Chronic liver disease from alcohol, non-alcoholic fatty liver disease and viral hepatitis is increasingly common, particularly in developing countries such as India. Portal hypertension is among its most serious consequences, giving rise to esophageal varices and, in severe cases, life-threatening variceal hemorrhage. Variceal bleeding is one of the principal causes of death in patients with cirrhosis, especially those with decompensated disease, so early detection and risk stratification of varices are essential to reduce morbidity and mortality [5,11].

Upper gastrointestinal endoscopy has long been the reference standard for detecting and grading esophageal varices, but its routine use is limited by its invasiveness, patient discomfort, and the infrastructure and expertise it requires, particularly in resource-limited settings. Hepatic venous pressure gradient measurement, although the gold standard for quantifying portal pressure, is likewise invasive. There is therefore growing interest in non-invasive tools — including the platelet count/spleen diameter ratio, transient elastography, and

Doppler ultrasound — to predict the presence of varices and the risk of bleeding [4,10].

The splenoportal index (SPI), derived from Doppler ultrasound, is the ratio of splenic venous volume flow rate to portal venous volume flow rate. As intrahepatic resistance rises in advancing cirrhosis, portal blood flow is redistributed and splenic venous flow increases relative to portal flow; an SPI greater than 1.0 has been proposed as a marker of high bleeding risk. Yu et al. first defined the index and reported that the risk of variceal bleeding was increased when splenic venous flow exceeded portal venous flow on Doppler sonography [1]. Doppler ultrasound additionally allows assessment of liver parenchyma, splenomegaly, ascites and collaterals, providing a comprehensive, repeatable and radiation-free evaluation of portal hypertension. This study was undertaken to calculate the SPI in patients with parenchymal liver disease, to correlate it with parameters of portal hypertension and with esophageal varices, and to determine its sensitivity and specificity for the risk of variceal bleeding.

Study design and population

This hospital-based cross-sectional study was conducted in the Department of Radiodiagnosis, Tirunelveli Medical College and Hospital, over a period of 18 months, after institutional ethics committee approval. The sample size of 125 was calculated based on an expected prevalence of esophageal varices in cirrhotic patients of 77%, with 20% relative precision and a 10% allowance for non-response. Written informed consent was obtained from every participant.

 Eligibility

Patients with diagnosed parenchymal liver disease, suspected portal hypertension, or biochemical parameters suggestive of cirrhosis were included. Patients unwilling to participate or to undergo endoscopy, and those with portal vein thrombosis or cavernous transformation, prior portosystemic shunt surgery or liver transplantation, or hepatofugal portal flow, were excluded.

 Doppler technique and SPI calculation

Examinations were performed on a real-time ultrasound system with a 3.5-MHz convex probe and colour Doppler capability (Mindray DC80). The probe was placed along the long axis of the main portal vein and splenic vein, with the sample gate positioned in the main portal vein at the level of its entry into the liver parenchyma and in the splenic vein where it exits the splenic parenchyma. Three angle-corrected Doppler waveforms (insonation angle <60°) were obtained from each vessel. Mean flow velocity and cross-sectional area of the main portal and splenic veins were measured. Volume flow rate was calculated as cross-sectional area × mean velocity, and the splenoportal index was derived as the splenic vein volume flow rate divided by the portal vein volume flow rate.

 Endoscopic correlation

All patients were referred to the Department of Medical Gastroenterology for upper gastrointestinal endoscopy. Esophageal varices were graded by the Japanese Research Society for Portal Hypertension classification (grade I: small and straight; grade II: tortuous, occupying less than one third of the lumen; grade III: large, occupying more than one third of the lumen). The SPI was correlated with the presence and grade of varices and with a history of hematemesis.

 Statistical analysis

Data were analysed using SPSS version 16.0. Continuous variables are expressed as mean ± standard deviation and categorical variables as frequencies and percentages. Associations were tested using the chi-square test, with p<0.05 considered statistically significant. Receiver operating characteristic analysis and a 2×2 contingency table were used to determine the diagnostic performance of an SPI threshold of 1.0 for predicting a history of hematemesis.

Demographic and clinical profile

Of the 125 patients, the largest group was aged 55–64 years (37.6%); the mean age was 49.6 ± 11.1 years. Ninety-one patients (72.8%) were male. Jaundice was the commonest presenting feature (52 patients), followed by hematemesis (41), melena (30) and abdominal distension (15); bleeding per rectum and leg swelling were each seen in 6 patients, and 4 patients were asymptomatic. Decompensated chronic liver disease of alcoholic and non-specific cause was the predominant etiology (Tables 1 and 2).

Table 1. Demographic profile of the study population (n = 125)

  Table 2. Etiology of underlying liver disease

Doppler hemodynamic parameters

The mean splenic vein velocity was 15.2 ± 9.0 cm/sec and the mean portal vein peak systolic velocity was 12.5 ± 6.4 cm/sec, with the majority of patients showing low portal velocities (72.0% ≤12 cm/sec), consistent with portal hypertension. The mean splenoportal index was 0.83 ± 0.64; 48 patients (38.4%) had an SPI greater than 1 (Table 3).

Table 3. Summary of Doppler hemodynamic parameters (n = 125)

Endoscopic findings

Endoscopy showed an almost even distribution across categories: grade 1 varices in 31 patients (24.8%), grade 2 in 31 (24.8%), grade 3 in 32 (25.6%) and no varices in 31 (24.8%). A history of hematemesis was present in 41 patients (32.8%) (Table 4).

Table 4. Distribution of endoscopic findings

 Association of the splenoportal index with hematemesis

The SPI was strongly associated with a history of hematemesis (p<0.0001). Among the 41 patients with prior hematemesis, 36 (87.8%) had an SPI greater than 1, whereas among the 84 without hematemesis, 72 (85.7%) had an SPI below 1. The mean SPI was markedly higher in patients who had bled (1.41 vs 0.55) (Table 5).

Table 5. Splenoportal index versus history of hematemesis

Chi-square p<0.0001.

 Association of the splenoportal index with variceal grade

The SPI also correlated strongly with endoscopic variceal grade (p<0.0001). Every patient with no varices or grade 1 varices had an SPI below 1, whereas 30 of 32 patients with grade 3 varices (93.8%) had an SPI above 1. Mean SPI values rose from 0.24 (no varices) and 0.35 (grade 1) to 1.33 (grade 2) and 1.31 (grade 3) (Table 6).

Table 6. Splenoportal index versus endoscopic variceal grade

Chi-square p<0.0001.

Other Doppler parameters and bleeding

Splenic vein volume was significantly associated with both hematemesis (mean 23.7 vs 12.3 cm³; p<0.0001) and variceal grade (p<0.0001), higher volumes accompanying more severe varices. Splenic vein velocity and portal vein volume also differed significantly between bleeders and non-bleeders, whereas portal vein peak systolic velocity did not (p=0.34), indicating it is not an independent predictor of bleeding (Table 7).

Table 7. Doppler parameters by history of hematemesis

 Diagnostic performance of the splenoportal index

Taking an SPI threshold of 1.0 to predict a history of hematemesis, the 2×2 analysis yielded high diagnostic accuracy. Sensitivity was 87.8% and specificity 85.7%, with a positive predictive value of 75.0% and a negative predictive value of 93.5%; overall accuracy was 86.4% (Tables 8 and 9).

 Table 8. Contingency table: SPI (threshold 1.0) versus history of hematemesis

Table 9. Diagnostic performance of the splenoportal index for esophageal variceal bleeding

Disease (hematemesis) prevalence in the cohort 32.8%. Predictive values are prevalence-dependent.

 Treatment and outcome

Endoscopic band ligation was the commonest intervention (52 patients), followed by supportive care (45) and pharmacological therapy with octreotide and beta-blockers (34). At follow-up, 65 patients (52.0%) improved, 50 (40.0%) remained stable, and 10 (8.0%) deteriorated.

In this cross-sectional Doppler study of 125 patients with parenchymal liver disease, the splenoportal index emerged as a strong non-invasive predictor of esophageal variceal bleeding, correlating significantly with both a history of hematemesis and endoscopic variceal grade. Using a threshold of 1.0, the SPI achieved a sensitivity of 87.8%, a specificity of 85.7% and an overall accuracy of 86.4%, with a particularly high negative predictive value of 93.5%. These findings are in keeping with Yu et al., who first defined the index and reported that a higher SPI is substantially associated with the risk of variceal bleeding in cirrhotic patients [1].

The cohort was predominantly middle-aged and older (mean age 49.6 years), reflecting the latency required for cirrhosis and significant portal hypertension to develop, and echoing reports that portal hypertensive complications cluster in the fifth and sixth decades [5,6]. The male predominance (72.8%) is consistent with the higher burden of alcohol-related and viral liver disease in men and with the male preponderance reported among patients presenting with variceal bleeding [7]. Decompensated chronic liver disease of alcoholic and non-specific cause was the leading etiology (66.4%), with viral hepatitis and NAFLD-related disease next; the comparatively low proportion of NAFLD contrasts with Western series in which metabolic liver disease is an increasingly common cause of portal hypertension, underscoring regional differences in the etiological spectrum [9].

Among the individual Doppler parameters, splenic vein volume was significantly associated with both bleeding history and variceal severity, higher volumes accompanying higher grades — consistent with progressive splenic venous congestion in advancing portal hypertension and with the emerging role of splenic volumetrics in non-invasive risk assessment [12]. Portal vein peak systolic velocity, by contrast, did not differ significantly between bleeders and non-bleeders, implying that although reduced portal velocity reflects general portal hemodynamics, it is not by itself a reliable predictor of bleeding [10,13]. The strength of the SPI lies precisely in combining splenic and portal hemodynamics into a single ratio, which captures the flow redistribution characteristic of advanced cirrhosis better than either vessel alone.

The graded relationship between the SPI and variceal severity was striking: patients with no varices or grade 1 varices uniformly had an index below 1, whereas the great majority of those with grade 3 varices had an index above 1. This clear separation, together with the high negative predictive value, suggests that a low SPI could reasonably identify patients at low risk of significant varices, in whom endoscopy might be deferred, while a high SPI flags patients who should be prioritised for endoscopic screening and prophylaxis. Such a strategy aligns with the broader movement in hepatology towards non-invasive risk stratification and away from routine invasive hepatic venous pressure gradient measurement [10,11].

The main advantages of the SPI are that it is non-invasive, inexpensive, radiation-free and repeatable, and that the same Doppler examination also depicts ascites, splenomegaly, collaterals and parenchymal change. These features make it particularly attractive in resource-limited settings where endoscopic capacity is constrained. Its limitations are those of Doppler ultrasound generally: operator dependence, sensitivity to respiration and patient positioning, and reduced accuracy in obese patients or in the presence of large collaterals or shunts, which can occasionally produce misleading values.

 LIMITATIONS

This was a single-centre, cross-sectional study, which limits causal inference, and Doppler and endoscopic assessments are subject to inter-observer variability. The reference standard for the diagnostic-accuracy analysis was a history of hematemesis rather than a prospectively observed bleeding event, and the SPI threshold of 1.0 was pre-specified rather than derived from the ROC curve of this cohort. Larger, prospective, multicentre studies with serial Doppler measurements and prospective bleeding outcomes would further validate the predictive value of the index.

The splenoportal index is a simple, non-invasive, Doppler-derived parameter that integrates splenic and portal venous hemodynamics. In this study it correlated strongly with a history of hematemesis and with advanced-grade esophageal varices, and showed high sensitivity, specificity and negative predictive value for variceal bleeding. By providing a reliable estimate of the severity of portal hypertension, the SPI offers clinicians an accessible tool to identify high-risk patients who would benefit from early endoscopic surveillance and prophylaxis. Compared with invasive methods such as hepatic venous pressure gradient measurement, it is simpler, more cost-effective and safer, making it especially valuable in resource-limited settings, where it may help prioritise patients for endoscopy and guide the timing of therapy.

 DECLARATIONS

Ethics approval and consent: Approved by the institutional ethics committee; written informed consent was obtained from all participants.

Conflicts of interest: The authors declare no conflicts of interest.

Funding: This research received no specific grant from any funding agency.

1. Yu IK, Choi BI, Park JH, Han JK, Yeon KM, Kim CW. Splenoportal index on color Doppler ultrasound: value as an index for predicting the risk of variceal bleeding in patients with postnecrotic liver cirrhosis. J Korean Soc Ultrasound Med. 1996;15:163–70.
2. Leung PS, editor. The gastrointestinal system: gastrointestinal, nutritional and hepatobiliary physiology. Dordrecht: Springer; 2014.
3. Debas HT. Gastrointestinal surgery: pathophysiology and management. New York: Springer; 2003.
4. Antil N, Sureka B, Mittal MK, Malik A, Gupta B, Thukral BB. Hepatic venous waveform, splenoportal and damping index in liver cirrhosis: correlation with Child-Pugh's score and oesophageal varices. J Clin Diagn Res. 2016;10(2):TC01–5.
5. Bosch J, Abraldes JG, Berzigotti A, García-Pagán JC. Portal hypertension and variceal bleeding: unresolved issues. Liver Int. 2008;28(2):140–5.
6. Sarin SK, Kumar A. Management of variceal hemorrhage. Trop Gastroenterol. 2010;31(3):163–9.
7. Blendis LM, Wong F. The hyperdynamic circulation in cirrhosis: an overview. Pharmacol Ther. 2003;99(2):139–44.
8. D'Amico G, Garcia-Tsao G, Pagliaro L. Natural history and prognostic indicators of survival in cirrhosis: a systematic review. J Hepatol. 2006;44(1):217–31.
9. Tsochatzis EA, Bosch J, Burroughs AK. Nonalcoholic fatty liver disease and portal hypertension. Hepatology. 2014;60(1):186–97.
10. Berzigotti A, Seijo S, Reverter E, Bosch J. Non-invasive prediction of portal hypertension. Gut. 2013;62(11):1667–75.
11. Garcia-Tsao G, Abraldes JG, Berzigotti A, Bosch J. Portal hypertensive bleeding in cirrhosis: risk stratification, diagnosis, and management. Hepatology. 2017;65(1):310–35.
12. Zardi EM, Di Matteo FM, Pacella CM, Sanyal AJ. The spleen and portal hypertension. Int J Clin Exp Med. 2015;8(8):11321–8.
13. Baik SK, Kim JW, Kim HS, et al. Doppler measurements of hepatic and portal venous blood flow for the prediction of large esophageal varices in liver cirrhosis. Korean J Hepatol. 2006;12(1):94–102.