On review of the very cellular bone marrow aspirate, marked erythroid hyperplasia was evident (M:E ratio 1:10) with a notable left shift in erythroid precursors, prominent proerythroblasts, relatively mild erythroid dysplasia and karyorrhectic forms (top images ×100 objective). There was no excess of myeloblasts but dysplastic neutrophils were present (top centre). Importantly, the marrow was not megaloblastic. The trephine biopsy sections showed 100% cellularity with marked erythroid hyperplasia with left shift (bottom left, H&E and bottom centre, glycophorin C, immunoperoxidase ×50). A proportion of proerythroblasts also showed expression of CD117 (bottom right, immunoperoxidase ×50). There was no excess of CD34+ cells. A diagnosis of pure erythroid leukaemia was made. Subsequent myeloid next generation sequencing identified an NRAS mutation (c.35G>A) and two mutations in WT1 (c.758A>G and c.759C>A).
Pure erythroid leukaemia is a rare neoplastic bone marrow disorder characterised by a proliferation of immature erythroid cells (>80% of bone marrow cells with at least 30% proerythroblasts) and no significant myeloblast component. Patients typically present with pancytopenia and the condition can appear de novo, can evolve from a myelodysplastic syndrome or can be therapy‐related. It is important that the erythroid hyperplasia is not attributed to haemolysis (the evidence in this patient was not convincing) or to a myelodysplastic syndrome. In the latter, erythroid hyperplasia and dysplasia is common but the marked left shift with predominance of proerythroblasts and myeloid hypoplasia is not seen. Furthermore, de novo pure erythroid leukaemia, as in this 40‐year‐old patient, has a more acute presentation rather than the more gradual onset of cytopenias typically associated with MDS. A reversion to primitive erythropoiesis with increased haemoglobin F levels can occur in erythroleukaemia. There is no specific cytogenetic abnormality associated with pure erythroid leukaemia; complex karyotypes with loss of chromosomes 5 and 7, 5q− and 7q− are common, whilst favourable cytogenetics is very rare. The prognosis is poor when the condition evolves from MDS or is therapy‐related (these cases being categorised differently), but may be more favourable and similar to other subtypes of AML if it arises as a primary condition (Santos et al. 2009).
Reference
1 Santos FPS, Faderl S, Garcia‐Manero G, Koller C, Beran M, O’Brien S et al. (2009) Adult acute erythroid leukaemia: an analysis of 91 patients treated at a single institution. Leukemia, 23, 2275–2280.
MCQ
1 Proerythroblasts in pure erythroid leukaemia often express:CD34CD61CD117E‐cadherin (CD234)Glycophorin A (CD235a)For answers and discussion, see page 206.
12 eactive mesothelial cells
A 53‐year‐old woman presented with chest pain and dyspnoea. On CT imaging a large mediastinal mass was identified associated with bilateral pleural effusions. On biopsy the mass was shown to be a myeloid sarcoma. There was an associated t(10;11)(p12;q14.2); (PICALM‐MLLT10). The full blood count was normal and a bone marrow aspirate showed no evidence of acute leukaemia. The patient was treated with two cycles of acute myeloid leukaemia induction chemotherapy with marked regression of the mass, but a left pleural effusion, though improved, persisted. A pleural fluid sample was aspirated due to concerns regarding persisting disease. A cytospin preparation is shown above (×50 objective). Note the population of large cells with vacuolated blue cytoplasm and an eccentric nucleus. These features are typical of pleural mesothelial cells; they do not represent a neoplastic population but might be interpreted as such by the inexperienced. In addition, note the small compact lymphoid cells, which are reactive T lymphocytes that are prevalent in reactive effusions (the diagonal line of cells, image right). Morphological and flow cytometry assessment of the fluid specimen identified no precursor myeloid cells. The patient completed a further two cycles of treatment and the effusion fully resolved. She remains well and disease free on follow‐up.
Mesothelial cells are often seen in pleural and ascitic taps. They are shed into the fluid environment when other disease processes cause pleural irritation or interfere with pleural lymphatic drainage. These cells have characteristic morphology as illustrated here, but do not express CD45 or haemopoietic lineage markers. They can on occasion be binucleated (image above ×50) and show significant size variation. They should not be mistaken for carcinoma or other non‐haemopoietic neoplasms. This is another good example of where a careful morphological assessment is necessary and careful correlation with other important points in the history is absolutely essential.
PICALM‐MLLT10 is most often found in T‐acute lymphoblastic leukaemia but also occurs in acute myeloid leukaemia and mixed phenotype acute leukaemia.
MCQ
1 Myeloid sarcoma:Can be the first sign of relapse of acute myeloid leukaemiaCan have a green colourCan be associated with t(8;21)(q22;q22.1)Is common in acute promyelocytic leukaemiaOccurs only at a single siteFor answers and discussion, see page 206.
13 Plasmablastic myeloma
A 76‐year‐old woman treated with multiple lines of therapy for multiple myeloma presented to clinic with generalised debility and recent heavy nose bleeds. Her paraprotein levels had been noted to be on the rise despite lenalidomide and dexamethasone therapy. The full blood count showed Hb 95 g/l, WBC 3.5 × 109/l, neutrophils 1.5 × 109/l and platelets 18 × 109/l. An IgG paraprotein quantitation was 69 g/l. It was clear that the disease was becoming refractory to therapy but a bone marrow aspirate was taken in view of the problematic thrombocytopenia and bleeding. The aspirate was markedly hypercellular and well over 90% of cells were large pleomorphic plasma cells, some showing prominent nucleoli (images above ×100 objective). Normal haemopoiesis was markedly reduced and megakaryocytes in particular were scarce.
The malignant plasma cells of multiple myeloma are usually easy to recognise allowing accurate quantitation using a manual differential count. The morphology of plasma cells in refractory myeloma, however, often changes with increasing pleomorphism, increasing cell size and multinuclearity. The cells can sometimes resemble those of a high‐grade lymphoma. In addition many of these patients start to shed plasma cells into the peripheral blood (also noted in this case, not shown). Despite their morphological abnormality, the lineage is indicated in this patient by the strongly basophilic cytoplasm and the paler paranuclear Golgi zone. These features are important in helping recognise plasma cells when the nuclear morphology is atypical (images below ×100). Note the variation in nuclear morphology with bilobed and even binucleated forms but the Golgi zone and intense blue cytoplasm are prominent; all of these cells are plasma cells. As an adjunct, note the cytoplasmic fragments and particles due to the intense fragility of plasma cells on handling.
It may be worth reassessing