Statistics from Altmetric.com
We congratulate the authors (Gut 2004;53:381–6) on gathering this large cohort of patients in this important area in familial adenomatous polyposis (FAP) but would like to express some reservations with regards to the study. Our first concern relates to the means of endoscopic assessment. Standard forward viewing endoscopy was used, whereas in clinical practice side viewing endoscopy is recommended as duodenal polyposis in FAP is more severe in the periampullary region1 and this is likely to be missed with standard endoscopy. This will therefore underestimate both adenoma staging and frequency. This matter is raised in their discussion where they describe side viewing endoscopy as unrealistic. We however feel this is unreasonable in an era where ERCP services are available in most hospitals, at least in the UK.
Furthermore, the need for appropriate endoscopy technique and biopsy protocols has been highlighted in a recent study2 which revealed understaging of duodenal disease when comparing biopsy specimens and resected specimens, in addition to the finding of invasive cancer in a number of specimens resected for “severe duodenal adenomatosis” (that is, Spigelman stage 3 with high grade dysplasia or stage 4). The need to operate before biopsy proven carcinoma is demonstrated by the high mortality rates from metastatic disease in those with duodenal carcinoma. Accurate staging and assessment for endoscopic or surgical intervention is, in our opinion, not possible by standard forward viewing endoscopy.
Our other concern relates to the quoted cancer incidence, which we feel must be biased. The cohort was not followed up from a young age, but rather the age range of the cohort at first endoscopy was 20–81 years. Both the authors’ own data and others have shown that there is an increased risk of stage 4 disease and cancer with increasing age. As such, those older patients in this group we would expect to be self selected and to have less severe disease. Those who were destined to develop severe duodenal disease or cancer may well have developed it prior to screening. Of note, the median age of those developing cancer was 52 years (range 26–58).
In addition, those 12 patients undergoing open duodenotomy and polypectomy are likely to be those with most advanced disease and a highest risk of malignant transformation, thus again biasing the likely natural incidence of duodenal carcinoma.
There are also few details with regard to medical intervention which may affect duodenal staging and disease progression. In the discussion, the authors mention that a few patients may have been on periodic sulindac but feel that this would have a negligible impact on their analysis. In the present era of selective COX-2 inhibition,3 our current clinical practice is to consider celecoxib treatment in those with stage 3 or 4 disease as well as those who have undergone surgical intervention. If this practice were followed using sulindac in the centres involved in this trial, then up to 91 (24%) patients in this cohort could have been exposed to non-steroidal anti-inflammatory drugs.
Conflict of interest: None declared.
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.