Utility of splenic transient elastography in assessing the presence of portal hypertension: A review

Abstract

Portal hypertension (PH) is a major complication of chronic liver disease, often leading to serious clinical consequences such as variceal bleeding, ascites, and splenomegaly. The current gold standard for PH diagnosis, namely, hepatic venous pressure gradient measurement, is invasive and not widely available. Transient elastography has emerged as a non-invasive alternative for assessing liver stiffness (LS), and recent studies have highlighted the potential role of splenic stiffness (SS) in evaluating PH severity. This narrative review summarizes the available evidence on the utility of splenic transient elastography in assessing PH. We evaluated its diagnostic accuracy, technical challenges, and clinical applications, particularly in distinguishing between cirrhotic PH (CPH) and non-cirrhotic PH (NCPH). A comprehensive literature search was conducted using the PubMed database, focusing on studies that assess splenic elastography in the diagnosis and prognosis of PH. This review compares splenic elastography with other non-invasive imaging modalities, including MR elastography and shear-wave elastography. Additionally, we examined the role of SS using elastography in predicting the presence of esophageal varices and its potential impact on reducing the need for endoscopic screening. Studies have demonstrated that splenic elastography correlates well with PH severity, with cut-off values ranging between 45 kPa and 50 kPa for significant PH detection. Splenic elastography, when combined with platelet count and LS measurements, improves diagnostic accuracy and risk stratification for the occurrence of variceal bleeding. Despite its clinical promise, technical challenges such as patient positioning, body habitus, and probe selection remain key limitations. Notably, splenic elastography may be particularly useful in diagnosing NCPH, where LS remains normal but PH is present. Splenic transient elastography is a valuable adjunct in the non-invasive assessment of PH. Its ability to predict varices, differentiate between CPH and NCPH, and reduce unnecessary endoscopies suggests that it should be incorporated into routine hepatology practice. Future research should focus on refining SS cut-offs, evaluating its cost-effectiveness, and integrating splenic elastography into clinical guidelines for PH management.

Key Words: Portal hypertension; Transient elastography; Splenic stiffness; Cirrhosis; Non-cirrhotic portal hypertension; Gastroesophageal varices; Liver fibrosis

Core Tip: Splenic transient elastography is a valuable adjunct in the non-invasive assessment of portal hypertension. Its ability to predict varices, differentiate between cirrhotic and non-cirrhotic portal hypertension, and reduce unnecessary endoscopies suggests it should be incorporated into routine hepatology practice.

INTRODUCTION

Portal hypertension (PH) is broadly defined as the clinical manifestation of elevated pressure within the portal venous system[1]. It is most commonly caused by liver cirrhosis, which may result from various underlying conditions such as viral hepatitis or metabolic dysfunction-associated steatotic liver disease[2]. However, PH is not limited to cirrhosis and can also occur in a range of non-cirrhotic conditions[3].

Evaluation of PH involves a combination of clinical history and ancillary diagnostic tools, including laboratory investigations and imaging studies. Clinical signs may include esophageal or gastric varices, palpable splenomegaly, ascites, portosystemic shunting (e.g., caput medusae), and hepatic encephalopathy[4]. Laboratory findings often include a serum-ascites albumin gradient > 1.1 g/dL - calculated by subtracting ascitic albumin from serum albumin - in addition to thrombocytopenia and mild leukopenia[5]. On imaging, PH may be suggested by features such as splenorenal or other portosystemic shunts, splenomegaly, dilated portomesenteric veins, and biphasic or reversed portal venous flow on Doppler ultrasound[6].

The causes of PH are diverse and are typically classified according to their anatomical relationship to the hepatic sinusoids - the microvascular spaces where nutrient-rich blood from the portal vein is processed by hepatocytes before draining into the hepatic veins[7]. An overview of these causes is presented in Figure 1. As noted, cirrhosis remains the most common etiology of PH.

Figure 1 Some causes of portal hypertension. IVC: Inferior vena cava; PV: Portal vein. Image created using Biorender.

PH is typically diagnosed through a combination of laboratory testing, endoscopic evaluation for varices, and imaging studies[8]. The gold standard for diagnosing PH is the invasive measurement of the hepatic venous pressure gradient (HVPG)[9]. This procedure involves inserting a catheter with a balloon tip via the internal jugular vein into the hepatic vein branches. The free hepatic venous pressure (FHVP) is measured without balloon inflation, followed by the wedged HVP (WHVP) with the balloon inflated. Then the HVPG is calculated as the difference between WHVP and FHVP, expressed in mmHg.

HVPG values[10] carry important prognostic implications: > 10 mmHg indicates clinically significant PH (CSPH); > 12 mmHg is associated with an increased risk of variceal rupture; > 16 mmHg correlates with higher mortality risk.

Despite its diagnostic accuracy, HVPG measurement has notable limitations. For instance, normal or only slightly elevated HVPG values may be observed in cases of pre-hepatic or post-hepatic PH, such as in splenomegaly-related PH or cardiac-related PH[10]. Furthermore, as an invasive procedure, HVPG carries risks including bleeding and other catheter-related complications[11]. Given these drawbacks, a multimodal approach combining clinical findings, laboratory values, imaging, and non-invasive measurements is often essential for accurate diagnosis.

Transient elastography (TE), which assesses liver stiffness (LS) in kilopascals (kPa), has become a widely used tool in the evaluation of liver fibrosis and cirrhosis. Its adoption has significantly impacted hepatology by reducing the need for liver biopsy in patients with chronic liver disease[12]. Moreover, recent guidelines recommend incorporating TE-derived LS values and platelet count in patients with known cirrhosis to help avoid screening endoscopy for esophageal varices[13]. In addition, the role of spleen stiffness (SS) has gained attention in these same guidelines, where lower SS values may be used to rule out the need for endoscopic screening. However, this recommendation is currently limited to specific etiologies, primarily viral hepatitis, and is not yet generalized to all causes of PH related to cirrhosis[13].

Given the relatively recent application of TE to spleen stiffness measurement, the aim of this review was to explore the available literature on SS, including the technical limitations of using equipment initially developed for LS estimation. We evaluated current data regarding the diagnostic performance of SS (sensitivity and specificity) for identifying complications of PH - particularly the presence of varices - across different liver disease etiologies. These findings were compared to existing guideline recommendations. We also examined the added diagnostic value of SS in distinguishing non-cirrhotic PH (NCPH) from cirrhotic PH (CPH). Finally, we highlighted potential areas for future research and advancement in the field. All included studies were identified through a literature search on the PubMed database. The review was conducted in a narrative format, with the goal of providing practical, clinically relevant insights for hepatologists managing patients with PH.

SPLEEN STIFFNESS – TECHNIQUES AND CHALLENGES

Several methods are available for measuring SS, of which the three most common are transient elastography, MR elastography, and shear-wave elastography. The differences between each of the methods are summarized in Table 1[14].

Table 1 Summary of different methods for estimation of splenic stiffness.

Modality	Feasibility	Availability	Cost	Operator dependence
Transient elastography	High	Widely available	Low	High
Shear-wave elastography	Moderate	Moderately available	Moderate	Moderate
Magnetic resonance elastography	Low	Limited to tertiary centers	High	Low

Despite the growing interest in SS as a non-invasive tool for evaluating PH, the optimal diagnostic cut-off remains controversial, largely due to differences in elastography modalities, study populations, and target outcomes. For instance, TE studies have proposed thresholds ranging from 26.5 kPa to 60 kPa[15] to detect significant PH, not considering the cut-off proposed by the Baveno VII guidelines at about > 50 kPa. In contrast, shear-wave elastography studies have reported lower thresholds, often between 12 kPa and 20 kPa[16], whereas magnetic resonance elastography (MRE) is associated with PH at values of about 10 kPa[17]. These discrepancies stem from differences in technical parameters (e.g., probe frequency, acquisition depth), measurement protocols, and the fact that SS is not yet fully standardized across platforms.

The latest Baveno VII guidelines[13] suggest the usage of TE for the purpose of SS measurements, which is the method we concentrate on here. However, the location and size of the spleen in the left-upper quadrant make such measurements difficult and somewhat limited in certain settings[18]. For instance, in early PH, the spleen may not be sufficiently enlarged to allow acoustic capture through the relevant intercostal space. While SS via TE is a promising non-invasive tool, several technical factors can affect its accuracy and feasibility. Obesity, for instance, precludes accurate measurements of SS, with some studies[19] demonstrating that already at a body mass index (BMI) > 24 kg/m³ the failure rates increase due to poor acoustic wave propagation and signal attenuation. In addition to this, the presence of ascites dampens the wave transmission interfering with the method´s accuracy[19]. In addition, spleen size smaller than roughly 93 mm[19] also hinders accurate measurements of SS.

Despite these disadvantages, reproducibility of SS measurements can be considered moderately high, with correlation coefficients ranging from 0.73 to 0.90 in different studies[20,21]. Simple maneuvers, such as telling patients to alternately inhale/exhale and hold their breath to see when the spleen is nearest to the intercostal spaces, may aid in allowing the spleen to be captured with the probe. Another maneuver is to place the patient in right lateral decubitus position such that the spleen is more prominent in its position in the left upper quadrant, having the effect of gravity pulling the spleen anteriorly, thus making it easier to get the spleen in the right acoustic window[22]. Furthermore, the use of an XL-probe may allow for easier acquisition of SS values especially in obese patients[23]. This may, however, lead to the underestimation of the actual SS with the XL-probe as compared to the M-probe, and a higher degree of uncertainty (expressed as interquartile range, IQR%) in the measurements obtained[24]. Lastly, the usage of a higher frequency (100 Hz) probe, compared to the standard “liver” probe (50 Hz), may also increase the accuracy of performed measurements and allow for higher (up to 100 kPa) values to be obtained for SS[25].

SPLEEN STIFFNESS-ASSOCIATION WITH DEVELOPMENT OF VARICES

The most devastating complication of PH is the formation of bleeding esophageal varices, which carry an estimated 6-week mortality of approximately 20% following a single bleeding episode[26]. The current Baveno VII consensus guidelines recommend the use of LS and SS measurements via TE for the non-invasive assessment of PH. While LS cut-offs are now well established, SS thresholds remain less well-defined and are not yet universally endorsed across all etiologies, with the aforementioned guidelines mainly focusing on their use in the viral hepatitis setting[13].

As summarized in Table 2[27-36], there appears to be a reasonable level of agreement across the literature regarding SS cut-offs for identifying significant PH as suggested by the identification of varices, irrespective of the underlying PH cause. Most studies report diagnostic thresholds in the range of 45–50 kPa, consistent with Baveno VII’s suggested value of > 50 kPa to rule in significant PH, although that recommendation is specific to viral etiologies. Reported sensitivities and specificities across studies generally cluster around 80%, indicating acceptable diagnostic performance for detecting esophageal varices.

Table 2 Survey of studies where splenic stiffness estimation using transient elastography for the presence of varices was performed[27-36].

Ref.	Pathologies studied	Cut-off (in kPa)	Sensitivity	Specificity
Stefanescu et al[27], 2011	HCV/ALD	46.4	83.5	71.0
Colecchia et al[28], 2012	HCV	55.0	96.0	85.0
Sharma et al[29], 2013	ALD/HBV/HCV/CLD	40.8	94.0	76.0
Calvaruso et al[30], 2013	HCV	50.0	65.0	61.0
Fraquelli et al[31], 2014	HBV/HCV/MPD	48.0	100	60.0
Zykus et al[32], 2015	ALD/HCV/CLD	50.7	78.1	77.1
Rewisha et al[33], 2016	HCV	43.2	92.9	84.5
Tseng et al[34], 2018	HBV/HCV/AIH/PBC/NAFLD/CLD	48.9	76.0	100
Ferreira-Silva et al[35], 2023	ALD	54.0	72.2	60.9
Lantinga et al[36], 2023	ALD/NAFLD/HCV/HBV/AIH	46.4	82.0	80.0

AIH: Autoimmune hepatitis; ALD: Alcoholic liver disease; CLD: Cryptogenic liver disease; HBV: Hepatitis B virus; HCV: Hepatitis C virus; MPD: Myeloproliferative disorder; NAFLD: Non-alcoholic fatty liver disease; PBC: Primary biliary cholangitis.

Importantly, combining SS measurements with platelet count appears to enhance diagnostic accuracy, with some studies reporting sensitivity and specificity approaching 90%[37]. This combined approach may further reduce the need for unnecessary screening endoscopies and allow for a more targeted use of non-selective beta-blockers, regardless of the etiology of PH.

Although some studies have reported receiver operating characteristic values, these were not consistently available due to methodological heterogeneity and varied patient populations. Therefore, to maintain coherence and comparability, this review focuses on SS cut-offs and corresponding sensitivity/specificity values, which are more immediately applicable in clinical settings. A formal meta-analysis was not performed, as it was beyond the scope of this narrative review; however, the convergence of SS thresholds within a narrow 45–50 kPa range across studies supports this interval as a practical and reproducible reference point for clinical use.

SPLEEN STIFFNESS AIDING IN PINPOINTING CAUSE OF PH

The most common cause of PH is, as previously mentioned, the presence of chronic liver disease, specifically liver cirrhosis. However, for patients from non-western countries, and especially those originating from Africa, infectious causes, such as schistosomiasis, or sequelae of prior abdominal infections, such as extrahepatic portal vein thrombosis, are the most common cause of PH, and are thus classified as NCPH[1]. Given the multitude of causes that are associated with NCPH as detailed in Table 3 below, the diagnosis may be elusive[38].

Table 3 Summary of some main causes of non-cirrhotic portal hypertension.

Group of causes	Example of specific disorders
Immunological disorders	CVID, CD, Organ transplant patients, autoimmune conditions (e.g., SLE)
Infections	Bacterial intestinal infections (esp. during childhood), helminths (e.g., schistosomiasis), HIV
Medications and toxins	Thiopurines, vitamin A, chemotherapeutic agents
Thrombotic and hematological conditions	Myeloproliferative syndromes; thrombophilias predisposing to MVT
Genetic disorders	Turner syndrome

CD: Crohn’s disease; CVID: Common variable immunodeficiency; HIV: Human immunodeficiency virus; MVT: Mesenteric vein thrombosis; SLE: Systemic lupus erythematosus.

The diagnosis of NCPH should be suspected In patients presenting with signs of elevated portal pressures (e.g., ascites, splenomegaly or variceal bleeds), with normal liver function tests (including prothrombin time/international normalized ratio and albumin), normal (or very slightly elevated) LS measures, and an otherwise negative work-up for liver cirrhosis (including liver biopsy). NCPH is specifically characterized, compared to CPH, by prominent splenomegaly on imaging/clinical examination alongside a relatively low LS but particularly elevated SS, when elastography is used[39]. This finding may even suggest the routine usage of SS estimations in patients with unclear PH where primary hepatic causes have been excluded.

Yet another clue for the presence of NCPH, especially when assessing for the location of varices, is the fact that there is usually a predominance of gastric varices compared to “lone” esophageal varices, which are otherwise the norm in CPH, with generally more severe thrombocytopenia[40]. The aforementioned findings are readily understandable especially when considering the venous drainage of the distal esophagus as compared to the stomach (Figure 2). This may reflect pre-portal vein causes, such as mesenteric or splenic venous pathologies; alongside the immunological/hematological basis for many of these conditions whereby the spleen is afflicted by these processes, causing increased platelet sequestration, alongside “left-sided” PH predominantly with gastric varices.

Figure 2 Portal circulation and related anatomy. The stomach’s venous drainage is into the splenic vein, while the esophageal drainage (especially lower esophagus) is into the portal vein (PV), which is the reason for preferential variceal development at the lower esophagus in cirrhotic portal hypertension (PH), whereas the non-cirrhotic PH there is mainly the development of gastric varices. DE: Distal esophagus; SMV: Superior mesenteric vein; SV: Splenic vein. Image created using Biorender.

LIMITATIONS OF SPLEEN STIFFNESS IN CLINICAL PRACTICE

Despite its growing clinical utility, SS measurement via TE has several limitations. One key issue is its limited sensitivity in early-stage PH, where structural and hemodynamic changes in the spleen ultimately leading to splenic congestion may not yet be significant enough to elevate stiffness values[30].

Coexisting splenic pathologies, such as lymphoma, infiltrative diseases, or congestive splenomegaly from cardiac dysfunction, can also independently increase SS, potentially leading to false-positive interpretations[31]. Moreover, technical challenges like obesity, ascites, and small spleen size can impair signal acquisition and measurement success[27,28]. These issues are particularly notable in patients with BMI > 24 kg/m² or spleens measuring < 93 mm in length[18]. While operator training and the use of XL-probes can help mitigate some of these difficulties, inter- and intra-observer variability remains a consideration, even with standardized protocols[23]. Notably, as mentioned earlier, SS may be elevated while LS remains normal - a valuable diagnostic clue rather than a confounder[39]. This differential pattern may aid in distinguishing NCPH from CPH when integrated with clinical and laboratory findings[41].

Together, these limitations emphasize the importance of interpreting SS values in the broader context of clinical presentation, LS, platelet count, and suspected disease etiology[42,43].

FUTURE RESEARCH PERSPECTIVES

There are several promising areas for future research where the application of splenic TE could provide significant clinical value. Given the simplicity and non-invasive nature of TE for assessing both LS and SS, its broader use in patients with newly diagnosed cirrhosis and PH, regardless of underlying etiology, may offer important insights into prognosis, particularly in predicting the development of esophageal varices. The combined assessment of LS and SS, especially when interpreted alongside platelet count, may further refine risk stratification and reduce the overuse of non-selective beta-blockers. In fact, this approach may support de-prescription of beta-blockers in patients with consistently low LS and SS values, minimizing unnecessary treatment in low-risk individuals.

Another area of growing interest is the use of LS and SS in patients with NCPH, or idiopathic PH. In this setting, TE could assist both surveillance strategies and therapeutic decisions, including beta-blocker use, particularly when variceal risk remains uncertain. Finally, establishing age-specific reference values for SS represents an important research need. Normative data from healthy individuals across different age groups would enhance the interpretability of SS values in clinical practice and ensure more individualized decision-making.

CONCLUSION

In conclusion, SS measurement using TE appears to be a valuable adjunct in the evaluation of significant PH regardless of the underlying etiology. While the Baveno VII consensus limits the > 50 kPa SS threshold to patients with viral hepatitis, our review demonstrates consistent diagnostic performance across a broader range of liver diseases, supporting its potential wider applicability. When used alongside LS, SS may enhance diagnostic accuracy and reduce the need for unnecessary endoscopic evaluations in patients with newly diagnosed cirrhosis. Clinically, a combination of high LS and high SS supports CPH, whereas low LS combined with elevated SS in the appropriate clinical setting may suggest NCPH, in which SS measurement may aid diagnostic confirmation. Further research is warranted to define SS's role in beta-blocker de-prescription and to support its integration into general clinical guidelines for PH management.