To the Editor: The trephine biopsy in Figure 1 demonstrates a hypercellular bone marrow (BM) with increased numbers of large megakaryocytes arranged in loose and dense clusters with abnormal features, including hypersegmented (staghorn-like) nuclei, and a mitotic form (arrow). (Haematoxylin-and-Eosin. Left: 10× objective lens. Right: 50× objective lens.) Reticulin deposition was not increased (MF-0). At first glance, this looks like a myeloproliferative neoplasm (MPN). MPNs like essential thrombocythemia (ET) and primary myelofibrosis (PMF) are characterized by atypical megakaryocyte proliferation and morphology 1. Large to giant megakaryocytes with staghorn nuclei predominate in ET and are typically dispersed throughout the BM, but may appear in loose clusters 1. PMF is characterized by loose or dense clusters of megakaryocytes with variation in size and nuclear lobulation, including hyper- and hypolobulation; hyperchromatic, bulbous or bare nuclei; and abnormal histotopography (paratrabecular, perisinusoidal and intrasinusoidal locations) 1. Moderate (MF-2) or severe (MF-3) myelofibrosis defines fibrotic PMF, whereas reticulin fibrosis is absent or minimal (MF-0 or MF-1) in pre-fibrotic PMF and ET 1. However, this BM biopsy was performed to investigate refractory thrombocytopenia in a 17-year-old girl who had failed multiple lines of treatment for primary HIV-associated thrombocytopenia (PHAT), including prednisone, dexamethasone, intravenous immunoglobulin, vincristine and danazol plus all-trans retinoic acid. At the time of the biopsy, she was receiving thrombopoietin receptor agonist (TPO-RA), romiplostim. Her platelet count (14 × 109/L) had shown no durable improvement after 8 months of romiplostim therapy. She experienced ongoing intermittent episodes of mucocutaneous bleeding, including heavy menstrual bleeding, which resulted in iron deficiency anaemia (haemoglobin 8.8 g/dL, ferritin 39 ug/L and transferrin saturation 10%). Other causes of secondary immune thrombocytopenia (ITP), including Helicobacter pylori, other viral infections, drugs (she was not taking trimethoprim-sulfamethoxazole) and autoimmune and lymphoproliferative disorders, had been ruled out. The patient had congenital HIV and was closely followed up for virological failure at an HIV Centre of Excellence in South Africa. Several months prior to the BM biopsy, drug-resistance testing revealed a K103N mutation, which conferred resistance to non-nucleoside reverse transcriptase inhibitors, nevirapine and efavirenz. Her antiretroviral therapy (ART) regimen was consequently changed from tenofovir/lamivudine/efavirenz to tenofovir/emtricitabine/atazanavir-ritonavir, resulting in a suppressed viral load (105 copies/µL) 2 months prior to the BM procedure. However, the virological suppression did not lead to recovery of the platelet count, which only improved transiently from 7 × 109/Lto 31 × 109/L, before dropping again to 14 × 109/L. The BM aspirate and trephine biopsy did not reveal an alternative cause for her thrombocytopenia, and it was concluded that she had refractory PHAT. While data on the use of TPO-RAs like romiplostim, eltrombopag and avatrombopag in PHAT are limited, these drugs are now widely used to treat chronic ITP and severe aplastic anaemia 2-5. Due to their efficacy in ITP, BM studies are rarely performed in patients receiving TPO-RAs. However, similar to this case, MPN-like megakaryocyte morphology has been described in patients treated with eltrombopag or romiplostim for primary ITP 6-8. The megakaryocyte proliferation observed here may be attributed to romiplostim's effects on megakaryopoiesis, which is non-clonal and relies on binding of an agonist to the TPO receptor 2. In contrast, clonal proliferation of megakaryocytes in MPNs results from TPO-independent constitutive JAK/STAT signalling due to somatic driver mutations in JAK2V617F, CALR or MPL 9. Notably, reticulin deposition was not increased in this patient's BM. BM fibrosis is another feature of MPNs that may be associated with TPO-RA therapy 1, 2, 8, 10, 11. However, this occurs in a small subset of patients, particularly with long-term use, and is usually reversible 2, 8. This case has several teaching points. The observation that severe thrombocytopenia persisted despite megakaryocyte response to romiplostim suggests that peripheral autoimmune platelet destruction, rather than impaired production, may be the predominant mechanism in PHAT. However, the anecdotal nature of a single case report limits the generalisability of these findings. Observational studies on the effect of ART on HIV-associated thrombocytopenia have yielded conflicting results, with some studies showing improvement subsequent to ART, while others have demonstrated similar findings to this case, where PHAT persisted despite ART initiation 12-14. Further research is needed to improve our understanding of the pathophysiology and treatment of PHAT. Finally, this case underscores the importance of integrating detailed clinical history, including patient age, medication exposure and other laboratory results, when interpretating BM morphology. This was a young HIV positive adolescent without splenomegaly who was on romiplostim therapy for thrombocytopenia. Her peripheral blood smear showed microcytic hypochromic anaemia with occasional large platelets with normal granularity, but no leucoerythroblastic reaction or tear drop cells. Taken together, these features made a diagnosis of an MPN highly unlikely. Whereas cytopenias frequently occur in PMF as a result of myelofibrosis, thrombocytosis is a defining feature of ET 1, ruling out the latter diagnosis. However, in a patient with rebound thrombocytosis on a TPO-RA, this BM histology could more easily have been confused with ET. Likewise, if this was an older adult with thrombocytopenia, pre-fibrotic PMF may potentially have been considered. However, MPNs are typically disorders of older adults and are extremely rare in children 15. Furthermore, while the BM histology was suggestive of an MPN, it was not typical of either PMF or ET. In the presence of abundant megakaryocyte clusters, especially dense clusters, a diagnosis of ET should be reconsidered in favour of pre-fibrotic PMF. However, although megakaryocyte clusters are typical of PMF, other characteristic features of PMF were not seen here. In cases where the significance of megakaryocyte hyperplasia remains unclear based on clinical history, and an MPN is considered, molecular studies may aid in confirming clonality. Although MPNs remain extremely rare in children, ITP and aplastic anaemia are comparatively well-known conditions that are increasingly being treated with TPO-RAs. To safeguard against erroneously diagnosing an MPN, pathologists should be cognisant of the effects of TPO-RAs on BM histology, especially megakaryocyte morphology. S.J.K. photographed the images, conceptualised, and wrote the manuscript. The author has nothing to report. Ethical clearance for this case study was obtained from the Health Sciences Research Ethics Committee of the University of the Free State (UFS-HSD2025/0295/3009). Permission was obtained from the Free State Department of Health. The patient's mother provided written informed consent, while the child provided written assent for publication of this case study. The author declares no conflicts of interest. The data that support the findings of this case study are available from the corresponding author, S.J.K., upon reasonable request.
Stephanie Juané Kennedy (Sat,) studied this question.