Meta-analysis of interstitial pneumonia in studies evaluating iodine-131-labeled lipiodol for hepatocellular carcinoma using exact likelihood approach.

PURPOSE

Iodine-131-labeled lipiodol is currently licensed for unresectable hepatocellular carcinoma with portal thrombosis. It is thought to be well tolerated. Cases of interstitial pneumonia have been reported, but their frequency (≈2%) has not been well estimated. Quantifying adverse drug event frequency requires an appropriate statistical approach because standard methods are biased.

METHODS

To estimate the frequency of interstitial pneumonia in patients with hepatocellular carcinoma receiving iodine-131-labeled lipiodol, we conducted a systematic review of English articles using MEDLINE and EMBASE. All types of articles were considered except case reports. Primary outcome measure was symptomatic interstitial pneumonia based on investigators' judgment. All pooled analyses were based on a random effects meta-analysis model using an exact likelihood approach based on the binomial within-study distribution.

RESULTS

Ten studies, including 142 patients, used low activity per dose, ranging from 0.3 to 1.1 GBq. No respiratory adverse event was noticed in these studies. Eighteen studies, including 542 patients, evaluated higher activity per dose, around 2.2 GBq; 24 cases of interstitial pneumonia were reported in these studies. Estimated frequency of interstitial pneumonia was 1.6% (95%CI, 0.4-6.4%) after one high dose and 4.1% (95%CI, 1.0-16.0%) after two or more high doses.

CONCLUSIONS

The frequency of interstitial pneumonia appears higher and more precise than previously estimated. The risk appears to be related to the number of injections and the dose level per injection. Generalized linear mixed models using the exact binomial within-study distribution initially described to summarize data on diagnostic evaluation could be relevant for drug-related adverse reaction frequency assessment.

Copyright © 2011 John Wiley & Sons, Ltd.