Manual method versus flow cytometry for diagnosing spontaneous bacterial peritonitis

ABSTRACT

Background

Diagnosing spontaneous bacterial peritonitis (SBP) requires a high clinical index of suspicion because the clinical presentation varies widely. Early detection and treatment of SBP are crucial for improving survival rates. Ascitic fluid neutrophil count ≥ 250 cells/mm³ is the only required criterion to diagnose SBP, regardless of culture results. The most frequently used manual ascitic neutrophil counting is difficult and time-consuming, and it often leads to frequent delays in obtaining results and a possible under- or overestimation of the neutrophilic count. The flow cytometric counting provides results in 30 minutes and is more accurate, with sensitivity and specificity approaching 100%.

Objective

This study aimed to assess the diagnostic efficacy of manual counting of neutrophils in comparison to flow cytometry for detecting SBP.

Patients and methods

In a cross-sectional study, we included 140 cirrhosis-related ascites patients who were clinically suspected to have SBP and admitted to the Alexandria Main University Hospital in the Department of Internal Medicine. Participants with ≥ 250 neutrophil cells/mm³ in ascitic fluid by flow cytometry were defined as SBP patients. In contrast, those with fewer than 250 neutrophil cells/mm³ in ascitic fluid using the flow cytometric method were defined as patients without SBP.

Results

At a cutoff value of 250 cells/mm³, the ROC curve showed that the manual ascitic fluid method had lower sensitivity, specificity, positive and negative predictive values (67.19%, 86.84%, 81.13%, and 75.86%, respectively, with AUC = 0.945, 95% CI, 0.91–0.97, p < 0.001) in diagnosing SBP than the flow cytometric method.

Conclusions

Using the flow cytometric method significantly improved diagnostic precision. Its use may be recommended to replace the traditional manual method, which may have unaccepted false-negative results.

KEYWORDS:

• Spontaneous bacterial peritonitis
• flow cytometry
• ascitic fluid neutrophils

1. Introduction

Ascites is a substantial complication of advanced liver dysfunction, which ranks as the principal reason for hospital admissions among patients with cirrhosis [1]. Around 50% of cirrhotic patients are diagnosed with spontaneous bacterial peritonitis (SBP) upon admission, whereas it develops in the remaining patients during their hospitalization [2]. Patients with SBP have a poor prognosis, with mortality up to 20% when treated [3] and higher mortality rates (˃ 80%) among untreated patients; so, the course and outcome depend on rapid diagnosis and prompt antibiotic therapy [4,5]. In patients with SBP, each hour of delay in appropriate antimicrobial therapy is associated with 1.86 times increased hospital mortality [6].

Given the nonspecific clinical manifestations of SBP, with only 50–68% presenting with fever and 49–72% complaining of abdominal pain, the diagnosis primarily relies on laboratory and microbiological evaluations [7]. Diagnostic paracentesis and subsequent analysis of ascitic fluid are deemed the definitive methods for confirming or excluding SBP in cirrhotic patients [8].

Even when sensitive methods are used, an ascitic fluid culture comes back negative in up to 60% of patients who have suggestive symptoms of SBP and a high ascitic neutrophil count. The presence of positive ascitic fluid culture is not a prerequisite for the initiation of antibiotic therapy, but its role is to support the diagnosis and guide the modification of empirical antibiotic therapy [9].

Presently, manual microscopy stands as the benchmark and most widely used technique for quantifying ascitic neutrophils. In some situations, it may be labor-intensive and time-consuming. Furthermore, it is susceptible to human error and interobserver variability. It may misclassify cells with complex morphologies, which results in an underestimation or overestimation of neutrophil counts [10].

This may result in higher mortality due to a delay in the diagnosis and treatment of these individuals, while the widespread use of needless empirical antibiotics will raise iatrogenic consequences and medical expenses [11].

Flow cytometry is a laser-based technique used to detect and analyze the chemical and physical characteristics of cells or particles. It is most commonly used to evaluate bone marrow, peripheral blood, and other body fluids such as ascitic, pleural, peritoneal, cerebrospinal, synovial, and peritoneal dialysis fluids [12]. It is the best technique for cell characterization, and it is specifically used for cell counting, cell sorting, determining cell function, and determining cell characteristics [13,14].

The flow cytometry protocol includes lysis of the RBCs only and incubation of other cells with fluorescently labeled monoclonal antibodies, which is why it is the routine method for cell characterization in hematology [15].

Certain studies support the use of automated flow cytometric differential counts as a rapid, accurate, and better reference method for neutrophils and other differential counts [10,16]. The EASL guidelines approved this technique for SBP diagnosis because it has a sensitivity and a specificity close to 100% [17–19]. The aim of the current study was to assess the diagnostic efficacy of manual counting of neutrophils in comparison to flow cytometry for detecting SBP.

2. Patients and methods

The sample size was calculated using PASS 2020 Software, “NCSS, LLC. Kaysville, Utah, USA, ncss.com/software/pass.” Based on the global prevalence of spontaneous bacterial peritonitis [10%] [19], a minimal total hypothesized sample size of 140 eligible patients was needed [20]. Numeric Results for the Chi-Square Test (Precision: 0.50, N: 145, Alpha: 0.05, Effect Size: 10% [5%–15%]).

Our study included patients with cirrhosis-related ascites who were clinically suspected to have SBP and admitted to the Alexandria Main University Hospital in the Internal Medicine Department. Clinical suspicion of SBP comprised either local manifestations of peritonitis, like tenderness in the abdomen, abdominal pain, vomiting, or ileus, or systemic inflammation signs, such as fever or hypothermia, chills, a change in white blood cell count, hepatic encephalopathy, worsening liver function, worsening renal function, and gastrointestinal hemorrhage. Ascitic fluid samples were obtained from all hospitalized patients with cirrhotic ascites [19].

2.1. Exclusion criteria

Patients with secondary bacterial peritonitis, tuberculous, malignant, or pancreatic ascites were excluded from the study. They were also not included if they had trauma, surgery, cancer, or any concurrent infections. Also, patients who had received immunosuppressive therapy or systemic antibiotic therapy within the previous month for any other infection were excluded from the study.

The protocol mandated comprehensive history collection, physical examination, abdominal ultrasonography, laboratory tests, routine diagnostic paracentesis, and ascitic fluid analysis.

All subjects were evaluated regarding the following:

1. Clinical evaluation of local manifestations of peritonitis or systemic inflammation signs.

   • Abdominal ultrasonography was used to assess liver echo pattern and size, the presence of cirrhosis, and

   • Routine laboratory investigations included a complete blood picture (CBC), serum aspartate and alanine aminotransferase (AST and ALT, respectively), serum albumin, serum bilirubin, and international normalized ratio (INR), serum creatinine and urea, sodium and potassium, and serum C-reactive protein (CRP).

   • Ascitic fluid sampling for:

     1. Chemical analysis (total proteins, glucose, and lactate dehydrogenase).

     2. Manual microscopic counting of neutrophils: to determine the total and differential leukocytic counts using the Bürker manual counting chamber [21]. The chamber was operated with a 0.5 µl sample [22].

     3. Flow cytometric counting of neutrophils using monoclonal antibodies: CD45 Antibody, anti-human, VioBlue®, HLA-DR Antibody, anti-human, PerCP, CD15 Antibody, anti-human, FITC, and CD16 Antibody, anti-human, PE, REAfinity^TM [17,23]. Immunofluore-scence on the viable neutrophils in ascitic fluid was analyzed using BD FACSCanto^TMII, USA. After examination of the forward and side scatter, the gating was done around the neutrophil population, which is CD45, CD15, and CD16 positive and HLA-DR negative. The data was reported as a percentage of the gated population. Absolute numbers of positive populations were also reported (Figure 1) Non-specific fluorescence and staining of non-viable cells were controlled through several methods such as using negative control cells, isotype controls, and specific staining.

        Figure 1. Flow cytometry dot plots of cell populations in the ascitic fluid. Using both side scatter and forward scatter, then gating on viable cells, we identified CD45 positive hematological cells, labeled macrophages (HLA-DR positive, CD15 positive), labeled granulocytes (HLA-DR negative, CD15 positive), and finally classified granulocytes according to CD16 into: neutrophils CD16 positive, and eosinophils CD16 negative.

        

     4. Ascitic fluid cultures: Ascitic fluid inoculation (10 ml) in 2 blood culture bottles was performed at the bedside in all patients.

We confirmed the SBP diagnosis if the number of polymorphonuclear neutrophils (PMN) in the ascitic fluid by flow cytometry was ≥ 250/mm³ and there were no other causes of peritonitis. The study included both culture-positive SBP cases and those with culture-negative neutrocytic ascites. The non-SBP group was made up of people whose ascitic fluid had less than 250 neutrophil cells per mm³ using the flow cytometric method [15].

2.2. Statistical analysis

The collected data was wrangled, coded, and analyzed using the SPSS software (Armonk, NY: IBM Corp., version 25.0) and MedCalc (version 18.2.1), with the Kolmogorov-Smirnov test verifying continuous data distribution normality. Categorical variables were expressed as frequency and percentage. Quantitative variables were described using either mean ± standard deviation (SD) (for normally distributed) or median with interquartile range (for non-normally distributed). We used Receiver Operating Characteristic (ROC) curve analysis to test the diagnostic performance of manual and flow cytometric counting methods of ascitic fluid neutrophils for their ability to discriminate between patients with and without SBP. Their AUCs were compared, and cutoff values were determined according to the Youden index. Statistical significance was considered at p < 0.05.

3. Ethical approval

The Ethical Committee of Alexandria University approved the study’s protocol on October 21 2021 (Serial No.: 0201652, IRB No.: 00012098). We ensured compliance with the 1975 Declaration of Helsinki provisions and Good Clinical Practice guidelines. Informed consent was obtained from all subjects included in the study.

4. Results

In the present study involving 140 hospitalized cirrhotic patients with ascites who were clinically suspected to have SBP. Fever was present in 23 patients (16.4%). Abdominal pain was present in 42 patients (30.0%). Hepatic encephalopathy affected 56 patients (40.0%). Variceal bleeding was the presenting complaint in 35 patients (25.0%), while 27 patients (19.3%) presented with HRS-AKI. All patients had high SAAG ascites. Table 1 presents clinical, laboratory, and demographic information.

Table 1. Clinical and biochemical characteristics of the studied population.

	Clinically suspected SBP (n = 140)
Age (y), Median (IQR)	60.0 (55.0–65.0)
Gender Males, n (%) Females, n (%)	93 (66.4) 47 (33.6)
Fever, n (%)	23 (16.4)
Abdominal pain, n (%)	42 (30.0)
HE, n (%) Grade 1,2 Grade 3,4	56 (40.0) 31 (22.1) 25 (17.9)
Variceal bleeding, n (%)	35 (25.0)
Child class B, n (%) C, n (%)	64 (45.7) 76 (54.3)
Child score, Median (IQR)	10.0 (8.25–11.0)
Ascites grade, n (%) Moderate Severe	107 (76.4) 33 (23.6)
HRS-AKI, n (%) AKI on admission AKI 48 h after admission	36 (25.7) 27 (19.3) 9 (6.4)
MELD, Median (IQR)	17.0 (14.0–20.0)
Serum Na (mEq/L), Mean ± SD	127.19± 5.28
Serum albumin (g/dL), Median (IQR)	2.69 (2.46–2.99)
Serum AST (U/L), Median (IQR)	52.0 (43.0–61.0)
Serum ALT (U/L), Median (IQR)	32.0 (26.0–39.0)
Serum total bilirubin (mg/dL), Median (IQR)	2.65 (1.73–3.7)
Serum total protein (g/dL), Mean ± SD	6.62± .82
INR, Median (IQR)	1.64 (1.38–1.99)
Hemoglobin (g/dL), Mean ± SD	10.31± 1.83
Platelet count (x10^3/mm^3), Mean ± SD	108.81± 32.45
White blood cell count (x10^3/mm^3), Median (IQR)	6.50 (4.89–9.02)
Manual Ascitic Neutrophils (cell/mm^3), Median (IQR)	187.0 (112.5–440.0)
Flow cytometric Ascitic Neutrophils (cell/mm^3), Median (IQR)	214.5 (140.25–628.0)
SBP diagnosed by manual method, n (%)	53 (37.9)
SBP diagnosed by flow cytometry, n (%)	64 (45.7)
Ascitic culture, n: Neutrophils ≥ 250 by flow cytometry, n (%) Neutrophils ≥ 250 by manual method, n (%)	14 14 (21.9) 11 (20.8)
Ascitic albumin (g/dL), Median (IQR)	0.42 (0.28–0.59)
Ascitic protein (g/dL), Median (IQR)	1.61 (1.26–2.05)
Ascitic glucose (mg/dL), Median (IQR)	121.5 (100.0–153.5)
Ascitic LDH (U/L), Median (IQR)	86.0 (69.25–103.75)

ALT, Alanine aminotransferase; AST, Aspartate aminotransferase; INR, International normalization ratio; HRS-AKI, Hepatorenal syndrome-Acute kidney injury; IQR, Interquartile range; K, Potassium; MELD, Model for End-Stage Liver Disease; Na, Sodium; SBP, Spontaneous bacterial peritonitis; SD, Standard deviation.

*Statistically significant at p ≤ 0.05.

Fifty three patients (37.9%) were identified with SBP by the manual method, while flow cytometry identified 64 patients (45.7%) to have SBP. The ascitic fluid culture was positive in 14 patients, of which 64.28% showed E. coli and the remaining 35.71% were Klebsiella species. All patients with positive cultures were classified as having SBP by flow cytometry, but three cases with positive ascitic cultures were missed by the manual method.

Both the manual method and flow cytometry assessment of ascitic fluid neutrophils showed a moderate degree of agreement (Cohen’s κ = 0.548). By plotting the ROC curve, we found that the sensitivity, specificity, PPV, and NPV of the manual ascitic fluid method when compared to the flow cytometric method were 67.19%, 86.84%, 81.13%, and 75.86%, respectively, in discriminating patients with SBP from patients without SBP at a cutoff value of 250 cells/mm³ (AUC = 0.945, 95% CI, 0.91–0.97, p < 0.001) (Table 2, Figure 2).

Figure 2. Receiver operating characteristics curve for the manual method of ascitic neutrophil counting in diagnosing spontaneous bacterial peritonitis.

Table 2. Diagnostic performance of the manual method of neutrophil counting compared to using flow cytometry markers in differentiation between SBP and non-SBP groups.

	Manual Ascitic Neutrophils
AUC (95% CI)	0.945 (0.91–0.97)
Cutoff point	≥250
Sensitivity% (95% CI)	67.19 (54.31–78.41)
Specificity% (95% CI)	86.84 (77.13–93.51)
PPV% (95% CI)	81.13 (70.19–88.71)
NPV% (95% CI)	75.86 (68.65–81.85)
Accuracy	77.86%
P	<0.001*

AUC, Area Under the Curve; CI, Confidence interval; NPV, Negative predictive value; PPV, Positive predictive value.

*Statistically significant at p ≤ 0.05.

5. Discussion

The PMN cell count in ascites is a key tool in the diagnosis of SBP and should be conducted immediately. Our study used flow cytometry to count PMN based on their immunophenotyping: CD45-positive, CD15- and CD16-positive, and HLA-DR-negative. The manual method was compared to the flow cytometric method. The manual method, which may have a high level of inter-observer variation, showed a low level of sensitivity (67.19%) and specificity (86.84%) at a cutoff of 250 PMNs/mm³. The manual counting chamber has been traditionally utilized to quantify cell concentrations in body fluids, but it presents notable analytical and logistical challenges [22]. Due to the Bürker manual chamber’s design, not all leukocytes are included in the count; for example, if the leukocyte concentration in the sample is 250 cells per µl, only approximately 125 cells can be counted. The largest source of error is the lack of reproducibility between technicians [22].

A review article on SBP diagnosis stated that the ascitic fluid cell count and differential are conducted manually without established quality control procedures [24]. Manual leukocyte differentiation has logistical challenges. The manual process could lead to misclassification of ascitic fluid leukocytes, resulting in either an overestimation or underestimation of neutrophil count [25].

Several laboratories utilized different automated cell counters to determine the PMN cell count in ascitic fluid. These counters had varying levels of sensitivity and specificity when compared to the traditional manual microscopic method [26–29]. The Technicon System H * 1 showed sensitivities above 94% and specificities above 97.7% [26,27]. The Coulter Gen-S and ADVIA systems had sensitivities above 98% and specificities above 92% [28]. The Cell-dyn 3700 demonstrated a sensitivity of 93.3% and a specificity of 98.2% [29]. The Sysmex XE-5000 had a diagnostic accuracy of 98% [30].

Some researchers have suggested using monoclonal antibody combinations for PMN counting by flow cytometry [31]. The flow cytometric test is simple to do and provides results in 30 minutes. This method is quicker and involves less manipulation compared to doing manual counting and differentiation. Immunology openly defines several thousand cells through counting [17]. This can reduce the risk of misclassifying cells and eliminate inter-observer variability [32]. The flow cytometric method accurately counts the total number of white blood cells and their different subpopulations, even when non-hematopoietic cells are present [17].

The current study concentrated on utilizing the flow cytometric approach to quantify ascitic fluid neutrophils in individuals with SBP, aiming to reach quick and precise measures to detect SBP, which can have the potential for better patient outcomes.

One of the early benchmark publications on SBP diagnosis that was released early in the eighties of last century indicated that the sensitivity and specificity of the manual approach for diagnosing SBP at a cutoff of ≥ 250 PMNs/mm³ were 100% and 94%, respectively, when compared to bacterial culture results [33,34]. Newer studies, on the other hand, show that ascitic fluid bacterial cultures are not very sensitive. EASL guidelines recommend diagnosing SBP when the neutrophil count is ≥ 250 cells/mm³ using either the widely used traditional manual counting or the newly presented flow cytometric method, as its sensitivity and specificity approach 100% [19].

The results of our study show that manual counting of ascitic fluid neutrophils is less accurate. Enhancing the sensitivity of the detection method is essential for prompt and precise diagnosis of SBP, which can greatly influence patient care and results. Flow cytometry is commonly used in diagnosing leukemia and lymphoma; however, there is little information on its application in counting ascitic fluid PMNs. Only two studies in the literature have investigated the potential use of flow cytometry for counting ascitic fluid cells to mitigate the drawbacks of the manual method [17,35]. Our findings are in line with these earlier investigations.

Van de Geijn et al. [17] conducted a study indicating that flow cytometry is suitable for cell counting in ascitic fluid. In this study, the combined use of the LH750 automated method (for counting ascitic fluid leukocytes) plus the traditional manual method (for differentiating and counting ascitic fluid PMN) was 100% sensitive and 65% specific to detect PMN at a cutoff of ≥ 250 cells/mm³, which was inferior in comparison to the faster and more accurate flow cytometric method. Osman H et al.‘s [35] study reported that flow cytometry demonstrated 100% sensitivity and specificity, whereas manual PMN count showed a sensitivity of 65.52% and a specificity of 90%.

In our study, the low sensitivity and less than-ideal specificity of the manual method point to the possibility of missing diagnosing patients with SBP, raising the risk of morbidity and mortality [1,6], the possibility of falsely positively diagnosing SBP with subsequent misuse of the resources, and an increase in the risk of antibiotic misuse and resistance [36].

In conclusion, the flow cytometric method for ascitic neutrophil counting can introduce an efficient strategy for diagnosing SBP. Its implementation in clinical settings could significantly improve the accurate diagnosis and management of this life-threatening condition, ultimately enhancing patient outcomes. However, the flow cytometric method has higher costs, and cost-effectiveness studies should also precede its implementation in daily practice in the future.

Acknowledgment

We acknowledge Dr. Salma Saeid, Clinical Pathology Department, Alexandria University, for the unlimited assistance throughout the study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Islam Moheb

Islam Moheb, Assistant lecturer of Internal Medicine and Hepatology, Faculty of Medicine, Alexandria University, Alexandria, Egypt.

Ehab Moustafa Hassouna

Ehab Mostafa Hassouna, Professor of Internal Medicine and Hepatology, Faculty of Medicine, Alexandria University, Alexandria, Egypt.

Reham Abdel Haleem Abo Elwafa

Iman Ezzat El-Gohary, Professor of Internal Medicine and Nephrology, Faculty of Medicine, Alexandria University, Alexandria, Egypt.

Reham Abdel Haleem Abo Elwafa, Professor of Clinical and Chemical Pathology, Faculty of Medicine, Alexandria University, Alexandria, Egypt.

Ahmed Kamal

Ahmed Kamal, Lecturer of Internal Medicine and Hepatology, Faculty of Medicine, Alexandria University, Alexandria, Egypt.