Proton Pump Inhibitor and Histamine–2 Receptor Antagonist Use and Risk of Liver Cancer in Two Population–Based Studies

K. T. Tran; Ú. C. McMenamin; B. Hicks; P. Murchie; A. P. Thrift; H. G. Coleman; L. Iversen; B. T. Johnston; A. J. Lee; C. R. Cardwell


Aliment Pharmacol Ther. 2018;48(1):55-64. 

In This Article


Primary Care Clinical Informatics Unit

Our nested case–control study in PCCIU included 434 cases of liver cancer and 2103 matched controls (Table 1). The median exposure period was 5.5 years (min 2.0, max 13.3) in cases and controls. Liver cancer cases were more likely than controls to smoke, consume high levels of alcohol, use aspirin, and have diabetes, liver diseases and peptic ulcer.

Overall, a greater proportion of liver cancer cases used PPIs compared with controls (33% vs 23%; Table 2). PPI use was associated with increased risk of liver cancer (unadjusted OR 1.74, 95% CI 1.38, 2.19; fully adjusted OR 1.80, 95% CI 1.34, 2.41). However, we found no evidence of a dose–response relationship with increased duration of exposure (1–11 PPI prescriptions fully adjusted OR = 1.92, 95% CI 1.33, 2.69 and 12 or more prescriptions OR = 1.66, 95% CI 1.13, 2.45). Associations were similar when exposure was based upon DDDs. There were stronger associations for omeprazole (fully adjusted OR 1.83, 95% CI 1.30, 2.56) than lansoprazole (fully adjusted OR 1.34, 95% CI 0.93, 1.93).

We found no association between H2RA use and risk of liver cancer (ever H2RA use: fully adjusted OR 1.21, 95% CI 0.84, 1.76), regardless of duration of use and type (Table 2).

In sensitivity analyses (Table 3), the association with PPI use was moderately attenuated after introducing 2 year (fully adjusted OR 1.65, 95% CI 1.15, 2.35) and 4 year lags (fully adjusted OR 1.60, 95% CI 0.97, 2.64). Associations were similar in other sensitivity analyses.

UK Biobank

Among 471 851 participants in the UK Biobank, we identified 182 liver cancer cases over a median follow-up of 5.6 years (range 1.0–8.6 years). Liver cancer cases were more likely than controls to be older, male, from deprived areas, smoke, consume alcohol more often, be overweight or obese, have diabetes, cirrhosis, hepatitis, and use statins and aspirin (Table 1).

Ever use of PPIs was associated with twofold increased risk of liver cancer (unadjusted HR 2.08, 95% CI 1.46, 2.96; fully adjusted HR 1.99, 95% CI 1.34, 2.94). The magnitude of the association with PPI use was greater for risk of IBDC (adjusted HR 3.12, 95% CI 1.72, 5.68) than for risk of HCC (adjusted HR 1.60, 95% CI 0.91, 2.83). The associations were similar by type of PPIs. A similar, though not statistically significant, association was observed between H2RAs and liver cancer risk (adjusted HR 1.70; 95% CI 0.82, 3.53) (Table 4).

Sensitivity analyses (Table 5) revealed the association between PPI use and liver cancer was slightly attenuated after introducing a 2 year lag (adjusted HR 1.79, 95% CI 1.15, 2.77) and more attenuated when a 4 year lag was introduced (adjusted HR 1.30, 95% CI 0.66, 2.55). This pattern was similar for HCC and IBDC.