Thrombotic risk in patients with immune haemolytic anaemia.

Abstract

We read with interest the paper of L'Acqua and Hod (2015) on thrombotic complications of haemolysis. Thromboembolic complications are frequently associated with haemolysis, but the thrombotic risk in autoimmune haemolytic anaemia (AIHA) is possibly underestimated. Recent guidelines do not suggest any prophylactic antithrombotic management in these patients (Crowther et al, 2011). Older literature includes only sparse reports on thrombotic events in AIHA. Crosby and Rappaport (1957) described a retrospective survey of 34 patients with primary AIHA, followed for 10 years, reporting 4 venous thrombosis, 2 of which were autopsy findings. Dausset and Colombani (1959) describe a retrospective study of 105 patients with primary warm AIHA and cold haemolytic anaemia, followed for 10 years, with 2 cases of deep venous thrombosis reported. To further assess the potential risk of thrombosis in AIHA, we carried out a literature search from 1964 to today. Medline bibliographic database was searched (search strategy: immune[All Fields] AND (“haemolytic anaemia”[All Fields] OR “anemia, hemolytic”[MeSH Terms] OR (“anemia”[All Fields] AND “hemolytic”[All Fields]) OR “hemolytic anemia”[All Fields] OR (“hemolytic”[All Fields] AND “anemia”[All Fields])) AND (“thrombosis”[MeSH Terms] OR “thrombosis”[All Fields])). Ninety-two papers were retrieved and the abstracts evaluated using the following inclusion criteria: studies on humans, English language, cohorts with more than 10 patients, reports of vascular events not reported as autopsy finding. Thirteen papers were evaluated and 6 studies were retained for analysis (Pullarkat et al, 2002; Hendrick, 2003; Bongarzoni et al, 2005; Barcellini et al, 2012; Birgens et al, 2013; Barcellini et al, 2014). A total of 60 thrombotic episodes were collected in 474 patients (Table 1); 44 events were venous and 16 arterial. The rate ratio of vascular complications in the considered patients was 54/474 [11·3%, 95% confidence interval (CI): 8·5–14·8%]. The largest reported cohort (Barcellini et al, 2014) described 33 patients with thrombotic events. These events were associated with a severe onset (79% of cases) in patients with haemoglobin level ≤80 g/l, P = 0·024, a higher median lactate dehydrogenase level (837 u/l vs. 750 u/l, P = 0·006) and splenectomy (8/33, 24% vs. 24/274, 8·8%, P = 0·014). The study did not evaluate a possible relationship between thrombotic events with the time elapsed from diagnosis or with remission/relapse in splenectomized patients. Detection of lupus anticoagulant was associated with an increased thrombotic risk in one study (relative risk: 7·50, 95% CI 1·25–45·2, P = 0·03; Pullarkat et al, 2002), but this finding was not subsequently confirmed (Bongarzoni et al, 2005; Barcellini et al, 2014). WAIHA CHAD (primary and secondary) WAIHA CHAD (primary) WAIHA CHAD (primary) WAIHA (primary) WAIHA CHAD (primary) Our literature review, therefore, suggests that patients with AIHA may present an hypercoagulable state, with a clinically significant rate of both venous and arterial vascular events. Of course, only multicentre cohort studies can estimate the precise thrombotic risk in patients with AIHA and the associated risk factors, such as age at diagnosis, concomitant lupus anticoagulant and/or antiphospholipid antibodies, severity of disease, previous splenectomy, immune-suppressive therapy, presence of cancer or other co-morbidities. Those AIHA patients who may eventually benefit from antithrombotic prophylaxis remains unresolved, together with the optimal duration of the anticoagulation after a thrombotic event. Despite all these uncertainties, clinicians should be aware of a thrombotic risk in AIHA patients and consider antithrombotic prophylaxis when concomitant risk factors, such as hospitalization, severe onset of disease or splenectomy, are present. Heparin or mechanical prophylaxis could be reasonable choices depending on the presence of bleeding risk factors. Ad hoc clinical prospective trials are required to demonstrate the cost-effectiveness of a prophylactic antithrombotic approach in all or selected patients with AIHA. MR designed the study, collected and interpreted the data and wrote the manuscript; FR wrote the manuscript and provided the final approval. The authors declare no potential conflicts of interest.