Fat malabsorption in cystic fibrosis patients

Dear Sir: We read with great interest the recent paper by Kalivianakis et al (1), who traced the gastrointestinal handling and postprandial partitioning of a labeled fatty acid to determine the extent to which the steatorrhea exhibited by patients with cystic fibrosis (CF) during habitual pancreatic enzyme replacement therapy (PERT) could be attributable to fat malabsorption. Malabsorption is a term widely used in clinical practice and is usually considered to be synonymous with increased stool losses. In clinical practice, there is often a failure to differentiate between dietary residue that is not absorbed as a result of poor digestion and thereby is not presented in a form that could be absorbed by the body (ie, maldigestion) from a failure in the absorptive capacity of the gastrointestinal tract (ie, malabsorption). We addressed this issue using 13C-labeled fatty acids and triacylglycerols in an attempt to improve the nutritional management of CF patients on the basis of our own reports of stool energy and lipid losses in CF patients during their habitual PERT (2, 3). Although we do not wish to comment on the study design, we do feel that the conclusions presented in Kalivianakis et al's study are not consistently supported by their own observations and warrant further examination. Kalivianakis et al concluded that the continuing fat malabsorption in CF patients receiving PERT was not due to insufficient lipolytic enzyme activity, but to incomplete intraluminal solubilization of long-chain fatty acids, reduced mucosal uptake, or both. Two lines of reasoning led to their conclusion. First, because no relation was observed between the recovery of label in breath after 1,3-distearyl,2[1-13C]octanoyl glycerol ingestion and gross lipid excretion in the stool, the authors concluded that increased stool lipid losses were probably not related to defective lipolysis (ie, maldigestion). However, in the absence of direct measurements of recovery of label in stool, the authors assumed that the sole determinant of differences in the recovery of label in breath was the extent of digestion within the gastrointestinal tract. In other words, once absorbed, the oxidation of the labeled fatty acid was uniform and consistent and not influenced by nutritional status and metabolic demands. Second, after ingestion of [13C]linoleic acid ([13C]LA), a strong relation was observed between the concentration of labeled fatty acid in the circulation 8 h later and gross lipid excretion in the stool. Justification that this finding differentiates between pancreatic insufficiency and deficient intestinal uptake is restricted to a single brief report (abstract) of lipid malabsorption studies conducted in rats (4). The assumption was that poor lipid absorption would be reflected in a delayed or lower concentration in the circulation and that factors that may influence the removal of labeled fatty acid from the circulation are unimportant in determining the concentration of fatty acid in the circulation. In the absence of control data, analysis was restricted to differences within the group of CF patients. We used a direct approach to determine the availability of dietary lipid on the basis of recovery of label in stool in patients with CF and compared the results with those from healthy volunteers to differentiate between maldigestion and malabsorption. Using this approach, we showed that the recovery of 13C label in the breath after oral administration of [1,1,1-13C]tripalmitin presented within an emulsion to patients with CF receiving habitual PERT was not related (inversely or otherwise) to the recovery of 13C label in the stool and did not reflect the extent to which labeled triacylglycerols are digested and absorbed (5). This may have been due, at least in part, to lower oxidation of the labeled fatty acids after absorption in CF patients than in control subjects. Furthermore, the recovery of label in stool after oral administration of nonesterified [1-13C]palmitic acid was paradoxically lower in patients with CF than in control subjects, implying that the availability of long-chain fatty acids after digestion was either the same as or possibly even greater in patients with CF (6). Therefore, we were surprised to find that despite measuring the excretion of 13C label in stool using gas chromatography–combustion isotope ratio mass spectrometry, the authors failed to take these observations into account when drawing their conclusions. As evident in their Table 2, stool [13C]LA expressed as a percentage of the administered dose was low and varied between 0.0% and 1.8%. In other words, because almost 100% of the labeled fatty acids were absorbed, there was no evidence of malabsorption. Note that such an apparently high availability of long-chain polyunsaturated fatty acids was also seen in healthy men (99–100% of administered dose) in the only other reported study in which the availability of [13C]LA was determined from stool losses (7). The absorption of [13C]LA in CF patients in Kalivianakis et al's study was also similar to what we observed for nonesterified [1-13C]palmitic and oleic acids (>98% of administered dose) in healthy volunteers (8). Our impression is that, having failed to show any association between [13C]LA excretion and total fat in stool, the authors overlooked the only direct measure of fatty acid availability. It is probably not surprising that no association was observed given that there was so little label in the stool yet stool lipid losses varied markedly. Therefore, although we do not dismiss the possibility that there may be problems associated with the intraluminal solubilization or reduced mucosal uptake of long-chain fatty acids, we feel that the authors could have reflected on this apparent dichotomy in their own observations.