Shear wave elastography in healthy patients: Pancreatic stiffness is less reliable than liver and spleen measurements

Abstract

BACKGROUND

Shear wave elastography (SWE) is a non-invasive ultrasound-based technique used to assess tissue stiffness, which reflects underlying pathological changes. While SWE has been widely applied for liver fibrosis evaluation, its application to other abdominal organs, such as the spleen and pancreas, is gaining interest. However, normal stiffness values and inter-system agreement remain poorly defined.

AIM

To assess the feasibility and agreement of liver, spleen, and pancreas stiffness using three SWE methods.

METHODS

This single-center observational study enrolled 50 healthy adult volunteers. Liver, spleen, and pancreas stiffness were assessed using three SWE methods: Point-SWE (p-QElaXto) and 2-Dimensional-SWE (2D-QElaXto) with Esaote MyLab 9, and 2D-SWE with SuperSonic Imagine. Feasibility, inter-operator reproducibility, and concordance among systems were evaluated. Stiffness was expressed as median kPa values, and technical reliability was assessed using the interquartile range/median ratio and stability index thresholds.

RESULTS

Liver and spleen stiffness assessment was feasible in > 98% of patients, while pancreas stiffness was measurable in 84%-88% depending on the SWE technique. Mean liver stiffness ranged between 3.9-4.7 kPa across techniques, spleen stiffness ranged from 19.4-23.0 kPa, and pancreas stiffness from 5.2-7.6 kPa. Inter-operator agreement was excellent for liver (intraclass correlation coefficient > 0.90) and good to moderate for spleen and pancreas (intraclass correlation coefficient from 0.43 to 0.90). Bland-Altman analysis confirmed good correlation but also systematic differences among devices, especially in pancreas measurements.

CONCLUSION

This is the first study to establish normal liver, spleen, and pancreas stiffness using MyLab 9 SWE integrated methods as compared to SuperSonic Imagine, with acceptable inter-technique agreement. Liver and spleen values matched existing guidelines; pancreas SWE showed more variability and reduced reproducibility.

Key Words: Shear wave elastography; Stiffness; Liver; Spleen; Pancreas; Healthy patients; Ultrasound; Elastography

Core Tip: In the literature, normal stiffness values for liver, spleen, and pancreas in healthy individuals are not well established. This is partly due to differences in equipment and cut-off values used, and partly due to inter-operator variability. In this study, we evaluated these parameters in 50 healthy individuals using three different methods (two of which were integrated into the same scan) to assess their correlation with clinical and technical features.

INTRODUCTION

Shear wave elastography (SWE) is a non-invasive ultrasound (US) technique that quantifies tissue stiffness and is increasingly applied to abdominal organs[1,2]. Establishing reference values in healthy individuals is crucial, since abnormal stiffness is associated with clinically relevant conditions such as liver fibrosis, portal hypertension, and pancreatic disease. In chronic liver disease, SWE-derived liver stiffness strongly correlates with fibrosis stage and disease progression[3-5]. Spleen stiffness has also emerged as a promising surrogate marker of portal hypertension[2,6,7]. However, reliable cut-offs still require robust normal ranges in healthy subjects and reproducible measurements across different systems. In the literature, liver and spleen normal stiffness values are not well established. There is a great variability of normal stiffness values, partly due to the availability of different equipment and the different cut-off employed, partly due to inter-operator variability. Liver stiffness normal values disagreement was reviewed and summarized in a range from 2.6 kPa to 7.5 kPa[8]. Spleen stiffness normal values were investigated in several studies, and their values ranged from 16.6 ± 2.5 kPa[1] to 18.14 ± 3.08 kPa[9].

The use of SWE is steadily increasing, also engaging other organs such as the pancreas. In fact, pancreatic stiffness has been linked to steatosis, metabolic syndrome, acute and chronic pancreatitis, and post-surgical outcomes[10]. Nevertheless, normal reference values remain poorly defined, and data are inconsistent across studies, with studies reporting normal values ranging from 0.52 m/second to 4.60 m/second[11,12]. A major limitation of SWE is the variability among US platforms and even across software versions of the same manufacturer. Recent consensus guidelines[13] have underlined the need for harmonization and for studies assessing inter-system reproducibility before stiffness thresholds can be widely adopted in practice.

Esaote MyLab 9 is one of the latest equipment introduced in 2017 by Esaote (Genova, Italy), and it is embedded with point-SWE (p-QElaXto) and 2-Dimensional-SWE (2D-QElaXto). To date, no published studies have established normal reference values for abdominal organs using this platform to the best of our knowledge. The aim of the present study was to assess the feasibility and agreement of liver, spleen, and pancreas stiffness measurements obtained with p-QElaXto and 2D-QElaXto on Esaote MyLab 9, compared with 2D-SWE on the SuperSonic Imagine Aixplorer (SSI). We sought to define normal stiffness ranges in healthy individuals and to evaluate their relationship with clinical and technical factors. Additionally, we examined inter-operator variability across the different SWE modalities.

MATERIALS AND METHODS

Study design

The study was designed as an observational single-center study according to the principles of the Declaration of Helsinki and the ethical standards of the local institutional research committee (Comitato Etico, Fondazione Policlinico Universitario Agostino Gemelli IRCCS). All subjects were made aware of the purpose of the study and agreed by signing the informed consent before taking part in the examination. We enrolled 50 healthy adult subjects (27 men and 23 women), primarily from hospital employees, residents, and students with no history of chronic diseases or other pathological conditions, in a period from January 2021 to June 2021.

The inclusion criteria were: (1) Age ≥ 18 years; no history of focal and/or diffuse disease of the analyzed organs; (2) No hematological disorders or other pathological conditions; (3) Negative viral serology for hepatitis B, hepatitis C and human immunodeficiency virus; (4) Laboratory tests within the normal range (in the previous 6 months), alcohol consumption not exceeding 30 g/daily for men and 20 g/daily for women; and (5) Suitable visualization and normal morphology of examined organs in US imaging. The exclusion criteria were: (1) Body mass index (BMI) > 30 kg/m²; (2) Moderate/severe steatosis at standard B-mode; and (3) Ongoing or previous infections, recent or past injury, important recent or past bleeding, immune system disorders, and pregnant or breastfeeding women. Subjects with mild hepatic steatosis or slightly overweight were retained, as low-grade steatosis is common in the general population and does not substantially affect shear-wave elastography. Moreover, standard B-mode US reliably detects moderate-to-severe, but not necessarily mild steatosis[14].

Equipment and technical approach

The examination consisted of stiffness evaluation through p-QElaXto, 2D-QElaXto, and SSI by an experienced operator (> 5 years of experience with different elastography techniques). A standardized protocol was used to minimize pre-compression, enforce breath-hold at end-expiration, and ensure uniform region of interest (ROI) placement. Moreover, to test the reproducibility of the SWE techniques, the first 10 patients enrolled in the study underwent additional elastography examination performed by a second operator, and median organ stiffness values were compared. To minimize measurement bias, operators performed SWE examinations independently and were blinded to prior results and to each other’s measurements when assessing interobserver reproducibility.

Liver, spleen, and pancreas stiffness were measured with the patient in the supine position and during breath-hold in neutral respiration, using standardized acoustic windows for each organ. For liver and spleen, intercostal approaches were used; for pancreas, the best available window (head, body, or tail) was selected according to individual anatomy. For each organ, up to 10 valid measurements were obtained in accordance with the European Federation of Societies for Ultrasound in Medicine and Biology and the World Federation for Ultrasound in Medicine and Biology (WFUMB recommendations, which indicate that a minimum of 5-10 acquisitions improves measurement reliability and reproducibility, especially in liver elastography. Full details of probe placement and acquisition protocols are provided in the Supplementary material.

SWE with Esaote was performed with a MyLab 9 equipment with convex broadband abdominal probe C1-8 MHz, obtaining up to 10 valid measurements with p-QElaXto and 5 valid measurements with 2D-QElaXto for every organ, considering median values in kPa and the ratio between interquartile range (IQR) and the median value (M) (directly provided by the software) for the analysis (Figure 1A-F, respectively). To further investigate whether a reduced number of acquisitions could still provide reliable results for p-QElaXto, we compared the mean values obtained from 10 measurements with those derived from the first 5 acquisitions. 2D-SSI was performed with Mach 30 Aixplorer (Aix-en-Provence, France) with a convex probe XC6-1 MHz, obtaining 5 valid measurements for every organ, recording median values in kPa and stability index (SI) (Figure 1G-I). For 2D-SWE systems, liver stiffness was measured in an ROI of 10 mm and 15 mm in a central area of the colorimetric map showing the best signal homogeneity. Moreover, for pancreas stiffness measurement, the ROI dimension was adjusted according to organ anatomy when using 2D-SWE techniques. All the SWE measurements for every single patient lasted no longer than 10 minutes for both elastography techniques.

Figure 1 Shear wave elastography in healthy patients. A-C: Point-shear wave elastography of liver (A), spleen (B), and pancreas (C); D-F: 2-dimensional-shear wave elastography of liver (D), spleen (E), and pancreas (F); G-I: 2-dimensional-SuperSonic Imagine Aixplorer of liver (G), spleen (H), and pancreas (I).

Technical feasibility was calculated after the exclusion of technical failures and unreliable measurements. Based on current literature, “technical failures” were defined as no successful measurement after 10 attempts for p-QElaXto and as the inability to measure stiffness in a homogeneous ROI of at least 10 mm for 2D-SWE techniques. “Unreliable measurements” were defined by an IQR/M ratio > 0.30 and SI < 80% for QElaxto and SSI techniques, respectively[15,16]. The remaining were considered “technically feasible”.

Statistical analysis

Individual patient measurements derived from multiple samples (e.g., liver stiffness derived from five or ten measurements) were expressed as medians. IQR for liver stiffness measurements and IQR/M were calculated at the individual patient level and were considered as an indicator for measurement quality. Pearson correlation was used to assess univariate relationships between continuous variables after verification of data normality with the Shapiro-Wilk test. Pearson’s correlation coefficients were classified as follows: 0-0.19, very weak correlation; 0.20-0.39, weak; 0.40-0.59, moderate; 0.60-0.79, strong; and 0.80-1.00, very strong. Interobserver reproducibility was estimated with a single-measure two-way mixed intraclass correlation coefficient (ICC) test for absolute agreement in a subgroup of patients. Based on the 95% confidence interval (CI) of the ICC estimate, values less than 0.5, between 0.5 and 0.75, between 0.75 and 0.9, and greater than 0.90 were considered indicative of poor, moderate, good, and excellent reliability, respectively[17]. In addition to the ICC used to assess inter-operator reproducibility, we also calculated the concordance correlation coefficient (CCC) and the coefficient of variation (CV%) to evaluate inter-technique agreement between the three SWE methods (2D-QElaXto, p-QElaXto, and SSI). The CCC was computed according to Lin’s method, combining measures of precision and accuracy, while the CV was calculated as the ratio between the standard deviation of paired differences and their mean value, expressed as a percentage. All comparisons were considered exploratory, and no adjustment for multiple testing was applied. Statistical analysis was performed by using the Statistical Package for the Social Sciences (SPSS 22.0; SPSS Inc., Chicago, IL, United States).

RESULTS

Study population

We enrolled a total of 50 patients in the study period. Organ stiffness measurements were obtained by the three different SWE techniques (2D-QElaXto, p-QElaXto, and SSI). Descriptive statistics of the “healthy” study population are reported in Table 1. Among the whole cohort, the mean BMI was 23 kg/m², with only two subjects showing a hyperechoic liver (both with mild liver steatosis and BMI > 25 kg/m²) and one subject with a hyperechoic pancreas (mild fatty infiltration) on US scanning. In two other patients, pancreas visualization on US was suboptimal due to intestinal gas, and therefore, pancreatic stiffness was not performed.

Table 1 Descriptive statistics of the “healthy” study population.

Variables	Total	Males	Females
Age (years)	37 ± 10	39 ± 11	35 ± 9
BMI (kg/m2)	23 ± 3	24 ± 2	21 ± 2
Spleen bipolar diam (cm)	10.8 ± 1.3	11.6 ± 0.9	9.8 ± 1.0
Spleen area (cm2)	33.3 ± 8.6	37 ± 8	29 ± 7
Pancreas body thick (cm)	1.1 ± 0.3	1.2 ± 0.3	1.1 ± 0.2

BMI: Body mass index.

Feasibility and reproducibility of organ stiffness measurements

The feasibility of the three SWE techniques was excellent, and successful measurements were obtained in all 50 subjects for liver and spleen stiffness assessment. In only one patient, IQR/M ratio was > 30% when performing liver stiffness with 2D-QElaXto with a ROI of 10 mm (but not with a ROI of 15 mm). In another single subject, spleen stiffness measurements were unreliable (IQR/M ratio > 30% or SI < 80%) with all three techniques. These unreliable measurements were probably due to the reduced spleen width (< 2.5 cm) found in this lean female volunteer (BMI 19 kg/m²). Moreover, in six other patients, spleen stiffness assessment with p-QElaXto was unreliable when performing only 5 measurements, but not when completing the evaluation with 10 measurements. In four of the aforementioned patients, spleen stiffness was also unreliable with 2D-QElaXto by using an ROI of 10 mm, but not with a ROI of 15 mm. To summarize, the feasibility and reliability for liver and spleen stiffness assessment were 100% if 10 measurements were performed with p-QElaXto and a ROI of 15 mm was selected when using the 2D-SWE technique.

When performing SWE of the pancreas, the feasibility was not quite as excellent as for the liver and spleen. Specifically, in six subjects undergoing pancreas evaluation with 2D-QElaXto, measurements were either unreliable or showed inhomogeneous filling of the colorimetric map (Figure 2). Assessment with p-QElaXto was unreliable in ten patients when only the first 5 measurements were considered, but if the other five measurements were added, IQR/M ratio was > 30% only in four patients. Pancreas stiffness with SSI was not feasible in four patients due to SI < 80% (three patients) or inhomogeneous filling of the colorimetric map (one patient). Eventually, if the two failed visualizations of the organ are included, the overall feasibility of pancreas stiffness was 84%, 88% and 88% for 2D-QElaXto, p-QElaxto with 10 measurements, and SSI, respectively.

Figure 2 2-dimensional-shear wave elastography of pancreas: Inhomogeneous filling of the colorimetric map suggests an unreliable measurement.

The ICC for liver stiffness measurement evaluated on 10 subjects was excellent for all three SWE techniques: 0.963 (95%CI: 0.850.99) for 2D-QElaXto (with a ROI of 15 mm), 0.906 (95%CI: 0.62-0.98) for p-QElaXto (10 measurements), and 0.928 (95%CI: 0.70-0.98) for SSI (with a ROI of 15 mm). Interoperator agreement was also very good for spleen stiffness measurements, but only for p-QElaXto and SSI, with a reliability of 0.929 (95%CI: 0.56-0.98) and 0.835 (95%CI: 0.38-0.96), respectively. 2D-QElaXto showed only a moderate agreement with an agreement of 0.519 (95%CI: -0.85 to 0.87). A similar trend was also observed for pancreas stiffness interobserver reproducibility: 0.831 (95%CI: 0.28-0.96) for p-QElaXto, 0.844 (95%CI: 0.41-0.96) for SSI, but only 0.434 (95%CI: -1.00 to 0.86) for 2D-QElaXto.

Liver stiffness measurement and concordance between SWE devices

The overall values of the median liver stiffness in the 50 healthy subjects ranged from 2.4 kPa to 8.1 kPa across the different SWE techniques (Table 2). Specifically, considering the whole study group, the mean liver stiffness value for 2D-QElaXto was 3.9 kPa regardless of whether a 10 mm or 15 mm ROI was used. For p-QElaX the mean liver stiffness values were 4.4 kPa for both 5 and 10 measurements. Mean liver stiffness obtained with SSI was 4.6 kPa and 4.7 kPa using a 10 mm and 15 mm ROI, respectively.

Table 2 Median liver stiffness values.

SWE technique	2D-QElaXto (10 mm ROI)	2D-QElaXto (15 mm ROI)	p-QElaXto (5 meas)	p-QElaXto (10 meas)	SSI (10 mm ROI)	SSI (15 mm ROI)
Mean liver stiffness (kPa)	3.9	3.9	4.4	4.4	4.6	4.7
Range (kPa)	2.4-6.5	2.7-6.8	3-0-8.1	3.0-8.0	3.5-7.3	3.5-7.1
SD	0.8	0.8	0.9	0.9	0.9	0.9
95%CI	3.66-4.11	3.69-4.15	4.13-4.66	4.12-4.65	4.41-4.89	4.44-4.93

2D-QElaXto: 2-dimensional-shear wave elastography; p-QElaXto: Point-shear wave elastography; SSI: SuperSonic Imagine; ROI: Region of interest; CI: Confidence interval.

Bland-Altman analysis showed that 2D-QElaXto average values were 0.8 kPa (CI: -1.9 kPa to 0.4 kPa) lower than those obtained with SSI, even in the presence of a significant correlation (r = 0.75; P < 0.001). A similar trend was observed for mean p-QElaXto values, which were only 0.3 kPa (CI: -1.6 kPa to 1.0 kPa) lower than those obtained with SSI (r = 0.71; P < 0.001) (Supplementary material). For the liver stiffness measurements, concordance was good (CCC = 0.67-0.80; CV < 12%), confirming the robustness of measurements across systems.

Spleen stiffness measurement and concordance between SWE devices

The mean value of spleen bipolar diameter was 10.8 ± 1.3 cm, while the splenic area and spleen width mean values were 33.3 ± 8.6 cm² and 3.5 ± 0.6 cm, respectively. The overall values of the median spleen stiffness in the whole group ranged from 9.8 kPa to 46.9 kPa for the various SWE techniques (Table 3). Specifically, the mean spleen stiffness value for 2D-QElaXto was 19.4 kPa with an ROI of 10 mm and 20.0 kPa with an ROI of 15 mm. For p-QElaX, the mean spleen stiffness values were 23.0 kPa and 22.9 kPa when performing 5 or 10 measurements, respectively. Mean spleen stiffness values obtained with SSI were 21.9 kPa using a 10 mm ROI and 21.7 kPa with a 15 mm ROI.

Table 3 Median spleen stiffness values.

SWE technique	2D-QElaXto (10 mm ROI)	2D-QElaXto (15 mm ROI)	p-QElaXto (5 meas)	p-QElaXto (10 meas)	SSI (10 mm ROI)	SSI (15 mm ROI)
Mean spleen stiffness (kPa)	19.4	20.0	23.0	22.9	21.9	21.7
Range (kPa)	9.8-26.7	11.7-32.6	13.5-46.9	13.1-36.1	13.4-35.9	13.5-33.8
SD	3.9	4.3	6.1	5.0	4.8	4.6
95%CI	18.27-20.60	18.72-21.20	21.10-24.86	21.45-24.34	20.57-23.33	20.37-23.04

2D-QElaXto: 2-dimensional-shear wave elastography; p-QElaXto: Point-shear wave elastography; SSI: SuperSonic Imagine; ROI: Region of interest; CI: Confidence interval.

Bland-Altman analysis showed that 2D-QElaXto average values were 1.7 kPa (CI: -7.4 kPa to 3.9 kPa) lower than those obtained with SSI, even in the presence of a significant good correlation (r = 0.79; P < 0.001). Regarding p-QElaXto, mean values were 1.2 kPa (CI: -6.4 kPa to 8.8 kPa) higher than those obtained with SSI (r = 0.68; P < 0.001) (Supplementary material). For the spleen stiffness measurements, CCC ranged from 0.61 to 0.73 (CV < 13%), consistent with moderate-to-good agreement.

Pancreas stiffness and variability between SWE devices

Pancreas stiffness evaluation was strongly influenced by organ visualization in every single subject. Consequently, a complete US assessment of the pancreas - including head, body, and tail - was achieved only in 23 subjects (46% of the whole population). SWE was therefore performed in the organ section with the best acoustic window. For 2D-SWE techniques (2D-QElaXto and SSI), the ROI dimension was adjusted according to organ anatomy (mean ROI size was 8.0 mm). The median pancreas stiffness in the whole group ranged from 2.7 kPa to 18.0 kPa for the various SWE techniques (Table 4). Particularly, the mean pancreas stiffness value for 2D-QElaXto was 5.2 kPa, while the mean values for p-QElaXto were 5.9 kPa and 6.1 kPa when performing 5 or 10 measurements, respectively. The mean pancreas stiffness value for SSI was 7.6 kPa.

Table 4 Median pancreas stiffness values.

SWE technique	2D-QElaXto	p-QElaXto (5 meas)	p-QElaXto (10 meas)	SSI
Mean pancreas stiffness (kPa)	5.2	5.9	6.1	7.6
Range (kPa)	2.9-15.0	2.8-17.4	3.1-18.0	2.7-17.2
SD	1.9	2.5	2.4	2.6
95%CI	4.63-5.80	5.03-6.70	5.40-6.83	6.86-8.41

2D-QElaXto: 2-dimensional-shear wave elastography; p-QElaXto: Point-shear wave elastography; SSI: SuperSonic Imagine; CI: Confidence interval.

Bland-Altman analysis showed that 2D-QElaXto average values were 2.2 kPa (CI: -6.3 kPa to 1.9 kPa) lower than those obtained with SSI, with only a moderate correlation (r = 0.56; P < 0.001). A similar trend was observed for mean p-QElaXto values, which were 1.5 kPa (CI: -6.4 kPa to 3.4 kPa) lower than those obtained with SSI (r = 0.49; P = 0.01) (Supplementary material). However, the correlation was stronger when comparing 2D-QElaXto with p-QElaXto average values (r = 0.75; P < 0.05). As opposed to liver and spleen, pancreatic stiffness measurements showed lower concordance (CCC = 0.35-0.66; CV 22%-32%), in line with the higher variability already observed in the Bland-Altman analysis.

DISCUSSION

Over recent years, ultrasonography and elastography techniques have garnered growing interest due to their capability for both qualitative and quantitative analysis of tissue stiffness, offering additional data compared to traditional semeiological examination. This heightened interest has led to the development of a wide array of US commercial devices, equipped with their own dedicated SWE technique. Our study aimed to assess the feasibility and agreement among three different SWE techniques (2D-QElaXto, p-QElaXto, and 2D-SWE SSI) to establish normal stiffness values for the liver, spleen, and pancreas, considering the substantial inter-system variability.

The different SWE techniques demonstrated excellent feasibility for the assessment of liver and spleen stiffness, whereas the feasibility for pancreatic evaluation was moderately lower. This limitation is primarily attributable to various anatomical and functional factors. Assessing pancreatic stiffness remains a technically demanding and clinically intriguing challenge. The retroperitoneal location increases depth and bowel-gas interposition, often limiting ROI size and reducing map homogeneity[18]. Proximity to the aorta may introduce pulsation-related artifacts[19]. Measurements are also sensitive to respiratory phase, with deep inspiration yielding higher values than free breathing[20]. Beyond acquisition constraints, variability is influenced by anthropometric factors such as BMI and gland depth[18,21]. Age and sex also play a role, affecting measured stiffness and inter-observer agreement[22]. In our study, a complete US evaluation of the pancreas was feasible in only 23 subjects, and, during 2D-SWE acquisitions, the ROI size was adapted to the specific anatomy of each patient. Simethicone-water-rotation protocols have been proposed to enhance pancreatic visibility[23]. However, their adoption in routine clinical settings is limited, and adequate pancreatic imaging remains challenging. In addition, both inter- and intra-operator variability were significantly higher for pancreatic stiffness assessment compared to liver and spleen, further confirming the technical challenges associated with evaluating the pancreatic gland. Bland-Altman analysis confirmed that for liver and spleen, the mean inter-device differences were small (< 2 kPa) and unlikely to affect interpretation in healthy subjects, whereas pancreatic measurements showed wider limits of agreement (up to ± 6 kPa), indicating lower reproducibility. Because of the limited sample size and the homogeneous demographic profile of our study population, no formal multivariate regression analysis could be reliably performed to identify predictors of pancreatic measurement unreliability. Future multicenter studies with larger and more heterogeneous cohorts may be helpful to quantitatively assess how anthropometric and technical parameters contribute to pancreatic SWE variability.

Concerning stiffness measurements, our findings aligned with the well-established reference values for liver elastography, while contributing novel data on spleen and pancreatic elastography, along with relevant technical considerations. In our study, liver stiffness values in healthy subjects measured with 2D-QElaXto and p-QElaXto were 3.9 kPa and 4.4 kPa, respectively, showing a good agreement with 2D-SWE SSI. These values agree with the Society of Radiologists in the US’s Consensus[24] and subsequent WFUMB Guidelines[13] “rule of four”, which states that the liver with a stiffness value equal to or less than 5 kPa (1.3 m/second) has a high probability of being normal. Furthermore, liver stiffness evaluation using different sizes of ROI for 2D-SWE techniques, 10 mm and 15 mm, showed no significant variation with similar values to other studies[8,9,25]. These findings support the robustness of the “rule of four” as a universal reference threshold that can be reliably applied across different elastographic equipment to exclude the presence of significant liver fibrosis.

Regarding the spleen stiffness, values ranged from 19.4 kPa (2D-QElaXto, 10 mm ROI) to 23.0 kPa (p-QElaXto, 5 measurements). Few studies have investigated spleen stiffness in healthy adults, and their reported values are consistent with our results[1,9,26]. For example, Leung et al[26] reported a mean spleen stiffness value of 17.3 ± 2.6 kPa in a cohort of healthy volunteers, with no statistically significant difference observed between male and female subjects. In recent years, efforts have focused on defining universal cut-off values for spleen elastography. Similar to liver elastography, these thresholds aim to rule out clinically significant portal hypertension (CSPH) or high-risk esophageal varices in patients with compensated advanced chronic liver disease (cACLD)[7,27,28]. Recently, the role of spleen stiffness in identifying or ruling out CSPH in cACLD patients was clarified by the 2024 WFUMB guidelines[13], which introduced a “Baveno VII−SSM dual-cutoff model”[13,29]. This model incorporates LSM, SSM and platelet count to improve diagnostic accuracy, establishing 21 kPa as the upper SSM threshold for excluding CSPH - slightly lower than the threshold proposed in Baveno VII[13]. It should be noted, however, that these thresholds refer to values obtained by transient elastography in cACLD patients, and no guideline-based cut-offs are currently available for SWE. The spleen stiffness values obtained in our cohort, together with the high reproducibility of liver and spleen measurements, support the clinical use of SWE for non-invasive assessment of liver fibrosis and for spleen-based evaluation of portal hypertension. Future studies should clarify whether spleen stiffness values in healthy individuals overlap with those of cACLD patients without CSPH.

Lastly, regarding pancreatic elastography, the mean stiffness values in healthy subjects in our cohort ranged from 5.2 kPa with 2D-QElaX to 6.1 kPa with SSI. These findings are in line with previously reported normal ranges. For example, Kaya et al[20] found median values around 4-5 kPa during free breathing, while Nosakova et al[21] reported a mean of approximately 6.1 kPa using 2D-SWE. Similarly, Gallotti et al[12] described mean velocities of 1.40 m/second (approximately 6 kPa) with ARFI in healthy volunteers, and Choudhury et al[30] reported mean values of 8.7-9.0 kPa in an Indian cohort, highlighting possible variability related to population and methodology. Other studies confirmed normal pancreas stiffness around 6-7 kPa in controls[31], whereas markedly higher values (23.8 ± 6.7 kPa) were reported by Durmaz et al[32]. Taken together, our data fall within the lower-to-mid portion of the spectrum of published reference values, supporting their reliability while underlining the considerable heterogeneity observed across studies depending on technique, respiratory phase, and anthropometric factors.

Although inter-system variability currently limits routine pancreatic SWE, emerging evidence indicates that stiffness measurements may provide additional value in detecting pancreatic steatosis, in correlation with metabolic syndrome and cardiovascular risk stratification[33,34], and in the evaluation of inflammatory changes in acute and chronic pancreatitis[11,33]. Latest evidence reports also a significant correlation between concomitant increase of pancreatic and splenic stiffness in patients with portal hypertension[35]. We acknowledge some potential limitations of our study. Firstly, the limited number of participants in our study may have affected the statistical power, potentially increasing the margin of error. Also, the recruitment of hospital staff and medical students may introduce a selection bias, as this group is relatively homogeneous in terms of age, health status, and BMI. To strengthen the validity of our findings, future multicenter studies should include a more demographically diverse cohort to improve generalizability. Secondly, it is important to note that our investigation was conducted within a single institution. Thirdly, measurements in our study were performed using a generation-specific device, and the normal values generated may be representative of this model and generation. However, it would be advisable to verify the concordance between different models and generations of the same manufacturer because the lack of harmonization across US platforms restricts comparability of SWE values, highlighting the importance of multicenter studies and standardized acquisition protocols for clinical adoption.

CONCLUSION

Our study provides preliminary reference stiffness values in healthy adults using three SWE techniques, with robust liver and spleen reproducibility that supports integration of SWE into multiparametric US assessment of chronic liver disease and portal hypertension. In contrast, pancreatic stiffness measurements remain more variable, highlighting the need for optimized acquisition protocols and standardization before SWE can be reliably applied in pancreatic routine clinical practice. Future multicenter studies with larger cohorts should focus on harmonizing reference values and technical parameters across different systems and manufacturers, enabling SWE to evolve into a truly reproducible, quantitative, and cross-platform non-invasive tool for abdominal organ evaluation.

ACKNOWLEDGEMENTS

Thanks to Fondazione Roma for its commitment to support our research.