Klebsiella pnuemoniae liver abscess following screening colonoscopy: a case report

A 72-year-old Sri Lankan man presented to the emergency department on 24 May 2025 with a 2-day history of increasing lethargy, general weakness and right upper quadrant abdominal pain. Symptoms began 3 days after a screening colonoscopy performed on 19 May 2025, prompted by a positive faecal immunochemical test (FIT) as part of the New Zealand bowel screening programme. The procedure was uneventful, revealing only small-mouthed diverticula and haemorrhoids, and no mucosal interventions such as biopsies or polypectomies were performed.

His past medical history included ischaemic cardiomyopathy with a left ventricular ejection fraction of 24%, previous coronary artery bypass grafting, inferior ST elevation myocardial infarction, peripheral vascular disease, type 2 diabetes mellitus and stable Parkinson’s disease.

View Figure 1–2.

On examination, he was septic with a fever (38.2°C) and mild right upper quadrant tenderness. Laboratory results showed a C-reactive protein (CRP) of 263mg/L, white cell count of 14.37×10^9/L, and normal liver function tests. A computed tomography (CT) scan revealed a large hepatic abscess in the right lobe, measuring 6.6x6.2x8.8cm with no evidence of colitis, perforation, diverticulitis or any other potential sources of infection (Figure 1). Hydatid and amoeba serology were negative, and blood cultures showed no growth. Ultrasound-guided aspiration of the hepatic abscess on 27 May 2025 yielded purulent material positive for Klebsiella pneumoniae (Figure 2).

The patient was admitted to the intensive care unit for cardiovascular support. He received broad-spectrum antibiotics (cefuroxime, metronidazole and gentamicin), later rationalised to a narrower spectrum antibiotic based on sensitivities. Following ultrasound-guided drainage of the abscess and antibiotic therapy, his condition improved with resolution of inflammatory markers, sepsis and fever.

Discussion

Population-based studies from North America have reported the incidence of pyogenic liver abscess (PLA) to be approximately 2.6 to 3.6 cases per 100,000 population. PLA remains a potentially serious condition, with in-hospital mortality rates ranging from 5.6 to 10%.^1,2 Klebsiella pneumoniae has been identified as the causative organism in about 27% of cases, and other commonly isolated pathogens include the Streptococcus milleri group (44%), Escherichia coli (16%) and anaerobic bacteria (20%).² While various underlying causes of PLA have been described in the literature—including diverticulitis³—approximately 56% of cases are classified as cryptogenic, with no identifiable cause.²

Risk factors for PLA include diabetes mellitus, immunosuppression, malignancies and Asian ethnicity—potentially related to the higher prevalence of hypervirulent Klebsiella pneumoniae K1 and K2 serotypes in this population.^4,5,6,7 In our patient, the combination of Asian ethnicity and underlying comorbidities—particularly diabetes mellitus—may have contributed to the development of PLA. Additionally, conditions such as end-stage renal disease, biliary tract infections, liver cirrhosis, gastrointestinal malignancies, appendicitis, diverticulitis and recent endoscopic retrograde cholangiopancreatography (ERCP) have all been implicated with increased risk of PLA.⁴

While the literature remains limited, there is growing recognition of a possible link between colonoscopies and PLA, particularly in high-risk patients. Although a colonoscopy is generally considered a safe procedure, several case reports have described PLA developing shortly after the procedure, typically in the setting of mucosal interventions.^5,8–11 A recent study also found that upper gastrointestinal endoscopy was significantly associated with subsequent PLA development, whereas lower gastrointestinal endoscopy was not.⁴ The true incidence of PLA following a colonoscopy remains undefined but appears to be extremely low, with most evidence arising from isolated case reports. Colonoscopy-specific risk factors include invasive mucosal interventions such as polypectomy or biopsy, which were absent here, suggesting that even non-invasive procedures may pose a risk in susceptible individuals. Notably, the clinical timeline observed in our patient—onset of symptoms within 3 to 5 days—mirrors other reported cases, where symptoms typically emerged within 4 to 7 days post-procedure.

The biological plausibility lies in the translocation of gut flora, including Klebsiella pneumoniae, across the colonic mucosa due to microtrauma, even without invasive interventions such as polypectomies.^11,12 In this patient, the presence of diverticula could have provided a nidus for bacterial translocation. Klebsiella pneumoniae, a gram-negative bacillus colonising the gastrointestinal tract, can then hematogenously spread to the liver via the portal vein, leading to abscess formation. While the temporal proximity suggests a potential link, the absence of mucosal disruption makes coincidental occurrence a plausible alternative, and our report aims to highlight this for clinical awareness rather than to imply causation.

Prevention strategies for post-colonoscopy infectious complications, including Klebsiella pneumoniae liver abscesses, remain limited and inadequately defined. Current guidelines from the American Society for Gastrointestinal Endoscopy (ASGE) do not recommend routine antibiotic prophylaxis for colonoscopy due to the low risk of infectious complications.¹³ The use of prophylactic antibiotics in high-risk individuals—such as those with diabetes, immunosuppression or undergoing complex endoscopic interventions—remains controversial, with no consensus guidelines currently in place. This case underscores the importance of vigilance for post-procedure infections in susceptible individuals, though the rarity precludes routine prophylactic measures. Alternative strategies warranting consideration include: risk stratification (based on factors such as diabetes status, ethnicity and colonic findings like diverticulosis) and optimising bowel preparation to reduce luminal bacterial load (particularly in patients with diverticular disease). Additionally, structured follow-up measures—such as post-procedure phone calls or routine blood tests in high-risk patients—could be explored as part of national screening programmes to enable earlier detection and management of rare but serious events like PLA. Further research is needed to evaluate the utility and cost-effectiveness of such interventions.

Conclusion

In conclusion, we present a patient with a Klebsiella pneumoniae PLA, diagnosed 5 days after undergoing an uncomplicated screening colonoscopy. Even though causality cannot be established, this case potentially highlights Klebsiella pneumoniae PLA as a rare but serious complication of colonoscopies, even in the absence of invasive mucosal interventions. Clinicians should maintain a high index of suspicion in patients presenting with fever or abdominal pain post-colonoscopy, particularly those with risk factors such as diabetes. Future studies are needed to establish the incidence, explore risk stratification and preventive strategies to enhance the safety of colorectal screening programmes.