Statistics from Altmetric.com
Editor,—In their very stimulating paper, Cavataioet al reported their experience concerning gastro-oesophageal reflux (GOR) associated with cows’ milk protein allergy (CMPA).1 In a series of 47 cases of GOR, they found a very high prevalence of CMPA: 14 cases (30%); they also stressed the possibility of recognising these cases from characteristic ‘phasic’ pH tracing.
We evaluated a series of 112 infants (65 boys and 47 girls; median age 3.7 months) with GOR confirmed by oesophageal 24 hour pH monitoring. Using the results of an elimination diet followed by an open challenge test, we found CMPA dependent GOR in 18 patients (16%), a rate significantly lower than that found by Cavataio et al, but similar to that of other authors.2Moreover, we failed to find any characteristic oesophageal pH-metric pattern distinguishing patients with GOR plus CMPA from patients with GOR only: one out of 18 cases with CMPA dependent GOR showed the oesophageal ‘phasic’ pH monitoring reported by Cavataioet al, and this pattern was also present in three out of 67 cases with GOR only.
It is difficult to explain the difference in these two sets of results. The increase in pH subsequent to feeding found by Cavataioet al (‘phasic’ tracing) may have been an non-specific event, evident for some reason only in those patients in whom CMPA was diagnosed. Besides, such a high prevalence of CMPA in patients with GOR is truly surprising and has never been found by us or other authors in Southern Italy,2 or by others in Europe.3 Staiano et al showed that in cases with CMPA dependent GOR, there is a significant increase in intestinal permeability2 that suggests an involvement of the entire intestinal tract. As a matter of fact, 10/18 of our cases with CMPA dependent GOR showed failure to thrive: a significantly higher number, than in cases with GOR only (55% v14/67=20%; p<0.05). Cavataio et al found difference neither in clinical symptoms, nor in the prevalence of failure to thrive between patients with GOR and/or CMPA, and this can cast doubts on whether CMPA was correctly diagnosed. We think that other studies are necessary to further evaluate the prevalence of CMPA in infants with GOR and to investigate the possibility of recognising these cases by a characteristic oesophageal pH monitoring pattern.
Drs Cavataio, Iacono, and Carroccio comment:
Milocco and colleagues base their comments on the results recorded in a group of infants that is not comparable with the one we studied. It is relevant to underline that all our CMPA diagnoses had been made after an elimination diet and double blind cows’ milk challenge; furthermore, in 31/39 patients with clear positive clinical reaction to cows’ milk challenge, we documented intestinal mucosa damage which did not exist 24 hours before the challenge. As we followed rigorous criteria in making the diagnosis of CMPA and as these criteria are the only ones accepted in the literature,1-4 we are surprised that there are doubts about this diagnostic procedure. On the contrary, Milocco and colleagues have based their CMPA diagnosis on a single open challenge and this creates great uncertainty as to the correctness of the diagnosis.
Moreover, the age of the infants studied by Milocco and colleagues is not specified, but on the single datum we have (median age), it does not seem comparable with that of the patients we studied, in fact, 50% of their infants were <3.7 months old and we think that in such young infants it is very probable that there are many cases of ‘physiological GOR’, without any clinical relevance1-5and, obviously, without any relation to CMPA.
They reported a low number of ‘phasic’ pH-metric tracings (4/85, but what about the remaining 27 patients with GOR? Is 112 really the total number?), but we think that they examined a majority of normal infants, probably with excessive physiological GOR and without GOR disease (did any of these infants undergo oesophageal endoscopy?).
Finally, it is noteworthy that another Italian group carried out a prospective study on GOR-CMPA association and found that 13/18 patients with GOR+CMPA and only 3/37 with GOR had a typical phasic pattern of the pH-monitoring tracing.1-6