In this study we analyzed the source, microbiological profile, and impact of infections among a large cohort of 150 patients with AP who developed fever during acute pancreatitis. We characterized the temporal patterns, distinctive features, microbiological profiles, and outcomes associated with infectious complications in patients with AP.
A large proportion of our patients had extra-pancreatic infections. This aligns with previous reported literature on infections in AP. [18,19] Among these extra-pancreatic infections, lower respiratory tract infections were most common followed by cholangitis and bacteremia.
The occurrence of definite infected pancreatic necrosis is lower in our study as compared to the previously reported literature. [20] Notably, we did not perform per protocol FNAC for suspected IPN. Performing FNAC for suspected infected pancreatic necrosis has been a matter of debate. Studies have shown that infected necrosis can be reliably diagnosed on basis of clinical signs. [13] At our center we do not routinely perform FNAC for pancreatitis and 24 patients received antibiotics on the clinical suspicion of infected necrosis.
Of the 54 patients with culture-confirmed infections, 36 (66.7%) had infections caused by multidrug-resistant (MDR) organisms. Among patients with infected pancreatic necrosis who had positive cultures, 10 out of 13 (77.9%) grew MDR organisms. These findings highlight the high prevalence of MDR infections, especially in infected pancreatic necrosis. Previous studies also reported a high incidence of MDR infections in AP. [12,20,21] This highlights the importance of local antibiograms in guiding empiric antibiotic therapy for suspected infections. Patients with MDR infections had a considerably higher occurrence of serious adverse events (SAE) compared to patients without MDR infections, aligning with previous studies. [12,22,23] The growing prevalence of MDR infections in pancreatitis is a matter of concern. The paucity of effective antimicrobial therapy for treatment of MDR infections poses a significant challenge. A small proportion of patients developed fungal infections. Patients with fungal infections had a significantly longer median duration of antibiotic use and most of them were admitted in ICU. This highlights the risk of developing fungal infections as a consequence of prolonged antibiotic exposure, which is a known risk factor for fungal infections. [24]
AP can be divided into two phases. In the initial 1-2 weeks, AP triggers a pro-inflammatory response, manifesting as SIRS. After the first 1-2 weeks, there is a shift from this pro-inflammatory state to an anti-inflammatory phase. It is during this second phase that the patient becomes vulnerable to the development of various infectious complications. [10] The current guidelines do not recommend routine antimicrobial prophylaxis in AP. However, our study and the study by Besselink et al highlights that ~ 50% of patients are at risk of infections in the first week of illness. [3] The challenge in the first week of illness is to discriminate fever secondary to sterile inflammation from fever secondary to infections. Cultures take time to become positive, and that wait may contribute to a poor outcome . Routine FNAC of suspected infected necrosis has its own challenges. Therefore, there is a need for surrogate parameters which can discriminate infection from inflammation in the first week of illness. In this study we explored two such parameters, NLR and procalcitonin.
The NLR has been studied in AP patients as a predictor of severity and adverse outcomes. [25-30] In our study, the NLR showed good accuracy in discriminating infections from sterile inflammation, with an AUROC of 0.70. At a cut-off of 6.2, the NLR had 77% sensitivity and 58% specificity for predicting infections. Further NLR >9 exhibited a specificity exceeding 80% in predicting the presence of infections. Given the high specificity associated with an NLR above 9, it would be reasonable to initiate antibiotic treatment promptly after obtaining blood cultures, since the false positive rate of approximately 20% is deemed acceptable. On the other hand, procalcitonin did not appear to be a good discriminant between infection and inflammation in the first week. Our study suggests that NLR could be a useful biomarker in differentiating infections from sterile inflammation in the first week of illness in AP. Analogous to our study, NLR has been employed as a discriminatory tool in patients with COVID-19, to differentiate fever driven by the SIRS from that caused by bacterial infections. [31] Moreover, in a study involving patients on haemodialysis, NLR in conjunction with C-reactive protein (CRP) levels was used to distinguish inflammation of infectious origin from non-infectious aetiology. [32] In another study, NLR was used to predict bacterial infections in patients with acutely decompensated cirrhosis. [4] NLR is a cost and time effective and easy to use daily practice biomarker.
The strength of our study is the large sample size, strict inclusion criteria, comprehensive microbiological data and quantitative analysis of NLR as a predictor of infections. To the best of our knowledge, ours is the first study to examine the ability of NLR in differentiating infections and sterile inflammation in AP.
We acknowledge certain limitations of this study including its single centre study design, and lack of post discharge outcome information.
In conclusion, this study focuses on the early occurrence of infections in AP and provides novel insights on using the NLR to distinguish between infectious and non-infectious fevers in AP. Additionally, it emphasizes the changing microbial profile, which has consequences for choosing antibiotics based on experience. Further multi-centric studies are necessary to confirm the diagnostic accuracy of the NLR and investigate additional biomarkers.