Statistics from Altmetric.com
Editor,—We read the commentary by Lindor (
) with great interest and would like to raise the following points.
Lindor is not surprised that in our study1 patients with primary biliary cirrhosis (PBC) who initially had less abnormal liver function tests responded more favourably to ursodeoxycholic acid (UDCA) than those who had initially greater abnormal liver function values. We believe this is interesting as it is known that patients with lower abnormal liver function tests respond less favourably (for example, chronic autoimmune hepatitis to treatment with glucocorticoids) and that values do not decrease in a linear manner. Furthermore, it is well known that UDCA in PBC does not cause normalisation of liver function tests in most patients, and to date there has been no extensive examination of full and incomplete responders. Only in one study was this area addressed but few liver parameters were studied and there was only a short follow up period.2
Lindor states that our finding of no correlation between the percentage of UDCA in serum bile acids and biochemical response is different from other reports. However, he quotes only one study.3 Based on data from the literature,4 we reported in our paper1 that it is improbable that a further increase in bile acid concentrations in serum and a shift from the more hydrophobic to a more hydrophilic bile acid pool could be responsible for a complete response to UDCA therapy. Further results are awaited.
Lindor speculates that in our study the high percentage of patients with early stage PBC could have been an artefact because there was no correlation between histological stage at entry and biochemical response. We started UDCA therapy for PBC in 1978/79. In that time we have had 120 patients under constant supervision and over this period of 21 years only three patients have undergone liver transplantation and two have died as a result of late stage liver disease. Furthermore, our patients underwent regular liver biopsies and some even laparoscopy. That we have seen no more deaths or complications can only be explained by the fact that patients were in the early stages of the disease and that they were treated continuously with UDCA.
Lindor says that improvement in liver histology in our patients treated with UDCA (p<0.05) differs from the overall experience in other studies. However, in our study, we discriminated between incomplete and complete responders whereas in other trials complete and incomplete responders were evaluated together and compared with an untreated group.
In addition, Lindor is surprised that the histological progression reported in our series, even in incomplete responders, was slow. Based on modelling studies of untreated patients with PBC, he stated that substantially more patients developed histological progression. The difference between the studies cited by Lindor and ours is that we studied patients treated long term and not untreated patients, and it is well known that UDCA retards histological progression,5as recently shown using the Markow model.
Our description of how the histological grading was performed was not sparse; it was presented carefully and in accordance with other studies. It is correct that the histological data are mentioned in a single sentence and are not tabulated or otherwise presented. But having been a pathologist myself, I am rather sceptical towards liver histology in a patchy disease. For example, in 1994 it was shown6 that in a focal disease such as PBC, nine liver biopsies were needed from one session to warrant a definitive histological diagnosis. As this is not possible for ethical reasons, histological findings should not be over interpreted. Clinical data, development of complications, outcome, etc, are more relevant.
The most important objection of Lindor is the question of the relation between normalisation of liver function tests and clinically relevant findings. This is in contrast with a statement by Lindor himself (personal communication, November 9, 1999, 5th Annual Meeting, AASLD, Dallas, Texas) where he told us that in his incomplete responders the disease progressed in 38% of patients and in full responders in only 5%. We believe our results are comparable. In common with Lindor, in full responders we found progression of the disease in 4% and in 11% of incomplete responders; in incomplete responders progression occasionally took place not only from one stage to the next, but to the next but one stage. As patient numbers were small in our study, we did not give percentage values. Hence it is clear from our results the significance of normalisation of liver function tests.
The most important findings in our study were that: (1) UDCA improved cholestatic indices in incomplete and full responders in a strictly parallel manner; (2) in incomplete responders, the curves levelled off after about 3–5 years and did not normalise; and (3) cholestatic indices in patients with anicteric early stages of PBC allowed differentiation between responders and incomplete responders. This parallelism of the curves may indicate that UDCA influences mainly cholestasis and that other reactions are secondary. Therefore, more potent choleretic compounds or a combination of various choleretic substances could further improve results in incomplete responders. As stated previously, we are about to conclude such a study and the results seem to support our hypothesis.