Helena Lomba-Viana

A follow up model for patients with atrophic chronic gastritis and intestinal metaplasia

Aim: To devise a follow up model for patients with gastric cancer associated lesions, such as atrophic chronic gastritis (ACG) and intestinal metaplasia (IM). Methods: Cohort study of 144 patients, followed for a minimum of one year, in whom at least two upper gastrointestinal endoscopic biopsies in flat gastric mucosa provided a diagnosis of ACG, IM, or low grade dysplasia (LGD). Results: Of those diagnosed with ACG at first endoscopic biopsy (entry biopsy), 12% progressed to LGD in outcome biopsy, as did 8% of those with type I IM, 38% with type II or III IM, and 32% with LGD. Type of IM at entry independently predicted progression to LGD and cancer. Type II and III IM had a higher rate of progression to LGD than type I IM, which showed an indolent behaviour similar to ACG. Patients with type II or III IM were at higher risk for development of dysplasia, and 7% of patients with type III IM at first biopsy progressed to high grade dysplasia (HGD), whereas no cases of ACG or type I/II IM progressed to HGD during the first three years. Conclusion: Patients with ACG or IM could possibly be allocated to different management schedules, based on differences in rate and proportion of progression to LGD or HGD. Less intensive follow up (two/three yearly with “serological evaluation” (pepsinogen)) may suit those with ACG or type I IM. Patients with type III IM may benefit from six to 12 monthly improved endoscopic examination (magnification chromoendoscopy).

Serum pepsinogen test for early detection of gastric cancer in a European country.

Aim To estimate the adherence of Western individuals to serum pepsinogen (PG) test and its accuracy in the detection of gastric cancer followed by upper gastrointestinal endoscopy. Methods Individuals from the northern region of Portugal, aged between 40 and 79 years, were invited to participate in a blood collection for the determination of serum PG values by ELISA method (Biohit kits). Participants were classified into two groups: positive (PG I ⩽70 ng/ml and PG I/ PGII ⩽3) and negative (all others). All participants with a positive test and a consecutive random sample of participants with a negative test were subjected to endoscopy with biopsy. All the participants (positive or negative) subjected to endoscopy were followed up over 5 years. Results From a total of 13 118 participants, 5326 were men (41%) with a median age of 60 years, and 446 (3.4%) had a positive test. Of these, 274 (61%) were subjected to endoscopy. We observed six gastric cancers, five intestinal and one diffuse type, and three early gastric cancers, representing one cancer per approximately 2200 PG tests or one cancer per 74 positive tests. From these 240 participants with a negative test, three patients with gastric cancer were diagnosed during follow-up (an estimated negative predictive value of 99%). In this study, the PG test showed an estimated sensitivity, specificity, positive predictive value, and negative predictive value of 67, 47, 2, and 99%, respectively. Conclusion Inhabitants of this high-risk region showed good adherence rate to a gastric cancer detection program based on a PG test followed by upper gastrointestinal endoscopy implemented for the first time. Accuracy estimates were similar to those in Japanese reports, indicating that this methodology could also be used effectively in Western countries with high rates of gastric cancer. Further formal cost-effective studies are however needed.

Associated primary tumors in patients with gastric cancer.

Goal To determine the prevalence of associated primary tumors in patients with gastric cancer. Study Retrospective study of 2,668 patients with gastric cancer observed at our department between July 1974 and December 1999. Associated tumors were diagnosed using Warren and Gates criteria, and included tumors that were not considered to be a metastasis, invasion, or recurrence of gastric cancer. Results Of all, 3.4% (n = 78) had primary tumors other than gastric cancer, 27% of which were synchronous (n = 21) and 73%, metachronous (n = 57). The mean follow-up time was 4 years (range, 1–13 years), and the male-to-female ratio was 1:1. The median age at diagnosis of gastric cancer was 67 years (range, 37–84 years), 69 years for patients with synchronous tumors versus 60 years for those with metachronous (p = 0.050). For at least half the patients the median time interval to metachronous cancer was 3 years (range, 1–22 years). Seventy-eight percent (n = 61) had two cancers; most were colonic (19%), uterine and ovarian (16%), and breast tumors (13%). Seventeen percent (n = 13) had three tumors: colon (46%), breast (23%), and skin (23%). Four percent (n = 3) had four tumors. One case with seven tumors was also observed [colon, breast (two tumors), uterus, skin, and stomach (two tumors)]. No statistically significant differences were found between synchronous and metachronous with regard to sex, gastric cancer location, and staging (TNM). Sixty-three percent (n = 49) died while under observation. Conclusions We found associated tumors in 3.4% of patients with gastric cancer. The most frequent associated tumors were breast and colon cancer. Surveillance for these tumors would be appropriate, at least in first years, after diagnosis of gastric cancer.

Should we exclude individuals from endoscopy based exclusively on the absence of alarm symptoms

TO THE EDITOR: We have read with great interest the paper by Voutilainen et al. (1), which reports on the relation between dyspepsia, alarm symptoms and several endoscopic findings, including gastric cancer. As the authors note, patient’s history and physical examination alone may be unreliable in diagnosing the aetiology of dyspepsia (2). After the exclusion of individuals undergoing follow-up for several reasons, such as those with lesions associated with gastric cancer, chronic atrophic gastritis or dysplasia, they analysed the data of 3378 patients. In this set, considering cancer diagnosis on the basis of the presence of alarm symptoms, the validity was 67.0% (2281 being those individuals with no alarm symptoms and no cancer and those with cancer correctly identified by the alarm symptoms/3378), with a positive predictive value of 1.1% (12/1104) but a very high negative predictive value, 99.8% (2269/2274). These estimates, even taking into consideration the biases of selection referred to by the authors, allowed them to conclude that alarm symptoms are a major indication for endoscopy. But, as Voutilainen et al. did stress, these results are the opposite of those in other studies, where alarm symptoms did not predict significant endoscopic findings (3). In fact, we would be very cautious about excluding patients without alarm symptoms from endoscopic examination. Even in this study, 30% (5 in 17) of cancer cases did not have alarm symptoms (false-negatives); and most symptomatic gastric cancers had metastasized at the time of endoscopy, with a short survival thereafter. Furthermore, we should also consider the diagnosis of other gastric lesions defined in this paper as gastropathy. Endoscopic observation shows a very high interobserver variability for several kinds of gastropathy, which may include lesions such as atrophy and intestinal metaplasia. In this study, 23 endoscopists performed endoscopic examinations, and gastropathy was identified in 1311 individuals. Considering again the accuracy of symptoms in the diagnosis of gastric lesions as gastropathy or cancer, validity decreases to 55.9% (1890/3378). The negative predictive value is now 62.5% (856/2274), with 41% falsely evaluated has having no gastric lesions (n = 632). This in fact could have a major impact on considering alarm symptoms as the only indication for prompt endoscopy, as gastropathy, as defined endoscopically, could reveal histological lesions for which diagnosis or follow-up might be worthwhile. We are following 136 patients (median follow-up time 3 years) with gastric lesions as severe as atrophic chronic gastritis, such as atrophy, metaplasia and dysplasia. In all of these patients, magnification chromoendoscopy (4) and serologic evaluation with pepsinogen I and II (5) were performed, as well as clinical questioning on symptoms. Not all of our patients were taking medication at the time of their visit, and none of our patients presented with alarm symptoms. Dyspepsia was the most commonly reported complaint. In patients with lesions as severe as low-grade dysplasia (n = 23), only 11% complained of dyspepsia. In this small sample, the negative predictive value of dyspepsia for prediction of endoscopic lesions was 66%. On the other hand, even in such a selective population, we found validity measures for diagnosis with serum levels of PGI and PGII, especially the PGI/II ratio, similar to others, including those in population-based studies (5). Thus, we must stress that alarm symptoms should not be over considered as the sole indication for endoscopy in a general population, as they do not accurately correlate with cancer or other lesions defined as gastropathy. Furthermore, although excluded from this study, also in patients with lesions associated with cancer, symptoms did not correlate at all with progression to cancer, at least not in the early forms. Other non-invasive ways to exclude patients from endoscopy, such as serum pepsinogen I and II, must be developed and implemented in this setting and in other settings such as in young patients without alarm symptoms in whom endoscopic examination was until now neither cost-effective nor readily acceptable, as Voutilainen et al. have stressed.

Colorectal carcinoma (CCR) recurrence — analysis and inducing factors

Purpose: To determine the characteristics of CCR that affected the pos-surgical recurrence and the most relevant aspects of the colic recurrence.

MUCOSA-ASSOCIATED LYMPHOID-TISSUE LYMPHOMA. SHOULD WE PERFORM A COLONOSCOPY IN MULTIORGAN INVOLVEMENT?

Mucosa‐associated lymphoid tissue (MALT) lymphoma is the most frequent non‐Hodgkin lymphoma in the gastrointestinal tract, but colon involvement has only been reported in multiorgan lymphoma. We present a rare case of a woman with MALT involvement of eye conjunctiva, tonsils, stomach, duodenum and colon. In selected cases like this, with multiorgan involvement, we recommend performing colonoscopy, with biopsies for immunohistochemistry with CD10 and cyclin D1 for differential diagnosis with other entities with different prognosis, such as follicular lymphoma, and mantle cell lymphoma, respectively.