In our study, we evaluated the feasibility of a new sampling method for splenic stiffness (SS) measurement by Quantitative Acoustic Radiation Force Impulse Elastography (Virtual Touch Tissue Quantification (VTTQ)).We measured SS in 54 patients with HCV-related cirrhosis of whom 28 with esophageal varices (EV), 27 with Chronic Hepatitis C (CHC) F1–F3, and 63 healthy controls. VTTQ-SS was significantly higher among cirrhotic patients with EV (3.37 m/s) in comparison with controls (2.19 m/s,
Chronic hepatitis C virus (HCV) infection represents a worldwide health concern, with 170 million chronically infected subjects, whose risk of developing cirrhosis within 20 years is estimated to be around 10% to 20% [
As variceal bleeding is a life-threatening condition, current guidelines recommend routine upper gastrointestinal endoscopy to be performed at the time of diagnosis in all patients with known cirrhosis for EV screening and grading. The endoscopic procedure should allow the identification of high-risk subjects, requiring prophylactic treatment with nonselective beta-blockers. Unfortunately, a number of reasons may limit the universal spread of endoscopic screening in cirrhotic patients. Firstly, universal endoscopic screening is expensive, time consuming, and invasive [
The aim of the present study was to compare the values of SS of cirrhotic subjects (with and without EV), measured by VTTQ, with those found among healthy controls and patients with HCV-related chronic hepatitis (fibrosis stages F1–F3), to evaluate the reliability of VTTQ-SS as a surrogate predictor of EV in a cohort of subjects with recently diagnosed HCV-related cirrhosis.
Between January 2009 and January 2011, 73 consecutive patients presenting at the Outpatient Infectious Diseases Unit of the Garibaldi Nesima Hospital in Catania were newly diagnosed as affected with HCV-related liver cirrhosis and considered for enrolment in the present study. In this stage, the diagnosis was based only on a clinical, virological, and ultrasound evaluation.
Subsequently, throughout a 6-month clinical followup, 19 patients were excluded from the study, 17 because of development of severe cirrhosis-related complications or comorbidities (7 patients had hepatocellular carcinoma, 4 severe sepsis, 3 spontaneous bacterial peritonitis, 2 colonic cancer, and 1 lung cancer) and 2 because of refusal to enter the study.
Thus, 54 patients were finally enrolled, 15 with a diagnosis of histologically confirmed Child-Pugh class A cirrhosis and 39 with a clinical diagnosis of Child-Pugh class B cirrhosis. None of the patients had a concomitant HBV and/or HIV coinfection, none had either autoimmune or metabolic disease. Alcohol abuse was excluded by questionnaire. Patients were not under treatment with beta-blockers, diuretics, antibiotics, interferon, or ribavirin by the time of study onset. Moreover, no patient had evidence of cardiac or renal insufficiency. All patients signed a written informed consent prior to study inclusion, in accordance with the Declaration of Helsinki. Patients were asked to hide their clinical features to the ultrasound operator.
Each patient underwent a consecutive 2-day protocol. On the first day, a complete clinical and biochemical evaluation was performed. Then, patients were examined with VTTQ by a single well-experienced operator (LR), blinded to biochemical and clinical data, at the private Outpatient Clinic “Ultrasuoni” in Catania. On the second day, all patients underwent upper gastrointestinal endoscopy. All endoscopic evaluations were performed by a single operator (blinded to biochemical and echographic results) at the Garibaldi Nesima Hospital in Catania.
We enrolled in the study 27 subjects with a virological and histological diagnosis of chronic hepatitis C (CHC), matched for gender and age distribution with the cirrhotic group. Patients with HBV or HIV coinfection, alcohol abuse, and those under antiviral treatment were excluded from the study. Written informed consent was obtained from each subject at enrolment. Biochemical tests and VTTQ were performed on the same day.
63 healthy adult volunteers, matched to cases on age, gender, and ethnic group, were selected as controls among the members of the staff of our institution. They all had normal liver function tests, negative HBsAg, HCV Antibodies, HIV Antibodies, and normal abdominal ultrasonography. All subjects gave their written informed consent prior to study enrolment.
B-mode standard ultrasonography scanning and quantitative ARFI elastography (VTTQ) were performed using a Siemens Acuson S2000 (Siemens AG, Erlangen, Germany) with a 4Cl transducer, as described in detail elsewhere [ the section, as the spleen was subdivided into three sections, subdiaphragmatic, intermediate, and caudal, and measurements were taken along the corresponding thickness diameter, the part of each section (external, central, and internal).
Sites for spleen stiffness measurement by Quantitative Acoustic Radiation Force Impulse Elastography (VTTQ). Cranial section: 1: external, 2 and 3: central, and 4: internal; intermediate section: 5: external, 6: central, and 7: internal; caudal section: 8: external, 9: central, and 10: internal.
The subdiaphragmatic section provided 4 measurements: 1 medial, 2 central, 1 lateral. The other two sections provided 3 measurements: 1 medial, 1 central, and 1 lateral. This difference was due to the fact that the subdiaphragmatic section was larger. We chose the best intercostal space to maximize the surface area of each section on the frontal plane. Figure
A summary of spleen sampling sites by Quantitative Acoustic Radiation Force Impulse Elastography (VTTQ).
Since the procedure could rely on some degree of subjective interpretation by the investigator, we conducted a prior double blind experiment on 16 randomly selected patients among those further enrolled in the study. Briefly, patients were examined for SS by VTTQ as described above by two independent operators within the same day: one operator (LR) was a well-experienced echographiste with a long-term (>4 years) training in ARFI technique, and the other operator (MP) was a young physician who received a prior 4-hour training in ARFI procedure.
Twenty-one randomly selected patients among those who entered the study were summoned again to the Outpatient “Ultrasuoni” Clinic four weeks after the first VTTQ-SS examination and submitted to a new VTTQ-SS evaluation by the same experienced operator (LR). The operator, who was still unaware of clinical and endoscopic data, could not access to the previous ARFI results.
Upper gastrointestinal endoscopywas performed using a flexible video gastroscope (Video Pentax Gastro).
EV were classified as follows. Grade 1: small straight varices; Grade 2: enlarged tortuous varices, occupying less than one third of the lumen; Grade 3: large, coil-shaped varices, occupying more than one third of the lumen.
EV were also classified on the basis of presence or absence of red wheals.
Statistical analysis was performed using statistical computing software R and Graphpad Prism 4. Continuous variables are expressed as median (interquartile range, IQR) and compared by nonparametric Kruskal-Wallis test. Categorical variables are presented as number of cases (percentage) and were compared by the
The best cut-off value to predict the presence of EV was determined by Kolmogorov-Smirnov index, that is, a natural generalization to continuous test of Youden index for binary test [
Pearson’s correlation coefficient was used to evaluate intraobserver agreement; Bland-Altman method was used to assess the agreement between VTTQ-SS measurements obtained by two different observers on the same patients [
Table
Demographic, clinical, echographic, and endoscopic characteristics of the 54 patients with HCV-related cirrhosis.
Variables |
|
---|---|
Age (years)** | 72 (63–79) |
Sex, male/female* | 25 (46.3)/29 (53.7) |
HCV RNA * 105 (IU/mL)● |
|
HCV genotype 1/2/3/4* | 38 (70.4)/4 (7.4)/12 (22.2)/0 ( |
Child-Pugh class A/B* | 15 (27.7)/39 (72.3) |
Total bilirubin (mg/dL)** | 1.3 (0.7–3.1) |
AST (IU/L)** | 57 (22–111) |
ALT (IU/L)** | 66 (31–138) |
Albumin (g/dL)** | 3.8 (2.5–4.1) |
Gamma globulins (g/dL)** | 2.9 (2.1–3.3) |
Platelet count * 103/ |
111 (34–212) |
International normalized ratio (INR)** | 1.29 (1–1.98) |
Echographic spleen diameter (cm)** | 13.9 (9.9–17.8) |
Esophageal varices (EV), yes/no* | 28 (51.8)/26 (48.2) |
EV | |
Grade 1* | 11 (39.3) |
Grade 2* | 11 (39.3) |
Grade 3* | 6 (21.4) |
Red wheals, yes/no* | 2 (7.1)/26 (92.9) |
*Data presented as
AST: aspartate aminotransferase; ALT: alanine aminotransferase; IQR: interquartile range.
Following upper endoscopy evaluation, 48.2% of cirrhotic patients showed no EV, 20.4% had grade 1 EV, 20.4% had grade 2, and 11% had grade 3. Red wheals were observed in 7.1% of patients with EV. The characteristics of the group of cirrhotic patients with EV and those without EV are shown in Table
Characteristics of cirrhotic patients with esophageal varices (EV) compared to cirrhotic patients without EV by Kruskal-Wallis test.
Cirrhotics with esophageal varices ( |
Cirrhotics without esophageal varices ( | |
---|---|---|
Age (years)** | 76 (63–81) | 68 (62–75) |
Sex, male/female* | 13 (46.4)/15 (53.6) | 12 (46.1)/14 (53.9) |
HCV RNA * 105 (IU/mL)● |
|
|
HCV genotype 1/2/3/4* | 18 (64.3)/2 (7.1)/8 (28.6)/0 (0) | 20 (77)/2 (7.7)/4 (15.3)/0 ( |
Child-Pugh class A/B* | 8 (28.5)/20 (71.5) | 7 (27)/19 (73) |
Total bilirubin (mg/dL)** | 1.9 (1–3.1)*** | 1.1 (0.7–2.8)*** |
AST (IU/L)** | 55 (27–111) | 58 (21–105) |
ALT (IU/L)** | 63 (39–136) | 69 (30–129) |
Albumin (g/dL)** | 3.4 (2.1–3.6) | 3.9 (2.7–4.1) |
Gamma globulins (g/dL)** | 3.0 (2.4–3.3) | 2.8 (2.1–3.1) |
Platelet count * 103/ |
96 (34–206)*** | 124 (65–212)*** |
International normalized ratio (INR)** | 1.11 (1–1.59) | 1.31 (1.23–1.97) |
Echographic spleen diameter (cm)** | 14.1 (11.6–17.8)*** | 11.8 (9.1–12.7)*** |
*Data presented as
AST: aspartate aminotransferase; ALT: alanine aminotransferase; IQR: interquartile range.
As for CHC patients, HCV genotype was 1a/b in 88% of patients, 2 in 3.7%, and 3 in 7.4%. 29.6% of them had F1 fibrosis, 40.8% had F2 fibrosis, and 29.6% had F3 fibrosis. Mean HCV-RNA plasma level was
VTTQ-SS values were higher in cirrhotic patients with EV, when compared with cirrhotic subjects without EV, CHC patients, and controls, as shown in Figure
Distribution of spleen stiffness median values measured by Quantitative Acoustic Radiation Force Impulse Elastography (VTTQ) among healthy controls, CHC patients and cirrhotic patients with or without esophageal varices (EV). Splenic stiffness (SS) was significantly higher among cirrhotic patients with EV (3.37 m/s) in comparison with controls (2.19 m/s,
To evaluate the variability associated with SS measurements, we divided patients into 3 subgroups according to SS median values (SS <2 m/s, 2-3 m/s, >3 m/s) and we found that the IQR was significantly higher among subjects with SS median values >3 m/s in comparison with those having SS <2 m/s (
Evaluation of interquartile range (IQR) variability according to spleen stiffness (SS) median values. IQR was significantly higher among subjects with SS median values >3 m/s in comparison with those having SS <2 m/s (
The best VTTQ-SS cut-off value to predict the presence of EV, as determined by Kolmogorov-Smirnov index, was 3.1 m/s. The diagnostic accuracy (AUROC) for the prediction of EV was 0.959 (95% confidence interval (CI) 0.91–1), Se 96.4%, Sp 88.5%, positive predictive value (PPV) 90%, negative predictive value (NPV) 96%, positive likelihood ratio (PLR) 8.36, and negative likelihood ratio (NLR) 0.04 (Figure
Receiver operating characteristic (ROC) curve of spleen stiffness measured by Quantitative Acoustic Radiation Force Impulse Elastography (VTTQ) for the prediction of esophageal varices. AUROC: area under the ROC curve; CI: confidence interval; NLR: negative likelihood ratio; NPV: negative predictive value; PLR: positive likelihood ratio; PPV: positive predictive value.
Figure
Fagan's Nomogram for esophageal varices (EV). Using the pretest probability for EV of 51.8% (the overall prevalence in our cohort), the posterior probability of having EV was 90%. LR: likelihood ratio; Prior Prob.: pretest (prior) probability; Posterior Prob.: posterior probability.
In our cohort of cirrhotic patients, SS measured by VTTQ showed a significant negative correlation with platelet count (
By Bland-Altman method, there was no significant difference between VTTQ-SS values of the 16 patients obtained from two different sonographers (Figure
Agreement between Quantitative Acoustic Radiation Force Impulse Elastography (VTTQ) data obtained by two different observers by Bland-Altman method.
Similarly, VTTQ-SS, as evaluated on 21 patients by the same operator in two different time points, was not significantly different (2.97 (IQR 2.52–3.23) m/s for the first measurement versus 2.66 (IQR 2.45–3.33) m/s for the second one). Pearson’s correlation coefficient was 0.933 (0.84–0.97).
PH is a common consequence of chronic liver diseases, leading to the formation of esophageal and gastric varices and other severe complications, such as portosystemic encephalopathy and sepsis [
So far, only few reports have assessed the possibility to use ARFI elastography of the spleen as a noninvasive diagnostic tool for the presence of EV [
In comparison with previous reports, our study showed a higher diagnostic performance of VTTQ in predicting EV. A possible explanation is the systematic and extensive sampling of the spleen, which may have been able to adequately represent the heterogeneity of SS distribution. Bota et al. [
In the study of Takuma et al. [
Our study has some limitations. Firstly, it was carried out on a relatively small number of subjects. Secondly, only few patients had large EV, so we were not able to identify a cut-off value for large varices. Nevertheless, our preliminary data seem to suggest that VTTQ would work much better and more safely in the selection of cirrhotic patients for endoscopic screening rather than to identify high-risk large-sized varices. Finally, in this preliminary study, we have intentionally omitted to evaluate LS, but it would be worthy for further research to evaluate both LS and SS to validate the combined algorithm suggested by Bota et al. [
In conclusion, our study suggests that VTTQ-SS could represent an easy-to-perform, noninvasive, reproducible diagnostic tool which can be performed altogether with the routine echographic exam in cirrhotic patients in order to limit and/or delay the indications of endoscopic screening for EV among patients with HCV-related cirrhosis. Further longitudinal studies are needed to confirm our data.
The authors declare that there is no conflict of interests regarding the publication of this paper.