Haematology. Barbara J. Bain

left) together with the ethnic origin of the patient, are indicative of a diagnosis of Southeast Asian ovalocytosis. This condition results from mutations in the SLC4A1 gene resulting in a variant band 3 protein which increases membrane stiffness. In the resting state there is normally minimal or absent haemolysis but this can be significant in the perinatal period. Carriers of this polymorphism have a survival advantage against all forms of malaria.

A second patient, a 37‐year‐old man, was referred for investigation of mild chronic thrombocytopenia; the full blood count showed Hb 140 g/l, WBC 5 × 10⁹/l and platelets 80 × 10⁹/l. There was no personal or family bleeding history and clinical examination was normal. The blood film showed prominent stomatocytes and macrothrombocytopenia but no neutrophil inclusions were present (centre images ×100). Routine biochemistry and serum lipid levels were normal. Serum phytosterol levels, however, were grossly elevated at up to 20 times the upper limit of normal. He was a compound heterozygote for mutations in the ABCG5 gene and these findings are in keeping with a diagnosis of hereditary phytosterolaemia, also known as sitosterolaemia. This is an under‐recognised disorder of plant lipid (phytosterol) absorption and excretion characterised by red cell stomatocytosis, macrothrombocytopenia and sometimes a mild haemolytic anaemia with splenomegaly. It is an important condition to be aware of as it can lead to abnormal lipid deposition in tissues (xanthelasma and xanthomas) and in blood vessels, leading to accelerated atherosclerosis. Patients may benefit from treatment with ezetimibe, which blocks absorption of phytosterols and may protect patients from the aforementioned complications.

      A third patient, a 60‐year‐old man with a historical diagnosis of hereditary spherocytosis treated with splenectomy at 18 years of age after an episode of parvovirus‐induced red cell aplasia, was referred for reassessment after developing life‐threatening pulmonary embolism following a long‐haul flight. His full blood count showed Hb 140 g/l, WBC 13 × 10⁹/l and platelets 507 × 10⁹/l. There was no personal or family history of thrombosis and physical examination was normal. There was no laboratory evidence of an inherited thrombophilia and no molecular evidence of a myeloproliferative neoplasm. His blood film showed well‐haemoglobinised red cells with prominent stomatocytes in addition to hyposplenic features (right images ×100). His historical pre‐splenectomy osmotic fragility test was traced and found to be normal as were red cell eosin‐5‐maleimide (EMA) binding studies by flow cytometry. Red cell membrane polyacrylamide gel electrophoresis showed no abnormality. The features here are in keeping with a diagnosis of overhydrated hereditary stomatocytosis, a mild, often well‐compensated haemolytic anaemia most often caused by mutations in the RHAG gene that result in loss of membrane stomatin, causing abnormal sodium and potassium homeostasis and increased intracellular water. Splenectomy significantly increases the risk of venous thrombosis; the patient is now on life‐long anticoagulation with rivaroxaban.

MCQ

      1 Stomatocytosis can be a feature of:Alcoholic excess and alcoholic liver diseaseHereditary xerocytosisHydroxycarbamide therapyTriose phosphate isomerase deficiencyZieve syndromeFor answers and discussion, see page 206.

25 Reactive lymphocytosis due to viral infection

      An 18‐year‐old ‘fresher’ medical student was referred to the university medical service with a fever, sore throat and cervical lymphadenopathy. His full blood count showed Hb 144 g/l, WBC 74 × 10⁹/l, neutrophils 7.4 × 10⁹/l, lymphocytes 66 × 10⁹/l and platelets 203 × 10⁹/l. The liver enzyme profile showed increased transaminases (AST 300 u/l, ALT 279 u/l) and a Monospot® test (which later became available) was positive. Epstein–Barr virus (EBV) was subsequently detected in his blood at 5606 copies/ml. The blood film caused some initial concern amongst the trainees due to the magnitude of lymphocytosis, but further careful morphological review from more senior members of the department was reassuring. Notably, the blood film shows a population of medium to large pleomorphic lymphocytes (all images ×100 objective), with some showing fine cytoplasmic granules. The cells have voluminous cytoplasm and tend to be indented by adjacent red cells (top images). Importantly, some cells have blastoid morphology, being larger with a prominent nucleolus and having more pronounced cytoplasmic basophilia. Flow cytometry identified a large population of CD8+, CD5+, CD2+, HLA‐DR+, CD7+/− cells which were not expressing precursor antigens, CD30 or CD25. The morphological diagnosis is a reactive lymphocytosis due to viral infection, in this case EBV infection. The flow cytometric studies are in keeping with this as expression of HLA‐DR is an activation marker of T cells and such cells frequently show some loss of CD7 expression. The activated cells are cytotoxic CD8+ T lymphocytes even though EBV targets the B‐lymphoid population.

Reactive lymphocytosis is relatively common in clinical practice and the morphological features described above are typical. Neoplastic disorders tend to have more monomorphic appearances, but of course a careful assessment of the clinical circumstances is always necessary. T‐cell neoplasms are remarkably variable in their presentation but the clinical history, morphology, laboratory investigations and flow cytometric findings were all consistent with a reactive process. Importantly, a wide range of viruses and other organisms can induce such a reaction; these include EBV, cytomegalovirus, human immunodeficiency virus (HIV), toxoplasma and pertussis. Most often the responsible organism is benign but if HIV is implicated it is important that this is identified. A positive test for a heterophile antibody is a pointer to EBV infection but not all cases are positive. Immunophenotyping is not generally indicated but was performed in this case because of initial concern about the possibility of a lymphoma.

MCQ

      1 Infectious mononucleosis due to the Epstein–Barr virus:Can be associated with pure red cell aplasiaCan be complicated by haemophagocytic lymphohistiocytosisCan be followed by aplastic anaemiaIs associated with a subsequent increased incidence of Hodgkin lymphomaPredisposes to B‐lineage acute lymphoblastic leukaemiaFor answers and discussion, see page 206.

26 Therapy‐related acute myeloid leukaemia with eosinophilia

      A 70‐year‐old man developed therapy‐related acute myeloid leukaemia (t‐AML) some years after starting chlorambucil therapy for chronic lymphocytic leukaemia. His FBC showed Hb 63 g/l, WBC 48.9 × 10⁹/l, neutrophils 1.0 × 10⁹/l, lymphocytes 5.5 × 10⁹/l, eosinophils 19.6 × 10⁹/l, monocytes 1.4 × 10⁹/l, myelocytes 0.4 × 10⁹/l, blast cells 21 × 10⁹/l and platelet count 11 × 10⁹/l. There were two nucleated red blood cells per 100 leucocytes. Many of the eosinophils were cytologically abnormal: 25% were hypogranular, 46% were vacuolated and 17% were both hypogranular and vacuolated (all images ×100 objective). Occasional eosinophils had fine basophilic granules (bottom centre) or non‐lobated nuclei. Neutrophils were cytologically normal. The myeloblasts were small and compact with little cytoplasm and prominent nucleoli (right images).

Clonal eosinophils in myeloid neoplasms tend to have more marked cytological abnormalities than those in reactive eosinophilia but there is considerable overlap. It is therefore essential for cytological abnormalities to be assessed together with other disease features. When there is an increase in blast cells, as in the current patient, or when there are clearly dysplastic features in other lineages, such as hypogranular neutrophils, the diagnosis is straightforward. However, if an increase in cytologically abnormal eosinophils occurs without other morphological clues, cytogenetic and molecular analysis may well be needed to confirm the diagnosis of a myeloid neoplasm.

      The images above (×100 objective) show