Bernard Soulier syndrome: a rare, frequently misdiagnosed and poorly managed bleeding disorder

Abstract

Bernard Soulier syndrome is a rare, congenital platelet bleeding disorder, with autosomal recessive inheritance. It is characterised by macrothrombocytopenia and platelet dysfunction, leading to mucocutaneous bleeding noted in early childhood. This entity poses an important diagnostic challenge, and blood smear and DNA sequencing are paramount for the correct diagnosis. Differential diagnosis includes May-Hegglin anomaly, Glanzmann Thrombasthenia and von Willebrand disease; it is also often misdiagnosed as idiopathic thrombocytopenic purpura. We report a 68-year-old man diagnosed with von Willebrand disease for three decades, admitted with gastrointestinal bleeding, anaemia and severe thrombocytopenia. Replacement with von Willebrand factor did not stop the haemorrhage, suggesting another aetiology for the bleeding disorder. Corticosteroids and intravenous immune globulin were also ineffective. Genetic sequencing showed a homozygous mutation in GP1BA gene, thus establishing the correct diagnosis.

Background

Bernard Soulier syndrome (BSS), first described in 1948 by Bernard and Soulier,¹ has an estimated prevalence of one in one million. It usually presents during childhood with epistaxis, gingival bleeding, spontaneous bruising and, less commonly, menorrhagia and gastrointestinal bleeding (GIB).^{2 3} Due to its rarity and non-specific spectrum of clinical manifestations, BSS is frequently reported as being misdiagnosed,^{2 4} often implicating unnecessary and unsuccessful treatment modalities, including splenectomy when idiopathic thrombocytopenic purpura (ITP) is assumed. A high index of suspicion is therefore imperative in order to avoid mismanagement of this giant platelet disorder.

Case presentation

We report a 68-year-old man with haemorrhagic blood dyscrasia noted at the age of 7, manifested by severe epistaxis and gingival bleeding after dental extractions. No history of first-degree consanguinity in his family was found. The patient has no siblings, and his two daughters, as well as his already deceased parents, had a negative history of easy bruising or any other manifestations suggestive of a bleeding disorder. Bleeding episodes persisted through adult life, and at the age of 37, the patient was referred to a haemotherapy department for aetiological investigation. At that time complete blood count (CBC) showed thrombocytopenia (baseline 30.000×10⁹/L), with giant platelets on peripheral smear. Bone marrow biopsy was normal and antiplatelet antibodies were negative. Repeated plasma levels of von Willebrand factor (VWF) activity were consistently within the normal range. Platelet aggregometry showed no response to ristocetin at different concentrations and normal aggregation with other agonists, namely adenosine 5′-diphosphate and collagen. A type 2B von Willebrand disease (T2BVWD) diagnosis was made based on these results; the patient maintained regular follow-up, with occasional transfusion support before minor interventions.

Thirty years after the diagnosis, he presented to the haemotherapy day hospital with a 2-week history of worsening exertional dyspnoea and melena. Initial CBC showed severe microcytic hypochromic anaemia (haemoglobin 5.4×10 g/L) and worsened thrombocytopenia (21.000×10⁹/L). Over the course of a month, he underwent multiple transfusions in the outpatient facility (total of 12 red blood cells and 2 platelet concentrates), as well as iron replacement (1500 mg of intravenous ferric carboxymaltose), with unsatisfactory increment on CBC. Furthermore, despite normal plasma levels of VWF antigen, the patient was also treated with factor replacement, considering the T2BVWD diagnosis, with no improvement. The persistence of severe refractory bicytopenia, with evidence of active GIB, led to admission in our intermediate care unit.

Investigations

On arrival, haemoglobin had lowered to a minimum of 4.2×10 g/L; haptoglobin and lactate dehydrogenase were normal, thus excluding a haemolytic component. Platelet manual count was also lower (8.000×10⁹/L). Coombs test was negative, as was serological testing for HIV and hepatitis B and C; leucocyte count, renal and hepatic function, fibrinogen level, prothrombin time and activated partial thromboplastin time were unremarkable. We initially administered prednisolone 1 mg/kg/day for 14 days, followed by intravenous immune globulin 1 g/kg/day for 2 days, with no effect on platelet count. We also insisted on VWF replacement, but were unsuccessful in controlling GIB. After 3 weeks of hospital admission, daily melena persisted despite tranexamic acid and octreotide perfusion trial, and frequent red blood cells transfusions were necessary to maintain haemoglobin above 7.0×10 g/L.

An extensive investigation was made in order to determine the source of GIB. Esophagogastroduodenoscopy showed no mucosal abnormalities, and total colonoscopy revealed partially digested blood on all segments, without active haemorrhage. Capsule endoscopy detected large clots and abundant fresh blood in the small intestine, starting at 2h07′, making it impossible to identify any bleeding lesions; CT angiography also failed to pinpoint a bleeding site. On day 30 the patient started high fever, rapidly progressing to septic shock, with neurological and cardiovascular dysfunction, requiring vasopressor support. Blood cultures isolated Enterococcus faecalis and Pseudomonas aeruginosa, most likely due to intestinal translocation, and he was treated with a 14-day course of piperacillin/tazobactam. Finally, on day 55, after raising the platelet count to 55.000×10⁹/L with four consecutive platelet concentrates, the patient underwent laparotomy with transverse enterotomy. Multiple punctiform clots were identified, as well as old blood extending from distal jejunum to the ileocecal valve, consistent with a non-active haemorrhagic enteropathy; no vascular malformations or other mucosal abnormalities were found. Given the absence of active bleeding and the extent of small intestine involved, segmental resection was not feasible. We maintained multiple daily platelet transfusions for 6 days after surgery, in order to keep counts above 50.000×10⁹/L. With this aggressive transfusion support GIB resolved, without additional need for red blood cells concentrates.

Differential diagnosis

T2BVWD is characterised by a qualitative defect in VWF that causes its enhanced binding to glycoprotein Ib platelet receptor.⁵ VWF antigen level may be normal in this subtype of VWD, as noted with our patient. However, the lack of response to repeated VWF replacement made us doubt the accuracy of the diagnosis established three decades earlier, since the mainstay of T2BVWD treatment is replacement with functionally adequate factor concentrates. Furthermore, despite being a common feature in T2BVWD, thrombocytopenia is rarely severe, and its contribution to persistent or major bleeding is not certain.^{5 6} Finally, platelet aggregometry showed no response to ristocetin at different concentrations; one of the diagnostic features of T2BVWD is an increased VWF-platelet interaction in the presence of ristocetin at low concentrations,⁵ and thus this diagnostic hypothesis was discarded.

Given the persistent thrombocytopenia, with platelet count significantly lower than the patient’s known baseline, ITP was also considered, but the absence of response to corticosteroids and immune globulin made it less plausible.

The patient’s lifetime bleeding history strongly suggested a congenital disorder. After ruling out the initial diagnosis of T2BVWD, we ordered a thrombosis and haemostasis genetic test, and a novel missense, homozygous variant (c.578T>C, p.Leu193Pro) was identified in GP1BA gene. According to the American College of Medical Genetics and Genomics and the Association for Molecular Pathology criteria, in silico analyses combined with other evidence data (population, functional and segregation) led to classification of this variant as likely pathogenic. Despite being a previously not described mutation, homozygous variants in this genetic region (N-terminal extracellular domain containing leucine-rich repeats) are associated with BSS.⁷ Based on this molecular study, associated with the previous platelet function tests showing a typically impaired aggregation in response to ristocetin, and presence of characteristic macrothrombocytopenia on blood smear, we considered this variant as clinically significant, and a definitive diagnosis of BSS was established.

Treatment

Throughout the hospital stay the patient remained haemodynamically stable (excepting septic shock), and hence, despite active GIB with persistent melena, no intensive platelet transfusion strategy was initially engaged. Platelet concentrates were administered occasionally, just to maintain counts above 10.000×10⁹/L, but with no haemostatic effect. We started aggressive platelet transfusion in the perioperative setting, aiming to continuously raise counts above 50.000×10⁹/L, and only then we managed to control GIB.

Outcome and follow-up

No postoperative complications were registered. After the suspension of platelet transfusions, the patient did not have other episodes of melena. Platelet count gradually returned to baseline (30.000×10⁹/L), haemoglobin stabilised at 10.0×10 g/L, and the patient was discharged on day 77. Despite having a diagnosis of a bleeding disorder (although incorrect) for most of his adult life, he was still an active bricklayer, and so we strongly advised him not to engage in high-risk activities. After 7 months, he maintains regular follow-up at the haemotherapy department, with no other haemorrhagic complications registered nor need for additional transfusion support.

Family studies showed that both his daughters are heterozygous for the same GP1BA variant. Their platelet count and size are unremarkable. Platelet function tests are also normal, in agreement with the autosomal recessive character of BSS, which explains their negative history for increased bleeding tendency.

Discussion

BSS is caused by a defect in, or absence of platelet membrane glycoprotein (GP)Ib-IX-V complex,^{2 3} the receptor for VWF, resulting in an adhesion dysfunction. The receptor’s quantitative and qualitative deficiency is attributable to multiple mutations in GP1BA, GP1BB and GP9 genes, which encode respectively for GPIbα, GPIbβ and GPIX subunits of the GPIb-IX-V complex.⁷ Furthermore, some reports have identified an association between variants in GP1BA and GP1BB and autosomal dominant macrothrombocytopenia⁸; this association could explain the typical morphological abnormalities of BSS.

BSS is extremely rare, with an estimated prevalence of one in one million. It usually presents as a paediatric bleeding disorder, most commonly with epistaxis, easy bruising and gingival bleeding. Its rarity, associated with non-specific clinical manifestations, often results in patients misdiagnosed as having other more frequent bleeding disorders, with consequent unnecessary and unsuccessful treatments. Our patient was initially diagnosed with T2BVWD, and as such received VWF replacement whenever a minor intervention was scheduled. On hospital admission we insisted on factor replacement, deferring aggressive platelet transfusion since he didn’t present with catastrophic haemorrhage. BSS and T2BVWD share indeed similar clinical features, as well as some laboratory findings, namely macrothrombocytopenia. However, as mentioned before, platelet aggregation in response to ristocetin is absent in the former and increased in the latter. Due to macrothrombocytopenia, many patients with BSS have been reported to be diagnosed with ITP,^{1 7} leading to treatment with corticosteroids, immune globulin and frequently, splenectomy. In fact, we also seriously pondered splenectomy, but had to delay the procedure on account of the septic complication; the patient, like others in the reports above, would have been submitted to an unnecessary and potentially harmful surgical intervention.

The lack of response to various therapeutic options, and the absence of a correct diagnosis, made us order a thrombosis and haemostasis genetic test. A previously not described variant in GP1BA was identified, ultimately establishing the BSS diagnosis. This entity typically poses a difficult workup challenge, since the combination of macrothrombocytopenia and response pattern to ristocetin is not pathognomonic. Flow cytometry is an accessible confirmatory test,^9–11 since it is able to detect deficiency of antigens CD42a (GPIX) and CD42b (GPIbα) on platelet surface. Genetic testing confirms the diagnosis⁷ and is useful for genetic counselling of family members, but may not be widely available. To date, more than 100 different variants in GP1BA, GP1BB and GP9 have been identified and associated with BSS.⁷ Genetic and molecular knowledge is still evolving, and it is likely that other variants are yet to be described, as verified in this report.

There is no specific treatment available for BSS, and management is based exclusively on supportive measures.^{2 9} If severe or uncontrolled bleeding occurs, platelet concentrates transfusion is first-line therapy. We tried intravenous antifibrinolytic therapy, as well as octreotide perfusion, with no success in controlling the GIB; multiple platelet transfusions, in the perioperative setting, was the only measure able to obtain haemostasis. However, the decision on platelet transfusion support should be taken with caution, since it is not a risk-free treatment. The major concern with repetitive platelet transfusion is platelet alloimmunisation, with development of anti-human leucocyte antigen (HLA) antibodies. Whenever possible, this cellular blood product should be leucocyte-reduced, as this measure has been proven to significantly reduce the incidence of HLA alloimmunisation.¹² Additionally, patients with BSS may have disease-specific immunisation concerns, besides HLA alloantibodies. Since patients with BSS often lack platelet membrane GPIb-IX-V receptor complex, transfusion of normal platelets that have this surface glycoprotein can potentially lead to development of anti-GPIb-IX antibodies.¹³ Alloimmunisation is an important cause of platelet transfusion refractoriness,¹⁴ and it must be promptly recognised, since alternative strategies are necessary when managing bleeding episodes. Antifibrinolytics may be useful in mild to moderate mucocutaneous bleeds.¹⁵ Although some individual reports may suggest a role for desmopressin in BSS, there are currently no clinical trials that can establish its effectiveness. Recombinant activated factor VII has been approved for treatment of bleeding episodes in Glanzmann Thrombasthenia, and although not yet formally indicated, some case reports have suggested its efficacy also in BSS, and treatment should be considered in patients with platelet refractoriness.^{13 15} Besides immunisation, other complications to be considered with platelet transfusions¹⁶ include allergic reactions, circulatory overload, transfusion-related acute lung injury and transfusion-associated graft-versus-host disease. Transfusion-transmitted infections, particularly bacterial, are also a potential risk, since the relatively high temperature of platelet storage (22°C–24°C) may facilitate bacterial proliferation.

Given the rarity of this disease, there are currently no guidelines nor quality evidence for the management of prophylactic platelet transfusions prior to surgical interventions in BSS. Furthermore, correlation between platelet count and haemorrhagic risk is difficult to establish, since patients with BSS also have a platelet adhesion dysfunction, in association with thrombocytopenia. Perioperative strategy should be multidisciplinary and tailored to each particular case. Transfusion history and previous documentation of immunisation/platelet refractoriness are determinant in selecting the treatment that better fits the patient’s profile. We followed the generally accepted recommendation¹⁷ of a minimum threshold of 50.000×10⁹/L platelets for any major intervention, other than neurosurgery or surgery involving the posterior segment of the eye, which requires a higher threshold of 100.000×10⁹/L.

With this report, we sought to describe a particularly difficult diagnostic workup, typical of this rare giant platelet disorder, and stress the utmost importance of a correct diagnosis, in order to avoid unnecessary, ineffective, and potentially harmful treatments. Although individual BSS case reports are scarce, misdiagnoses (particularly as ITP,¹⁸ but also to a lesser extent as other bleeding disorders) are undeniably frequent. Furthermore, the proportion of misdiagnoses not recognised posteriorly (due to incomplete investigation or low physician awareness) is probably significant, meaning the prevalence of patients with BSS wrongly diagnosed is not negligible. A high suspicion is paramount, and consideration of alternative diagnoses is mandatory in the setting of lack of response to treatment directed to more common bleeding disorders.

Finally, another prominent feature of this report was the uncharacteristic diagnosis of BSS in late adulthood. BSS is typically considered a paediatric bleeding disorder, since most diagnoses are made at birth or during childhood. Nonetheless, as exemplified with our patient, some cases may remain undetected or misdiagnosed for a relatively long time. Therefore, BSS should be considered in the differential diagnosis of adults with long-term, unspecified bleeding diatheses.

Patient’s perspective.

It is somewhat frightening to think that for all these years, if I were to have a serious bleeding, I could potentially not receive the adequate type of treatment. It has been a long and harsh hospital stay, but I’m glad my illness could be correctly diagnosed.

Learning points.

• Bernard Soulier syndrome (BSS) is frequently misdiagnosed as idiopathic thrombocytopenic purpura and other more common bleeding disorders, resulting in unnecessary and ineffective therapies; correct identification of this entity is imperative, since transfusion with platelets with normal membrane glycoprotein (GP)Ib-IX-V receptor complex is the only effective treatment for severe bleeding episodes.

• Platelet concentrates should however be reserved for patients undergoing major surgery or with life-threatening and uncontrolled haemorrhage, as repetitive platelet transfusions can potentially lead to alloimmunisation and platelet refractoriness.

• Despite being considered a paediatric disorder, typically diagnosed at birth or during early childhood, BSS should also be considered in adults with bleeding tendency of undetermined nature.

• Flow cytometry is an easily accessible diagnostic test, able to confirm the absence or reduced expression of platelet surface CD42a (GPIX) and CD42b (GPIbα).

• Genetic counselling is recommended for patients with BSS. Likewise, family members should be offered function and genetic testing in order to determine carrier status.

Ethics statements

Patient consent for publication

Obtained.