Splenic rupture was first reported as a complication of bacterial endocarditis in 1919.1 The possibility of splenic rupture should be evaluated in patients with prior or active endocarditis presenting with signs and symptoms of hypovolaemic shock. We describe a case of splenic rupture occurring as a complication of subacute bacterial endocarditis.

A 42-year-old male presented with erythema, oedema, and bilateral leg pain three weeks after aortic valve replacement for subacute bacterial endocarditis (Streptococcus viridans). History was significant for intravenous drug use and aortic valve stenosis/insufficiency. Examination was significant for a mechanical S2 heart sound with 3/6 systolic murmur and a blanching petechial rash with pitting oedema of both legs. There was no abdominal pain or distension, palpable splenomegaly, or other stigmata of endocarditis. The white cell count was 8300/mm3 without bands; the haematocrit was 30.0%. The international normalised ratio (INR) on warfarin 5 mg/day was 3.2. A transoesophageal echocardiogram revealed thin stranding on the prosthetic valve suggestive of early vegetation, possibly causing embolisation to the legs resulting in his symptoms at presentation to the hospital. However, Doppler ultrasound of the legs failed to demonstrate deep venous thrombosis.

He was admitted for two weeks’ intravenous antibiotics, although blood cultures subsequently showed no growth. He was continued on the same dose of warfarin. The hospital course was uneventful until the tenth morning, when he complained of back and abdominal pain and a feeling of bloating. Examination found him to be diaphoretic, hypotensive, tachypneic, hypothermic, tachycardic, and hypoxaemic, but no abdominal tenderness was elicted or splenomegaly appreciated. Two haematocrit values were 20.1 and 18.2% and the INR 2.4. Nasogastric aspiration and stool haemoccult testing were negative for blood.

Several hours after resuscitative efforts began, progressive abdominal distension prompted paracentesis and a peritoneal lavage, both returning dark, bloody fluid. Interpretation of abdominal computed tomography (CT) was that the spleen was small but normal, and a large amount of intraperitoneal blood attributed to a ruptured viscus was present. Throughout resuscitative efforts, the patient was given packed red cells, platelets, fresh blood plasma, fluids, cryoprecipitate, vitamin K, and vasoconstricting agents to stabilise his vital signs and correct his coagulopathic state prior to undergoing exploratory laparotomy, but he developed disseminated intravascular coagulation and mutli-organ failure. Resuscitation was discontinued after 20 hours by family request.

On autopsy, there was approximately 4 l of blood in the peritoneal cavity. The spleen weighed 1100 g with multiple capsular disruptions extending into the parenchyma (Figure 1). Infarcts were visualised underneath intact portions of the splenic capsule. Serial macroscopic sections revealed multiple infarctions and cystic areas with evidence of recent haemorrhage (Figure 2). Microscopic sections contained multiple areas of infarction with haemorrhage but no other abnormality. Splenic rupture was attributed to the development of endocarditis-related infarcts that subsequently haemorrhaged. Retrospectively, the CT images of the spleen demonstrated subtle irregularities that could have represented small areas of rupture with haemorrhage.

Physiological trauma (eg, severe retching) can sometimes injure the spleen, but this mechanism was discounted in light of the autopsy findings. There was no history of recent, blunt abdominothoracic trauma, a frequent cause of splenic rupture.

Anticoagulation with warfarin undoubtedly complicated resuscitative efforts. Splenic rupture has been reported in patients being anticoagulated with warfarin or heparin,2–4 and this complication has been reported in patients undergoing thrombolytic therapy.4

A recent review5 and retrospective case series6 assessing endocarditis-related complications did not mention splenic rupture, but clinicians should have a high degree of suspicion in patients with a history of bacterial endocarditis presenting in shock and with evidence of intraperitoneal haemorrhage. Timely diagnosis and surgical management are crucial to reduce the mortality and morbidity of this complication.

Author information: James Bradley Summers, Physician; Joseph Kaminski, Physician; A Martin Chandler, Physician, Department of Internal Medicine, University of South Alabama, Mobile, AL, United States

Correspondence: Dr James Bradley Summers, Department of Internal Medicine, University of South Alabama, Mobile, AL 36617, USA. Fax: +1 251 471 7882; email: orotic2001@aol.com