Statistics from Altmetric.com
Editor,—We read with interest the paper by Christieet al (Gut 1997;41:513–17) discussing screening of patients for gastric cancer below the age of 55 in the context of an open access endoscopy service. Based on their data the authors suggest that gastric cancer is rare below the age of 55 and presents with suspicious symptoms and signs in the overwhelming majority of cases. Thus they advocate a lower age limit of 55 for screening “uncomplicated” dyspepsia.
There are however a number of fundamental flaws in the design and interpretation of this study. The data are retrospective and incomplete and based on the study of only 25 patients (the under 55 group) out of 319 with gastric cancer. There is no comparison of this group with the remaining 296 patients over 55 with regard to pattern of presentation or symptoms.
Perhaps more importantly there is no mention of the number of patients who actually presented for endoscopy or were picked up on the open access service. Surely the only way to set protocols for this service would be to analyse the data from it. In contrast this study seems merely to describe 25 patients below the age of 55 presenting in the Gloucester region, drawn from a pathology database, most of whom had advanced gastric cancers.
We would hope that better awareness of the importance of early referral and increased use of diagnostic endoscopic facilities should result in more patients of all ages presenting with early disease. Based on our experience1 and that of centres such as Leeds,2 we would advocate open access endoscopy in anyone with new dyspeptic symptoms over the age of 40.
Finally and perhaps most worryingly the authors state in their discussion that early detection does not necessarily mean improved survival. Those of us involved in the treatment of gastric cancer realise that the only significant factor that is going to improve survival from this disease is early detection and treatment. There is overwhelming and irrefutable evidence to support this. In the UK the detection and treatment of early gastric cancers has led to a five year survival rate of over 90% in these patients.3 In Japan, where mass radiological screening of the over 40s, ready access to endoscopy and population awareness of the disease has meant detection of early gastric cancer in more than half of all gastric cancer cases and again a five year survival of over 90% in these patients.4 Early detection does mean improved survival.
Until we have adequate prospective data from a large open access endoscopy unit we cannot agree with the interpretation and findings of this study and urge other centres to continue to endoscope symptomatic patients under the age of 55.
Editor,—Most of the arguments raised by Karat and colleagues were covered in our paper. However, because we consider that the epidemiological principles underlying our paper are important, we think it is worth reiterating them.
The data are retrospective but there are no comparable prospective data. Unlike other studies ours was of a defined population based on postcodes. The importance of this for determining presentation characteristics and natural history of disease cannot be overemphasised. Most other studies emanate from referral centres that receive selected patients. The resulting referral bias invariably influences the type of patient seen and interpretations made. Thus we believe our study is a better reflection of the real world.
We went to great lengths to ensure completeness (e.g. searching several databases) of patient capture. Our incidences are comparable to OPCS data and therefore we think it unlikely that many, if any, patients were missed. The small number (25 in seven years) of patients aged less than 55 just emphasises how rare the disease is in this age group.
Karat et al advocate better awareness of the importance of early referral to improve the chance of finding early disease. A point we stressed in our paper was that referral by general practitioners and subsequent investigation were not significantly delayed.
We did not determine the source of patients (clinic or open access) because general practitioners use clinic referrals rather than open access endoscopy for all sorts of reasons unrelated to symptoms. There would be too many confounders to make a meaningful exploration of differences in these groups. The point about the open access service in Gloucester is that it has been in operation for 20 years. The local doctors are relatively experienced in its use. The experience of its effects in the past 10 years in our district is relevant to experience in the next 10 years in other districts (the majority) that have introduced an open access service more recently.
In the UK there is no “overwhelming and irrefutable evidence” that early detection improves survival from gastric cancer. Only a randomised controlled trial can give this level of certainty and none has been done in the UK. The case series quoted from Leeds was not based on a geographically defined population and it is subject to the biases of all screening and early detection studies: lead and length bias.
We can sympathise with the sentiments of Karat et al: watching patients with gastric cancer die when intuitively you believe an earlier diagnosis would have helped is not easy. However, we would urge them to consider the following questions. Is there evidence that patients with curable gastric cancer are more likely to have uncomplicated dyspepsia than age and sex matched controls? If not, why investigate on the basis of symptoms? What opportunities are missed (e.g. early detection of colorectal cancer) if a huge amount of diagnostic resource is directed towards detecting the small number (<200 per annum) of gastric cancers in the under 55s. Could we use our limited financial resource in a more effective way?
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.