Cirrhosis of liver is a major cause of morbidity and mortality and is commonly complicated by portal hypertension. One of the most clinically important consequences of portal hypertension is the development of esophageal varices, which may lead to life-threatening upper gastrointestinal bleeding.[1] Variceal bleeding is associated with significant mortality, need for hospitalization, blood transfusion, endoscopic intervention, and long-term prophylactic therapy.[1,2] Upper gastrointestinal endoscopy remains the gold standard for diagnosis and grading of esophageal varices. It also helps in identifying high-risk varices that require primary prophylaxis with non-selective beta blockers or endoscopic variceal ligation.[1,2] However, routine endoscopic screening of all cirrhotic patients may be difficult in resource-limited settings because endoscopy is invasive, costly, uncomfortable for patients, and not universally available.[2] Because of these limitations, several studies have evaluated non-invasive predictors of esophageal varices and large esophageal varices. Simple clinical, hematological, biochemical, and ultrasonographic markers may help identify patients at high risk and prioritize them for early endoscopy.[3,4] Such predictors may be particularly useful in primary care and secondary care settings, where access to endoscopy may be delayed. Thrombocytopenia is one of the most frequently studied non-invasive markers of portal hypertension and esophageal varices. In cirrhosis, low platelet count may occur due to hypersplenism, splenic sequestration, reduced thrombopoietin production, and advanced liver disease.[5] Platelet count alone, however, may not fully represent portal hypertension because it can be affected by multiple mechanisms. Therefore, combined parameters such as platelet count/spleen diameter ratio have also been proposed as useful predictors of esophagealvarices.[5,6] Splenomegaly is another important manifestation of portal hypertension. Increased splenic diameter reflects splenic congestion and hypersplenism due to elevated portal pressure.[7] Ultrasonographic measurement of spleen size is simple, non-invasive, widely available, and routinely performed in cirrhotic patients. Several studies have shown that splenic diameter and platelet count/spleen diameter ratio may help predict the presence of esophageal varices and large esophagealvarices.[6,7] Portal vein diameter is also an important ultrasonographic marker of portal hypertension. Dilatation of the portal vein may reflect increased portal venous pressure and altered splanchnic hemodynamics.[4,8] Previous studies have reported that portal vein diameter, spleen size, platelet count, and Child-Pugh class may be associated with esophageal varices in patients with cirrhosis.[8,9] Indian studies have also evaluated non-invasive markers for prediction of large esophageal varices. Sarangapani et al. reported that platelet count, palpable spleen, splenic size, portal vein size, and platelet count/spleen diameter ratio were useful predictors of large esophageal varices.[9] Kumar et al. also observed that platelet count and ultrasonographic parameters such as spleen size and portal vein size were significantly associated with large esophageal varices in cirrhotic patients.[10] Although endoscopy remains essential for definitive diagnosis, non-invasive parameters may help identify high-risk cirrhotic patients who require urgent endoscopic evaluation. Therefore, the present study was undertaken to evaluate clinical, hematological, and ultrasonographic predictors of large esophageal varices in patients with cirrhosis of liver, with special emphasis on platelet count, splenic diameter, and portal vein diameter. Aim To evaluate non-invasive clinical, hematological, and ultrasonographic predictors of large esophageal varices in patients with cirrhosis of liver. Objectives 1. To compare demographic, clinical, hematological, biochemical, and ultrasonographic parameters between cirrhotic patients with large and small esophageal varices. 2. To assess the diagnostic performance of platelet count, splenic diameter, and portal vein diameter in predicting large esophageal varices among patients with cirrhosis.

Study Design and Setting This was a hospital-based observational analytical study conducted in the Department of Gastroenterology, Apollo Institute of Medical Sciences, over a period of one year ( June 2023 to May 2024). Study period: June 2023 to May 2024. Duration: One years Study Population A total of 250 patients with cirrhosis of liver were assessed for eligibility. Among them, 59 patients were excluded based on predefined exclusion criteria. Finally, 191 patients were included in the final analysis. Patients were classified according to endoscopic findings into two groups: Group 1: Patients with large esophageal varices Group 2: Patients with small esophageal varices Inclusion Criteria 1. Patients diagnosed with cirrhosis of liver based on clinical, biochemical, radiological, and/or endoscopic findings. 2. Patients aged 18 years and above. 3. Patients who underwent upper gastrointestinal endoscopy for screening or evaluation of esophageal varices. 4. Patients willing to participate in the study. Exclusion Criteria 1. Patients with previous gastrointestinal bleeding who were already on beta blockers or had undergone endoscopic variceal ligation. 2. Patients presenting with active gastrointestinal bleeding. 3. Patients with portal vein thrombosis. 4. Patients with hepatocellular carcinoma. 5. Patients with incomplete clinical, laboratory, ultrasonographic, or endoscopic data. Data Collection A structured proforma was used to record demographic details, clinical history, etiology of cirrhosis, presence and severity of ascites, history of repeated paracentesis, hepatic encephalopathy, hepatorenal syndrome, and spontaneous bacterial peritonitis. The etiology of cirrhosis was classified as alcohol-related, hepatitis B virus-related, hepatitis C virus-related, cryptogenic, autoimmune, NASH-related, Budd-Chiari syndrome, or other causes. Clinical Assessment All patients underwent detailed clinical examination. Ascites was classified as absent, mild to moderate, or severe. Child-Turcotte-Pugh score and MELD score were calculated for all patients to assess liver disease severity. Laboratory Investigations The following laboratory parameters were recorded: 1. Hemoglobin 2. Total leukocyte count 3. Platelet count 4. Prothrombin time 5. International normalized ratio 6. SGOT 7. SGPT 8. Serum albumin 9. Total bilirubin 10. Serum creatinine 11. Serum sodium Platelet count <1 lakh/mm³ was considered as thrombocytopenia for predictor analysis. Ultrasonographic Assessment All patients underwent abdominal ultrasonography. The following ultrasonographic parameters were recorded: 1. Splenic diameter 2. Portal vein diameter 3. Presence of ascites 4. Other relevant features of portal hypertension, wherever available Splenic diameter >120 mm and portal vein diameter >13 mm were considered significant cut-off values for prediction of large esophageal varices. Endoscopic Assessment Upper gastrointestinal endoscopy was performed in all patients. Esophageal varices were classified as small or large according to endoscopic appearance. Large esophageal varices were defined as varices that were large, protruded significantly into the esophageal lumen, or were considered clinically significant/high-risk on endoscopic assessment. Outcome Variable The primary outcome variable was the presence of large esophageal varices. Predictor Variables The main predictor variables included platelet count <1 lakh/mm³, splenic diameter >120 mm, and portal vein diameter >13 mm. Other variables included age, gender, etiology of cirrhosis, ascites, Child-Pugh class, MELD score, hemoglobin, TLC, prothrombin time, INR, serum albumin, bilirubin, liver enzymes, and history of repeated paracentesis. Statistical Analysis Data were entered and analyzed using appropriate statistical software. Categorical variables were expressed as frequency and percentage. Continuous variables were expressed as mean ± standard deviation and median where applicable. The chi-square test or Fisher’s exact test was used to compare categorical variables. The independent t-test or Mann–Whitney U test was used to compare continuous variables between patients with large and small varices, depending on data distribution. Univariate logistic regression analysis was performed to identify factors associated with large esophageal varices. Variables that were clinically relevant or statistically significant on univariate analysis were entered into multivariate logistic regression analysis to identify independent predictors. Diagnostic performance of significant non-invasive predictors was assessed using sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. Receiver operating characteristic curve analysis was performed to determine area under the curve for platelet count, splenic diameter, and portal vein diameter. A p-value of less than 0.05 was considered statistically significant.

A total of 250 patients with cirrhosis were assessed for eligibility. Among them, 59 patients were excluded due to predefined exclusion criteria. The excluded patients included 29 patients with previous gastrointestinal bleed on beta blocker or endoscopic variceal ligation, 14 patients with active gastrointestinal bleed, 11 patients with portal vein thrombosis, and 5 patients with hepatocellular carcinoma. Finally, 191 patients were included in the final analysis. Patient Flow Screening status Number of patients Total patients assessed 250 Excluded patients 59 Previous GI bleed on beta blocker or EVL 29 Active GI bleed 14 Portal vein thrombosis 11 Hepatocellular carcinoma 5 Patients included in final analysis 191 The present cohort included 148 males, 77.5%, and 43 females, 22.5%. The mean age of the study population was 43.8 ± 12.6 years. Alcohol-related cirrhosis was the most common etiology, observed in 106 patients, 55.5%, followed by cryptogenic cirrhosis in 31 patients, 16.2%, HBV-related cirrhosis in 25 patients, 13.1%, autoimmune-related cirrhosis in 12 patients, 6.3%, NASH-related cirrhosis in 10 patients, 5.2%, HCV-related cirrhosis in 6 patients, 3.1%, and Budd-Chiari syndrome in 1 patient, 0.5%. Out of 191 patients, 132 patients, 69.1%, had large esophageal varices, while 59 patients, 30.9%, had small esophageal varices. Flow diagram showing total patients assessed, excluded patients with reasons for exclusion, and final classification into large and small esophageal varices groups. Table 1: Comparative Analysis of Demographic Characteristics According to Type of Varices Variable Large varices n=132 Small varices n=59 Total n=191 p-value Gender 0.421 Female 26, 19.7% 17, 28.8% 43 Male 106, 80.3% 42, 71.2% 148 Etiology of cirrhosis Alcohol 74, 56.1% 32, 54.2% 106 0.873 Autoimmune 6, 4.5% 6, 10.2% 12 0.142 Budd-Chiari syndrome 1, 0.8% 0, 0.0% 1 0.502 Cryptogenic 20, 15.2% 11, 18.6% 31 0.552 HBV 18, 13.6% 7, 11.9% 25 0.739 HCV 5, 3.8% 1, 1.7% 6 0.438 NASH 8, 6.1% 2, 3.4% 10 0.439 Ascites 0.647 No ascites 15, 11.4% 10, 16.9% 25 Mild to moderate 4, 3.0% 2, 3.4% 6 Severe 113, 85.6% 47, 79.7% 160 Child-Pugh class 0.502 A 14, 10.6% 4, 6.8% 18 B 38, 28.8% 17, 28.8% 55 C 80, 60.6% 38, 64.4% 118 Most patients in both groups were males, and alcohol-related cirrhosis was the most common etiology. No statistically significant difference was observed between large and small varices groups with respect to gender, etiology of cirrhosis, ascites severity, or Child-Pugh class. Bar diagram showing the proportion of patients with large and small esophageal varices among the final study population. Table 2: Clinical and Hematological Characteristics According to Type of Varices Parameter Large varices n=132 Mean ± SD Median Small varices n=59 Mean ± SD Median p-value Age, years 44.2 ± 12.9 45.0 42.9 ± 11.8 42.0 0.510 NS Hemoglobin, g/dL 8.7 ± 2.0 8.6 9.5 ± 1.9 9.4 0.013 S TLC, cells/mm³ 7800 ± 5250 6500 10650 ± 7400 8700 0.003 S Platelet count, /mm³ 102000 ± 98000 88000 168000 ± 111000 142000 <0.001 S Prothrombin time, seconds 17.3 ± 6.4 15.6 17.7 ± 5.6 16.4 0.680 NS INR 1.61 ± 0.60 1.42 1.64 ± 0.56 1.50 0.750 NS Splenic diameter, mm 125.6 ± 6.2 126.0 112.8 ± 10.9 113.0 <0.001 S Portal vein diameter, mm 16.3 ± 1.8 16.0 12.8 ± 1.4 13.0 <0.001 S CTP score 10.4 ± 2.2 10.0 10.8 ± 2.0 11.0 0.220 NS MELD score 18.4 ± 5.6 18.0 18.9 ± 5.2 19.0 0.550 NS Patients with large esophageal varices had significantly lower hemoglobin, TLC, and platelet count compared with patients with small varices. Splenic diameter and portal vein diameter were significantly higher in patients with large varices. Age, prothrombin time, INR, CTP score, and MELD score were not significantly associated with variceal size. Patients with large esophageal varices had significantly lower platelet count compared with patients with small esophageal varices. Mean splenic diameter and portal vein diameter were significantly higher among patients with large esophageal varices compared with small esophageal varices. Table 3: Biochemical Parameters According to Type of Varices Parameter Large varices n=132 Mean ± SD Small varices n=59 Mean ± SD p-value SGOT, U/L 78.4 ± 40.2 75.6 ± 38.5 0.660 NS SGPT, U/L 51.8 ± 28.0 50.6 ± 25.9 0.780 NS Serum albumin, g/dL 2.58 ± 0.55 2.62 ± 0.50 0.630 NS Total bilirubin, mg/dL 3.6 ± 2.9 3.4 ± 2.6 0.650 NS Serum creatinine, mg/dL 1.02 ± 0.42 0.98 ± 0.38 0.500 NS Serum sodium, mEq/L 132.6 ± 5.4 133.1 ± 5.0 0.470 NS No statistically significant difference was observed between patients with large and small varices with respect to SGOT, SGPT, serum albumin, total bilirubin, serum creatinine, or serum sodium. This suggests that routine biochemical parameters alone were not reliable predictors of large esophageal varices in the present study. Adjusted odds ratio plot showing portal vein diameter >13 mm, splenic diameter >120 mm, and platelet count <1 lakh/mm³ as independent predictors of large esophageal varices. Risk Factors for Large Esophageal Varices On univariate analysis, repeated paracentesis, platelet count <1 lakh/mm³, portal vein diameter >13 mm, and splenic diameter >120 mm were significantly associated with large esophageal varices. Hepatic encephalopathy, hepatorenal syndrome, spontaneous bacterial peritonitis, and Child-Pugh class were not significantly associated with large varices. Table 4: Univariate Analysis for Predicting Large Esophageal Varices Parameter Large varices n=132 Small varices n=59 Total Odds ratio, 95% CI p-value Hepatic encephalopathy 31, 23.5% 10, 16.9% 41 1.50, 0.68–3.32 0.309 NS Hepatorenal syndrome 32, 24.2% 9, 15.3% 41 1.78, 0.79–4.01 0.162 NS Spontaneous bacterial peritonitis 12, 9.1% 5, 8.5% 17 1.08, 0.36–3.22 0.890 NS Repeated paracentesis 72, 54.5% 17, 28.8% 89 2.97, 1.53–5.73 0.001 S Platelet count <1 lakh/mm³ 82, 62.1% 15, 25.4% 97 4.81, 2.43–9.53 <0.001 S Portal vein diameter >13 mm 128, 97.0% 16, 27.1% 144 86.00, 27.26–271.29 <0.001 S Child-Pugh class A 14, 10.6% 4, 6.8% 18 1.63, 0.51–5.19 0.403 NS Splenic diameter >120 mm 115, 87.1% 7, 11.9% 122 50.25, 19.65–128.54 <0.001 S Portal vein diameter >13 mm showed the strongest association with large esophageal varices on univariate analysis. Splenic diameter >120 mm and platelet count <1 lakh/mm³ were also significantly associated with large varices. Repeated paracentesis showed a moderate but statistically significant association. Table 5: Multivariate Logistic Regression Analysis for Predicting Large Esophageal Varices Parameter p-value Adjusted odds ratio 95% confidence interval Portal vein diameter >13 mm <0.001 S 58.20 9.80–345.40 Splenic diameter >120 mm <0.001 S 31.60 7.90–125.60 Platelet count <1 lakh/mm³ 0.003 S 9.20 2.10–39.60 After adjustment for other clinical and laboratory variables, portal vein diameter >13 mm, splenic diameter >120 mm, and platelet count <1 lakh/mm³ remained independent predictors of large esophageal varices. Portal vein diameter >13 mm emerged as the strongest independent predictor. Table 6: Diagnostic Values of Significant Predictors for Diagnosing Large Esophageal Varices Diagnostic parameter Splenic diameter >120 mm Portal vein diameter >13 mm Platelet count <1 lakh/mm³ Sensitivity, % 87.12 96.97 62.12 Specificity, % 88.14 72.88 74.58 Positive predictive value, % 94.26 88.89 84.54 Negative predictive value, % 75.36 91.49 46.81 Accuracy, % 87.43 89.53 65.97 Portal vein diameter >13 mm had the highest sensitivity and overall accuracy for predicting large esophageal varices. Splenic diameter >120 mm had the highest specificity and positive predictive value. Platelet count <1 lakh/mm³ was significant but had comparatively lower diagnostic accuracy. Comparison of sensitivity, specificity, and accuracy of splenic diameter >120 mm, portal vein diameter >13 mm, and platelet count <1 lakh/mm³. Table 7: ROC Analysis of Non-Invasive Predictors for Large Esophageal Varices Predictor Optimal cut-off AUC Sensitivity, % Specificity, % Interpretation Portal vein diameter >13 mm 0.925 96.97 72.88 Excellent predictor Splenic diameter >120 mm 0.912 87.12 88.14 Excellent predictor Platelet count <1 lakh/mm³ 0.681 62.12 74.58 Fair predictor ROC analysis showed that portal vein diameter had the highest AUC, followed closely by splenic diameter. Platelet count had only fair discriminative ability. These findings indicate that ultrasonographic parameters, especially portal vein diameter and splenic diameter, are better non-invasive predictors of large esophageal varices than platelet count alone. Area under the curve comparison showing portal vein diameter as the best predictor, followed by splenic diameter and platelet count. Table 8: Comparison of Present Study with Previous Studies Ref. Study Sample / design Key predictors assessed Important numerical findings Comparison with present study Present study Present cohort 191 cirrhotic patients; 132 large varices and 59 small varices Platelet count, splenic diameter, portal vein diameter Large varices: 69.1%; PVD >13 mm AUC 0.925; splenic diameter AUC 0.912; platelet count AUC 0.681 PVD >13 mm was the strongest predictor, followed by splenic diameter and platelet count. 11 Bishnoi et al. Cirrhotic patients with large and small varices PVD, splenic diameter, platelet count PVD >13 mm AUC 0.929; splenic diameter AUC 0.922; platelet count AUC 0.684 Very close to present study; both showed PVD and splenic diameter as excellent predictors. 12 Kumar et al., 2020 50 cirrhotic patients Platelet count, splenomegaly, ascites, USG parameters Male 76%; alcohol 64%; splenomegaly significant, p=0.004; platelet count significant, p<0.0002 Supports present study; platelet count and splenic enlargement were significant predictors. 13 Cherian et al., 2011 229 cirrhotic patients CTP class, platelet count, spleen diameter, PVD Large varices 35.4%; platelet <90,000 OR 2.7; spleen >160 mm OR 3.1; CTP B/C OR 3.8 Supports platelet and spleen size findings; differs because CTP class was not significant in present study. 14 Thomopoulos et al., 2003 184 cirrhotic patients Platelet count, spleen size, ascites EV present in 50%; large varices 17.9%; platelet count, spleen size, and USG ascites were independent predictors Supports platelet and spleen association; ascites was not significant in present study. 15 Kothari et al., 2019 202 male alcoholic cirrhosis patients Platelet count, spleen bipolar diameter, PC/SD ratio, APRI, FIB-4 Large EV 65.84%; spleen BPD significant, p=0.0034; PC/SD AUC 0.656 for EV; PC AUC 0.79 for bleeding Large EV prevalence close to present study; present study showed stronger AUC for PVD and splenic diameter. 16 Sarangapani et al., 2010 106 chronic liver disease patients Platelet count, spleen size, PVD, PC/SD ratio EV 72.6%; large EV 41.1%; platelet 157725 vs 202781, p=0.02; spleen 149 vs 111 mm, p=0.0001; PVD 13.9 vs 11.3 mm, p=0.001; PC/SD AUC 0.883 Strongly supports present study; platelet count, spleen size, and PVD were significant predictors. 20 Giannini et al., 2003 Cirrhotic patients Platelet count/spleen diameter ratio PC/SD ratio ≤909 proposed as non-invasive predictor of esophageal varices Supports combined platelet–spleen approach; present study separately confirms platelet and splenic diameter association. 22 Ying et al., 2012 Meta-analysis Platelet count/spleen diameter ratio PC/SD ratio showed useful diagnostic performance for esophageal varices Supports non-invasive prediction approach in cirrhosis. 24 Deng et al., 2015 Systematic review/meta-analysis APRI, AAR, FIB-4, FI, Lok, Forns, FibroIndex Non-invasive scores showed variable accuracy Present study suggests simple ultrasound parameters may perform better than routine serum-based scores.

Esophageal varices are an important complication of portal hypertension in patients with cirrhosis, and large esophageal varices are clinically significant because of their increased risk of bleeding. Although upper gastrointestinal endoscopy remains the gold standard for detection and grading of varices, it is invasive, resource-intensive, and may not be readily available in all clinical settings. Therefore, identification of simple, reproducible, and non-invasive predictors of large esophageal varices is clinically useful for selecting high-risk patients who require early endoscopic evaluation and prophylactic therapy. In the present study, 191 patients with cirrhosis were included in the final analysis. Large esophageal varices were present in 132 patients, 69.1%, while small varices were present in 59 patients, 30.9%. This proportion of large varices was higher than that reported by Cherian et al., who studied 229 patients and found large varices in 81 patients, 35.4%.[13] It was also higher than the study by Thomopoulos et al., where large esophageal varices were present in 33 of 184 patients, 17.9%.[14] In contrast, the finding was closer to Kothari et al., who studied 202 male patients with alcoholic cirrhosis and reported large esophageal varices in 133 patients, 65.84%.[15] The higher frequency of large varices in the present study may be related to the advanced liver disease profile of the cohort, with most patients belonging to Child-Pugh class B or C. Alcohol-related cirrhosis was the most common etiology in the present study, observed in 106 patients, 55.5%. This was comparable with Kumar et al., who reported alcohol as the cause of cirrhosis in 64% of patients.[12] Kothari et al. also studied an exclusively alcoholic cirrhosis cohort and reported that 65.84% of patients had large esophageal varices.[15] Similarly, Sarangapani et al. reported alcohol as the cause in 62 of 106 patients in a South Indian cohort.[16] These findings indicate that alcohol-related cirrhosis remains a common background etiology in Indian studies evaluating non-invasive predictors of esophageal varices. In the present study, large varices were not significantly associated with age, gender, etiology of cirrhosis, ascites grade, or Child-Pugh class. This partly differs from Cherian et al., who reported that Child-Pugh class B/C was independently associated with large varices, with an odds ratio of 3.8.[13] Kumar et al. also observed that the proportion of large esophageal varices increased with worsening Child-Pugh class, as 56% of Child-Pugh class C patients had large varices compared with 39.1% of Child-Pugh class B patients.[12] In the present study, Child-Pugh class did not emerge as a significant predictor, possibly because most patients were already in advanced disease categories, reducing the ability of Child-Pugh class to discriminate between large and small varices. Among hematological parameters, the present study showed that patients with large varices had significantly lower platelet count compared with patients with small varices: 102000 ± 98000/mm³ versus 168000 ± 111000/mm³, p<0.001. Platelet count <1 lakh/mm³ was significantly associated with large varices on univariate analysis, with an odds ratio of 4.81, and remained significant on multivariate analysis, with an adjusted odds ratio of 9.20. This finding is consistent with Kumar et al., who found platelet count to be the only significant hematological marker associated with large esophageal varices, with p<0.0002.[12] Cherian et al. also reported platelet count <90,000/µL as an independent predictor of large varices, with an odds ratio of 2.7.[13] Sarangapani et al. similarly reported significantly lower platelet counts among patients with large varices compared with those with none or small varices: 157725/mm³ versus 202781/mm³, p=0.02.[16] The relationship between thrombocytopenia and large varices is biologically plausible. Portal hypertension causes splenic congestion and hypersplenism, which leads to sequestration and destruction of platelets. In addition, reduced thrombopoietin production in chronic liver disease may further contribute to thrombocytopenia. However, platelet count alone may be influenced by several mechanisms other than portal hypertension, which explains why it was less accurate than ultrasonographic parameters in the present study. Ultrasonographic parameters showed the strongest association with large esophageal varices in the present study. Mean splenic diameter was significantly higher in patients with large varices than small varices: 125.6 ± 6.2 mm versus 112.8 ± 10.9 mm, p<0.001. Splenic diameter >120 mm was strongly associated with large varices on univariate analysis, with an odds ratio of 50.25, and remained independently significant on multivariate analysis, with an adjusted odds ratio of 31.60. This agrees with Sarangapani et al., who reported larger spleen size in patients with large varices: 149 mm versus 111 mm, p=0.0001.[16] Cherian et al. also found spleen bipolar diameter >160 mm to be an independent predictor of large varices, with an odds ratio of 3.1.[13] Kumar et al. reported clinically palpable spleen in 52% of patients and found a significant association between splenomegaly and large varices, with p=0.004.[12] Portal vein diameter was the strongest predictor in the present study. Mean portal vein diameter was significantly higher in the large varices group than in the small varices group: 16.3 ± 1.8 mm versus 12.8 ± 1.4 mm, p<0.001. Portal vein diameter >13 mm had the highest univariate odds ratio of 86.00 and remained the strongest independent predictor on multivariate analysis, with an adjusted odds ratio of 58.20. This finding is strongly supported by Sarangapani et al., who reported larger portal vein size in patients with large varices: 13.9 mm versus 11.3 mm, p=0.001, and found portal vein size >13 mm to be a significant predictor of large varices.[16] Cherian et al. also found portal vein diameter >13 mm to be significantly associated with the presence of varices, with an odds ratio of 2.421.[13] In contrast to platelet count and ultrasonographic parameters, routine biochemical markers did not significantly predict large esophageal varices in the present study. SGOT, SGPT, serum albumin, total bilirubin, creatinine, and sodium did not show significant association with variceal size. This finding is similar to Kumar et al., who reported that serum bilirubin, serum albumin, SGOT, and SGPT were not significantly associated with large esophageal varices.[12] Sarangapani et al. also observed that although patients with large varices showed features suggestive of more advanced disease, many biochemical parameters did not achieve statistical significance.[16] These findings suggest that routine biochemical tests alone are less reliable than platelet count and ultrasonographic markers for identifying large varices. On multivariate analysis in the present study, portal vein diameter >13 mm, splenic diameter >120 mm, and platelet count <1 lakh/mm³ remained independent predictors of large esophageal varices. The strongest predictor was portal vein diameter >13 mm, followed by splenic diameter >120 mm and platelet count <1 lakh/mm³. This is broadly comparable with Cherian et al., who identified Child-Pugh class B/C, platelet count <90,000/µL, and spleen diameter >160 mm as significant independent predictors of large varices.[13] However, unlike Cherian et al., portal vein diameter emerged as an independent predictor in the present study. This difference may be due to variation in study population, disease severity, etiological profile, and the fact that the present cohort included only patients with esophageal varices, whereas Cherian et al. included patients with no varices, small varices, and large varices. The diagnostic performance analysis in the present study showed that portal vein diameter >13 mm had the highest sensitivity and accuracy for predicting large esophageal varices. It showed 96.97% sensitivity, 72.88% specificity, 88.89% positive predictive value, 91.49% negative predictive value, and 89.53% accuracy. Splenic diameter >120 mm showed 87.12% sensitivity, 88.14% specificity, 94.26% positive predictive value, and 87.43% accuracy. Platelet count <1 lakh/mm³ showed lower diagnostic performance, with 62.12% sensitivity, 74.58% specificity, and 65.97% accuracy. This pattern indicates that ultrasonographic parameters, particularly portal vein diameter and splenic diameter, were superior to platelet count alone. ROC analysis further supported these observations. Portal vein diameter had the highest AUC of 0.925, followed by splenic diameter with an AUC of 0.912. Platelet count had a lower AUC of 0.681. These findings are comparable with the study model by Bishnoi et al., where portal vein diameter >13 mm had an AUC of 0.929, splenic diameter had an AUC of 0.922, and platelet count had an AUC of 0.684.[11] Sarangapani et al. reported an AUC of 0.883 for platelet count/spleen diameter ratio at a cut-off of 909, with sensitivity and specificity of 88.5% and 83%, respectively.[16] Kothari et al. reported that platelet count/spleen diameter ratio had an AUC of 0.656 for predicting esophageal varices, while platelet count and platelet count/spleen diameter ratio showed better AUCs of 0.79 and 0.78, respectively, for predicting variceal bleeding.[15] The platelet count/spleen diameter ratio has been widely discussed as a useful non-invasive marker for esophageal varices. Giannini et al. proposed a cut-off of ≤909 as an accurate non-invasive marker for predicting esophageal varices, and this was later validated in follow-up studies.[20,21] Sarangapani et al. also found that a platelet count/spleen diameter ratio <909 had good diagnostic performance for large varices, with 88.5% sensitivity, 83% specificity, 83.3% positive predictive value, and 90.5% negative predictive value.[16] Although the present study did not primarily analyze platelet count/spleen diameter ratio, the significant association of both platelet count and splenic diameter supports the same pathophysiological principle. Systematic reviews and meta-analyses have also evaluated non-invasive scores for predicting esophageal varices. Ying et al. reported that platelet count/spleen diameter ratio has useful diagnostic performance in cirrhosis.[22] Chawla et al. also reviewed the role of platelet count/spleen diameter ratio and concluded that it may help reduce unnecessary endoscopy in selected patients.[23] Deng et al. evaluated non-invasive indices such as APRI, AAR, FIB-4, Lok, Forns, and FibroIndex and found variable diagnostic accuracy for predicting esophageal varices.[24] Compared with these composite scores, the present study supports the use of simple ultrasound-based parameters, especially portal vein diameter and splenic diameter, which are readily available, inexpensive, and routinely assessed in cirrhotic patients. Overall, the present study demonstrates that portal vein diameter >13 mm, splenic diameter >120 mm, and platelet count <1 lakh/mm³ are useful non-invasive predictors of large esophageal varices. Among these, portal vein diameter >13 mm emerged as the most sensitive and accurate predictor, while splenic diameter >120 mm showed the highest specificity and positive predictive value. These findings support the use of routinely available ultrasonographic and hematological parameters for identifying cirrhotic patients at high risk for large esophageal varices, particularly in settings where immediate endoscopy may not be feasible.

The present study showed that portal vein diameter, splenic diameter, and platelet count are useful non-invasive predictors of large esophageal varices in patients with cirrhosis of liver. Among the evaluated parameters, portal vein diameter >13 mm emerged as the strongest independent predictor, with the highest sensitivity and overall diagnostic accuracy. Splenic diameter >120 mm also showed excellent predictive ability, particularly with high specificity and positive predictive value. Platelet count <1 lakh/mm³ was significantly associated with large esophageal varices but had lower diagnostic accuracy compared with ultrasonographic parameters. Routine biochemical parameters such as SGOT, SGPT, serum albumin, serum bilirubin, serum creatinine, and serum sodium were not significantly associated with variceal size. Therefore, simple and routinely available parameters such as portal vein diameter, splenic diameter, and platelet count may help identify cirrhotic patients at high risk for large esophageal varices and prioritize them for early endoscopic evaluation. Clinical Implications 1. Portal vein diameter and splenic diameter are simple ultrasonographic parameters that can be assessed during routine abdominal ultrasound in cirrhotic patients. 2. Portal vein diameter >13 mm may be used as an important warning marker for the presence of large esophageal varices. 3. Splenic diameter >120 mm and platelet count <1 lakh/mm³ can provide additional supportive evidence for identifying high-risk patients. 4. These non-invasive parameters should not replace endoscopy but may help prioritize patients who need urgent endoscopic screening. Strengths of the Study 1. The study included a clinically relevant sample of cirrhotic patients undergoing evaluation for esophageal varices. 2. The study focused on simple, routinely available, and low-cost parameters. 3. Both hematological and ultrasonographic predictors were evaluated. 4. Multivariate analysis was used to identify independent predictors of large esophageal varices. 5. Diagnostic performance was assessed using sensitivity, specificity, positive predictive value, negative predictive value, accuracy, and ROC analysis.

1. Garcia-Tsao G, Sanyal AJ, Grace ND, Carey WD. Prevention and management of gastroesophageal varices and variceal haemorrhage in cirrhosis. Am J Gastroenterol. 2007;102(9):2086-2102. 2. Jensen DM. Endoscopic screening for varices in cirrhosis: findings, implications, and outcomes. Gastroenterology. 2002;122(6):1620-1630. 3. Stefanescu H, Procopet B. Noninvasive assessment of portal hypertension in cirrhosis: liver stiffness and beyond. World J Gastroenterol. 2014;20(45):16811-16819. 4. Zhang W, Wang L, Wang L, et al. Liver stiffness measurement, better than APRI, FibroIndex, FIB-4, and NBI gastroscopy, predicts portal hypertension in patients with cirrhosis. Cell BiochemBiophys. 2015;71(2):865-873. 5. Giannini EG, Botta F, Borro P, Risso D, Romagnoli P, Fasoli A, et al. Platelet count/spleen diameter ratio: proposal and validation of a non-invasive parameter to predict the presence of esophageal varices in patients with liver cirrhosis. Gut. 2003;52(8):1200-1205. 6. Giannini EG, Zaman A, Kreil A, Floreani A, Dulbecco P, Testa E, et al. Platelet count/spleen diameter ratio for the non-invasive diagnosis of esophageal varices: results of a multicenter prospective validation study. Am J Gastroenterol. 2006;101(11):2511-2519. 7. Ying L, Lin X, Xie ZL, Hu YP, Shi KQ. Performance of platelet count/spleen diameter ratio for diagnosis of esophageal varices in cirrhosis: a meta-analysis. Dig Dis Sci. 2012;57(6):1672-1681. 8. Cherian JV, Deepak N, Ponnusamy RP, Somasundaram A, Jayanthi V. Non-invasive predictors of esophageal varices. Saudi J Gastroenterol. 2011;17(1):64-68. 9. Sarangapani A, Shanmugam C, Kalyanasundaram M, Rangachari B, Thangavelu P, Subbarayan JK. Noninvasive prediction of large esophageal varices in chronic liver disease patients. Saudi J Gastroenterol. 2010;16(1):38-42. 10. Kumar P, Singh K, Joshi A, Thakur P, Mahto SK, Kumar B, Pasricha N, Patra BR, Lamba BMS. Evaluation of non-invasive marker of esophageal varices in cirrhosis of liver. J Family Med Prim Care. 2020;9(2):992-996. 11. Bishnoi, et al. Can non-invasive parameters help us to predict large esophageal varices? International Journal of Contemporary Medical Research. 2020;7(10):J2-J6. 12. Kumar P, Singh K, Joshi A, Thakur P, Mahto SK, Kumar B, Pasricha N, Patra BR, Lamba BMS. Evaluation of non-invasive marker of esophageal varices in cirrhosis of liver. J Family Med Prim Care. 2020;9(2):992-996. doi:10.4103/jfmpc.jfmpc_854_19. 13. Cherian JV, Deepak N, Ponnusamy RP, Somasundaram A, Jayanthi V. Non-invasive predictors of esophageal varices. Saudi J Gastroenterol. 2011;17(1):64-68. 14. Thomopoulos KC, Labropoulou-Karatza C, Mimidis KP, Katsakoulis EC, Iconomou G, Nikolopoulou VN. Non-invasive predictors of the presence of large oesophageal varices in patients with cirrhosis. Dig Liver Dis. 2003;35(7):473-478. 15. Kothari HG, Gupta SJ, Gaikwad NR, Sankalecha TH, Samarth AR. Role of non-invasive markers in prediction of esophageal varices and variceal bleeding in patients of alcoholic liver cirrhosis from Central India. Turk J Gastroenterol. 2019;30(12):1036-1043. doi:10.5152/tjg.2019.18334. 16. Sarangapani A, Shanmugam C, Kalyanasundaram M, Rangachari B, Thangavelu P, Subbarayan JK. Noninvasive prediction of large esophageal varices in chronic liver disease patients. Saudi J Gastroenterol. 2010;16(1):38-42. 17. Garcia-Tsao G, Sanyal AJ, Grace ND, Carey WD. Prevention and management of gastroesophageal varices and variceal haemorrhage in cirrhosis. Am J Gastroenterol. 2007;102(9):2086-2102. 18. Chalasani N, Imperiale TF, Ismail A, Sood G, Carey M, Wilcox CM, et al. Predictors of large esophageal varices in patients with cirrhosis. Am J Gastroenterol. 1999;94(11):3285-3291. 19. Madhotra R, Mulcahy HE, Willner I, Reuben A. Prediction of esophageal varices in patients with cirrhosis. J Clin Gastroenterol. 2002;34(1):81-85. 20. Giannini EG, Botta F, Borro P, Risso D, Romagnoli P, Fasoli A, et al. Platelet count/spleen diameter ratio: proposal and validation of a non-invasive parameter to predict the presence of esophageal varices in patients with liver cirrhosis. Gut. 2003;52(8):1200-1205. 21. Giannini EG, Zaman A, Kreil A, Floreani A, Dulbecco P, Testa E, et al. Platelet count/spleen diameter ratio for the non-invasive diagnosis of esophageal varices: results of a multicenter prospective validation study. Am J Gastroenterol. 2006;101(11):2511-2519. 22. Ying L, Lin X, Xie ZL, Hu YP, Shi KQ. Performance of platelet count/spleen diameter ratio for diagnosis of esophageal varices in cirrhosis: a meta-analysis. Dig Dis Sci. 2012;57(6):1672-1681. 23. Chawla S, Katz A, Attar BM, Gupta A, Sandhu DS, Agarwal R. Platelet count/spleen diameter ratio to predict the presence of esophageal varices in patients with cirrhosis: a systematic review. Eur J Gastroenterol Hepatol. 2012;24(4):431-436. 24. Deng H, Qi X, Guo X. Diagnostic accuracy of APRI, AAR, FIB-4, FI, King, Lok, Forns, and FibroIndex scores in predicting the presence of esophageal varices in liver cirrhosis: a systematic review and meta-analysis. Medicine (Baltimore). 2015;94(42):e1795. 25. Stefanescu H, Procopet B. Noninvasive assessment of portal hypertension in cirrhosis: liver stiffness and beyond. World J Gastroenterol. 2014;20(45):16811-16819.