2018 Section 6 - Laryngology, Voice Disorders, and Bronchoesophalogy

S. E. Attwood et al.

cally signi fi cant changes in haemoglobin, iron, ferritin, total iron-binding capacity or vitamin B 12 levels. The results for vitamin B 12 are of particular relevance because some studies have described vitamin B 12 de fi - ciency emerging over time in patients receiving PPI treatment, allegedly as a result of impaired intrinsic factor secretion and/or its malabsorption in a non- acidic environment. However, published studies have shown inconsistent results. 3 The divergent results observed when assessing vitamin B 12 levels in relation to PPIs are probably a consequence of small study sizes and/or a lack of adequate controls. Homocysteine levels may act as an early marker of vitamin B 12 de fi - ciency; hence, we measured these levels throughout the LOTUS study – no effects connected with acid sup- pression were observed. Concerns have also been raised in the literature about an increased incidence of osteoporotic fractures during PPI therapy. Some published studies have reported a small-to-moderate increase in the risk of osteoporotic fractures (hip, wrist and spine) associated with PPI use, and others have reported no association. No consistent pattern between duration of PPI therapy and fracture risk has been shown. 1, 2, 21 – 25 There are no human data from randomised controlled clinical trials indicating an effect of PPIs on bone metabolism, or on bone fracture risk. The SOPRAN and LOTUS studies were not designed to evaluate such parameters. Nevertheless, no clinically signi fi cant changes in ionised calcium, vitamin D or ALP levels were found, and only isolated fractures from different locations were reported, which were evenly distributed across the PPI and ARS treatment groups. Gastric acid acts as a barrier to infection of the upper gastrointestinal tract. Accordingly, some concern has been raised that long-term hypochlorhydria may lead both to enteric infections and to respiratory infections caused by microaspiration. Enteric infections including Campylobacter and Salmonella , which are acid-sensitive, have been found in patients taking PPIs. 26, 27 It has been suggested that C. dif fi cile might be more common in hospitalised patients taking PPI therapy. 28 The risks seem greater when there are multiple comorbidities or when patients have recently been prescribed antibiotics. Observations from the SOPRAN and LOTUS studies did not indicate any difference in the incidence of enteric infections between the treatment groups, suggest- ing that the magnitude of the problem has to be better de fi ned in well-controlled post-marketing surveillance studies.

and LOTUS studies, and concluded that ‘ Based on every- thing now known at the agency, the reported difference in the frequency of heart attacks and other heart-related problems seen in the earlier analyses of the two small long-term studies does not indicate the presence of a true effect ’ . The fi nal statement in this review reads: ‘ FDA ’ s Safety Reviews of Prilosec and Nexium Find No Evidence of Increased Rates of Cardiac Events ’ (FDA web site 13 ). An issue which has attracted considerable interest is the interaction between PPIs and the anti-platelet agent clopidogrel. 1, 4 No signal of concern emerged in our study, but no conclusions can be drawn owing to the small number of patients actually exposed. However, data from recent randomised studies on the co-prescrip- tion of PPIs with clopidogrel question whether any such interaction is clinically relevant 14, 15 and this matter has recently been comprehensively reviewed. 5 An early signi fi cant concern of PPI therapy raised as early as the 1980s is related to the rise in serum gastrin associated with reduced gastric acid secretion and the consequent increase in carcinoid tumour development in female rats receiving extremely high doses of omep- razole during an entire life time. 16 However, our study and many other studies of PPI use in humans over more than 20 years have failed to con fi rm any PPI- induced gastric carcinoids of enterochromaf fi n-like (ECL) cells. 17 An important observation in our study is that the increase in gastrin in both the SOPRAN and LOTUS studies was seen during the fi rst year, after which it stabilised. It is well established that gastrin has a trophic effect on ECL cells. It has also been reported that chromogranin can serve as a marker for the prolif- eration of ECL cells. In the LOTUS study, the increases in gastrin levels were closely paralleled by an increase in chromogranin A. The increase in chromogranin A levels does not, however, raise any safety concerns. 6 The lack of clinically signi fi cant ECL cell proliferation in man may re fl ect that gastrin levels observed are at least 10- fold lower than those reported in female rats and there is also a species difference in terms of the susceptibility of the target endocrine cells. 6, 16 Other concerns raised in relation to the safety of PPIs include their possible interference with vitamin B 12 , folate or iron absorption, 18 – 20 with subsequent development of macro- or microcytic anaemia. In the SOPRAN study, folate levels showed only marginal changes over time with no associated clinical effects. Available data from the studies also showed no clini-

Aliment Pharmacol Ther 2015; 41: 1162 – 1174 ª 2015 John Wiley & Sons Ltd

185