HCP Live
Contagion LiveCGT LiveNeurology LiveHCP LiveOncology LiveContemporary PediatricsContemporary OBGYNEndocrinology NetworkPractical CardiologyRheumatology Netowrk

Endocarditis and splenic abscess

Surgical Rounds®, July 2006, Volume 0, Issue 0

Erica Giblin, General Surgery Resident, Department of Surgery Michael Ganey, Clinical Fellow in Surgery, Department of Surgery, Baystate Medical Center, Springfield, MA

Erica Giblin, MD General Surgery Resident

Department of Surgery Michael Ganey, MD Clinical Fellow in Surgery

Department of Surgery

Baystate Medical Center

Springfield, MA

Infective endocarditis is associated with systemic embolization in 22% to 50% of cases.1 Splenic abscess is a documented complication of embolization from infective endocarditis. It is a rare condition, with fewer than 600 cases reported in the literature.2 While rarely seen, it is associated with a high mortality rate if left untreated. With appropriate management, the mortality rate is reduced to 14%.1 Patient history is a key factor in diagnosis because intravenous (IV) drug abuse poses a significant risk in both the development of infective endocarditis and splenic abscess.3 Therefore, the diagnosis of splenic abscess should be considered in patients who have a history of IV drug abuse and present with abdominal pain and fever. We report the case of a patient with a history of IV drug abuse who presented with abdominal pain, fever, and an elevated white blood cell (WBC) count. The patient was found to have a splenic abscess secondary to embolization from endocarditis.

Case report

A 40-year-old man with a history significant for IV heroin abuse and hepatitis C came to the emergency department because of left-sided chest pain and dyspnea. He was febrile to 103°F and had a WBC count of 27,000/µL. A chest radiograph showed left lower-lobe pneumonia (Figure 1). The patient was discharged from the emergency department but returned later that evening with increasing dyspnea and left-sided chest pain radiating down into his left upper abdomen. A second chest radiograph demonstrated no change in the left lower-lobe pneumonia. The patient underwent a contrast-enhanced computed tomography (CT) scan of the abdomen and pelvis, which showed an enlarged spleen (15 cm) with a large, irregularly marginated area of increased attenuation and a rim of hypoattenuation (Figure 2). Inflammatory changes were noted around the inferior margin of the spleen and around the adjacent perisplenic fascia.

Streptococcus viridans

The patient was admitted to the hospital, IV broad-spectrum antibiotics were initiated, and a hematology consultation was obtained. CT angiography was performed, which confirmed the diagnosis of splenic infarction. The splenic vessels were patent to the hilum without any thrombus, embolus, or evidence of volvulus (Figure 3). The patient was observed for 48 hours while the hematology workup was pending. Subsequent blood cultures revealed . These findings, along with the patient’s history of IV drug abuse, suggested endocarditis as an etiology. The patient underwent a transthoracic echocardiogram, which demonstrated a perturbing, mobile left atrial mass.

The patient’s WBC count remained elevated at around 30,000/µL despite appropriate antibiotic coverage. A contrast-enhanced CT scan of the abdomen was repeated, which demonstrated interval splenic enlargement with gas evident within the spleen (Figure 4). He was brought to the operating room and a transesophageal echocardiogram (TEE) was performed, which showed no evidence of a left atrial mass. There was a mobile, echogenic density located below the septal leaflet of the tricuspid valve in the region of the membranous intraventricular septum that measured less than 1 cm in length.

After the intraoperative TEE, the patient underwent a splenectomy. A left oblique subcostal incision was used, and upon entering the intra-abdominal cavity, the posterior and lateral attachments of the spleen were taken down, bringing the spleen up into the operating field. The spleen was necrotic and contained multiple foci of yellow and green abscesses over its contour. The surrounding splenic capsule was thickened and inflamed. The spleen was transected at the hilum, and the splenic vessels were ligated with 2-0 absorbable sutures. The splenic vessels did not appear to be patent and were nonpulsatile. Several bleeding points from the short gastric vessels were ligated with two figure-of-eight sutures. The intra-abdominal cavity was copiously irrigated, and a 10-mm, flat Jackson-Pratt drain was placed in the left upper quadrant.

Postoperatively, the patient did well initially and his fever and WBC count decreased; however, on postoperative day 5, his WBC count increased to 35,000/µL. A CT scan of his abdomen showed a 4.6 x 9.5 x 8.0-cm loculated fluid collection in the left upper quadrant (Figure 5). The tip of the surgically placed Jackson-Pratt drain was just outside of this collec-tion. An interventional radiologically guided drain was placed and withdrew 50 cubic centimeters of purulent material, which on culture grew no organisms. Over the next several days, the patient had increasing amounts of serous drainage from his Jackson-Pratt drain. The drained fluid had amylase levels of 69,741 U/L. The patient was told to take nothing by mouth, total parenteral nutrition was initiated, and octreotide was started. Over the course of 1 week, the drainage output slowed dramatically. The octreotide was stopped, and the patient’s diet was steadily advanced. The Jackson-Pratt drain was removed on postoperative day 18. The patient was discharged to home with only the interventionally placed drainage tube remaining. He was also given a 7-day course of oral clindamycin. At 1-week follow-up, the drainage tube was removed.


Splenic abscesses are uncommon, with autopsy reports documenting an incidence between 0.14% and 0.70%.4 Five distinct causative factors have been identified: metastatic hematogenous infection, hemoglobinopathy, immunosuppression, contiguous infection, and trauma.3

Splenic abscesses are typically caused by a primary focus of infection, with endocarditis being a classic example. Endocarditis has a 10% to 20% incidence of associated splenic abscess.5 Patients with endocarditis develop splenic abscesses secondary to bacteremia-induced seeding of splenic tissue or through the development of sterile and septic microemboli.3 Most emboli present 2 to 4 weeks after antimicrobial therapy has been initiated.





Gram-positive aerobic bacteria such as (16%–20%), (6%–22%), and (11%–16%) are the most common isolates from splenic abscesses.3 Initial antibiotic coverage should be broad because up to 36% of abscesses are polymicrobial.3



With the spread of human immunodeficiency virus (HIV), the number of immunosuppressed patients presenting with a splenic abscess has increased since the 1980s. Fungal organisms such as mycobacteria, , and cause 8% of cases in immunosuppressed individuals.5 In addition to HIV infection, patient demographics involving IV drug abuse, chemotherapy, and transplantation have altered the management and clinical setting of splenic abscesses. IV drug abuse has been well documented as a route of infection. The first citation regarding IV drug abuse as an inciting factor of splenic abscess was by Fry and colleagues in 1978.6 In a more recent series of 181 patients with splenic abscesses followed by Phillips and associates, 29% of patients had a history of concomitant IV drug abuse.3 IV drugs act as a conduit for bacteremia and resultant hematologic seeding to the spleen; however, most of these patients present with simultaneous endocarditis at the time of diagnosis.

Splenic abscesses can be difficult to diagnose because symptoms are often nonspecific. The triad of fever (90%), leukocytosis (88%), and left upper quadrant pain (56%) is the most common indication.7 The best radiologic modalities for detecting splenic abscesses are contrast-enhanced abdominal CT scanning and magnetic resonance imaging (MRI). Both of these imaging studies have 90% to 95% sensitivity and specificity.4

Recurrent positive blood cultures and evidence of ongoing sepsis or enlargement of splenic defects on CT scanning or MRI are suggestive of splenic abscesses. It has been shown that splenic abscesses respond poorly to antibiotics alone. The mortality rate is as high as 50% when antimicrobial therapy is the only treatment used for a pyogenic splenic abscess.8

Fungal splenic abscesses respond much more favorably to antifungal therapy. This is because these abscesses are a resultant manifestation of a disseminated infection.7 The mortality rate with antifungal therapy alone is 6.2%.4 Percutaneous drainage of a splenic abscess has been performed successfully and may be an alternative to splenectomy in patients who are poor surgical candidates.9 Ooi and colleagues demonstrated that CT-guided percutaneous drainage is amenable to unilocular abscess without septations or particularly viscous contents.9 Risk assessment is a crucial component of selecting patients for percutaneous drainage, as noted by Faught and associates.8 Contraindications to percutaneous drainage include uncorrectable coagulopathy; multiple large, septated abscesses; abscess rupture with bleeding; the need for surgical correction of associated intra-abdominal pathologies; diffuse ascites; or no safe route for drainage.8 Percutaneous drainage is an appealing option for patients who are unable to tolerate surgery or require temporary stabilization before surgical management. The failure rate of percutaneous drainage has been shown to be up to 60%, with these patients requiring splenectomy.2 The mortality rate of splenectomy after percutaneous drainage is around 14%, which does not differ appreciably from splenectomy as first-line therapy (17%).5

Open splenectomy for splenic abscesses has a mortality rate of 0% to 17% and morbidity rates of 28% to 43%.7 Splenectomy is the definitive procedure for most patients compared with medical management or percutaneous drainage. Laparoscopic splenectomy is an acceptable alternative to open splenectomy for splenic abscesses. Carbonell and colleagues demonstrated that although laparoscopic splenectomy was more difficult than the standard procedure secondary to the effects of perisplenitis, it is a safe treatment in experienced hands and associated with shorter hospital stays.2


Splenic abscess is a rare condition that should be considered in at-risk populations, such as those with a history of IV drug abuse and immunosuppressed individuals. In these susceptible patient populations, the hallmark presenting signs include fever, abdominal pain, and leukocytosis. CT scanning has made diagnosis increasingly straightforward. Patients with fungal pathogens may be amenable to antimicrobial therapy alone. Percutaneous drainage is an alternative to splenectomy in patients with a minimally loculated and liquified collection. For most patients, however, splenectomy remains the gold standard of treatment for splenic abscess.


1. Bayer AS, Bolger AF, Taubert KA, et al. Diagnosis and management of infective endocarditis and its complications. Circulation. 1998;98(25):2936-2948.

2. Carbonell AM, Kercher KW, Matthews BD, et al. Laparoscopic splenectomy for splenic abscess. Surg Laparosc Endosc Percutan Tech. 2004;14(5):289-291.

3. Phillips GS, Radosevich MD, Lipsett PA. Splenic abscess: another look at an old disease. Arch Surg. 1997;132(12): 1331-1336.

4. Nelken N, Ignatius J, Skinner M, et al. Changing clinical spectrum of splenic abscess. A multicenter study and review of the literature. Am J Surg. 1987;154(1):27-34.

5. Green BT. Splenic abscess: report of six cases and review of the literature. Am Surg. 2001;67(1):80-85.

6. Fry DE, Richardson JD, Flint LM. Occult splenic abscess: an unrecognized complication of heroin abuse. Surgery. 1978;84(5): 650-654.

7. Ooi LL, Leong SS. Splenic abscesses from 1987 to 1995. Am J Surg. 1997;174(1):87-93.

8. Faught WE, Gilbertson JJ, Nelson EW. Splenic abscess: presentation, treatment options, and results. Am J Surg. 1989;158(6): 612-614.

9. Ooi LL, Nambiar R, Rauff A, et al. Splenic abscess. Aust N Z J Surg. 1992;62(10):780-784.