Interventional procedure overview of transvenous obliteration for gastric varices

Analysis

Study design issues: This systematic review and meta-analysis evaluated the efficacy and safety of BRTO for bleeding or at-risk gastric varices. The meta-analysis was done to Meta-analyses Of Observational Studies in Epidemiology (MOOSE) guidelines. All studies included were uncontrolled and all but 1 were retrospective. The efficacy outcomes were:

Meta-analyses were done to estimate the pooled rates using the DerSimonian-Laird random effect for the studies, where the model assumes heterogeneity between the studies.

Study population issues: Of the people whose aetiology was provided, 124 (13.7%) had underlying hepatitis B virus, 467 (51.4%) had underlying hepatitis C virus, 205 (22.6%) had cirrhosis attributable to alcohol, and 112 (12.3%) had cirrhosis of other aetiology. Most studies assessed both patients with acutely bleeding and at-risk gastric varices without sub-stratifying their results. Of the studies that did specifically assess people with at-risk varices without active bleeding, rates of clinical success, technical success, and oesophageal varices recurrence were similar to the overall pooled rates. Two people (0.2%) had BRTO primarily for hepatic encephalopathy, with an additional 18 treated for hepatic encephalopathy in addition to gastric varices. Of studies citing the number of people with underlying hepatocellular carcinoma at the time of BRTO, 271 (31.8%) people had underlying hepatocellular carcinoma. Among studies citing the Child-Pugh classification, 424 people (42.8%) were class A, 452 (45.6%) were class B, and 115 (11.6%) were class C.

Key efficacy findings

Key safety findings

Analysis

Study design issues: This systematic review and meta-analysis compared the efficacy and safety of BRTO versus TIPS for bleeding or at-risk gastric varices caused by portal hypertension. The meta-analysis was reported according to the Preferred Reporting Items for Systematic review and Meta-Analysis (PRISMA) statement. One study was an RCT (21 total patients) and 6 were cohort studies. Two studies were available only as conference abstracts. Risk of bias in all studies was assessed by the Newcastle-Ottawa quality assessment scale (6 of the studies were rated as 'high' quality). However, note that the Newcastle-Ottawa quality assessment scale is not designed to assess RCTs. The outcomes included mortality at 1 year, technical success, rebleeding, haemostasis, post-procedure complications, and hepatic encephalopathy. A definition for technical success was not reported.

Pooled ORs with CIs for the outcomes of interest were synthesised by meta‐analysis using a random‐effects model. Heterogeneity between the included studies was estimated using the inconsistency index (I²). Heterogeneity of less than 30%, 30% to 50%, 50% to 75%, and more than 75% was classified as low, moderate, substantial, and considerable heterogeneity, respectively.

Key efficacy findings

Key safety findings

Analysis

Study design issues: This systematic review and network meta-analysis compared the outcomes of several techniques (including BRTO) for secondary prophylaxis of gastric variceal bleeding. The meta-analysis was reported according to the PRISMA statement.

Only 1 study that used BRTO was included in the network meta-analysis – Luo, 2021 (included in this overview as study 4). Endoscopic cyanoacrylate injection was the common comparator across all RCTs. All comparisons were based on the results of 1 or 2 RCTs. Risk of bias in individual studies was assessed by using the Cochrane Risk of Bias assessment tool version 2.0. Studies were assessed as at low to moderate risk of bias. Due to the objective outcomes, the authors felt that single-blinded or unblinded trials were not at high risk of bias.

Outcomes were:

For pairwise comparisons, RR and 95% CIs were estimated. Heterogeneity was assessed using the I² statistic, with values greater than 50% suggesting substantial heterogeneity. Network meta-analysis was performed based on a random-effects consistency model following a multivariate meta-regression approach. Relative ranking of the interventions using P scores were calculated. P scores range between 0 when a treatment is certainly the worst and 1 when a treatment is certainly the best. The authors then classified the interventions based on P scores into the following 4 categories: P scores >0.75, 0.50 to 0.75, 0.25 to 0.50, and <0.25 corresponding to high, moderate, low, and very low mean probability of the intervention for being ranked the best, respectively. The Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach to rate the certainty of evidence of estimates derived from the network meta-analysis was used.

Study population issues: Viral hepatitis was the main underlying cause of cirrhosis (56.27%), followed by alcohol (24.19%). People with cirrhosis had Child-Pugh class A (34.22%), class B (44.09%), and Class C (21.69%). The location of gastric varices was GOV1 in 20.35%, GOV2 in 57.73%, and IGV1 in 21.92%.

Key efficacy findings

Key safety findings

Safety findings were not reported.

Analysis

Follow up issues: Five people (3 in the cyanoacrylate injection group and 2 in the BRTO group) had poor adherence to the treatment protocol. Three people could be contacted and the other 2 could not be contacted. It is not reported in which group the people who could not be contacted were in.

Study design issues: This RCT compared BRTO with cyanoacrylate injection for the secondary prophylaxis of gastric varices rebleeding after primary haemostasis. People in stratum 1 had treatment for 3 to 5 days. An emergency endoscopy was done within 24 hours, and cyanoacrylate injection was administered to control acute bleeding without obliterating all GVs. People in stratum 1 were then randomised on day 6 to either BRTO or endoscopic cyanoacrylate injection. People in stratum 2 were randomised after screening gastroscopy. The blinding of participants was done using sealed opaque envelopes. The primary outcome was gastric variceal rebleeding and all-cause rebleeding. Secondary outcomes included all-cause mortality and complications, including worsening of oesophageal varices.

Comparisons between the 2 groups were done using the student t-test for normally distributed continuous data and the chi-squared test or Fisher's exact test for categorical data. Gastric variceal rebleeding, all-cause rebleeding, and mortality were estimated using the Kaplan–Meier method and compared with the log-rank test. Primary analyses were conducted on an intention-to-treat basis. All tests were 2-tailed and p<0.05 was considered statistically significant.

Study population issues: There were no statistically significant differences between the baseline characteristics of the treatment groups. The aetiology of cirrhosis was hepatitis B infection in 39 people (61%), alcohol in 8 (13%), autoimmune liver disease in 5 (8%), and unspecified 'others' in 12 (19%). The mean Child-Pugh score for the BRTO group was 7.2 ± 2.2 (class B) and for the cyanoacrylate group was 7.4 ± 1.7 (class B). The mean MELD score was 10.0 ± 2.3 in the BRTO group and 10.7 ± 2.9 in the cyanoacrylate group. The location of gastric varices was GOV2 in 42 people (66%) and IGV 1 in 22 people (34%). A total of 42 people (66%) had concomitant oesophageal varices. No people were prescribed non-selective beta-blocker therapy.

Key efficacy findings

Key safety findings

Analysis

Follow up issues: Of 177 people who successfully completed BRTO, 151 people had the endoscopic or radiologic follow-up examinations.

Study design issues: This case series reported the outcomes of BRTO for bleeding gastric varices. People were identified retrospectively from medical records. Outcomes included:

Procedure-related complications were defined as any untoward adverse events during or immediately after BRTO, which required active treatment or prolonged hospitalisation.

Quantitative variables were compared using Student's t -test, and qualitative variables were compared using Fisher's exact test. Kaplan–Meier analyses were applied to examine the time to the first episode of re-bleeding, or time to death. The log–rank test was used to examine the variation of rebleeding episodes and survival rates. All p-values were 2-tailed, and p<0.05 was considered statistically significant. People who were lost to follow up were right-censored at the time of drop out. People who failed to BRTO were excluded for the analysis of the rebleeding and mortality rates.

Study population issues: The aetiology of cirrhosis was hepatitis B in 90 people (49%), hepatitis C in 23 (13%), chronic alcohol abuse in 50 (27%), others (including autoimmune hepatitis, primary biliary cirrhosis, cryptogenic) in 13 (7%), and combined aetiologies in 7 (4%). According to the Child-Pugh classification, 48 people were class A (26%), 103 people were class B (56%), and 32 people were class C (17%). The location of gastric varices was GOV1 in 42 people (23%), GOV2 in 80 people (44%), and IGV1 in 61 people (33%). The sizes of gastric varices were grade 1 in 18 people (10%), grade 2 in 40 (22%), grade 3 in 124 (68%), and undetermined in 1. A total of 131 people (72%) had concomitant oesophageal varices at the time of BRTO. Fifty people (27%) had hepatocellular carcinoma.

Key efficacy findings

Key safety findings

Analysis

Study design issues: This retrospective case series evaluated the long-term outcomes of BRTO for gastric varices. People were identified retrospectively from medical records. Outcomes included:

The cumulative survival rates and aggravation rates of oesophageal varices were determined using the Kaplan–Meier method. The Cox's proportional hazards model was used to estimate the significance of independent variables. p<0.05 was statistically significant.

Study population issues: There was no difference in baseline characteristics between bleeding and prophylactic cases except gender (male/female) (p=0.041). The aetiology was hepatitis B in 4 people (4%), hepatitis C in 54 (54%), alcohol in 24 (24%), and unspecified 'others' in 18 (18%). There were 39 people (39%) classified as Child-Pugh grade A, 48 (48%) as grade B, and 13 (13%) as grade C. Fifty people (50%) had grade F2 variceal size and 50 (50%) had grade F3. A total of 68 people (68%) had concomitant oesophageal varices at the time of BRTO. A total of 61 people (61%) had hepatocellular carcinoma.

Key efficacy findings

Key safety findings