Elderly individuals exhibit numerous distinct physical characteristics compared to their younger counterparts. Ageing is intricately linked to numerous alterations in organ functionality, encompassing an augmentation in frailty, a decrement in activities of daily living, a reduction in mobility, and cognitive decline (Vincent and Creteur, 2022). A significant disadvantage among the elderly population is the diminished barriers to pathogens, resulting in a heightened susceptibility to infection. The incidence of sepsis increases with advanced age. Furthermore, aging results in diverse physiological changes within the immune system, which are collectively referred to as immunosenescence, thereby increasing the susceptibility of elderly patients to fungal colonization and infections (Akinosoglou et al., 2023). Consequently, elderly patients constitute a distinct subgroup within the septic patient population, necessitating tailored management strategies.

Intra-abdominal candidiasis (IAC) poses a significant threat to life, with estimates primarily stemming from immunocompromised individuals. Despite its relatively low cumulative incidence of 1.84 cases per 1,000 intensive care unit (ICU) admissions (Bassetti et al., 2019), IAC is associated with a significant mortality rate of up to 60%, highlighting its severity and importance in critical care settings (Vergidis et al., 2016). However, there is a paucity of data specifically focusing on the elderly population, who are potential immunodeficient hosts with age-related immune dysregulation. Except for the established risk factors, the high rate of comorbidity, malnutrition, and polypharmacy, which are common in the elderly, could intervene the clinical management of IAC. The incidence and mortality of IAC increases with advanced age (Bassetti et al., 2015). However, the risk factors and techniques guiding the rapid recognition for IAC in elderly septic patients are lacking.

The diagnosis of IAC poses a significant challenge in elderly septic patients with intra-abdominal infection (IAI), often leading to excessive or delayed utilization of antifungal agents. Accurate diagnosis of IAC necessitates the isolation of Candida species via traditional mycological culture techniques, albeit with the potential for prolonged turnaround times. In consideration of the detrimental outcomes stemming from delayed treatment, antifungal agents are frequently administered prior to obtaining culture results, taking into account clinical risk factors and the comprehensive context of the patient’s condition. While research is ongoing for certain biomarkers and clinical risk factors, such as (1,3)-β-D-glucan (BDG), the currently available clinical scoring systems are insufficient in identifying patients with a high risk of IAC (Leroy et al., 2016; Bailly et al., 2017). Candida spp. can affect immune status by decreasing immunorecognition or overactivating the inflammatory response (Romani, 2011). T cells constitute a crucial component of the adaptive immune system, and play a pivotal role in generating immunological memory against invading fungi. Previous investigations conducted by our research team have demonstrated a strong association between CD8+ T-cell impairment and the development of invasive candidiasis (IC) among ICU patients. This finding underscores the potential utility of lymphocyte subtyping as a rapid diagnostic tool with significant clinical implications for the early identification of IAC (Zhang et al., 2019; Zhang et al., 2022).

Based on our current understanding, prior investigations into IAC have exhibited limitations in scope and comprehensiveness, often focusing narrowly on clinical manifestations (Bassetti et al., 2022; Novy et al., 2023). Furthermore, the extant literature has not yet arrived at a definitive consensus regarding the effectiveness of lymphocyte subtyping in identifying IAC, and at present, there exist no established models capable of swiftly stratifying elderly patients into distinct high- and low-risk categories. With the objective of enhancing the diagnostic precision of IAC and enabling the prompt commencement of antifungal therapy, this study aimed to identify elderly septic patients with an elevated risk of developing IAC. Additionally, we assessed the potential value of lymphocyte subtyping in the early prediction of IAC and formulated a nomogram that integrates lymphocyte subtyping and clinical parameters for the rapid diagnosis of IAC.

Based on our current understanding, this study represents the initial effort to elucidate the risk factors associated with IAC among elderly septic patients. In this study, we demonstrated that new biomarkers, T cell-count, CD28+CD8+T cell-count and CD38+CD8+T cell-count, were significantly altered in elderly septic patients with IAC. Moreover, we developed a nomogram for the early diagnosis of IAC in elderly septic patients. The nomogram, based on gastrointestinal perforation, need for RRT, T cell-count, CD28+CD8+T cell-count and CD38+CD8+T cell-count, had a good identifiable value for IAC, with an AUC of 0.840.

Over the past decade, a substantial accumulation of evidence has been documented, suggesting the involvement of Candida species as a significant contributor to the pathogenesis of IAI (Bassetti et al., 2022). Presently, Candida is widely acknowledged as a leading causative agent of IAI, owing to its strong correlation with significant morbidity, mortality, and escalated healthcare expenditures (Klingspor et al., 2015). Furthermore, any delays in initiating antifungal therapy have the potential to exacerbate mortality rates. In contrast, the excessive prescription of antifungal agents poses significant risks, including exposure to drug toxicity and the promotion of resistance (Azoulay et al., 2012). Prompt identification of IAC is paramount. However, this approach is constrained by nonspecific clinical manifestations, prolonged waiting times for culture results, and logistical challenges encountered during sampling of potential infection sites within deep tissue. Despite the advent of novel diagnostic biomarkers, such as BDG, their widespread accessibility remains to be established, and they are prone to false-positive outcomes. Given these considerations, there is a pressing need for strategies that aim to provide timely antifungal therapy to elderly septic patients who are at a heightened risk of developing IAC.

Multiple risk models have been documented for forecasting IC among patients admitted to the ICU. However, a significant proportion of these investigations rely heavily on retrospective data analysis (Playford et al., 2016), or they narrowly focus on candidaemia, despite the fact that IAC also comprises a substantial segment of IC. Unfortunately, our understanding of the risk factors associated with IAC has predominantly stemmed from studies encompassing patients with candidaemia or those experiencing nonpostoperative IAI (Dupont et al., 2015; Xia and Wang, 2019). The aforementioned studies predominantly concentrated on community-acquired infections, displaying unique features that may not be directly applicable to septic patients in the ICU, who are inherently predisposed to IAC. In contrast, the current study provides a more precise depiction of routine clinical practice, as it exclusively includes elderly septic patients admitted to the ICU, reflecting the significant proportion of patients afflicted by IAC.

The current study has identified various clinical risk factors, all of which have been previously implicated in the diagnosis of IAC (Bassetti et al., 2022). Gastrointestinal perforation results in the disruption of the gastrointestinal barrier, thereby establishing a conducive environment that encourages the proliferation of Candida cells within the peritoneal cavity. Additionally, breaches in mucosal barriers increase the likelihood of translocation of the Candida species into the bloodstream (Pemán et al., 2017). The impact of need for RRT on risk of IAC is likely multifactorial. There is evidence of innate and adaptive immune dysregulation in the setting of renal failure. Moreover, patients those with haemodialysis catheters are also at elevated risk of IAC owing to the loss of integrity of the normal anatomical defences (Lass-Florl et al., 2024).

The present study is believed to be the first prospective study to use lymphocyte subtyping assessing the quantitative changes in host immune status and the potential role of these immune parameters in the diagnosis of IAC in elderly septic patients. We found a significant correlation between T-cell count, CD28+CD8+ T-cell count and CD38⁺CD8⁺ T-cell count and the presence of IAC in elderly septic patients. As is known to all, the ability of the host to survive IAC requires a well-coordinated response by the innate and adaptive immune systems; both of which are often impaired in patients with invasive candidiasis (Zhang et al., 2019). Most patients with IAC are immunosuppressed due to aging and underlying disease. Although CD4+T cells have a central role in coordination of the host defenses against fungi, the cytotoxic effect of CD8+T-cells seems ultimately to be essential for termination of the infection. We have previously demonstrated that CD8+T immunity is impaired in response to invasive candidiasis through the mammalian target of rapamycin signaling pathway (Zhang et al., 2021). T-cell activation is carefully regulated by expression of positive and negative co-stimulatory molecules that prevent unbridled T-cell function. We detected significant alterations in the T cell compartment in elderly IAC patients. CD28 acts as a classic co-stimulating molecule on the surface of T cells, implicates in a wide array of T-cell responses, including T-cell proliferation, antigen-stimulated differentiation, activation and cytokine production. Our previous study identified CD28+CD8+T cell-count as an independent risk factors for invasive candidiasis and early mortality in critically ill patients (Zhang et al., 2019), which is consistent with the present study. CD38 is a surface glycoprotein existing on many immune cells. As a marker of cell activation, it is associated with many infectious diseases. In certain infections, CD38+CD8+T cells undergo a rapid increase after the infection once the immune control of acute phase is achieved (Cao et al., 2011; Piedra-Quintero et al., 2020). However, persistency of immune activation and CD38⁺ expression may reflect the failure of the host’s immune system to fully suppress pathogens. We demonstrated that CD38+CD8+T cells were more prominent in IAC patients. It is possible that CD38+CD8+T cells represent overactivated, potentially exhausted, and less efficient T cells in Candida control, as suggested by SARS-CoV-2+ and HIV studies (Rebillard et al., 2021). These results of the present study provide evidence that T-cell count, CD28+CD8+ T-cell count and CD38+CD8+ T-cell count, which are quickly available through lymphocyte subtyping, are important for the early diagnosis of IAC in elderly septic patients in clinical practice, facilitating timely intervention and improving patient outcomes.

Although numerous reports have established a notable correlation between all five risk factors and the diagnosis of IAC, current research indicates that none of these factors alone serves as a definitive indicator for identifying patients with a high risk of developing IAC, thus justifying the commencement of antifungal therapy (Bassetti et al., 2022; Lass-Florl et al., 2024). Given the accessibility and swiftness in acquiring data on all five factors, the formulation of a predictive nomogram that incorporates these factors, encompassing clinical risk factors and lymphocyte subsets, emerges as a promising strategy for the early detection of IAC. This nomogram demonstrated excellent discriminative ability, with an AUC of 0.840 in the training cohort, and its validity was further confirmed in the testing cohort (AUC = 0.783). Furthermore, the Candida score is extensively employed for diagnosing IC in patients admitted to the ICU (León et al., 2009). However, its reliance on fungal culture results, which typically require at least 2 days to obtain, does not expedite the diagnostic timeline. Our current research underscores the Candida score’s limited diagnostic efficacy in the context of IAC in the elderly septic patients. Conversely, our proposed nomogram exhibits superior discriminatory capability compared to the Candida score. This finding suggests that the nomogram offers enhanced diagnostic precision for the early detection of IAC. Additionally, the positive net benefits demonstrated by DCA within the predictive model indicate its promising clinical utilization.

The nomogram was a reliable tool for clinical decision-makers to predict the risk of IAC in elderly septic patients. Based on the nomogram scores, we have delineated three distinct risk groups. The forest plot analysis reveals that the nomogram possesses remarkable predictive prowess and significant potential for clinical utilization. In clinical settings, patients belonging to the low-risk category, scoring between 67 and 106 points, exhibit a conspicuously low incidence of IAC and do not require antifungal therapy. Conversely, the high-risk group, scoring between 116 and 159 points, demonstrates a significantly elevated incidence of IAC and may benefit from the initiation of antifungal agents. Patients classified as moderate risk, scoring between 106 and 116 points, may require further assessment, and the necessity for antifungal therapy remains indeterminate. The utilization of advanced diagnostic modalities, including next-generation sequencing, has the potential to provide valuable insights for this patient cohort. Notably, the nomogram score exhibits a robust positive correlation with the actual risk of IAC. The parameters encompassed within the nomogram are readily accessible, economically feasible, and reliable in clinical practice. Therefore, the nomogram can be widely utilized and serves as a practical tool for physicians to assess the risk of IAC among septic patients at the bedside. Further research is crucial to validate the efficacy of the nomogram in guiding the initiation or discontinuation of antifungal therapy. Since BDG has been shown to be useful in identifying IAC in critically ill patients, we established a new nomogram by independent predictors plus BDG positivity on the basis of clinical validity. Although no statistical significance was observed, the nomogram constructed by independent predictors plus BDG positivity had higher AUC value in predicting IAC. This result demonstrated the potential role of BDG in diagnosing IAC. In future researches, fungal biomarkers such as BDG and Candida PCR should be considered for optimization of the predicting model.

The strength of this study is the concentration on elderly septic patients in the ICU, as this demographic group is especially susceptible to infections due to their weakened immune system. Candida spp. are typically the resident microflora that colonize the gastrointestinal tracts. However, as individuals get older, metabolic disorders can weaken the mucosal barriers, allowing these colonized Candida strains to penetrate and invade deeper into the body’s organs (Gong et al., 2020). To the best of our knowledge, this study is the first to highlight the characteristics of IAC in the elderly septic population. However, it is worth noting that there are indeed several limitations in this study. Firstly, it is noteworthy that the data were procured specifically from one hospital located in China, and the study was confined to ICU patients only. Because the study is a single-center cohort, our study is limited in its generalizability due to selection bias. Consequently, the extrapolation of the results to the broader elderly population should be approached with caution. Despite this inherent limitation, our validation efforts have yielded remarkably satisfactory AUC values and calibration metrics in both the training and testing cohorts, surpassing the Candida score in terms of its discriminatory capabilities. Secondly, there exists the potential risk of misclassification between intra-abdominal candidiasis (IAC) and non-IAC due to the sensitivity of peritoneal culture. So rigorous adherence to a standardized protocol has been implemented. This protocol emphasizes the direct inoculation of peritoneal samples into an optimized culture medium, with cultures being stored for a maximum of 14 days to facilitate detection of delayed positivity. By following this protocol, the aforementioned risk of misclassification is mitigated to a considerable degree. Thirdly, although the nomogram presented in our study effectively facilitates early and precise diagnosis of IAC among elderly septic patients, it is crucial to acknowledge that our analysis was inherently limited in terms of controlling for certain variables. Notably, we were unable to account for specific factors such as the type of abdominal surgery undergone by the patients or the exact site of perforation or leakage. These variables have been extensively documented in previous research to have a substantial impact on the frequency of Candida isolation from abdominal fluid samples. Fourthly, we focused on a single snapshot of lymphocyte subtyping at the onset of IAI to formulate a model for the rapid diagnosis of IAC. What we should note is that static measurement of immune markers can fail to capture the nuances of immunological progression in critically ill patients. Future research should conduct a time-series analysis of lymphocyte subtyping to better capture the trajectory of the immune response in IAC. Given the preliminary nature of our findings, it is crucial to acknowledge that the clinical utility of our nomogram, while promising, necessitates further validation. To comprehensively assess its clinical performance, we advocate for the conduct of larger-scale, multi-center, prospective studies applying this model to a diverse patient cohort. Such endeavors will not only solidify the evidence base but also justify the allocation of resources towards more exhaustive longitudinal research endeavors. It is imperative to view our current results as a stepping stone towards a more definitive understanding of the nomogram’s potential to inform clinical decision-making.

To summarize, our findings indicate a significant correlation between T-cell count, CD28+CD8+ T-cell count and CD38+CD8+ T-cell count and the presence of IAC in elderly septic patients. By creating a predictive nomogram incorporating factors such as gastrointestinal perforation, RRT and the lymphocyte subtyping markers, clinicians can promptly assess an elderly patient’s risk of developing IAC early on the infection process. This tool holds the potential to enhance the prognosis for elderly septic patients with IAC by facilitating timely intervention and management strategies.