Portal hypertension is the main determinant of prognosis in cirrhosis of the liver. The gold standard for its evaluation is hepatic venous pressure gradient measurement. Because it is an invasive test performed only at highly specialized centers, noninvasive tools for the detection of portal hypertension have been developed. One of the most relevant is vibration-controlled transient elastography (FibroScan®; Echosens, Paris, France). Liver stiffness measurement is the most widely utilized and validated noninvasive tool, although more recent studies suggest spleen stiffness measurement correlates better with the hepatic venous pressure gradient. In the present narrative review, a comprehensive search of the literature was carried out on the PubMed/MEDLINE, Embase, and Cochrane CENTRAL databases, complemented by manual reference screening, to critically analyze the available evidence on splenic elastography in the noninvasive evaluation of portal hypertension. The following topics are addressed: the clinical impact of portal hypertension evaluation, the role of liver elastography as a noninvasive strategy, and the different applications of splenic elastography in the evaluation of portal pressure, including the diagnosis of clinically significant portal hypertension and severe portal hypertension, the prediction of esophageal varices and high-risk varices, clinical decompensation and recompensation risk assessment, and the usefulness of spleen stiffness measurement in monitoring the response to beta blocker therapy.
La hipertensión portal es el principal factor determinante del pronóstico en la cirrosis hepática. El estándar de oro para su evaluación es la medición del gradiente de presión venosa hepática. Sin embargo, se trata de una prueba invasiva que se realiza únicamente en centros de alta especialización y, por este motivo, se han desarrollado herramientas que permiten la detección no invasiva de hipertensión portal. Una de las más relevantes es la elastografía de transición a vibración controlada (FibroScan®; Echosens, París, Francia). La medición de la rigidez hepática es la herramienta no invasiva más ampliamente utilizada y validada, aunque estudios más recientes sugieren que la medición de la rigidez esplénica tiene una mayor correlación con el gradiente de presión venosa hepática. En esta revisión narrativa, se llevó a cabo una búsqueda bibliográfica exhaustiva en PubMed/MEDLINE, Embase y Cochrane CENTRAL, complementada con cribado manual de referencias, con el objetivo de analizar de forma crítica la evidencia disponible sobre la elastografía esplénica en la evaluación no invasiva de la hipertensión portal. Abordando los siguientes tópicos: impacto de la evaluación de la hipertensión portal, papel de la elastografía hepática como estrategia no invasiva y las diferentes aplicaciones de la elastografía esplénica en la evaluación de la presión portal, incluyendo el diagnóstico de hipertensión portal clínicamente significativa e hipertensión portal grave, predicción de várices esofágicas y várices de alto riesgo, evaluación del riesgo de descompensación clínica y recompensación y su utilidad en la monitorización de la respuesta al tratamiento betabloqueante.
Portal hypertension (PHT) is the primary hemodynamic consequence of liver cirrhosis (LC) and the main cause of most of its complications: ascites, variceal bleeding, and hepatic encephalopathy (HE). The development of said complications marks the transition to decompensated cirrhosis (DC), significantly increasing morbidity and mortality and the need for liver transplantation (LT).1–3 The hepatic venous pressure gradient (HVPG) is the gold standard for the direct evaluation of portal pressure (PP), but its invasive nature, the need for specialized infrastructure, and the risk of complications have limited its applicability in routine clinical practice.3,4
With the advent of precision medicine, determining the risk of clinical decompensation, i.e., the presence of clinically significant portal hypertension (CSPH), gastroesophageal varices (GEV), and high-risk varices (HRV),2,5 has become one of the primary aims of individualized treatment in patients with compensated cirrhosis (CC). Consequently, numerous noninvasive tools for detecting PHT have been developed and are supported by growing scientific evidence. The most widely validated technique is vibration-controlled transient elastography (VCTE) or transient elastography (TE) (FibroScan®; Echosens, Paris, France). Liver stiffness measurement (LSM) performed with this tool has shown its usefulness across different clinical scenarios, including the diagnosis of CSPH and GEV.3,6,7 More recently, splenic elastography has emerged as a promising tool, given that the spleen more accurately reflects hemodynamic changes secondary to a sustained increase in PP. Spleen stiffness measurement (SSM) has shown its usefulness in the diagnosis of CSPH, severe PHT (defined as a HVPG ≥ 12 mmHg), GEV, and HRV, as well as in clinical decompensation risk evaluation and beta blocker therapy response monitorization (Fig. 1) positioning it as a viable alternative for risk stratification and monitoring of PHT in patients with LC.
Usefulness of liver and splenic elastography in portal pressure evaluation.
cACLD: compensated advanced chronic liver disease; dACLD: decompensated advanced chronic liver disease; PHT: portal hypertension; CSPH: clinically significant portal hypertension; LSM: liver stiffness measurement; HVPG: hepatic venous pressure gradient; SSM: spleen stiffness measurement.
The aim of the present narrative review was to analyze and synthesize the available evidence on the role of splenic elastography in noninvasive PHT assessment, including its usefulness in diagnosing CSPH, GEV screening, clinical decompensation risk, beta blocker therapy management, and recompensation monitorization.
MethodologyA narrative review of the literature was carried out on the usefulness of splenic elastography in PHT assessment. A comprehensive bibliographic search was conducted using the PubMed/MEDLINE, Embase, and Cochrane CENTRAL databases, encompassing publications from 1954 to 2025. The search strategy combined MeSH terms and free-text keywords related to “cirrhosis”, “decompensated cirrhosis”, “compensated chronic liver disease”, “portal hypertension”, “clinically significant portal hypertension”, “liver elastography”, “splenic elastography”, and “esophageal varices”. Study selection was based on thematic relevance and clinical value, prioritizing those that directly addressed splenic elastography in the context of PHT. Manual screening of the reference lists of relevant articles was carried out to complement the electronic search. Given the narrative nature of this review, no formal systematic review criteria were applied, but rather, the available evidence was critically summarized and structured to provide a comprehensive and up-to-date overview of the theme.
Portal hypertension evaluation impact: invasive diagnosisLC has traditionally been classified into two stages: CC and DC, the latter defined by the development of any complication related to PHT, including ascites, gastrointestinal bleeding, HE, and jaundice.1 CC has a mean survival time that varies from 10 to 15 years and decreases to 2–4 years, once there is an episode of clinical decompensation.8,9 This universally accepted classification may oversimplify the course of liver disease,10,11 and so a multi-stage model has been proposed that that includes several subgroups with distinct prognoses (Fig. 1).5,9,12
PHT is the main determinant of prognosis in LC,5 making PP evaluation essential; the gold standard for its assessment is HVPG measurement. The HVPG represents the difference between hepatic sinusoidal capillary pressure, or wedged hepatic venous pressure, and systemic pressure, or free suprahepatic vein (SHV) pressure, and is determined by catheterization of the right or middle SHV in a liver hemodynamics laboratory.4 A value between 1 and 5 mmHg is considered normal.2
The HVPG is a robust prognostic indicator in both CC and DC. The risk of complications progressively increases as the PP increases. Once the HVPG reaches a threshold of ≥ 10 mmHg, there is an elevated risk for developing esophageal varices, clinical decompensation, and hepatocellular carcinoma, and so the condition is considered CSPH.2,3,13–15 In addition, there is an approximate 11% increase in the risk of developing hepatic clinical events or decompensation for every increase of 1 mmHg above 10 mmHg.14 An HVPG > 12 mmHg is associated with a risk of both acute and recurrent variceal bleeding.16 Likewise, a value ≥ 16 mmHg is associated with greater mortality in patients with CC and DC,17–19 as well as with short-term mortality in candidates for elective non-hepatic abdominal surgery.20 In patients with acute variceal bleeding, HVPG values ≥ 20 mmHg have been associated with failed bleeding control and increased mortality,21,22 whereas an HVPG > 22 mmHg has been linked to in-hospital mortality in patients with severe alcoholic hepatitis.23
On the other hand, improvement in the HVPG reduces the risk of complications, in such a way that the decrease in the HVPG ≥ 20% or an HVPG < 12 mmHg significantly reduces the risk of variceal bleeding and death.24
Despite the abovementioned prognostic implications, HVPG measurement is a costly invasive technique with limited access, and so its performance in routine clinical practice is not very efficient.25 Similarly, esophagogastroduodenoscopy (EGD) for GEV screening is an expensive test and sometimes uncomfortable for patients, which has led to the implementation of noninvasive methods for evaluating the presence of PHT. Different elastography techniques are among them, and include acoustic radiation force impulse (ARFI), two-dimensional shear wave elastography (2D-SWE), magnetic resonance elastography, and TE (FibroScan®; Echosens, Paris, France). TE measures tissue stiffness using a probe that emits a vibration and induces a shear wave that propagates through the tissue. The speed of the wave propagation is directly proportional to the stiffness of the tissue.26 Because it is the most validated and widely available noninvasive test (NIT) it is exclusively discussed below.
The role of liver elastography as a noninvasive strategyLSM by TE (FibroScan®; Echosens, París, France) has significantly broadened the ability to detect patients with advanced liver disease, enabling patients with CC (now also called compensated advanced chronic liver disease [cACLD]) to be identified and classified. An LSM > 15 kPa defines patients with cACLD, whereas an LSM < 10 kPa (in the absence of other clinical data) rules it out; additional confirmatory tests are required when values are between 10 and 15 kPa.2
In such settings, early detection of CSPH is essential for determining the risk of potentially severe complications and implementing timely therapeutic measures. Beta blocker use in patients with CSPH has been shown to reduce the risk of developing a first clinical decompensation.27 In contrast, in patients without CSPH, beta blockers as primary prophylaxis are not useful because the hyperdynamic circulation that characterizes advanced liver disease has not yet developed in those patients.28
LSM is the most validated and widely utilized NIT for evaluating the presence of PHT.6,7,29–34 LSM has been shown to be closely correlated with the HVPG,29–34 in such a way that the presence of CSPH may also be determined noninvasively through LSM.3 A value ≥ 25 kPa predicts CSPH in 90% of patients with cACLD due to alcohol-associated liver disease (ALD), viral hepatitis, and metabolic dysfunction-associated steatotic liver disease (MASLD), without obesity.6 In patients with an LSM < 25 kPa, the ANTICIPATE model may be applied. In said model, values between 20 and 25 kPa, together with a platelet count < 150 × 109/L or an LSM between 15 and 20 kPa and a platelet count < 110 × 109/L determine at least a 60% risk of CSPH across most LC etiologies.7 In addition to LSM, the “ANTICIPATE-NASH” also incorporates body mass index (BMI) and platelet count for estimating the risk of CSPH. It may be applied to patients with MASLD and obesity (BMI > 30 kg/m2), in whom the positive predictive value of LSM is barely above 60%.6
In addition to its performance in diagnosing cACLD and CSPH, LSM also makes it possible to identify patients at low risk of esophageal varices, reducing the need for endoscopic procedures. According to the Baveno VI criteria, patients with LSM < 20 kPa (assessed by TE) and a platelet count >150 × 109/L have a risk of HRV below 5%, and so need not undergo a yearly screening EGD.2 Those criteria have been extensively validated in subsequent studies, reporting an area under the receiver operating characteristic (AUROC) curve of (95% CI 0.87−0.92) and a HRV missed rate of only 0.6%.35
Splenic elastography in portal pressure evaluationSplenomegaly is a frequent finding in LC with PHT and the increase in volume was previously considered solely a consequence of passive splenic congestion associated with elevated PP and resistance to splenic venous outflow.36 However, more recent studies have shown that the spleen actively participates in maintaining PHT, with a decrease in splenic arterial resistance and an increased contribution to portal blood flow,37,38 in addition to hyperplasia and hyperactivation of the splenic lymphoid tissue, fibrogenesis, and angiogenesis, which together lead to increased splenic stiffness.39 As in the liver, changes in splenic stiffness can also be quantified through different elastography techniques, including TE (FibroScan®; Echosens, Paris, France).
Different clinical trials have shown that, compared with other NITs (including LSM), SSM correlates more strongly with the HVPG in more advanced stages of LC,40,41 suggesting that the changes in splenic density more accurately reflect the dynamic component of PHT. In early stages, PHT is primarily determined by the fixed component (liver fibrosis), and as LC progresses, the dynamic component, derived from the increased portal flow secondary to splanchnic vasodilation, hyperdynamic circulation, and increased resistance to portal flow by the portosystemic collaterals, becomes increasingly relevant.29,38,42
Concurring with those results, numerous clinical trials have shown that the correlation between LSM and the HVPG weakens, once the latter exceeds the threshold of ≥ 10−12 mmHg,29,33,41 the cutoff point at which most severe complications develop. Moreover, LSM correlates poorly with GEV grade29,31,40,41,43,44 or a history of variceal bleeding.29,41 In that context, LSM could underestimate PHT severity and the risk of variceal bleeding, and so SSM may be a more suitable tool for evaluating PP changes, and unlike LSM, SSM is not affected by liver disease etiology, liver congestion, inflammation, or cholestasis.45
Splenic TE has been shown to be useful in different clinical scenarios, particularly in detecting CSPH and severe PHT and in predicting GEV, HRV, response to beta blocker therapy, and the risk of clinical decompensation.
Diagnosis of CSPH and severe portal hypertensionSSM has also been shown to be a useful tool for detecting patients with CSPH. In a prospective clinical trial that included 100 patients with LC due to chronic hepatitis C virus (HCV) infection, Colecchia et al.40 reported that the correlation between the HVPG and SSM was stronger (r = 0.885, p = 0.0001) than with other NITs, including LSM, the liver stiffness-spleen size-to-platelet ratio score (LSPS), and the platelet count/spleen diameter (PC/SD) ratio; the SSM was superior to the LSPS and PC/SD ratio (AUROC = 0.96) and similar to the LSM for diagnosing CSPH. The cutoff point was an SSM > 52.8 kPa for diagnosing CSPH (sensitivity: 76.9%, specificity: 97.1%) and <40 kPa (sensitivity: 98.5%, specificity: 74.3%) for ruling it out. Subsequently, Zykus et al.46 corroborated those results in patients with LC due to other etiologies, including ALD, with a cutoff point of >47.6 kPa for diagnosing CSPH with 77.7% accuracy, although it was not superior to LSM.
The capacity of SSM to predict CSPH was evaluated in a meta-analysis that included 9 clinical trials, reporting a high correlation with the HVPG (r = 0.72), with sensitivity of 0.88 (95% CI: 0.7−0.96) and specificity of 0.84 (95% CI: 0.72−0.92).47
Because the spleen is stiffer than the liver, the standard 50 Hz probe (initially utilized in most clinical trials, with a measuring limit of 75 kPa) may overestimate spleen stiffness.44,48 Thus, a new shear wave between 5 and 100 Hz, with a stiffness range between 6 and 100 kPa and an adjusted measuring depth between 25 and 55 mm, has recently been developed.48 Stefanescu et al.44 showed that this new 100 Hz probe correlates more strongly with the HVPG than the standard probe, with a cutoff point >34.15 kPa (AUROC 0.81, 95% CI 0.67−0.95) for detecting CSPH, results that were later corroborated. Odriozola et al.49 reported that the SSM had elevated accuracy for ruling out and diagnosing CSPH, with the cutoff points <40.9 and >49.9 kPa (AUROC: 0.95), utilizing a spleen-specific 100 Hz probe in patients with MASLD. Afterwards, Zhang et al.50 conducted a prospective study that included 185 patients with cACLD, reporting a positive predictive value (PPV) of 98.0% with 98.8% specificity for diagnosing CSPH, using a cutoff point for SSM > 50 kPa and the spleen-specific 100 Hz probe. The SSM > 50 kPa was more beneficial, compared with other algorithms based on elastography.
The addition of SSM to different models has shown increased diagnostic capacity.51–53 Dajti et al.51 reported that incorporating an SSM > 40 kPa into the Baveno VII diagnostic algorithm (LSM ≤ 15 kPa + platelet count >150 × 109/L to rule out CSPH and LSM > 25 kPa to diagnose CSPH) reduced the gray zone from 40%–60% to 7%–15%, improving the identification of patients at high risk for clinical decompensation. Those authors also developed a nomogram based on LSM, platelet count, and SSM for estimating individual CSPH risk, with an AUROC of 0.94, superior to the ANTICIPATE model. The more recently developed NICER model, which includes SSM, LSM, platelet count, and BMI, has been shown to outperform the ANTICIPATE ± NASH model for predicting CSPH in patients with cACLD of different ideologies, including viral hepatitis, MASLD, ALD, and metabolic dysfunction and alcohol-associated liver disease (MetALD) (AUC 0.906 [0.864−0.946] vs 0.863 [0.810−0.916]; p = 0.012).52 Lastly, a recent post hoc analysis from the PREDESCI clinical trial that included more than 170 patients with CSPH, utilizing the Baveno VII/American Association for the Study of Liver Diseases (AASLD) criteria for diagnosing CSPH (LSM ≥ 25 kPa or LSM > 20 kPa combined with thrombocytopenia), reported that up to 37% of the patients can have an indeterminate result despite presenting with a high risk for decompensation. To address that limitation, a model based on EGD in inconclusive cases was developed, reducing the gray zone to only 22%, correctly classifying 86% of the patients with CSPH. Subsequently, a fully noninvasive algorithm combining the Baveno VII criteria with the SSM was evaluated that performed similarly to the EGD-based model (86% vs 85% specificity, 92% vs 91% PPV, 28% vs 31% gray zone), suggesting that the real-world clinical diagnosis of CSPH should be based on the combination of different methods, such as LSM, SSM, platelet count, and endoscopy.53
Due to its close correlation with the HVPG, SSM is useful for diagnosing severe PHT (HVPG ≥ 12 mmHg), with cutoff points that range from 44.95 kPa, with the spleen-specific 100 Hz probe (AUROC 0.78, 95% CI 0.677−0.88)44 to 50−55 kPa, with the standard probe.40,46 The meta-analysis by Song et al.47 corroborated those results, reporting an elevated diagnostic accuracy for diagnosing severe PHT, with sensitivity of 0.92 (95% CI: 0.82−0.96) and specificity of 0.79 (95% CI: 0.72−0.85). However, the strong correlation between the SSM and HVPG has been described to have a ceiling effect, given that it is lost, once the HVPG reaches the threshold of ≥19 mmHg (r = 0.05, p = 0.81).41Tables 1 and 2 summarize the evidence on SSM utility in the diagnosis of CSPH and severe PHT.
Spleen stiffness measurement in clinically significant portal hypertension.
| Authors | Type of study | Population | Tool | Cutoff point for diagnosing CSPH | Cutoff point for ruling out CSPH | Other results |
|---|---|---|---|---|---|---|
| Colecchia et al., 2012 40 | Prospective | 100 patients with HCV-induced LC | SSM through TE (FibroScan) compared with noninvasive tools (LSM, LSPS, and PC/SD) and the HVPG | ≥52.8 kPa (S: 76.9%, Sp: 97.1%) | <40 kPa (S: 98.5%, Sp: 74.3%) |
|
| Zykus et al., 201546 | Prospective | 107 patients with LC due to HCV, ALD, and cryptogenic cirrhosis | SSM through TE (FibroScan) compared with LSM and the HVPG | 47.6 kPa (S: 77.3%, Sp: 79.2%, PPV: 92%, diagnostic accuracy: 77.7%) | – |
|
| Stefanescu et al., 202044 | Prospective | 216 patients, HCV, HBV, ALD, others | SSM (100 Hz) vs SSM (50 Hz), other NITs (LSM, LSPS, Lok index, PSR, Fib4, APRI) and Baveno VI and HPVG criteria | 34.15 kPa (AUROC 0.81, 95% CI, 0.67−0.95) with SSM (100 Hz) | – | – |
| Odriozola et al., 202349 | Retrospective | 85 patients, 60 with MASLD | SSM (100 Hz) | <40.9 kPa (S: 95%, Sp: 73.3%, PPV: 64%, NPV: 96.7%) | >49.9 kPa (S: 80%, Sp: 96.7%, PPV: 92.3%, NPV: 90.6%) | |
| Zhang et al., 202550 | Prospective | 185 patients with cACLD due to HBV, HCV, ALD, PBC | SSM (100 Hz) | >50 kPa (Sp: 98.8%, PPV: 98.0%). | – | – |
| Dajti et al., 202251 | Retrospective | 114 patients with cACLD due to viral hepatitis, ALD, MASLD, and others | SSM and Baveno VII algorithm (LSM ≤ 15 kPa + platelet count >150 × 109/l rule out CSPH and LSM > 25 kPa diagnose CSPH) | >40 kPa | – |
|
| Jachs et al., 202452 | Prospective | 407 with cACLD, viral MASLD, ALD, and MetALD | NICER (SSM, LSM, platelet count, and BMI) model | ≥ 55 kPa | ≤ 25 |
|
| Dajti et al., 202553 | Retrospective | 195 with cACLD and CSPH and/or GEV | Algorithm based on the Baveno VII-SSM criteria | >40 kPa (Sp: 86%, PPV: 92%) | <21 kPa (S:100%, NPV:100%) |
|
ACLD: acute-on-chronic liver disease; ALD: alcohol-related liver disease; AUROC: area under the receiver operating characteristic; CSPH: clinically significant portal hypertension; GEV: gastroesophageal varices; HBV: hepatitis B virus; HCV: hepatitis C virus; HRV: high-risk varices; HVPG: hepatic venous pressure gradient; LC: liver cirrhosis; LSPS: liver stiffness-spleen size-to platelet ratio score; MASLD: metabolic dysfunction-associated steatotic liver disease; NIT: noninvasive test; NPV: negative predictive value; PBC: primary biliary cholangitis; PC/SD ratio: platelet count to spleen diameter ratio; PHT: portal hypertension; PPV: positive predictive value; S: sensitivity; PSR: platelet-to-spleen volume ratio; Sp: specificity; SSM: spleen stiffness measurement; TE: transient elastography.
Spleen stiffness measurement in the diagnosis of severe portal hypertension.
| Authors | Type of study | Population | Tool | Cutoff point for diagnosing severe PHT | Cutoff point for ruling out severe PHT | Other results |
|---|---|---|---|---|---|---|
| Colecchia et al., 2012 40 | Prospective | 100 patients with HCV-induced LC | SSM through TE (FibroScan) compared with noninvasive tools (LSM, LSPS, and PC/SD) and the HVPG | 55.0 kPa (S: 72.2%, Sp: 97.8%), | 41.3 kPa (S: 98.1%, Sp: 67.4%) |
|
| Zykus et al., 201546 | Prospective | 107 patients with LC due to HCV, ALD, and cryptogenic cirrhosis | SSM through TE (FibroScan) compared with LSM and the HVPG | 50.7 kPa (S: 78.1, Sp: 77.1, PPV: 86.2, diagnostic accuracy 77.7%) | ||
| Stefanescu et al., 2020 44 | Prospective | 216 patients, HCV, HBV, ALD, others | SSM (100 Hz) vs SSM (50 Hz), other NITs (LSM, LSPS, Lok index, PSR, Fib4, APRI) and Baveno VI and HPVG criteria | 44.95 kPa (AUROC 0.78 95% CI 0.677−0.88) |
ACLD: acute-on-chronic liver disease; ALD: alcohol-related liver disease; AUROC: area under the receiver operating characteristic; GEV: gastroesophageal varices; HBV: hepatitis B virus; HCV: hepatitis C virus; HRV: high-risk varices; HVPG: hepatic venous pressure gradient; LC: liver cirrhosis; LSPS: liver stiffness-spleen size-to platelet ratio score; MASLD: metabolic dysfunction-associated steatotic liver disease; NIT: noninvasive test; NPV: negative predictive value; PC/SD ratio: platelet count to spleen diameter ratio; PHT: portal hypertension; PPV: positive predictive value; S: sensitivity; PSR: platelet-to-spleen volume ratio; Sp: specificity; SSM: spleen stiffness measurement; TE: transient elastography.
SSM has demonstrated its superiority to other noninvasive tests in diagnosing GEV and HRV,40,44,54,56 with cutoff points for GEV diagnosis that range from ≥ 40.8–55 kPa,40,41,55,57 and values < 41.3 kPa for ruling out their presence.40 Those results have been corroborated in a systematic review and meta-analysis that reported sensitivity of 78% (95% CI 75%–81%), specificity of 76% (95% CI 72%–79%), a positive likelihood ratio (LR+) of 3.4 (95% CI 2.3–4.9), a negative likelihood ratio (LR–) of 0.2 (95% CI 0.1−0.4), and a diagnostic odds ratio (OR) of 19.3 (95% CI 7.5–49.8) for the presence of GEV of any grade.58 As with the diagnosis of CSPH, combining NITs improves diagnostic accuracy. Stefanescu et al.57 demonstrated that the sequential strategy using LSM followed by SSM, when the former was >19 kPa and the latter > 55 kPa, predicted the presence of GEV more accurately (89.5%).
SSM has been shown to be more useful in evaluating the risk of HRV and variceal bleeding, compared with LSM, i.e., SSM can differentiate between small and large varices and estimate bleeding risk.41,44,54 The spleen-specific probe (100 Hz) is reported to be superior to the standard probe (50 Hz) for diagnosing large varices,44,54 even in patients with advanced hepatocellular carcinoma (BCLC B/C), in which a SSM ≤ 40 kPa reaches a variceal miss rate of 4.8%, significantly outperforming the Baveno VI criteria and yielding results comparable to the Baveno VII criteria.59 The reported cutoff points for HRV diagnosis range from 40.1 to 46.4 kPa with the standard probe (50 Hz)44,60,61 to 41.3 kPa for the spleen-specific probe (100 Hz).44
Despite the elevated sensitivity of the Baveno VI criteria for diagnosing HRV, their application has reduced the need for endoscopy by only 26%. An attempt has been made at expanding the criteria but an acceptable variceal miss rate (< 5%) has not been achieved,35 resulting in the proposal of the combined NIT models for improving variceal risk stratification.44,62 Colecchia et al.62 showed that by combining the SSM with a ≤ 46 kPa cutoff point and the Baveno VI criteria, an additional 22.5% of endoscopies would have been avoided, with a final value of 43.8% spared procedures, and only a 5% HRV miss rate, compared with the SSM ≤ 46 kPa or Baveno VI criteria alone. Stefanescu et al. later demonstrated that a sequential strategy combining SSM (100 Hz) < 41.3 kPa with the Baveno VI criteria reduced the need for endoscopy by 38%, compared with the Baveno VI criteria alone (p < 0.001), with a HRV miss rate of only 4.7%.44 A recent real-world, multicenter, retrospective study validated the results of the study by Stefanescu et al.,44 reporting that the sequential application of the Baveno VI criteria, followed by the SSM with the spleen-specific probe (100 Hz) and a > 43 kPa cutoff point, may increase the number of patients in whom EGD can be safely avoided (66.2% vs 71.1%), without surpassing the 5% HRV miss rate.63Tables 3 and 4 summarize the evidence on SSM usefulness in the diagnosis of GEV and HRV.
Spleen stiffness measurement in esophageal variceal prediction.
| Authors | Type of study | Population | Tool | Cutoff point for diagnosing GEV | Cutoff point for ruling out GEV | Other results |
|---|---|---|---|---|---|---|
| Colecchia et al. 201240 | Prospective | 100 patients with HCV-induced LC | SSM through TE (FibroScan) compared with noninvasive tools (LSM, LSPS, and PC/SD) and the HVPG | 55 kPa (LR– 0.029), | <41.3 kPa (LR– 0.029) |
|
| Stefanescu et al., 2011 57 | Prospective | 135 patients with LC due to HCV or ALD | SSM by TE compared with LSM and PC/SD | 46.4 kPa (AUROC 0.78) |
| |
| Sharma et al., 201341 | Prospective | 174 patients with HCV, HBV, ALD, cryptogenic cirrhosis | SSM by TE, LSM, HPVG, LSPS, PC/SD | ≥ 40.8 kPa (S:94%, Sp: 76%, PPV: 91%, NPV: 84%, AUROC 0.89 and 86% diagnostic accuracy) |
| |
| Nagai et al., 202254 | Cross-sectional | 123 CLD due to HCV, HBV, ALD, MASLD, idiopathic PHT, 55 with LC | SSM (100 Hz) compared with SSM (50 Hz), other NITs (APRI, FIB-4, PC/SD, LSPS), HVPG and endoscopy |
|
ACLD: acute-on-chronic liver disease; ALD: alcohol-related liver disease; AUROC: area under the receiver operating characteristic; CLD: chronic liver disease; GEV: gastroesophageal varices; HBV: hepatitis B virus; HCV: hepatitis C virus; HRV: high-risk varices; HVPG: hepatic venous pressure gradient; LC: liver cirrhosis; LSPS: liver stiffness-spleen size-to platelet ratio score; MASLD: metabolic dysfunction-associated steatotic liver disease; NIT: noninvasive test; NPV: negative predictive value; PC/SD ratio: platelet count to spleen diameter ratio; PHT: portal hypertension; PPV: positive predictive value; S: sensitivity; PSR: platelet-to-spleen volume ratio; Sp: specificity; SSM: spleen stiffness measurement; TE: transient elastography.
Spleen stiffness measurement in the prediction of high-risk esophageal varices.
| Authors | Type of study | Population | Tool | Cutoff point for diagnosing HRV | Cutoff point for ruling out HRV | Other results |
|---|---|---|---|---|---|---|
| Colecchia et al., 201862 | Retrospective | 498 patients with cACLD of different etiologies (most were HCV) | Combined SSM (50 Hz) and Baveno VI criteria model | ≤ 46 kPa (AUROC 0.847) |
| |
| Stefanescu et al., 202044 | Prospective | 216 patients, HCV, HBV, ALD, others | SSM (100 Hz) vs SSM (50 Hz), other NITs (LSM, LSPS, Lok index, PSR, Fib4, APRI) and Baveno VI and HPVG criteria | < 41.3 kPa |
| |
| Hirooka et al., 202160 | Retrospective | 349 HCV and HBV | SSM (50 Hz) and HVPG | 45 kPa (S: 92.1%; Sp: 64.6%; PPV:47.3%; NPV: 95.9%, AUROC; 0.85) | ||
| Tanaka et al., 202161 | Prospective | 292 HCV and HBV | SSM by TE (FibroScan), endoscopy and HVPG | 46.4 kPa (AUROC: 0.88) |
| |
| Nagai et al., 202254 | Cross-sectional | 123 CLD due to HCV, HBV, ALD, MASLD, idiopathic PTH 55 with LC | SSM (100 Hz) compared with SSM (50 Hz), other NITs (APRI, FIB-4, PC/SD, LSPS), HVPG and endoscopy | 43.8 kPa (S: 93.3%, Sp: 82.0%, PPV: 70%, AUROC: 0.9) | SSM (100 Hz) was superior to SSM (50 Hz) for diagnosing HRV (AUROC = 0.941). | |
| Vanderschueren et al., 202463 | Retrospective | 201 patients with ACLD due to ALD, MASLD, cholestatic and viral disease | Sequential model with the Baveno VI criteria and SSM (100 Hz) | 43 kPa (S: 96.7%, Sp: 40 %, PPV: 46.1%, NPV: 95.8%) |
ACLD: acute-on-chronic liver disease; ALD: alcohol-related liver disease; AUROC: area under the receiver operating characteristic; HBV: hepatitis B virus; HCV: hepatitis C virus; HRV: high-risk varices; HVPG: hepatic venous pressure gradient; LC: liver cirrhosis; LSPS: liver stiffness-spleen size-to platelet ratio score; MASLD: metabolic dysfunction-associated steatotic liver disease; NIT: noninvasive test; NPV: negative predictive value; PC/SD ratio: platelet count to spleen diameter ratio; PHT: portal hypertension; PPV: positive predictive value; S: sensitivity; PSR: platelet-to-spleen volume ratio; Sp: specificity; SSM: spleen stiffness measurement; TE: transient elastography.
In a prospective clinical trial that included 92 patients with LC secondary to HCV infection, with a 2-year follow-up, Colecchia et al.64 showed that SSM may be useful in determining the clinical decompensation risk. They reported that SSM was an independent predictor of the development of decompensation, with a low risk of developing it in the following 2 years in patients with SSM values < 54 kPa (negative predictive value [NPV] 0.975, 95% CI 0.86−0.99). They also developed a predictive model based on SSM and the model for end-stage liver disease (MELD) that was as accurate as the HVPG for predicting the development of clinical decompensation, with a cutoff point of 0.76 for identifying patients at high risk of complications (PPV = 0.875, 95% CI 0.47−0.997), making it a valid alternative to HVPG. Subsequent studies have confirmed those results. Gaspar et al.65 conducted a prospective study on 242 patients, most of whom presented with cACLD secondary to ALD (62%). They found that the SSM can be an excellent tool for predicting liver-related complication risk (AUROC 0.823, 95% CI 0.742−0.904), utilizing a cutoff point of 50 kPa for determining the increase in the risk of liver decompensation.
In addition to determining the risk of clinical decompensation, the SSM may also be useful in evaluating recompensation. In a recent prospective clinical trial, Gülcicegi et al.66 reported that serial SSM application was a useful marker for evaluating recompensation. The SSM revealed a significant decrease in spleen stiffness on day 3, compared with the initial evaluation (−18.5 kPa, −21.53 %; p = 0.0002), with no additional decrease on day 5 (−17.63 kPa, −21.23 %; p = 0.0326).
Response to treatment with beta blockersClinical trials have shown that noninvasive monitoring of changes in splenic stiffness through SSM by TE is useful for evaluating response to beta blocker therapy. The study conducted by Marasco et al.67 demonstrated that changes in SSM correlated with changes in the HVPG (r = 0.78, p < 0.0001), identifying responders to beta blocker therapy as those who had a decrease in SSM ≥ 10% (sensitivity: 100%: specificity: 60%, PPV: 100%; NPV: 90%). A subsequent observational descriptive study that included 41 patients who were candidates for receiving beta blocker therapy as primary and secondary variceal bleeding prophylaxis, reported that an SSM ≥ 74 kPa had 100% sensitivity and 60% specificity, as well as an NPV of 100%, for predicting poor acute response to beta blocker therapy. The same cutoff point determined 87% sensitivity, 71% specificity, and an NPV of 71% for predicting poor chronic response, utilizing HVPG measurement for defining both acute and chronic responses to beta blockers.68 More recently, a study by Giuffrè et al.69 showed that a decrease in SSM > 11.5% at 3 months from the start of treatment was associated with response to beta blockers, with 86.4% diagnostic accuracy and an AUROC of 0.89. However, only EGD was used as a reference standard in that study.
Limitations of the SSMDespite the widely mentioned advantages, there could be certain limitations in carrying out SSM in routine clinical practice, such as accessibility and failure rate. Several studies have reported a failure rate for SSM from 13% to 17%,41,54,61,62 with particular difficulty in measuring small spleens with an anteroposterior diameter under 4 cm.57 Therefore, a precise ultrasound (mode B) examination of the spleen before performing TE is mandatory, to ensure that the ultrasound beam remains inside the splenic parenchyma.40 This last consideration appears to be less relevant once the 100 Hz splenic probe is used, configured with the specifications designed for measuring the spleen.44,48 The implementation of an external 3D printing device may also increase the success rate from 35% to 76.9% in patients with small spleens (<100 mL).61 Some reports suggest that the quantification of the controlled attenuation parameter (CAP) in the spleen may serve as a reliability parameter when carrying out the SSM and that patients with a low CAP (≤118 dB/m) have a better HVPG and SSM correlation.60
ConclusionsSplenic elastography is a useful and reliable tool for evaluating the PP in patients with CC and DC. The most recent scientific evidence has shown that this technique closely correlates with the HVPG, predicts CSPH, severe PHT, and GEV, and may also be useful in evaluating the response to treatment with beta blockers. Consequently, splenic elastography is a valuable method for risk stratification, clinical follow-up, and therapeutic decision-making in patients with chronic liver disease.
FundingNo sponsorship of any kind was received to carry out this article.
The authors declare no conflict of interest
