Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
We read with great interest the article by Sifrim and colleagues (Gut 2005;54:449–54) addressing the issue of the importance of weakly acidic reflux (as measured by 24 hour ambulatory pressure and pH impedance monitoring) in patients with chronic cough. The presence of asthma, postnasal drip, or use of angiotensin converting enzyme inhibitors was ruled out so that an association with gastro-oesophageal reflux was probable. They found that only 15% of cough bursts were preceded by reflux episodes, which in 4% of cases were weakly acidic and therefore not detectable by conventional automated analysis of oesophageal pH tracings. Although the temporal relationship between acid or weakly acidic reflux and cough was highly significant, it could not be demonstrated in most episodes. In this respect their findings are in agreement with a previous study by Laukka and colleagues,1 who also used manometry for a more accurate timing of cough. Taken together, the results of the two studies suggest that the pathogenetic mechanisms usually proposed as a link between reflux and cough (micro/macroaspiration of refluxate into the airways and vagally mediated cough reflex) may be involved, at best, in a minority of cough episodes.
A different explanation has therefore to be formulated, not requiring a strict temporal relationship between reflux and cough. We suggest that such an explanation is represented by a low cough threshold induced by repeated reflux.
We have previously demonstrated that patients with reflux oesophagitis present a tussive response to minute amounts of inhaled capsaicin.2 In our series, the temporal relationship between irritant inhalation and cough bursts was so reproducible and immediate that a casual association with a single reflux was unlikely. Rather, our finding strongly suggests the presence of a reduction in cough threshold, related to the cumulated effect of repeated oesophageal acid exposures. These lead to nociceptor sensitisation, which has been reported to be reversible by proton pump inhibitor treatment.3 In fact, in our series of oesophagitis patients, only five days of omeprazole treatment produced a striking improvement in tussive reactivity, so that a dose of capsaicin 15-fold higher was required to elicit cough.4 Once this acid driven mechanism is activated, a variety of otherwise subliminal stimuli (smoke, pollution, etc) can cause cough.
For this reason, the temporal association of reflux with cough does not appear to be a reliable criterion to diagnose reflux associated cough, even using state of the art methodology. Future studies will clarify whether variations in cough threshold after proton pump inhibitor therapy has a role in identifying among patients with cough those in whom the symptom is due to reflux.5
We are thankful to Dr Benini and colleagues for their comments on our article (Gut 2005;54:449–54). We appreciate their reading of our manuscript, which led them to question the relevance of the temporal relationship between reflux and cough. However, we disagree with them with regard to their statement that “the temporal association of reflux with cough does not appear to be a reliable criterion to diagnose reflux associated cough, even using state of the art methodology”.
Our main finding was that there is a subgroup of patients with chronic cough that have a time association with weakly acidic reflux, either alone or together with acid. The majority of these subjects did not have an increased total number of weakly acidic reflux episodes but they had a positive symptom association probability (SAP), suggesting hypersensitivity to other aspects of the refluxate, either volume or other non-acidic gastric juice components.
In a recent study using the same techniques, we found that 25% of patients with persistent cough despite treatment with proton pump inhibitors (PPI) had a positive SAP for weakly acidic reflux.1 These data confirm that finding a temporal relation between reflux and cough is a reliable criterion for the diagnosis of reflux induced cough, although perhaps not the sole criterion.
It is true that in the remaining patients, the majority of cough episodes were not time associated to reflux. In some of these patients, cough may have had nothing to do with reflux and in others the mechanism for reflux cough might not have been related to a direct time association. We agree with Benini et al that a lower cough threshold induced by repeated reflux possibly features as an additional mechanism of reflux induced cough.
However, we do not think that determining variations in cough threshold after PPI therapy will enable us to identify those patients, for several reasons:
The hypothesis of a lower cough threshold due to repeated reflux is not only true for acid reflux but can also be applied to weakly acidic reflux. In this case, nociceptor sensitisation might not be reversible with PPI treatment which does not reduce significantly the number of postprandial reflux events with pH >4.2
It is well known that patients with reflux oesophagitis present a tussive response to minute amounts of inhaled capsaicin. However, exhibiting a lower threshold to inhaled capsaicin is not only limited to patients with reflux but is equally present in patients with asthmatic cough and idiopathic cough. Moreover, other authors in a larger series of patients3 have shown that cough sensitivity to capsaicin did not improve significantly in most patients with asthma or GORD, despite adequate medical treatment. Their conclusion was that the discriminative value of the capsaicin test to differentiate healthy subjects from patients with asthma or reflux was poor.
It is unlikely that variations in cough threshold after PPI therapy will be a reliable criterion to detect patients with cough in whom the symptom is due to reflux. It has recently been shown in an animal model of allergic airway inflammation that PPI can directly decrease antigen induced cough reflex hypersensitivity by a mechanism other than acid suppression.4 This suggests that amelioration of cough threshold with PPI treatment is not necessarily diagnostic for reflux.
Our study presented, for the first time, evidence of a temporal association between weakly acidic reflux and chough. Clinical studies and outcome data will ultimately define the clinical relevance of this time association.
Conflict of interest: None declared.
Conflict of interest: None declared.