Peter B. Neame

Elevated Platelet-Associated IgG in the Thrombocytopenia of Septicemia

The mechanism of thrombocytopenia, a frequent complication of septicemia, is obscure, but indirect evidence suggests that the immune system may be involved. To investigate this possibility, we quantitated platelet-associated IgG on platelets obtained from 44 patients during 46 episodes of septicemia. Thrombocytopenia occurred in 21 of 46 episodes (46 per cent). Platelet-associated IgG was elevated in eight of 11 episodes of gram-negative septicemia and thrombocytopenia (47.3 +/- 11.7 fg of IgG per platelet [mean +/- S.E.]) and in one of 20 patients with gram-negative septicemia and normal platelet counts (5.9 +/- 1.1) (P less than 0.001). Elevated levels occurred in eight of 10 patients with gram-positive septicemia and thrombocytopenia (55.3 +/- 14.7 fg of IgG per platelet) and in none of 11 patients with gram-positive septicemia and normal platelet counts (5.6 +/- 1.7) (P less than 0.001). Serial testing during the thrombocytopenia and recovery showed an inverse relation between the platelet count and platelet-associated IgG. Thrombocytopenia in some patients with septicemia may be related to the binding of IgG to platelets.

Morphology of acute promyelocytic leukemia with cytogenetic or molecular evidence for the diagnosis: characterization of additional microgranular variants.

Early diagnosis of t(15;17) acute promyelocytic leukemia (APL) is essential because of the associated disseminated intravascular coagulation and the unique response of the disease to all‐trans retinoic acid (ATRA) therapy. Early diagnosis depends primarily on morphological recognition. The French‐American‐British (FAB) classification, however, does not describe all morphological variations that occur in APL. In 25 cases with evidence of APL confirmed by cytogenetic and/or molecular analysis, we found a heterogeneous morphological group. The most common form of APL was heterogeneous and consisted of various combinations of cells in which hypergranular cells and some cells with multiple Auer rods were obvious. In some cases, one cell predominated. This led to the description of five subcategories. These included the classical FAB M3 with hypergranular cells and multiple Auer rods; the FAB variant with hypogranular bilobed cells; the basophilic cell type of McKenna et al. [Br. J. Haematol 50:201, 1982]; and two additional subtypes, one consisting of differentiated promyelocytes and a few blast cells (M2‐like), and the other consisting largely of blast cells and a few early promyelocytes (M1‐like). Immunophenotyping revealed a pattern of CD33 and/or CD13 positivity, and CD14 and HLA‐DR negativity in 96% of cases. CD2 was positive in the FAB variant and in the subtype with basophilic cells, but negative with other subtypes. Three out of five cases with basophilic cell predominance [McKenna et al.: Br J Haematol 50:201, 1982], and one out of two M2‐like cases, responded to ATRA therapy. Awareness of the heterogeneity and the atypical morphologic subtypes found in t(15;17) APL will contribute to improved recognition and early institution of ATRA therapy. Am. J. Hematol. 56:131–142, 1997.

Autologous peripheral blood progenitor cells are a potential source of parvovirus B19 infection

BACKGROUND: Parvovirus B19 is a cause of delayed red blood cell (RBC) engraftment after marrow transplantation (BMT). The diagnosis of parvovirus infection requires serologic and DNA testing in the context of clinical disease and characteristic marrow morphologic findings; however, the source of infection is often difficult to determine.

Platelet fragments do not contribute to elevated levels of platelet associated IgG

Summary. Most assays that measure platelet associated IgG (PAIgG) relate the IgG associated with the test platelets to the platelet count. This could lead to a systematic error if platelet fragments were present in the washed platelet sample but not counted. To address this issue, we studied platelets from patients with idiopathic thrombocytopenic purpura (ITP), thrombocytopenia complicating cardiopulmonary bypass, and laboratory synthesized platelet fragments (freezethawed) using electron microscopy. Scanning electron microscopic (SEM) examination of the platelet samples demonstrated appreciable numbers of fragments only in the freeze‐thawed specimens. Yet, ‘fragments’ could be seen in all specimens using transmission electron microscopy (TEM). Most of these ‘fragments’ proved to be artefacts: we found that the ratio of ‘fragments’ to intact platelets observed in the TEM specimens was similar to the estimated ratio of ‘fragments’ to platelets that would have been generated had the specimens been sectioned at 90° to the plane of the actual section.

Factors XI and XII are low in subjects with liver disease

We prospectively measured levels of factors XI and XII in parallel with other coagulation factors in 39 unselected patients with liver disease and in 20 control subjects. Mean levels of factors XI and XII in subjects with liver disease were significantly reduced, being 58% and 61%, respectively, compared with 100% and 94% in controls. Reductions in levels of factors XI and XII were most pronounced in those subjects with low serum albumin. The partial thromboplastin time (APTT) reflected low levels of either factor XI or XII and was most prolonged when both were low, but cause and effect was not demonstrated. Low levels of these factors may explain previous reports of poor response of APTT to infusions of prothrombin complex concentrates. Finally, these low levels strongly suggest that factors XI and XII are produced in the liver.We prospectively measured levels of factors XI and XII in parallel with other coagulation factors in 39 unselected patients with liver disease and in 20 control subjects. Mean levels of factors XI and XII in subjects with liver disease were significantly reduced, being 58% and 61%, respectively, compared with 100% and 94% in controls. Reductions in levels of factors XI and XII were most pronounced in those subjects with low serum albumin. The partial thromboplastin time (APTT) reflected low levels of either factor XI or XII and was most prolonged when both were low, but cause and effect was not demonstrated. Low levels of these factors may explain previous reports of poor response of APTT to infusions of prothrombin complex concentrates. Finally, these low levels strongly suggest that factors XI and XII are produced in the liver.

Prolonged complete remission in two cases of acute promyelocytic leukaemia treated with atra alone

The use of all-trans retinoic acid (ATRA) therapy for induction for acute promyelocytic leukaemia (APL) is well established and leads to a complete remission rate of > 70%. However, remissions are short and consolidation with chemotherapy is required. Continuous treatment with ATRA has been associated with a decrease in plasma concentrations of the drug, which may lead to resistance. Recently, a case has been reported successfully treated with ATRA alone, followed by maintenance with intermittent ATRA and continuous low-dose methotrexate and 6-mercaptopurine (Sanz et al, 2000). Here, we report two cases of APL with prolonged remissions after treatment with ATRA alone. An 81-year-old woman presented with pancytopenia due to APL in October, 1992. Cytogenetic analysis revealed t(15;17) as the sole abnormality, and the rearrangement was confirmed by Southern blot analysis. She was started on ATRA at 45 mg ⁄m. Chemotherapy was not given because of her age and history of cardiovascular disease. Her leucocyte count increased to 25 · 10 ⁄ l on day 1. On day 10, bilateral pleural effusions and culture negative fever developed, which was treated successfully with dexamethasone. The dose of ATRA was decreased to 25 mg ⁄m on day 30 because of abnormal liver function tests, and she remained on this dose for another 95 d. Bone marrow on day 73 showed morphological remission with persistence of t(15:17). No further bone marrow examinations were performed, but her blood counts remained normal, and reverse transcription-polymerase chain reaction (RT-PCR) for the PML ⁄RAR transcript on peripheral blood was negative, with normal blood counts assessed 4 and 6 years after diagnosis. She died of unrelated cardiovascular complications in November 1999, 7 years after diagnosis. The second case involved a 76-year-old woman who presented in September 1997 with extensive mucosal bleeding. Laboratory investigations showed pancytopenia and disseminated intravascular coagulation due to APL, and RT-PCR confirmed the presence of the short isoform of the hybrid transcript generated by t(15;17). She was treated with ATRA at 45 mg ⁄m for 149 d. She declined treatment with chemotherapy. Her coagulation parameters normalized by day 2, and the leucocyte count reached a peak of 40 · 10 ⁄ l on day 12. She developed a cough and dyspnoea, which responded to dexamethasone. By discharge on day 18, she was feeling well with normal blood counts except for platelets at 60 · 10 ⁄ l; by day 35, the platelet count was normal. On day 58, bone marrow showed morphological remission, with normal cytogenetics and negative RT-PCR. Further RT-PCR analyses on peripheral blood have remained negative, most recently performed in July 2001. The patient remains alive and well with normal blood counts as of September 2001, 4 years after diagnosis. The mechanism by which our patients achieved prolonged complete remissions is unclear. It is unlikely that we observed spontaneous remissions as they are rare and of short duration (Enck, 1985). It is likely that ATRA was important in inducing the remissions. ATRA in supraphysiological doses causes differentiation of the promyelocyte by dissociation of a transcriptional corepressor complex, that exerts a dominant-negative effect on normal RAR-a-regulated gene transcription (Melnick & Licht, 1999). Continuous treatment leads to decreased plasma concentration and clinical resistance. However, leukaemic cells from relapses can continue to differentiate in vitro, though not in vivo. Moreover, the efficacy of liposomal ATRA suggests that the intracellular concentration of ATRA is critical, as use of this formulation maintains plasma concentrations of tretinoin and response in relapsed patients, but the response does not correlate with plasma concentration (Estey et al, 1996). Furthermore, the use of liposomal ATRA alone can induce prolonged molecular remissions (Estey et al, 1999). Thus, maintaining adequate intracellular concentrations of retinoids in the relevant target cell may permit ongoing differentiation. The prolonged complete responses in our patients indicate that the metabolism of ATRA in some patients may be endogenously altered to permit persistent and clinically effective intracellular concentrations of the ligand. Alternatively, the unusual response to ATRA may indicate that the disease pathogenesis in our patients may have differed from the norm, as it is clear from animal models that the translocation is necessary but not sufficient to produce leukaemia, and that other cytogenetically silent lesions must contribute (Zimonjic et al, 2000). The mechanisms mediating induced differentiation by ATRA have been intensively investigated (Melnick & Licht, 1999), but only recently has the means of cell death been elucidated (Altucci et al, 2001). It is conceivable that differing clinical outcomes to ATRA treatment may be due to intrinsic variations in leukaemic stem cells in their susceptibility to retinoidinduced, TRAIL ⁄DR5-mediated paracrine apoptosis. British Journal of Haematology, 2002, 117, 768–774