrosis of the bone marrow. Physicians are left with the choice that is often made based on detailed discussions with the patients about the benefits and risks of these drugs. In this respect, one of the ‘advantages’ of rituximab is the short-term treatment compared with the TPO-agonists, which is only noted to maintain a safe platelet count as long as the medication is continued; while discontinuation can lead to platelet count drop to pretreatment levels, with an associated increased risk of bleeding. Complete remission of ITP with very short course of TPO-agonists has been published recently in one case series (5). This study reported sustained remissions with the use of both romiplostim and eltrombopag, although they confirmed a definite response in only three patients, receiving romiplostim. Two observations are quite important to draw similarities between these three cases and our patient. Firstly, all of them required only very small doses of the TPOagonist, typically 1–2 lg/kg. Secondly, before the onset of remission, there were wide fluctuations in the platelet count, especially a sudden increase before the counts stabilised. This may reflect changes in innate TPO cycling, which could have caused the remission. Ghadaki and colleagues argue that this spontaneous remission reflects immune tolerance to platelets by increasing the exposure to platelet antigens, similar to that of individuals with haemophilia who