Procalcitonin and presepsin for detecting bacterial infection and spontaneous bacterial peritonitis in cirrhosis: A systematic review and meta-analysis

Abstract

BACKGROUND

Diagnosing bacterial infections (BI) in patients with cirrhosis can be challenging because of unclear symptoms, low diagnostic accuracy, and lengthy culture testing times. Various biomarkers have been studied, including serum procalcitonin (PCT) and presepsin. However, the diagnostic performance of these markers remains unclear, requiring further informative studies to ascertain their diagnostic value.

AIM

To evaluate the pooled diagnostic performance of PCT and presepsin in detecting BI among patients with cirrhosis.

METHODS

We performed a systematic search of the MEDLINE, EMBASE, and Scopus databases for studies that evaluated the diagnostic role of PCT and presepsin from inception to June 2024. Sensitivity and specificity values were pooled using a random effects model. BI was diagnosed based on clinical manifestations, physical examination, laboratory data, and radiological findings.

RESULTS

Of the 6639 articles retrieved, 28 met the inclusion criteria and included 4287 patients with 1789 cases of BI (41.7%). The bivariate pooled sensitivity and specificity estimates of PCT for BI diagnosis were 0.73 [95% confidence interval (CI): 0.64-0.81] and 0.83 (95%CI: 0.79-0.87), respectively. The diagnostic odds ratio (DOR) of PCT was 17.21 (95%CI: 9.57-30.95). Presepsin showed a pooled sensitivity of 0.75 (95%CI: 0.60-0.86), specificity of 0.80 (95%CI: 0.68-0.88), and DOR of 12.33 (95%CI: 5.10-29.83) for diagnosing BI. The pooled sensitivity and specificity of PCT for diagnosing spontaneous bacterial peritonitis (SBP) were 0.76 (95%CI: 0.67-0.84) and 0.87 (95%CI: 0.78-0.92), respectively. The positive likelihood ratio of PCT was 5.57 (95%CI: 3.34-9.29), which was sufficiently indicative of SBP. The DOR of PCT was 29.50 (95%CI: 12.30-70.80).

CONCLUSION

PCT and presepsin have high sensitivity and specificity for detecting BI in patients with cirrhosis. Furthermore, PCT has good diagnostic value as a rule-in test for SBP diagnosis.

Key Words: Biomarker; Cirrhosis; Infection; Meta-analysis; Presepsin; Procalcitonin; Soluble CD14

Core Tip: Our meta-analysis highlights the comparison between serum procalcitonin (PCT) and presepsin in the detection of bacterial infection (BI) among patients with cirrhosis. While both biomarkers demonstrated comparable sensitivity and specificity, PCT exhibited greater diagnostic ability for BI.

INTRODUCTION

Bacterial infections (BI) commonly occur in patients with cirrhosis, resulting in poor outcomes, including the development of cirrhotic complications, septic shock, acute-on-chronic liver failure (ACLF), multiple organ failures, and mortality[1,2]. BI is observed in 20%-30% of hospitalized patients, with and without ACLF[3]. Patients with cirrhosis are susceptible to BI because of internal and external factors. The major internal factors are changes in gut microbial composition and function, bacterial translocation, and cirrhosis-associated immune dysfunction syndrome[4,5]. External factors include alcohol use, proton-pump inhibitor use, frailty, readmission, and invasive procedures. Spontaneous bacterial peritonitis (SBP), urinary tract infection, pneumonia, and primary bacteremia are the common BIs in hospitalized patients with cirrhosis[6].

Early diagnosis and adequate empirical antibiotic therapy are two critical factors that improve the prognosis of BI in patients with cirrhosis. However, early detection of BI in cirrhosis is challenging due to subtle clinical signs and symptoms, low sensitivity and specificity of systemic inflammatory response syndrome criteria, and low sensitivity of bacterial cultures. Thus, effective biomarkers need to be identified for the early detection of BI. Several biomarkers have been evaluated, but their efficacy in detecting BI is unclear.

Procalcitonin (PCT) is a precursor of the hormone calcitonin, which is secreted by parafollicular cells of the thyroid gland[7]. In the presence of BI, PCT gene expression increases in extrathyroidal tissues, causing a subsequent increase in serum PCT level[8]. Changes in serum PCT are detectable as early as 4 hours after infection onset and peaks between 8 and 24 hours, making it a valuable diagnostic biomarker for BI. Several studies have demonstrated the favorable diagnostic accuracy of PCT in the diagnosis of BI in individuals with cirrhosis[9-13] and without cirrhosis[14-16]. Since 2014, two meta-analyses have been published on the diagnostic value of PCT for SBP and BI in patients with cirrhosis[17,18]. Other related studies have been conducted since then[10-12,19-33].

Serum presepsin has recently emerged as a promising biomarker for diagnosing BI. This biomarker is the N-terminal fraction protein of the soluble CD14 g-negative bacterial lipopolysaccharide–lipopolysaccharide binding protein (sCD14-LPS-LBP) complex, which is cleaved by inflammatory serum protease in response to BI[34]. Presepsin levels increase within 2 hours and peaks in 3 hours[35]. This is useful for detecting BI since presepsin levels increase earlier than serum PCT. Several studies have been conducted to explore the diagnostic potential of serum presepsin in identifying BI in patients with cirrhosis[25,28,32,36]. No meta-analysis has evaluated the performance of serum presepsin in diagnosing BI in patients with cirrhosis. Therefore, the current study evaluated the diagnostic value of serum PCT and presepsin for BI in patients with cirrhosis.

MATERIALS AND METHODS

Our systematic review and meta-analysis was performed with adherence to PRISMA guidelines[37].

Protocol registration

The objectives and methodologies of this meta-analysis were predefined in a protocol registered with PROSPERO. The registration was accepted on June 19, 2024, under the number CRD42024555777.

Search strategy and search terms

A comprehensive search of the MEDLINE, EMBASE, and Scopus databases was conducted for relevant studies published from inception until June 2024. The following MeSH terms and free text were used: “liver cirrhosis”, “end-stage liver disease”, “hepatic insufficiency”, “acute-on-chronic liver failure”, “biomarkers”, “presepsin”, “procalcitonin”, “BIs”, “bacteremia”, “peritonitis”, “pneumonia”, “cellulitis”, “cholangitis”, “cholecystitis”, “urinary tract infections”, “sepsis”, and “SBP”. The search was limited to publications written in the English language. Additional relevant studies were manually identified by searching the bibliographies of the retrieved publications.

Selection process

The title and abstract of the studies were screened in the first round, and potentially relevant studies were retrieved for full-text review in the second round. This step examined studies that assessed the diagnostic accuracy of PCT and presepsin levels for diagnosing BI and SBP. The inclusion criteria were the population of adult patients with cirrhosis (aged > 18 years), cases with BI as an outcome, data on results of serum PCT or presepsin test, data on calculations of sensitivity or specificity or sufficient data to construct a 2 × 2 contingency table, and studies published as a full-text article. Case reports, case series, review articles, and clinical guidelines were excluded. Two independent reviewers (Wejnaruemarn S and Thanapirom K) performed the article selection. The resolution of any disagreements between the reviewers was achieved through a consensus meeting[38]. The study selection process is shown in Figure 1.

Figure 1 Flow chart showing the number of articles (n) included during the identification and selection processes.

Data extraction

Data extraction and quality control were performed by two reviewers (Wejnaruemarn S and Thanapirom K). Data extracted from published studies included author names, publication year, country, study design, sample size, age, sex, etiology, liver disease severity, BI and SBP diagnosis, cutoff value, testing systems, sensitivity, and specificity of the biomarkers. Disagreements were resolved by consensus.

Statistical analysis

Pooled sensitivity, specificity, likelihood ratio, and diagnostic odds ratio (DOR) with 95% confidence interval (CI) were calculated based on data from the included studies[39]. Summary receiver operating characteristic (sROC) analysis was used to assess the discriminating ability of the diagnostic test. The estimation of summary DORs was performed using either random (DerSimonian-Laird) or fixed (Mantel-Haenszel) effects models, depending on whether inconsistency index (I²) exceeded or decreased below 50%[40]. The I² was calculated to evaluate heterogeneity between studies[41]. Sensitivity analysis of crucial diagnostic measures in the presence of heterogeneity was performed. Statistical analyses were performed using OpenMeta (Stata Corp, College Station, TX, United States) by Susantitaphong P. All statistical tests were two-sided, and statistical significance was defined as P value < 0.05.

Quality assessment

The quality of selected studies was assessed by two independent reviewers (Wejnaruemarn S and Thanapirom K) using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2)[42]. Discrepancies between the reviewers were resolved through a consensus meeting.

Risk of bias assessment

Publication bias was evaluated by visual inspection using Begg’s funnel plot.

RESULTS

Search process

The comprehensive search retrieved 6639 studies (Figure 1). During the initial screening of titles and abstracts, 6490 irrelevant studies, case reports, and reviews were eliminated, leaving 149 studies that were considered potentially relevant. Of these, 121 studies were excluded after a full-text review, primarily due to the absence of reported outcomes of interest, unavailability of full-text articles, or the lack of relevant biomarkers (Figure 1). Finally, 28 studies that investigated serum PCT and presepsin levels in patients with cirrhosis and BI were eligible for inclusion in the analysis.

Characteristics of the studies

The study included 1789 (41.7%) cases of suspected BI out of 4287 cirrhosis cases. Ten studies examined SBP as the primary outcome, whereas all other studies investigated BI as the primary outcome. SBP developed in 458 out of 964 patients (47.5%) in 10 studies. Four studies used standard PCT cutoff values (0.5 ng/mL), whereas the remaining 24 studies used the optimized threshold with the highest sensitivity and/or specificity. Of the 28 studies, 22 focused on PCT testing, 4 examined PCT and presepsin tests, and 2 exclusively evaluated presepsin tests. Five of the 28 studies adopted a case-control design, while the remainder used a cohort design. The sample size for each study ranged from 20 to 386 patients. Table 1 summarizes characteristics of the included studies and patients. All articles included in our analysis employed well-established inclusion and exclusion criteria. SBP was defined as an ascites sample containing polymorphonuclear cells 250 cells/mm³ with or without culture positivity or compatible clinical symptoms. Other sites of BI were identified based on relevant clinical symptoms and/or positive culture results.

Table 1 Characteristics of included studies.

Ref.	No. of patients/infection	Mean age	Underlying liver disease	Study design	Biomarker	PCT testing system	Cut-off (PCT = ng/mL, presepsin = pg/mL)	Outcome	PCT sensitivity/specificity (%)	Presepsin sensitivity/specificity (%)
Viallon et al[9]	61/21	N/A	Cirrhosis	Cohort	PCT	LUMItest	0.76	SBP	95/98	N/A
Spahr et al[50]	20/10	N/A	Cirrhosis	Cohort	PCT	LUMItest	0.50	SBP	50/100	N/A
Connert et al[51]	127/36	57	Cirrhosis	Cohort	PCT	LUMItest	0.58	Infection	92/78	N/A
Elefsiniot et al[52]	106/16	54.7	Acute and chronic liver disease	Cohort	PCT	LUMItest	0.75	Infection	61/99	N/A
Li et al[13]	98/27	N/A	Cirrhosis	Cohort	PCT	N/A	0.49	Sepsis	81.5/87.3	N/A
Papp et al[53]	368/139	56.4	Cirrhosis	Cohort	PCT	Liaison Diasorin	0.15	Infection	72/84	N/A
Lazzarotto et al[54]	64/24	54.31	Cirrhosis	Cohort	PCT	ADVIA	1.10	Infection	67/90	N/A
Yuan et al[55]	84/42	48.31	Chronic severe hepatitis B	Cohort	PCT	Liaison Diasorin	0.48	SBP	95/79	N/A
Marciano et al[19]	103/29	N/A	Cirrhosis	Cohort	PCT	Roche Diagnostics	0.098	Infection	97/82	N/A
Qu et al[33]	324/125	45.19	Chronic liver disease	Cohort	PCT	Roche Diagnostics	0.53	Infection	94.4/79.9	N/A
Cai et al[10]	96/61 (SBP), 129/94 (infection)	N/A	Cirrhosis	Case-control	PCT	VIDAS	2 (SBP), 0.5 (infection)	SBP, infection	68.8/94.2 (SBP), 92.5/77.1 (infection)	N/A
Asadi Gharabaghi et al[20]	33/8	N/A	Cirrhosis	Case-control	PCT	N/A	0.50	SBP	75/92	N/A
Papp et al[32]	216/75	57.5	Cirrhosis	Cohort	PCT, presepsin	Liaison Diasorin	0.39 (PCT), 844 (presepsin)	Infection	53.3/94.4	60/84.4
Abdel-Razik et al[11]	79/52	58.4	Decompensated cirrhosis	Cohort	PCT	RayBio	0.94	SBP	94.3/91.8	N/A
Wu et al[21]	88/40	N/A	Decompensated cirrhosis	Case-control	PCT	Cobas	0.78	SBP	77.5/60.4	N/A
Zuwala-Jagiello et al[22]	126/51	58	Cirrhosis	Cohort	PCT	LUMItest	0.20	Infection	69.9/81.2	N/A
Wang et al[12]	259/109	51.4	Cirrhosis	Cohort	PCT	Roche Diagnostics	0.88	SBP	75.2/87.5	N/A
Tao et al[23]	156/78	N/A	Cirrhosis	Cohort	PCT	Wuhan Mingde	0.63	Infection	90.23/72.35	N/A
Khedher et al[24]	92/60	63	Decompensated cirrhosis	Cohort	PCT	VIDAS	0.50	Infection	45/96.6	N/A
Fischer et al[25]	114/53	60	Decompensated cirrhosis	Cohort	PCT, presepsin	Artron	0.47 (PCT), 1444.5 (presepsin)	Infection	59/76	50/86
Guo et al[26]	154/54	N/A	Cirrhosis	Cohort	PCT	Roche Diagnostics	0.51	Sepsis	77.8/93	N/A
Chen et al[31]	143/94	47	ACLF by APASL criteria	Case-control	PCT, presepsin	N/A	0.765 (PCT), 404.5 (presepsin)	Sepsis	55.3/81.6	96.8/59.2
Lin et al[27]	386/169	N/A	ACLF by APASL criteria	Cohort	PCT	N/A	1.01	Infection	42.6/78.8	N/A
Novelli et al[56]	70/7	49.5	Decompensated cirrhosis	Cohort	Presepsin	N/A	669.5	Sepsis	N/A	99.1/98.1
Ferrarese et al[28]	251/127 (PCT), 278/127 (presepsin)	57.4	Cirrhosis	Cohort	PCT, presepsin	N/A	0.86 (PCT), 741.31 (presepsin)	Infection	53.5/71.8	66/63
El-Hassib et al[29]	199/101	N/A	Cirrhosis	Case-control	PCT	N/A	0.95	SBP	68.3/72.4	N/A
Verma et al[30]	45/14	53.8	Decompensated cirrhosis	Cohort	PCT	N/A	2	SBP	50/100	N/A
Igna et al[36]	365/134	59	Decompensated cirrhosis	Cohort	Presepsin	N/A	980	Infection	N/A	80.2/82.5

N/A: Not applicable; PCT: Procalcitonin; SBP: Spontaneous bacterial peritonitis.

Methodological quality of included studies

The QUADAS-2 criteria were used to assess the risk of bias and the applicability of the selected studies. Figure 2 shows the comprehensive assessment of the methodological quality of the studies. All patients were verified using the same reference standard across all studies. None of the included studies provided explanations for patient withdrawals or reported uninterpretable results. Several studies blinded the assessing physicians to the index test while verifying the outcomes based on reference standards. PCT or presepsin was not used as diagnostic criterion for SBP or systemic BI.

Figure 2 Methodological evaluation of the included studies according to the Quality Assessment of Diagnostic Accuracy Studies-2 criteria. A: Percentage of studies grouped according to risk of bias; B: Percentage of studies grouped according to concerns regarding applicability.

Sixteen studies failed to report that patients were consecutive or randomized, raising suspicion for selection bias. For the index test, 25 studies selected a test threshold with optimized sensitivity and/or specificity based on the receiver operating characteristic curve, which could overestimate the test performance.

As for applicability, all presepsin studies used a commercial kit (PATHFAST^® presepsin analyzer, Mitsubishi Chemical Medience Corporation, Tokyo, Japan). Five PCT studies used an immune-luminometric assay kit (LUMItest PCT, BRAHMS-Diagnostica, Berlin, Germany), 4 used an electrochemiluminescence immunoassay kit (Roche Diagnostics, Mannheim, Germany), 9 did not declare the method of PCT testing, and the rest used various other kinds of commercial testing kits. Regarding reference standards, all studies used ascitic profile to diagnose SBP.

Performance of serum biomarkers in diagnosing BI and SBP

The pooled sensitivity and specificity of serum PCT for the diagnosis of BI were 0.73 (95%CI: 0.64-0.81) and 0.83 (95%CI: 0.79-0.87), respectively. PCT had a low positive likelihood ratio (PLR, 4.25; 95%CI: 3.30-5.48) for rule-in diagnosis and a low negative likelihood ratio (NLR, 0.25; 95%CI: 0.16-0.41) for rule-out diagnosis. The DOR of PCT for BI was 17.21 (95%CI: 9.57-30.95) (Table 2).

Table 2 Summary of diagnostic performance of procalcitonin and presepsin in bacterial infection diagnosis and spontaneous bacterial peritonitis diagnosis.

Variable	Studies (n)	Sensitivity (95%CI)	Specificity (95%CI)	PLR (95%CI)	NLR (95%CI)	DOR (95%CI)	I2
For bacterial infection
PCT	17	0.73 (0.64-0.81)	0.83 (0.79-0.87)	4.25 (3.30-5.48)	0.25 (0.16-0.41)	17.21 (9.57-30.95)	89.90
Presepsin	6	0.75 (0.60-0.86)	0.80 (0.68-0.88)	3.51 (2.20–5.59)	0.24 (0.10-0.56)	12.33 (5.10-29.83)	87.69
For spontaneous bacterial peritonitis
PCT	10	0.76 (0.67-0.84)	0.87 (0.78-0.92)	5.57 (3.34-9.29)	0.21 (0.12-0.37)	29.50 (12.30-70.80)	69.34

I² for sensitivity. DOR: Diagnostic odd ratio; NLR: Negative likelihood ratio; PCT: Procalcitonin; PLR: Positive likelihood ratio.

Serum presepsin had a pooled sensitivity and specificity of 0.75 (95%CI: 0.60-0.86) and 0.80 (95%CI: 0.68-0.88), respectively, for the diagnosis of BI. The PLR ratio (3.51; 95%CI: 2.20-5.59) and NLR (0.24; 95%CI: 0.10-0.56) were low for rule-in and rule-out diagnoses, respectively. The DOR of presepsin for BI was 12.33 (95%CI: 5.10-29.83). The sROC curves for PCT and presepsin are shown in Figure 3.

Figure 3 Summary receiver operating characteristic curves summarizing the overall diagnostic accuracy of procalcitonin and presepsin for diagnosing bacterial infections and spontaneous bacterial peritonitis. A: Procalcitonin for diagnosing bacterial infections; B: Presepsin for diagnosing bacterial infections; C: Procalcitonin for diagnosing spontaneous bacterial peritonitis.

The analysis for SBP diagnosis revealed that the pooled sensitivity and specificity of PCT were 0.76 (95%CI: 0.67-0.84) and 0.87 (95%CI: 0.78-0.92), respectively. The PLR of PCT showed good diagnostic value (5.57; 95%CI: 3.34-9.29) and was sufficiently high to establish serum PCT as a rule-in test, whereas the NLR was moderate (0.21; 95%CI: 0.12-0.37). The DOR of PCT for SBP was 29.50 (95%CI: 12.30-70.80).

Forest plots illustrating the pooled sensitivity and specificity of PCT and presepsin in diagnosing BI and SBP are shown in Figure 4. For diagnosing BI using PCT, the I² values of sensitivity, specificity, PLR, NLR, and DOR were 89.90, 69.73, 77.28, 94.06, and 86.83, respectively. For presepsin in BI, I² values of sensitivity, specificity, PLR, NLR, and DOR of presepsin were 87.69, 88.73, 87.00, 96.25, and 86.07, respectively. For PCT in SBP, I² of sensitivity, specificity, PLR, NLR, and DOR of PCT were 69.34, 75.95, 75.26, 79.21, and 75.83, respectively. The P values for sensitivity, specificity, PLR, NLR, and DOR were all < 0.001, indicating considerable heterogeneity across the included studies.

Figure 4 Forest plots of sensitivity and specificity of procalcitonin and presepsin for diagnosing bacterial infections and spontaneous bacterial peritonitis. A: Procalcitonin for diagnosing bacterial infections; B: Presepsin for diagnosing bacterial infections; C: Procalcitonin for diagnosing spontaneous bacterial peritonitis in patients with cirrhosis. FN: False negative; FP: False positive; TN: True negative; TP: True positive.

Sensitivity analysis based on the primary outcome of the presepsin study was performed by excluding two studies that used sepsis as a primary outcome. The rest of the studies using BI as a primary outcome were included in the analysis. Sensitivity and specificity decreased to 0.65 (95%CI: 0.54-0.78) and 0.79 (95%CI: 0.71-0.88), respectively.

We constructed a Fagan nomogram to evaluate post-test probability using the likelihood ratio of each biomarker. Using a pretest probability of 41.7% prevalence for BI, PCT had PLR of 4.25 and NLR of 0.25, with a positive post-test probability of 75.2% and negative post-test probability of 15.2%. Presepsin had a PLR of 3.51 and NLR of 0.24, with a positive post-test probability of 71.5% and negative post-test probability of 14.7%. For SBP, the pretest probability was 47.5%. PCT had a PLR of 5.57 and a NLR of 0.21. The positive post-test probability was 83.4%, and negative post-test probability was 16% (Figure 5).

Figure 5 Diagnostic performance of procalcitonin and presepsin for diagnosing bacterial infections and spontaneous bacterial peritonitis. A: Procalcitonin for diagnosing bacterial infections; B: Presepsin for diagnosing bacterial infections; C: Procalcitonin for diagnosing spontaneous bacterial peritonitis among patients with cirrhosis: Fagan nomogram. NLR: Negative likelihood ratio; PLR: Positive likelihood ratio; prob: Probability; SBP: Spontaneous bacterial peritonitis.

The presence of publication bias was demonstrated using an asymmetrical funnel plot of the assessment of the sensitivity and specificity of PCT for diagnosing BI and SBP (Figure 6). Because of the limited number of studies (less than 10), a funnel plot was not constructed for presepsin.

Figure 6 Funnel plot for studies. A: Sensitivity of procalcitonin for diagnosing bacterial infection; B: Specificity of procalcitonin for diagnosing bacterial infection; C: Sensitivity of procalcitonin for diagnosing spontaneous bacterial peritonitis; D: Specificity of procalcitonin for diagnosing spontaneous bacterial peritonitis.

DISCUSSION

This meta-analysis included 28 studies that evaluated the diagnostic performance of serum PCT or presepsin for BI and SBP in patients with cirrhosis. The results of this meta-analysis showed that PCT had a sensitivity of 0.73, specificity of 0.83 and DOR of 17.21 in diagnosing BI. The sensitivity and specificity of PCT were 0.76 and 0.87, respectively, in identifying SBP. Furthermore, PCT demonstrated a robust PLR and was established as an appropriate marker for the rule-in diagnostic test for SBP diagnosis but not for BI. The Fagan nomogram demonstrated that PCT for SBP diagnosis enhanced positive post-test probability to 83.4%. To our knowledge, this meta-analysis is the first to investigate the diagnostic role of serum presepsin in diagnosing BI in patients with cirrhosis. The sensitivity, specificity, and DOR of presepsin in the diagnosis of BI were 0.75, 0.80, and 12.33, respectively. The pooled sensitivity and specificity of PCT and presepsin in diagnosing BI were comparable.

Prior meta-analyses have shown that serum PCT can potentially detect BI and SBP. However, further informative studies are required to ascertain the diagnostic capacity of serum PCT given the limited number of studies on this topic[17,18]. Over the last few years, evidence has accumulated on the diagnostic value of serum PCT for BI[10,22,24-28,31-33], particularly for SBP[10-12,21,29,30], in patients with cirrhosis. However, the sensitivity and specificity varied widely among the studies. The results of the present study are comparable to the previous meta-analysis conducted by Lin et al[17], which included 10 studies published between 2000 and 2013. They found that PCT demonstrated good diagnostic performance with a sensitivity of 0.79, specificity of 0.89, and DOR of 24.00 in diagnosing BI in cirrhosis. They also compared the diagnostic accuracy between PCT and C-reactive protein (CRP), concluding that both biomarkers exhibited high accuracy in differentiating BI from other noninfective causes of inflammation in patients with cirrhosis. Although CRP has been used as an infection indicator in various contexts, it is a general marker of inflammation[43]. The baseline CRP level in cirrhosis is elevated. For BI, patients with more severe liver dysfunction exhibited decreased response in CRP levels[44,45], indicating that CRP has limited prognostic value for infection in cirrhosis. Thus, in this meta-analysis, we excluded CRP from further comparison. Our meta-analysis demonstrated greater heterogeneity than that reported by Lin et al[17]. We hypothesized that the heterogeneity is attributed to variations in the PCT cutoff values used in the included studies. In our study, 4 of 17 studies used a PCT cutoff of 0.5 pg/mL, whereas the remaining studies used the optimal cutoff value. This contrasts with the previous meta-analysis, where 7 of 10 studies used a PCT cutoff of 0.5 pg/mL.

For SBP diagnosis using serum PCT, 10 studies were analyzed, revealing a sensitivity of 0.76, specificity of 0.87, and DOR of 29.5 for identifying SBP. The meta-analysis conducted by Yang et al[18] included 7 studies published between 2000 and 2014 and reported a sensitivity of 0.82, specificity of 0.86, and DOR of 22.6 for PCT. The results of our meta-analysis were comparable in terms of sensitivity and specificity but obtained a higher DOR. Furthermore, our study demonstrated a robust PLR value of 5.57, which established serum PCT as a suitable marker for a rule-in diagnostic test for SBP diagnosis but not for BI (PLR = 4.25). The advantage of our meta-analysis to the previous one was the integration of the Fagan nomogram. The Fagan nomogram demonstrated that using PCT for SBP diagnosis enhanced positive posttest probability to 83.4%. These findings thus validate the diagnostic ability of serum PCT for SBP in patients with cirrhosis. Furthermore, Yang et al[46] evaluated the diagnostic value of serum PCT for bacterial peritonitis and found that serum PCT exhibited diagnostic sensitivity and specificity for SBP. The key difference between our and Yang et al’s study is that they included patients with cirrhosis and peritoneal dialysis, whereas our study exclusively focused on patients with cirrhosis[46]. Furthermore, Yang et al[46] conducted a comprehensive systematic search across MEDLINE, EMBASE, Scopus, China Biology Medicine Database, China National Knowledge Infrastructure Database, and Cochrane databases, our study only included a search of the MEDLINE, EMBASE, and Scopus databases.

To our knowledge, this meta-analysis is the first to investigate the diagnostic role of serum presepsin and compare its efficacy with serum PCT in detecting BI in patients with cirrhosis. Four of six studies included in our meta-analysis conducted a head-to-head comparison between presepsin and PCT; three demonstrated comparable diagnostic efficacy[25,28,32], while one study demonstrated the superior efficacy of presepsin to that of PCT[31]. In noncirrhotic critically ill patients admitted to the intensive care unit, the diagnostic accuracy of presepsin and PCT in identifying BI was similar (presepsin: Pooled sensitivity of 0.84 and specificity of 0.73; PCT: Pooled sensitivity of 0.80 and specificity of 0.75)[47]. The sensitivity, specificity, and DOR values obtained for presepsin in our study for the diagnosis of BI were 0.75, 0.80, and 12.33, respectively, which were comparable to those obtained for the noncirrhotic population in the previous meta-analysis. The comparison between presepsin and PCT in our meta-analysis showed that only the pooled sensitivity of presepsin was higher than that of PCT, while the specificity and DOR of presepsin were lower than those of PCT in diagnosing BI in patients with cirrhosis. Our current findings further demonstrated that serum PCT was more effective than presepsin in diagnosing BI, as indicated by a higher DOR. The direct comparison between PCT and presepsin may be limited by the small pooled presepsin sample size in this meta-analysis. This difference may be because of the significant difference in their production mechanism, although both biomarkers are upregulated in response to BIs and inflammation. PCT is upregulated in parenchymal cells outside the thyroid gland by bacterial lipopolysaccharides and proinflammatory cytokines, such as interleukin-6 and tumor necrotic factor-alpha. This specific response to BIs makes PCT a reliable marker. Serum presepsin is released from activated monocytes and macrophages by proteolytic cleavage, which is triggered by bacterial endotoxins and systemic inflammation. This may reduce the specificity of presepsin for BIs. Figure 7 illustrates the differences in the response mechanisms of serum PCT and presepsin to BI. Regarding their kinetics, presepsin has a shorter half-life of 8 hours and peaks at approximately 3 hours after the onset of BI[34], making it useful for early detection. However, PCT has a longer half-life of 24 hours and peaks later at 8-24 hours[48]; thus, it provides a better diagnosis for ongoing BI.

Figure 7 Mechanism of serum procalcitonin and presepsin in response to bacterial infection. IL-1: Interleukin-1; IL-6: Interleukin-6; LBP: Lipopolysaccharide binding protein; LPS: Lipopolysaccharide; mCD14: Membrane-CD14; sCD14: Soluble CD14; sCD14-ST: Soluble CD14-subtype; TNF-α: Tumor necrotic factor-alpha.

Sensitivity analysis, which excluded two studies that primarily investigated sepsis, showed that the sensitivity and specificity of serum presepsin decreased. Thus, serum presepsin may be a better diagnostic biomarker for sepsis than for BI, considering that it has also been associated with the severity of sepsis in adult patients[49]. However, presepsin cutoff values in each of the included studies ranged from 400 to 1444 pg/mL, depending on the optimal cutoff value that would have optimized the sensitivity and specificity for that study. Consequently, substantial heterogeneity was introduced into the meta-analysis.

This meta-analysis demonstrated several strengths and limitations. Regarding the strengths, we collected a greater number of articles compared with previous prior meta-analyses. The primary limitation was the significant heterogeneity arising from variations in the cutoff values used across the included studies, causing differing levels of sensitivity and specificity values for each study[50-55]. This could reduce the reliability of the findings. Second, we also analyzed case-control studies, which may overestimate the efficacy of the diagnostic test by removing patients with an uncertain diagnosis. Third, several included studies did not select patients consecutively or randomly, which may potentially cause selection bias. Fourth, the studies included in the meta-analysis used various commercial kits for measuring PCT, potentially leading to variable results. Several studies failed to clarify the testing methods used, limiting the reproducibility of their findings. Fifth, our study did not analyze the performance of these biomarkers in patients with BIs diagnosed solely through positive bacterial culture because of insufficient data in the included studies, which generally diagnosed BIs based on a combination of clinical symptoms, laboratory findings, and culture results. Further subgroup analysis could provide insights into the mechanistic pathways directly linked to bacterial pathogens. Finally, while presepsin shows potential as a biomarker, the limited number of studies and small pooled sample size compared with those of PCT may have diminished the statistical power of the meta-analysis[36,56]. This could produce underpowered results, which would not accurately reflect the true diagnostic accuracy of the test. Compared with previous studies, the current meta-analysis incorporates more recent and specific data. Thus, future large prospective studies with standardized cutoff values are needed to minimize heterogeneity and improve clinical applicability in real-world settings.

CONCLUSION

Our meta-analysis highlights the comparison between serum PCT and presepsin in detecting BI among patients with cirrhosis. While both biomarkers demonstrated comparable sensitivity and specificity, PCT exhibited greater diagnostic ability for BI, as indicated by a higher DOR. Furthermore, PCT demonstrated significant diagnostic efficacy as a rule-in test for diagnosing SBP.

ACKNOWLEDGEMENTS

The authors would like to thank the Excellence Center in Liver Diseases and Center of Excellence in Liver Fibrosis and Cirrhosis, King Chulalongkorn Memorial Hospital for their contribution to the success and completion of this study.