Statistics from Altmetric.com
Editor,—I read with interest the recent paper by Lankisch et al(Gut1998;42:551–4). The authors studied faecal elastase 1 and faecal chymotrypsin in 30 patients with chronic pancreatitis and concluded that the estimation of faecal elastase 1 is not helpful in diagnosing mild and moderate exocrine pancreatic insufficiency and that this enzyme is not superior to faecal chymotrypsin in the diagnosis of exocrine pancreatic insufficiency. As various studies have already been published on this subject, I believe that it is useful to report the results of these studies, mainly because they clearly show that Lankisch et al’s conclusions cannot be substantiated. Domínguez-Muñoz and colleagues,1 in 20 patients with chronic pancreatitis (six mild, four moderate and 10 severe), found that the sensitivity of faecal elastase 1 was 0%, 100% and 100%, and that of faecal chymotrypsin 0%, 75% and 70%, in the three groups, respectively. Glasbrenner and colleagues2 studied 63 patients with chronic pancreatitis (19 mild, 14 moderate and 30 severe) and found that faecal elastase 1 had a sensitivity of 47%, 79% and 100%, and faecal chymotrypsin a sensitivity of 32%, 36% and 63%, respectively. Löser et al,3 in 44 patients with chronic pancreatitis (eight mild, 14 moderate and 22 severe), found that faecal elastase 1 had a sensitivity of 63%, 100% and 100%, and faecal chymotrypsin a sensitivity of 25%, 50% and 86%, respectively. Finally, in 44 patients with chronic pancreatitis (nine mild, 13 moderate and 22 severe) we found4 a sensitivity of 22%, 77% and 100% for faecal elastase 1, and of 11%, 54% and 77% for faecal chymotrypsin. In the four studies combined, 171 patients with chronic pancreatitis were studied and the sensitivity of faecal elastase 1 was 38%, 87% and 100% in mild (n=42), moderate (n=45) and severe (n=84) chronic pancreatitis, respectively, whereas the sensitivity of faecal chymotrypsin was considerably lower, at 21%, 49% and 74% respectively. Based on these numbers, I believe that it can be clearly stated that faecal elastase 1 has a good sensitivity in patients with moderate and severe pancreatic insufficiency and that it is clearly superior to faecal chymotrypsin in the diagnosis of exocrine pancreatic insufficiency. Lankisch et al’s conclusion may have resulted from the small number of patients included in their study.
Editor,—We are grateful for the interest Professor Gullo has taken in our recently published paper. He disagrees with one of our conclusions, namely that the estimation of faecal elastase 1 is not distinctly superior to the usual faecal chymotrypsin estimation.
His disagreement is based on four different studies which he has combined for meta-analysis. We doubt whether these four studies may be used for a meta-analysis as faecal elastase 1 was studied using a variety of different methods. In Glasbrenner et al’s study1-1 the faecal elastase test was done in patients in whom the diagnosis of chronic pancreatitis was based on endoscopic retrograde cholangiopancreatography (ERCP), and in the study by Domínguez-Muñoz and colleagues1-2 the diagnosis was based on ERCP and computed tomography. Löser and colleagues1-3 tested, as we did, against the secretin-pancreozymin test, which is the gold standard for testing exocrine pancreatic function.1-4 1-5 The tests which Professor Gullo used are presently unknown as his study has yet to appear in print.
The upper limit of normal for a faecal chymotrypsin test is 3 U/g stool. Concentrations between 3 and 6 U/g stool usually raise a suspicion of exocrine pancreatic insufficiency. In our study, both concentrations were used. It would be interesting to know what the results of the other studies1-1-1-3 would have been if they had used the upper level of normalcy. However, even if a correct analysis was to undermine in a larger number of patients our modest conclusion that faecal elastase 1 is not distinctly superior to the faecal chymotrypsin estimation, the criticism does not concern the main message of our paper.
We concluded that faecal elastase 1 is particularly helpful in detecting severe exocrine pancreatic insufficiency. Several studies, including ours, have shown that the severe form may be detected with a sensitivity of between 73 and 100%.1-1-1-3 1-6-1-8 However, a similar conclusion can be reached about various other indirect pancreatic function tests.1-4 1-5
We concluded that the test was not helpful in the mild to moderate form where the sensitivity was 40% and 33%. This is in agreement with other studies scoring a sensitivity from 0 to 47%1-1 1-2 1-8for the mild form of chronic pancreatitis or exocrine pancreatic insufficiency. Only Löser and colleagues1-3 claim a sensitivity of 63% for mild and 100% for moderate cases. It should be noted that in these groups the incidence of patients with diabetes mellitus was 25% and 50%, which is unusually high and may indicate that these patients had more severe exocrine pancreatic insufficiency than believed. Furthermore, this group used bicarbonate output rather than bicarbonate concentration plus enzyme output for grading exocrine pancreatic insufficiency. In our experience, about a quarter of patients with abnormal bicarbonate concentration still have a normal bicarbonate output, and some of Löser et al’s patients may have belonged to a more severe group according to our grading system. If they had used our system, they probably would have obtained a lower sensitivity for faecal elastase 1 in their moderate group.
Additionally, the specificity of the test for differentiating steatorrhoea of pancreatic and non-pancreatic origin is not sufficient as most patients with non-pancreatic malabsorption had false positive results in one study.1-6 The test may also show falsely low results in patients with diarrhoea and requires lyophilisation of stool samples, but even then very low concentrations of elastase 1 can be found, yielding indecisive information in most cases.1-9
Thus, the two major problems for the clinician—that is, the diagnosis of mild exocrine pancreatic insufficiency/chronic pancreatitis and differentiating pancreatic steatorrhoea/diarrhoea from non-pancreatic steatorrhoea/diarrhoea, cannot be solved with this new, and certainly not cheap, faecal enzyme test, and that was the key message of our paper.
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.