I
Volume 69 August 1976
617
Section of Radiology President Margaret Snelling FRCS
Meeting 17 October 1975
Short Papers Isch2emic and Evanescent Colitig
The etiology of ischemic colitis itself is far from certain. In the relatively small number of patients on whom selective arteriography is done, by K C Dewbury1 MB DMRD (King's College Hospital, occlusive vascular disease or low flow states are rarely seen. Arteriography has little or no part London SE5) to play in either the diagnosis or management of In 1971 Miller and others described a syndrome ischmmic colitis, which predominantly affects the similar to ischemic colitis. The presentation is territory of the inferior mesenteric artery. This is acute, usually with pain, followed by bloody in contrast to mesenteric ischaemia or infarction diarrhoea. Radiological abnormalities typical of in the superior mesenteric artery territory, in the acute stage of ischaemic colitis were reported, which occlusions or low flow state$ are demonand these with the signs and symptoms rapidly strated in the majority of patients. returned to normal. This syndrome affects young In the absence of more extensive arteriography, patients, usually in their 20s, and it was considered sufficiently well defined for introduction the pathology of both ischemic and evanescent colitis is unlikely to be fully resolved. In the of the descriptive name 'evanescent colitis'. meantime, since the age and severity are the only With the aim of distinguishing evanescent apparent differences between the two conditions, from ischmmic colitis, the findings in 3 recent there seems no justification for their separation. It would be more logical to regard evanescent cases of each are summarized: colitis as one end of the spectrum. The clinical presentation in all 6 cases was very similar, with a short history of abdominal pain REFERENCE Miller W T, De Poto D W, Scholl H W & Raffensperger E C and vomiting, followed by bloody diarrhoea. The (1971) Radiology 100, 71-78 differential diagnosis from intussusception is important (and may be difficult) as the patients with colitis do not benefit from laparotomy.
The ages of the 3 patients classified as ischemic colitis were 36, 71 and 88. One patient had an associated carcinoma. In all cases the initial barium enema at the acute stage showed spasm, mucosal cedema ('thumbprints') and mucosal ulceration. All progressed on follow-up barium enemas to stricture formation. The ages of the 3 patients classified as evanescent colitis were 21, 22 and 26. The findings were initially similar to those in ischmmic colitis but in all cases returned to normal within a month. 'Present address: Royal South Hants Hospital, Southampton
Plain Abdominal X-ray in Acute Colitis by Clive I Bartram MB FFR (Diagnostic Radiology Department, St Bartholomew's Hospital, London ECIA 7BE) Halls & Young (1964) suggested that changes in the fmcal residue and haustral pattern usefully predicted the extent of active mucosal lesions.
Assessment of the colon on the plain radiograph may be made from the feecal residue, lumen width, mucosal edge and haustral pattern,
618 Proc. roy. Soc. Med. Volume 69 August 1976 where shown by intracolonic air. Radiographs in 69 patients without colonic disease and in 22 prior to colectomy were examined. A normal profile was found to consist of an extensive feecal residue, often total or limited to the right side. Only one patient had none. Half had sufficient intracolonic air to outline part of the colon, usually in the left side. The mean width of the transverse colon was 5.5 cm, similar to a previous study (Hywel Jones & Chapman 1969). The mucosal edge was smooth with sharp haustral clefts. The patients with ulcerative colitis underwent colectomy due to a failure to respond to medical treatment. Most had severe and extensive disease. In 19 the proximal limit of fecal residue correctly indicated the extent of active mucosal lesions. Three had no significant intracolonic air. The mean luminal width was 5.8 cm. No active mucosal disease was found where the mucosal edge was smooth and the haustral clefts sharp. Where there was active disease but no ulceration, the mucosal edge was fuzzy or finely irregular with widened clefts or absent haustration. Course irregularity of the mucosal edge and absent haustration was associated with marked ulceration. Where extensive mucosal destruction was present 'mucosal islands' (Brooke & Sampson 1964) were seen. Nine patients had very severe disease with transmural inflammation. Four had mucosal islands pathologically which were not seen radiologically. Four had sealed localized perforations which also could not be detected radiologically. The most active area was also not visible in 2 patients due to the absence of intracolonic air. However, there was marked dilatation of relatively normal proximal colon. Three had mucosal changes indicating early acute dilatation but the colonic widths were not above the mean for this group (7.2 cm). - It is suggested that the feecal residue is a useful index to show the extent of disease. Where intracolonic air is present the mucosal state can be accurately assessed. However, severe mucosal -lesions may occur without any intracolonic air to indicate their presence on the plain radiograph. Also sealed perforations may be impossible to diagnose. Acute dilatation cannot be distinguished in its early stages from severe mucosal disease, but must be considered when there is evidence of the latter and of dilatation of the colon develop-
ing. REFERENCES Brooke B N & Sampson P A (1964) Lancet ii, 1272 Halls J & Young A C (1964) Proceedings of the Royal Society of Medicine 58, 859-860 Hywel Jones J & Chapman M (1969) Gut 10, 562-564
2
The Angiographic Diagnosis of Insulinoma by H Herlinger MD DMRD
(St Jameses University Hospital, Leeds, LS9 7TF) A firm clinical diagnosis of insulinoma may present difficulties even in the presence of high fasting levels of plasma insulin. Surgery without prior angiography has led to some 25 % of insulinomas being missed, mostly when situated in the head of the pancreas or in the uncinate process. The typical angiographic appearance of an insulinoma is that of gradual tumour opacification via an inconspicuous supplying artery, the opacification persisting through the parenchymal and venous phases (Fig 1). Within the tumour contrast medium slowly diffuses from small arterial channels into blood spaces where it is only gradually replaced by non-opacified blood; venous drainage channels are not seen in these circumstances. Undue abdominal compression during angiography was seen to interfere with contrast medium diffusion into the tumour, and should be avoided (Fig 2). More rarely the feeding artery terminates in a capillary network within the insulinoma; tumour opacification is then not so obvious and a venous drainage channel may draw attention to it (Fig 3). Photographic subtraction is most useful in the identification of such an insulinoma. Occasionally, an insulinoma is almost avascular, and this is due to degenerative changes, fibrosis or hyalinization. One malignant insulinoma was encountered. As described by Fujii et al. (1974), it showed poor opacification because of encasement and occlusion of arteries within the tumour. Secondaries in the liver showed abundant neovascularity. This paper is based on a series of 15 patients who had angiograms done in the investigation of their hypoglycemia. Insulinomas were present in 8 of the patients. - One insulinoma was missed (1 x 2 cm), almost certainly because of overlay by opacified gastric mucosa and the omission of gas distension of the stomach during angiography. There were two false positive diagnoses: A spleniculus was mistaken for an insulinoma and -so was a focal area of pancreatitis due to a perforating gastric ulcer (Fig 4). The appearance of the spleniculus could be compared with that of a small insulinoma in the same position in another patient (Fig 5). The spleniculus shows a draining vein which is wider and more distinct than would be found in an insulinoma. The smallest insulinoma in this series was 6 mm in diameter and could only be shown by means of
Volume 69 August 1976
617
Section of Radiology President Margaret Snelling FRCS
Meeting 17 October 1975
Short Papers Isch2emic and Evanescent Colitig
The etiology of ischemic colitis itself is far from certain. In the relatively small number of patients on whom selective arteriography is done, by K C Dewbury1 MB DMRD (King's College Hospital, occlusive vascular disease or low flow states are rarely seen. Arteriography has little or no part London SE5) to play in either the diagnosis or management of In 1971 Miller and others described a syndrome ischmmic colitis, which predominantly affects the similar to ischemic colitis. The presentation is territory of the inferior mesenteric artery. This is acute, usually with pain, followed by bloody in contrast to mesenteric ischaemia or infarction diarrhoea. Radiological abnormalities typical of in the superior mesenteric artery territory, in the acute stage of ischaemic colitis were reported, which occlusions or low flow state$ are demonand these with the signs and symptoms rapidly strated in the majority of patients. returned to normal. This syndrome affects young In the absence of more extensive arteriography, patients, usually in their 20s, and it was considered sufficiently well defined for introduction the pathology of both ischemic and evanescent colitis is unlikely to be fully resolved. In the of the descriptive name 'evanescent colitis'. meantime, since the age and severity are the only With the aim of distinguishing evanescent apparent differences between the two conditions, from ischmmic colitis, the findings in 3 recent there seems no justification for their separation. It would be more logical to regard evanescent cases of each are summarized: colitis as one end of the spectrum. The clinical presentation in all 6 cases was very similar, with a short history of abdominal pain REFERENCE Miller W T, De Poto D W, Scholl H W & Raffensperger E C and vomiting, followed by bloody diarrhoea. The (1971) Radiology 100, 71-78 differential diagnosis from intussusception is important (and may be difficult) as the patients with colitis do not benefit from laparotomy.
The ages of the 3 patients classified as ischemic colitis were 36, 71 and 88. One patient had an associated carcinoma. In all cases the initial barium enema at the acute stage showed spasm, mucosal cedema ('thumbprints') and mucosal ulceration. All progressed on follow-up barium enemas to stricture formation. The ages of the 3 patients classified as evanescent colitis were 21, 22 and 26. The findings were initially similar to those in ischmmic colitis but in all cases returned to normal within a month. 'Present address: Royal South Hants Hospital, Southampton
Plain Abdominal X-ray in Acute Colitis by Clive I Bartram MB FFR (Diagnostic Radiology Department, St Bartholomew's Hospital, London ECIA 7BE) Halls & Young (1964) suggested that changes in the fmcal residue and haustral pattern usefully predicted the extent of active mucosal lesions.
Assessment of the colon on the plain radiograph may be made from the feecal residue, lumen width, mucosal edge and haustral pattern,
618 Proc. roy. Soc. Med. Volume 69 August 1976 where shown by intracolonic air. Radiographs in 69 patients without colonic disease and in 22 prior to colectomy were examined. A normal profile was found to consist of an extensive feecal residue, often total or limited to the right side. Only one patient had none. Half had sufficient intracolonic air to outline part of the colon, usually in the left side. The mean width of the transverse colon was 5.5 cm, similar to a previous study (Hywel Jones & Chapman 1969). The mucosal edge was smooth with sharp haustral clefts. The patients with ulcerative colitis underwent colectomy due to a failure to respond to medical treatment. Most had severe and extensive disease. In 19 the proximal limit of fecal residue correctly indicated the extent of active mucosal lesions. Three had no significant intracolonic air. The mean luminal width was 5.8 cm. No active mucosal disease was found where the mucosal edge was smooth and the haustral clefts sharp. Where there was active disease but no ulceration, the mucosal edge was fuzzy or finely irregular with widened clefts or absent haustration. Course irregularity of the mucosal edge and absent haustration was associated with marked ulceration. Where extensive mucosal destruction was present 'mucosal islands' (Brooke & Sampson 1964) were seen. Nine patients had very severe disease with transmural inflammation. Four had mucosal islands pathologically which were not seen radiologically. Four had sealed localized perforations which also could not be detected radiologically. The most active area was also not visible in 2 patients due to the absence of intracolonic air. However, there was marked dilatation of relatively normal proximal colon. Three had mucosal changes indicating early acute dilatation but the colonic widths were not above the mean for this group (7.2 cm). - It is suggested that the feecal residue is a useful index to show the extent of disease. Where intracolonic air is present the mucosal state can be accurately assessed. However, severe mucosal -lesions may occur without any intracolonic air to indicate their presence on the plain radiograph. Also sealed perforations may be impossible to diagnose. Acute dilatation cannot be distinguished in its early stages from severe mucosal disease, but must be considered when there is evidence of the latter and of dilatation of the colon develop-
ing. REFERENCES Brooke B N & Sampson P A (1964) Lancet ii, 1272 Halls J & Young A C (1964) Proceedings of the Royal Society of Medicine 58, 859-860 Hywel Jones J & Chapman M (1969) Gut 10, 562-564
2
The Angiographic Diagnosis of Insulinoma by H Herlinger MD DMRD
(St Jameses University Hospital, Leeds, LS9 7TF) A firm clinical diagnosis of insulinoma may present difficulties even in the presence of high fasting levels of plasma insulin. Surgery without prior angiography has led to some 25 % of insulinomas being missed, mostly when situated in the head of the pancreas or in the uncinate process. The typical angiographic appearance of an insulinoma is that of gradual tumour opacification via an inconspicuous supplying artery, the opacification persisting through the parenchymal and venous phases (Fig 1). Within the tumour contrast medium slowly diffuses from small arterial channels into blood spaces where it is only gradually replaced by non-opacified blood; venous drainage channels are not seen in these circumstances. Undue abdominal compression during angiography was seen to interfere with contrast medium diffusion into the tumour, and should be avoided (Fig 2). More rarely the feeding artery terminates in a capillary network within the insulinoma; tumour opacification is then not so obvious and a venous drainage channel may draw attention to it (Fig 3). Photographic subtraction is most useful in the identification of such an insulinoma. Occasionally, an insulinoma is almost avascular, and this is due to degenerative changes, fibrosis or hyalinization. One malignant insulinoma was encountered. As described by Fujii et al. (1974), it showed poor opacification because of encasement and occlusion of arteries within the tumour. Secondaries in the liver showed abundant neovascularity. This paper is based on a series of 15 patients who had angiograms done in the investigation of their hypoglycemia. Insulinomas were present in 8 of the patients. - One insulinoma was missed (1 x 2 cm), almost certainly because of overlay by opacified gastric mucosa and the omission of gas distension of the stomach during angiography. There were two false positive diagnoses: A spleniculus was mistaken for an insulinoma and -so was a focal area of pancreatitis due to a perforating gastric ulcer (Fig 4). The appearance of the spleniculus could be compared with that of a small insulinoma in the same position in another patient (Fig 5). The spleniculus shows a draining vein which is wider and more distinct than would be found in an insulinoma. The smallest insulinoma in this series was 6 mm in diameter and could only be shown by means of
