Javed Iqbal Kazi ( Department of Pathology, Basic Medical Sciences Institute, Jinnah Postgraduate Medical Centre, Karachi. )
Syed Mahmood Alam ( Department of Pathology, Basic Medical Sciences Institute, Jinnah Postgraduate Medical Centre, Karachi. )
Abdul Moeed Kazi ( Department of Medicine, Jinnah Postgraduate Medical Centre, Karachi. )
Aftab Anwar ( G.I. Consultants, 5/15, Rimpa Plaza, M.A. Jinnah Road, Karachi. )
Ziauddin Shamsi ( G.I. Consultants, 5/15, Rimpa Plaza, M.A. Jinnah Road, Karachi. )
A correlation between endoscopic and histological diagnosis was studied in 240 patients undergoing upper gastrointestinal endoscopy for dyspeptic symptoms. The biopsies from one or all the three sites were taken from oesophagus (158), stomach (193) and duodenum (146) according to the presenting symptoms of the patients. There was a better correlation between endoscopic and histological diagnosis when a positive finding was seen by the endoscopists as compared to when the endoscopy revealed no abnormality. One out of six benign looking ulcers in oesophaçjus and 1 out of 9 benign looking gastric ulcers turned out on endoscopy, to be malignant on subsequent histology. Inflammation of oesophagus, stomach and duodenum were the most commonly missed lesions on endoscopy (JPMA 40: 281, 1990).
The use of flexible fiberoptic gastroscope is now a part of routine gastroenterological practice. Endoscopic examination is recommended with negative Barium x-rays, in patients with persistent dyspeptic symptoms. Barium x-rays may not be the final answer in lesions like reflux oesophagitis, bile gastritis, ulcers of oesophagus and stomach and mass lesions1. Any patient experiencing upper abdominal symptoms might reasonably be asked to have an endoscopic examination2. An appreciable number of endoscopic examinations specially done in patients with persistent non-ulcer dyspeptic symptoms are normal and many lesions may not be detected without histology. The purpose of this study was to find out the correlation between the endoscopic and histological diagnosis and to determine the prevalence of upper gastrointestinal lesions in dyspeptic patients in Karachi.
PATIENTS AND METHOD
Endoscopically directed biopsies were taken from 240 patients referred for dyspeptic symptoms. The endoscopie diagnoses were recorded at the time of endoscopy, and multiple biopsies were taken from oesophagus, stomach and duodenum in different combinations according to the presenting symptoms of the patients and the endoscopic findings. Out of 240 patients, oesophageal biopsies were taken in 158, gastric biopsies in 193 and duodenal biopsies in 146 cases. The endoscopic biopsies were flattened on a piece of filter paper and fixed in 10% buffered formalin. They were processed for routine paraffin embedding and trimming. Serial sections were examined in light microscope with the help of H&E, reticulin and PAS stains. The histological diagnosis was made without prior knowledge of endoscopic findings. The diagnostic criteria were used as described3,4.
Table-I shows the distribution of histological lesions. and their correlation with endoscopic diagnosis in 158 oesophageal biopsies. Fifty showed no significant changes, 33 reflux oesophagitis, 8 benign ulcers, 42 hyperplastic epitheliurn and 25 carcinomas. There was 100% correlation between the two diagnostic methods when carcinoma was diagnosed on endoscopy. One benign’ looking ulcer turned out to be malignant on subsequent histology. A single case of carcinoma in situ presented as inflammation on endoscopy. There was a poor correlation between the two diagnostic methods when no significant changes were seen on endoscopy.
Table-IT shows correlation in 193 gastric biopsies, which was good when positive diagnosis was made on endoscopy and poor in negative endoscopy. The lesion most frequently missed was chronic gastritis. One benign looking ulcer was histologically malignant and one large dirty ulcer suspected to be malignant was benign on histology. A distinct and rare clinico-pathological entity eosinophilic gastritis was only diagnosed on histological examination.
Table-III shows the correlation in 146 duodenal biopsies. The most commonlymissed lesion was duodenitis. A case with a polypoidal mass in the duodenum was diagnosed as schwannoma on histology. The nodularity of mucosa seen on endoscopy was found to be produced by brunner gland hyperplasia.
The oesophageal lesions revealed abetter correlation between the two methods with a positive endoscopy and a poor correlation with negative endoscopic findings. All 23 cases of carcinoma diagnosed on endoscopy were subsequently confirmed by histology. Inflammatory changes endoscopically labelled as oesophagitis could only be found in 32 cases. It was on subsequent histology only that of these 32 cases, 21 were classified as reflux oesophagitis, one case on histology turned out to be carcinoma in situ. The most commonly missed oesophageal lesion on endoscopy was reflux oesophagitits and hyperplastic epithelium both being intra-epithelial lesions. According to Pope5 a negative endoscopic finding did not rule out the possibility of presence of refiux oesophagitis. According to Morson and Dawson3 in the early stages of reflux oesophagitis the oesophageal mucosa may look normal on endoscopy. The present study reaffirmed that a few oesophageal lesions are likely to be missed on endoscopic examination alone. In the gastric lesions there existed no correlation between the two diagnostic methods specially when endoscopy was negative, the subsequent histology in most cases revealed gastritis with variable degrees of atrophy and activity classified according to Whitehead4. Thus because of the poor correlation the endoscopists should convey the endoscopic findings such as erythema, oedema, and exudate to the pathologist alongwith the biopsy rather than making a diagnosis on endoscopy alone. Similar observations have been reported by Morson and Dawson3. Histological examination of the ulcerative gastric lesions is the most valuable diagnostic method as apparently benign gastric ulcer may in fact be malignant which can only be confirmed by histology. In our series one out of nine cases of benign looking gastric ulbers turned out to be malignant which is comparable to the fincUng of Cohn -Jones2 where 10% of benign ulcers were diagnosed as malignant on subsequent histology. The diagnostic correlation in the duodenal lesions showed similar trends as in oesophagus and stomach. A poor correlation existed between the two methods when endoscopy was normal. Duodeñitis was the commonest lesion missed on endoscopy. Our findings of correlation in the two diagnostic methods in duodenal lesion was in sharp contrast to findings of Owen6, where, when endoscopy was normal, histology was also normal in 99% cases, and when duodenum was abnormal endoscopically, lesions could be confirmed only in 70% cases The study of Paoluzi et al. 7 revealed that endoscopic and histological findings were in agreement in 82% of patients but endoscopic judgement of inflammation was false positive in 17 of 54(35%) observations of histologically normal mucosa. The observation made in our study points to the importance of biopsy for the histological examination not only in ulcerative and mass lesion of upper gastrointestinal tract but also in cases with negative endoscopic findings. Our understanding of the prevalence of upper gastrointestinal mucosal lesions in dyspeptic patients has greatly improved by routine endoscopic and histological examination, as compared to when endoscopic examination alone was done without routine biopsies8.
The guidance given by Professor N.A. Jafarey and Dr. S.J. Zuberi is greatly acknowledged.
1. Gowen, G.F. Endoscopic biopsy for improved accuracy in upper gastrointestinal tract diagnosis. Arch. Surg., 1982; 117: 485.
2. Cohn-Jones, D.G. Endoscopy in the diagnosis of gastrointestinal disorders. Practitioner, 1983; 227: 607.
3. Morson, B.C. and Dawson, I.M.P. Gastrointestinal pathology. 2nd ed. Oxford, Blackwell, 1979, p. 21.
4. Whitehead, It Mucosal biopsy of the gastrointestinal tract. 2nd ed. Philadelphia, Saunders, 1979. p. 11.
5. Pope, C.E. Gastrointestinal disease pathophysiology, diagnosis, management. 3rd ed. Philadelphia, Saunders, 1983, p.449.
6. Owen, D.A. Pathology of the esophagus, stomach and duodenum. New York, Churchill Livingstone, 1984, p. 69.
7. Paoluzi, P., Pallone, F., Palazzesi, P., Marcheggiano, A. and Iannoni, C. Frequency and extent of bulbar duodenitis in duodenal ulcer, endoscopic and histological study. Endoscopy, 1982; 14: 193.
8. Zuberi, S. J. and Samad, F. Gastroscopy in upper gastrointestinah symptoms. JPMA., 1979; 29: 87.