HAROLD ELLIS D.M., M.Ch., F.R.C.S - Review of general surgery 1981 - Postgraduate Medical Journal
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Postgrad Med J: first published as 10.1136/pgmj.58.680.325 on 1 June 1982. Downloaded from http://pmj.bmj.com/ on October 27, 2021 by guest. Protected by
Postgraduate Medical Journal (June 1982) 58, 325-345
REVIEW ARTICLE
Review of general surgery 1981
HAROLD ELLIS
D.M., M.Ch., F.R.C.S.
Surgical Unit, Westminster Medical School, S. W.J
Introduction with a high-normal serum cholesterol. Important
This year's review has an undoubted intra-abdominal side effects were diarrhoea and significant hepato-
and malignant flavour. However, it does represent, toxicity in 3 % of the patients. The trial, carried out
once again, a survey of those publications which I in 10 treatment centres, cost nearly 11 million
have found of particular interest and which inevi- dollars! (Schoenfield and Lachine, 1981).
tably, therefore, reflect my own surgical hobbies. Better results can be achieved if ursodeoxycholic
Topics include the decline in the surgery of peptic acid is used. This does not seem to damage the liver
ulcer, the management of gastrointestinal haemor- nor cause diarrhoea and its efficiency is as good as
rhage, non-operative treatment of gallstones, in- that of chenodeoxycholic acid (Dowling, 1981).
teresting studies on large bowel cancer, attempts to Data on recurrence when treatment is discon-
preserve the traumatized spleen and efforts to tinued are scanty, but a preliminary report suggests
copyright.
rationalize and humanize the management of breast that as many as half the patients have new stones
cancer. We consider the hazards of smoking in within 2 years of discontinuing treatment (Ruppin
peripheral arterial disease, the prognosis of aortic and Dowling, 1981). Unfortunately, there is no
aneurysm, the prophylaxis of deep vein thrombosis escaping from the fact that the great majority of
and surgery on the kidney removed from the patient patients who require treatment for their gall stones
and put onto a work bench. Finally, a variety of the still have to be submitted to cholecystectomy (Smith
wide range of experimental studies are presented and Sherlock, 1981).
which are being carried out by surgeons in the labor- A note of warning is sounded by Irving (1981)
atory, the wards and the operating theatres, and who reports a female of 58 years treated over a
which demonstrate the constant quest of today's 5-year period for dissolution of stones using cheno-
surgeons to improve the quality of service they deoxycholic acid. Further pain led to cholecystec-
deliver to their patients. tomy at which three large calculi were found, to-
gether with a fundal polypoid carcinoma. He points
Gastroenterology out that carcinoma of the gall-bladder is associated
Dissolving gall stones with 1-2% of cholecystectomy specimens and ac-
In 1972, the first hope of a non-surgical treatment counts for 600 deaths annually in the United King-
was raised when the primary bile acid, chenodeoxy- dom. Stones are present in up to 98 % of these cases.
cholic acid, was shown to be capable of dissolving Infected bile may degrade bile salts into carcinogens;
cholesterol stones; 3 years later it was joined by moreover, bile is altered by chenodeoxycholic acid
ursodeoxycholic acid (Leading Article, 1981a). The with increase of the chenodeoxycholate and litho-
American National Cooperative Gallstone Study cholate conjugates which are implicated in experi-
has now published the results of its trial of the use of mental carcinogenesis.
chenodeoxycholic acid; 916 patients with radio- Motson (1981) has reviewed the related topic of
lucent stones were randomly allocated to high dose, the management of gall stones retained in the com-
low dose or placebo treatment. After 2 years there mon bile duct after cholecystectomy. There is no
was complete gall-stone dissolution in 13 5 % of the known solvent for pigment stones. Bile acids are
high dose group, 5 2 % of the low dose patients and not of value because of the prolonged period neces-
0-8 % of the placebo group. Complete dissolution sary for dissolution, but mono-octanoin (a medium
was most likely in patients who were thin (but this chain monoglyceride), has been advocated as a
might be related to the fact that dosage was not result of in vitro studies at the Mayo Clinic. Motson
weight-related), those with small stones and those points out that merely flushing out the duct with
0032-5473/82/0600-0325 $02.00 © 1982 The Fellowship of Postgraduate MedicinePostgrad Med J: first published as 10.1136/pgmj.58.680.325 on 1 June 1982. Downloaded from http://pmj.bmj.com/ on October 27, 2021 by guest. Protected by
326 H. Ellis
saline through the T-tube at a pressure of 30 cm of X-rays disclosed a gastric ulcer or suggested a
water may be successful in about 50% of cases. If neoplasm. Of the 100 patients undergoing routine
this simple treatment is not effective, 6 weeks later endoscopy, the cause of haemorrhage was duodenal
stone extraction can be attempted by means of a ulcer in 22, gastric ulcer in 18, oesophageal varices
steerable catheter introduced into the T-tube tract in 20, the Mallory-Weiss syndrome in 16, gastritis in
under X-ray control. If this fails, endoscopic 6, gastric cancer in 2, a normal examination in 4 and
sphincterotomy can be performed. no definite diagnosis established in 12. Interestingly
enough when the two groups were compared there
Gastrointestinal haemorrhage were no significant differences in overall hospital
Bleeding from the gastrointestinal tract remains a deaths (11 in the routine endoscopy group v. 8 in the
common cause of emergency admission to hospital. selected group), recurrence of bleeding (33 v. 32),
In spite of increased accuracy in diagnosis, as a numbers of transfusions required, duration of hos-
result of emergency endoscopy (which has virtually pital stay or re-admissions to hospital, incidence of
eliminated the need for less reliable urgent barium further haemorrhage or frequency of surgery. These
meal examinations), there remains a hard core authors conclude that endoscopy should not be a
mortality of between 5-10% in this country (Leading routine procedure in patients with upper gastro-
Article, 1981). A valuable report from Cardiff intestinal bleeding but should be used on a more
(Mayberry et al., 1981) studied a series of 583 selective basis (Peterson et al., 1981).
patients admitted between 1972 and 1978 to the The endoscope as a therapeutic instrument-to
University Hospital of Wales. There were 60 deaths control ulcer bleeding by argon laser coagulation-
(10-3 %). Most of the patients who died were elderly, has now been submitted to a controlled trial (Swain
with a mean age of 75 years, and no less than three- et al., 1981); 52 patients with a bleeding vessel in the
quarters of these had additional serious medical con- base of a peptic ulcer visualized at endoscopy were
ditions, for example malignancy, cardiovascular randomized between conservative treatment and
disease or respiratory problems. Of 565 of these coagulation (both groups receiving cimetidine).
patients available for detailed analysis, it is interest- Eight of 24 coagulated ulcers re-bled compared with
ing that the site of bleeding was identified in only 381 17 in 28 controls. Seven in the latter group died
copyright.
patients. In 184 the source of haemorrhage was not after re-bleeding but there were no deaths in the
established, but most of these had only a small treated group. This significant reduction in bleeding
haemorrhage which was not fully investigated. How- and mortality is most encouraging. However,
ever, no source of bleeding was identified in 17 Vallon and his colleagues (1981) present a random-
patients who died, and these were chiefly elderly. Of ized trial of endoscopic argon laser photocoagu-
the 381 patients with an identified source of haemor- lation in bleeding peptic ulcers; the series comprised
rhage, duodenal ulcer headed the list, with 154 28 patients with active bleeding ulcers and 108 who
patients. Gastric ulcer occurred in 98, carcinoma of had suffered recent haemorrhage. It was interesting
the stomach in 34 and oesophageal varices in 19. that there was overall no statistical difference in the
This low incidence of variceal haemorrhage is, of re-bleeds, need for surgery or mortality in those
course, in contrast to the much higher incidence patients treated by argon photocoagulation and
reported from many centres in the U.S.A. where those treated on routine lines.
alcoholic cirrhosis is a much commoner problem. Apart from the elderly patient suffering from some
Acute erosions were found in 27 of the patients, and other serious medical or surgical problem, there are
13 of these were localized in the oesophagus. Less two groups of patients with severe upper gastro-
common causes included the Mallory-Weiss syn- intestinal haemorrhage who are a particular cause of
drome (6 patients), two patients with the Zollinger- anxiety. The first is the seriously ill patient who
Ellison syndrome, and single examples included develops haemorrhage from stress ulceration, the
Henoch-Schonlein purpura, hereditary telangiec- second the cirrhotic with bleeding oesophageal
tasia, an antral haemangioma, a Meckel's diverti- varices.
culum and Hodgkin's disease of the stomach. Basso and his colleagues (1981) present an impor-
Eighty-four of the patients were submitted to emer- tant study of the prophylaxis of stress ulceration in
gency surgery (15%) and there were 11 deaths. high risk patients. They studied 168 cases in the in-
A recent American series, in contrast, contained tensive care unit with one or more of the following
no less than 26% of patients with alcoholic liver high risk factors for the development of stress ulcer-
disease (Peterson et al, 1981). This interesting study, ation:-neurosurgery, head injury, severe burns,
from a Veterans Administration Hospital in Dallas, toxic shock, multiple trauma, renal failure, hypo-
included 100 patients randomly assigned to routine tension and serious post-operative complications.
endoscopy and 106 where endoscopy was only These patients were randomized into 60 who received
carried out if recurrent bleeding occurred or if cimetidine 200 mg 6 hourly, among whom therePostgrad Med J: first published as 10.1136/pgmj.58.680.325 on 1 June 1982. Downloaded from http://pmj.bmj.com/ on October 27, 2021 by guest. Protected by
Review of general surgery 1981 327
were no examples of stress haemorrhage, 52 who 2 later deaths and 5 patients who re-bled. Among the
received an antacid (Maalox) 10 ml each hour by 20 shunt patients there were 7 deaths, no fewer than
mouth, among whom there was one haemorrhage, 8 examples of encephalopathy, of which 4 were
and 56 control patients, who received no specific severe, 4 re-bled and 2 died later. This study gives
treatment, among whom there were 8 haemorrhages. some idea of the very serious prognosis of these
The difference between the two treated groups and extremely poor risk patients.
the controls was statistically significant. All the Because of the simplicity of the procedure, there
patients who bled were subsequently treated with is no doubt that more and more surgeons are turn-
cimetidine. In 8 the haemorrhage ceased, although ing to injection sclerotherapy via the oesophagoscope.
one patient died subsequently, and the remaining In most cases the haemorrhage can be con-
case was treated by angiographic embolization. It is trolled satisfactorily even though the patient is likely
interesting that of a further 638 patients admitted to to die of the other consequences of his cirrhosis
the intensive care unit without the risk factors listed (Lewis, Chung and Allison, 1981; McDougall,
above, there were no examples of haemorrhage. Westaby and Williams, 1981).
This study certainly underlines the value of either Terblanche and his colleagues (1981) of Cape
alkali or cimetidine as prophylactic agents in the Town, who are among the pioneers of this technique,
high risk group. report control of haemorrhage in 95 % of 66 cases
Bleeding oesophageal varices are treated con- with no deaths from continued variceal bleeding.
servatively in the first instances but some 30-40 % of However death rate per hospital admission was no
these continue to bleed. Recent years have seen in- less than 28 %, reflecting the very poor prognosis of
creasing dissatisfaction with the results of portal- the majority of these cirrhotic bleeders. Yassin and
systemic shunting operations, which have a high Sherif (1981) report equally effective control of
mortality, may not control haemorrhage and have a haemorrhage in 20 cases with hepatic schistosomia-
high incidence of post-shunt encephalopathy. Sur- sis complicated by bleeding varices.
geons are turning more and more either to injection An ingenious new approach to the problem of
of the varices through the oesophagoscope or fibre- bleeding varices is presented by Taylor and Neilson
copyright.
optic endoscope, or oesophageal transection. John- (1981). It is known that the passage of an electric
ston (1981) reports from Belfast on a series of 170 current across a blood vessel precipitates a thrombus
patients treated by injection for acute bleeding at the site of the positive electrode. These authors
episodes. The haemorrhage was controlled in no less have used this principle to construct a system of
than 93 % of patients. However the overall hospital longitudinal flexible gold plated strip electrodes
mortality was 18% and, if patients with extra- around the oesophageal component of a Sengstaken
hepatic block are excluded, this rose to 25% for tamponade tube. This was used in 8 poor-risk
patients with cirrhosis. Moreover the veins gradually patients in whom the tube was used to control oeso-
recanalize so that there is the risk of further haemor- phageal haemorrhage. A current of 7 volts was
rhage. For this reason, the Belfast group have re- applied for 90 min and the tube kept inflated within
ported on the value of oesophageal transection using the oesophagus for a total of 24 hr. In all the
the stapling gun in poor-risk patients in whom any patients the haemorrhage was satisfactorily con-
form of shunt was contra-indicated (Graham et al., trolled, although 4 patients subsequently re-bled and
1981; Johnston, 1981). This operation was carried in 2 this proved fatal. Certainly in a condition as
out on 60 patients, 18 as emergencies with 6 deaths serious as this, a further development of this ingeni-
(33%). In the other 42 cases operation was per- ous device might be of value.
formed at least 48 hr after the last haemorrhage and Massive bleeding from the large bowel is fortun-
5 of these died, giving a 12% mortality for elective ately far less common and 75% of such cases can
operation in these relatively poor risk patients. usually be managed successfully by conservative
Among the 49 patients who survived to leave hos- means. When haemorrhage continues, however,
pital, there have been 7 late deaths. Thirty-three emergency surgery may be necessary. The common-
patients had their operations more than 2 years ago est cause is diverticular disease and right-sided
and two-thirds are still alive. colonic diverticula have a high incidence of haemor-
Osborne and Hobbs (1981) report an interesting rhagic complications (Roberts and Thomas, 1981).
comparison between two matched groups of patients The introduction of selective mesenteric angio-
with bleeding oesophageal varices, 20 of whom graphy has demonstrated an interesting group of
were treated by transection and 20 by mesocaval non-neoplastic lesions of intestinal blood vessels
shunt using an 18 mm Dacron graft between the variously termed angiodysplasia, vascular ectasia
superior mesenteric vein and the portal vein. Of the and arterio-venous malformations. These have a
20 transections using the stapling gun, there were 5 predilection for the right colon and occur most fre-
early deaths, 2 mild examples of encephalopathy, quently in elderly patients. The demonstration byPostgrad Med J: first published as 10.1136/pgmj.58.680.325 on 1 June 1982. Downloaded from http://pmj.bmj.com/ on October 27, 2021 by guest. Protected by
328 H. Ellis
angiography of a right-sided colonic bleeding site or enough this figure varied between the centres from
an angiodysplastic area allows segmental resection 2-4% to 66-9%! The change in pattern is certainly
to be performed instead of the previously advocated not confined to the United Kingdom. Fineberg and
'blind' sub-total colectomy. Max and his colleagues Pearlman (1981) note that over the past 20 years the
(1981) report that the colonoscope may be of value incidence and severity of peptic ulcer has declined in
in the diagnosis of these vascular malformations. In the U.S.A. and the mortality has dropped by two-
a series of 42 patients with angiodysplastic mal- thirds since its peak in 1962. In August 1977, cime-
formations of the gastrointestinal tract, 18 had tidine was approved by the FDA and there was a
normal colonoscopic appearances but in 8 of these large decline in the number of operations for duo-
the lesion was in the small intestine and out of reach denal ulcer in 1978. Admittedly there was a rise in
of the instrument. In 3 patients the examination was 1979, but the numbers were significantly below the
incomplete but in 14 the lesion was identified. These predicted trend following the figures from 1966 to
authors suggest that it may be possible, in the future, 1977.
to treat at least some of these bleeding lesions via Pyloric stenosis as a result of chronic duodenal
colonoscopic coagulation and thus avoid surgical ulceration is now becoming quite an uncommon
intervention in often elderly and ill patients. disease. Indeed, although it has long been a hobby
Other, rarer, causes of massive colonic haemor- of mine, I have only operated on one case this year.
rhage include carcinoma, ulcerative colitis, radiation Dunn, Thomas and Hunter (1981) confirm that very
proctitis and coagulation disorders due to chemo- conservative surgery is effective in such patients. In
therapy, immunosuppression or anticoagulation. 15 examples of duodenal ulcer with stenosis, surgi-
These rarely present diagnostic problems. cal management combined highly selective vago-
tomy with dilatation of the strictured pylorus. In two
Duodenal ulcer cases the stricture tore during dilatation and re-
Surgeons in this country will confess that they are quired an omental patch. The follow-up results were
operating far less frequently upon patients with excellent in 14 of the patients and the single recur-
duodenal ulcer than ever before in living memory. rence was treated successfully with cimetidine. Six
The reason is partly due to the introduction of of the patients had postoperative barium studies 1-3
copyright.
powerful new pharmacological agents of the hist- years later and in all there was good emptying.
amine-antagonist group, but also to an actual While on the subject of pyloric stenosis, an inter-
decline in the incidence of duodenal ulcer, quite un- esting case report by Keenan (1981) from Northern
related to treatment. Indeed, during the past 30 Ireland describes a male of 78 years diagnosed at
years the hospital admission rates have fallen pro- laparotomy as having a pyloric obstruction due to a
gressively for both perforated and non-perforated carcinoma of the pylorus with peritoneal seedlings
ulcer. Barker and his colleagues (1981) have studied and a gastro-jejunostomy was performed. Biopsy of
the figures obtained from the Hospital In-Patient one of the nodules showed caseating tuberculosis
Enquiry for the years 1963-73 with the main diag- which responded well to specific therapy. No such
nosis of perforated duodenal ulcer. This was used case has been reported in the indigenous population
as an indicator of ulcer incidence because, in com- of the United Kingdom for the last 20 years,
parison with rates for non-perforated ulcer, they are although examples have been recorded in India and
less liable to be influenced by changes in hospital Africa.
admission practises. Among men in the north and
west regions of England and Wales, for example, the Long-term results of highly selective vagotomy
rates fell from 22-7 per 100 000 in 1963 to 16-3 in Surgeons are cynics. Many of us wondered if the
1973. In the south and east regions, the rates fell by a magnificent early results reported for highly selec-
similar amount from 14-7 to 10-2. The rates among tive vagotomy in the treatment of duodenal ulcer
women were lower than among men and the regional were too true to last. Blackett and Johnston (1981)
variations were small; there was also little variation from Leeds, the very home of highly selective vago-
in incidence throughout the 11-year period. It is tomy in this country, now report the follow-up
interesting that for both men and women urban studies of 433 patients submitted to elective highly
rates are higher than rural ones. selective vagotomy between 1969 and 1980. In 233
Wyllie and his colleagues (1981) carried out a patients followed up for 5-12 years (12% of whom
survey of surgical practice in 6 centres in this coun- were lost to follow up) the incidence of recurrence
try, considering the number of patients submitted to was 10-7 %. One patient presented with a perforation,
surgery for duodenal ulcer 5 years before and 4 years 4 with haemorrhage and 30 with epigastric pain.
after the introduction of cimetidine. On average, Asymptomatic patients were not endoscoped and so
there had been a 39% reduction in the number of asymptomatic recurrence would have been missed.
patients submitted to vagotomy but interestingly Nine of these patients were treated by re-operationPostgrad Med J: first published as 10.1136/pgmj.58.680.325 on 1 June 1982. Downloaded from http://pmj.bmj.com/ on October 27, 2021 by guest. Protected by
Review of general surgery 1981 329
and the remainder were controlled with cimetidine. sented 48 hr or more after becoming strangulated.
When the 35 patients with recurrence were compared In developing countries, not surprisingly, strangu-
with patients without recurrence, no pre-operative lated inguinal hernias remains a very common cause
factors could be identified that might be used to of intestinal obstruction (Ajao, 1981). Of 273 obstruc-
predict relapse and this particularly applied to assess- tions admitted to the University Hospital in Benin,
ment of pre-operative acid output. Thus, contrary to Nigeria, Chiedozi (1981) found that 163 were stran-
some previous reports, no evidence was found that gulated hernias. No fewer than 55 of those contained
patients who are hypersecretors of acid should be gangrenous bowel.
treated by vagotomy combined with antrectomy. Although the vast majority of small bowel ob-
The only factor which was found to influence the structions are due to adhesions or strangulated
incidence of recurrent ulceration strongly was the hernia and the bulk of large bowel obstructions due
surgeon who performed the operation. From to cancer of the colon, diverticular disease or volvu-
University College, Storey and his colleagues (1981) lus, the surgeon must be prepared to come across an
report a follow-up of 93 patients treated by highly extraordinary variety of conditions which result in
selective vagotomy from 5-9 years previously. The mechanical intestinal obstruction. Cathcart and his
confirmed recurrence rate was no less than 16-1 %. colleagues (1981) review that interesting rarity of
In addition, there have been patients with transient mid-gut non-rotation and point out that it is often
recurrent ulceration and a group with persistent forgotten that this may occur in the adult patient.
dyspeptic symptoms but in whom no ulcer has been Berardi (1981) presents a collective review of the 109
demonstrated. Although the operation was com- papers dealing with paraduodenal hernias and points
mended for its lack of side effects, the high incidence out that there is general acceptance that these repre-
of recurrent or persistent symptoms led the authors sent anomalies of peritoneal development rather
to some reservations about its general application than herniation into the paraduodenal fossae. Vinard
in the treatment of chronic duodenal ulceration. and his colleagues (1981) report obstructions due to
intra-mural haematomas in the small intestine in
Intestinal obstruction patients on long-term anticoagulant therapy and
copyright.
Intestinal obstruction remains a common and Daniell (1981) reports a previously undescribed con-
dangerous surgical emergency. Approximately 1750 dition in which a patient's prolapsed colostomy
deaths from this cause are notified each year in itself contained 60 cm of gangrenous small intestine
England and Wales, of which 750 are due to strangu- between the two walls of the colonic prolapse.
lated external hernias. Recovery followed resection of the gangrenous small
The aetiology of common causes of intestinal intestine and of the necrotic colostomy. From West-
obstruction has changed quite dramatically in the minster Hospital we have recorded a female patient
active life of senior surgeons alive today. In the first of 43 who presented with a strangulated femoral
three decades of the century, strangulated hernias hernia, and then 14 days later developed intestinal
accounted for a very high percentage of the total obstruction once again, this time from a strangulated
cases but the incidence has dropped considerably in left obturator hernia. An empty obturator hernia sac
the Western world in more recent years, no doubt was present on the right side. This combination of
because of the considerable enthusiasm with which bilateral obturator hernia and a femoral hernia has
these hernias are subjected to elective repair. not been previously reported (Watkins, Ellis and
Adhesions, in contrast, have become more and Leach, 1981).
more common and this can be attributed, in turn, to Two rare examples of large bowel obstruction are
the enormous increase in the frequency with which reported-Michowitz and his colleagues (1981), of
abdominal surgery is now being performed (Ellis, endometriosis of the colon, and Anseline (1981),
1981a). In the 1930s, about half the cases of intes- who records the 57th published example of intestinal
tinal obstruction were due to strangulated hernia but obstruction due to a gall stone impacted in the colon.
a recent review of 405 patients with mechanical The unusual emergency of transverse colon volvulus
small intestinal obstruction in New York (Bizer is reviewed by Anderson and his colleagues (1981).
et al., 1981) showed that 74% were due to adhesions Intussusception is a particularly fascinating cause
and only 8 % to strangulated hernia. of obstruction, of course particularly associated
This is not to say that strangulated external hernia with paediatric abdominal emergencies. Raudkivi
is not still a serious problem. Andrews (1981), in a and Smith (1981) review a 16 year experience of 98
review of 195 examples of strangulated external paediatric cases in Auckland, New Zealand. A
hernia in one hospital in North West England over a barium enema was performed in 67 cases but reduc-
5-year period, notes a mortality in adult patients of tion was only effected in 13. Eighty-five laparotomies
11%, which is almost unchanged over the last 30 were carried out with only one death. It is inter-
years and he finds that even these days 50% pre- esting that gangrene was already present in 17 cases,Postgrad Med J: first published as 10.1136/pgmj.58.680.325 on 1 June 1982. Downloaded from http://pmj.bmj.com/ on October 27, 2021 by guest. Protected by
330 H. Ellis
demonstrating that even in modern times diagnosis as a routine for acute mechanical obstruction (Ellis,
is often late. Although in centres where there is an 1981b; Hofstetter, 1981).
enthusiastic radiologist, barium enema reduction is Even when the abdomen is opened, the surgeon
highly effective and safe, these authors review 1371 may have great difficulty in determining whether or
cases of barium enema attempted reduction in 11 not a strangulated loop of bowel is viable. There has
published series (including their own) and point out been much interest over the years, particularly in the
that there is only a 62% success rate overall. U.S.A., in aids to making this important differential
It should be remembered, of course, that intus- diagnosis. Marfuggi and Greenspan (1981) report
susception may also occur in adult patients. Nag- their experimental and clinical studies using fluores-
orney, Sarr and McIlrath (1981) review the 48 adult cein injected intravenously followed by viewing of
subjects treated with this condition at the Mayo the bowel in ultraviolet light. This technique has
Clinic between 1955 and 1978. Twenty-four involved been used from time to time over many years but
the small and 24 the large intestine. In adults there is these authors found it highly reliable in a rabbit
usually an underlying cause and in the colon this is experimental model and found that it was better
usually a malignant tumour. Indeed, 15 of the colonic than clinical judgement when used on 20 patients.
cases had an underlying malignant cause compared Shah and Andersen (1981) advocate the use of a
with only 7 in the small intestine group. These Doppler ultrasound and found, in dogs, that if flow
authors therefore advocate immediate resection of was consistently heard on the anti-mesenteric border,
adult large bowel intussusception without any the bowel remained viable after mesenteric venous
attempt at initial surgical reduction. In the case of occlusion. Moreover, the compromised intestine did
the small intestine, it suffices to reduce the intussus- not develop strictures during a 6 month follow-up
ception and to carry out a limited surgical resection period. These authors present four clinical cases
of the underlying cause unless a malignant lesion is where Doppler ultrasound was found to be useful in
clinically suspected or obvious at the time of lapar- supplementing the usual methods of determining
otomy. These same authors (Sarr, Nargorney and small bowel viability. Bulkley and his colleagues
McIlrath, 1981) also point out that intussusception (1981) report on 28 patients operated on for acute
may occur immediately postoperatively or following intestinal ischaemia and compared clinical judge-
copyright.
recent abdominal surgery and report a series of cases ment against Doppler ultrasound and fluorescein.
in adults. This complication was particularly liable The findings were checked against microscopic
to occur when a Miller Abbott tube had been em- examination of the resected bowel. Clinical judge-
ployed. They point out that, although well recog- ment was accurate in 89 % of cases and this was not
nized in paediatric cases, there have been few reports improved upon when the Doppler was employed.
of this postoperative emergency in adult patients. However the fluorescein test was always correct.
While on this subject, Foldes and Fontaine (1981) These authors point out that the surgeon tends to err
record a great rarity-a jejuno-ileal intussusception clinically on the side of resection.
due to a schwannoma. Surgeons in this country are rarely tempted to
Published mortality figures throughout the West- use ancillary methods and most of us would teach
ern world demonstrate that the very high mortality that, faced with bowel of questionable viability,
(in the region of 25 % of all cases) which was typical 'when in doubt, take it out'!
of publications in the 1920s and 1930s, has now been
reduced to the region of 10-15% (Ellis, 1981b). This Pseudo-obstruction
improvement has undoubtedly been due to a combin- Pseudo-obstruction can be defined as a condition
ation of improved anaesthesia, better knowledge of in which the physical and radiological findings are
fluid and electrolyte replacement, efficient blood identical to those associated with mechanical ob-
transfusion and the introduction of antibiotics. The struction and yet no organic cause is found either at
major factors influencing survival rate in an adverse contrast radiology, laparotomy or autopsy. The
manner are strangulation of the bowel with gang- pseudo-obstruction may be isolated to a single seg-
rene, delay in treatment with gross fluid and electro- ment of the intestine or be part of a more generalized
lyte disturbance, and extremes of age; mortality is process involving most or all of the gastrointestinal
especially high in infants and the elderly. tract. These cases may proceed to such gross colonic
Although hundreds of papers have been published distension that there may be caecal necrosis or
on the differential diagnosis between simple and perforation.
strangulated obstruction, attempts at such a differ- Careful study of the details of many patients pur-
ential diagnosis are little more than an academic ported to be examples of pseudo-obstruction of the
exercise. Indeed, most experienced surgeons will intestine reveal that there is, in fact, an underlying
point out the very real dangers of attempting such cause even though this is not a mechanical obstruc-
diagnostic accuracy and advocate early laparotomy tion. Golladay and Byrne (1981), in an extensivePostgrad Med J: first published as 10.1136/pgmj.58.680.325 on 1 June 1982. Downloaded from http://pmj.bmj.com/ on October 27, 2021 by guest. Protected by
Review of general surgery 1981 331
collective review, describe many such causes, which incidentally, polyposis coli has only been reported
must be excluded before a diagnosis of true idiopa- on three occasions in black South Africans. Multiple
thic pseudo-obstruction can be made. These include synchronous cancers were absent in the black
collagen disease, electrolyte disturbance (particu- patients compared with 6% in the whites and diver-
larly hypokalaemia), cathartic abuse, lead poison- ticula were absent in the resected specimens from
ing, congestive heart failure (where associated use black patients although present in 13% of resected
of diuretics and consequent hypokalaemia may colonic cancers in the white population. How far
aggravate oedema of the intestinal wall), sepsis, this is a genetical difference or related to diet is the
spinal trauma and drug induced obstruction (includ- subject for further study, although the diet of the
ing ganglion blocking agents, antidepressants, chlor- black African is low in meat and fat and consists
promazine and anti-Parkinsonian drugs). Batalis, mainly of bread and maize.
Muers and Royle (1981) report an example in a From China, Ming-Chai et al. (1981) report that
patient with myxoedema who responded to intrave- colorectal cancer is a common tumour. In some
nous triiodothyronine. They postulate an autonomic areas endemic for schistosomiasis the prevalence
neuropathy or deposition of mucopolysaccharides rate is 44 per 100000. These authors studied 60
in the bowel wall as possible causes. surgical specimens resected from patients with schis-
Hanks and his colleagues (1981) review 30 case tosomal large bowel disease but without malignant
histories of chronic primary intestinal pseudo- change and found that rather more than a third had
obstruction in 21 reports published over the last 20 mild to severe epithelial dysplasia, very much like
years. The age of onset ranged from 1-60 years with the malignant potential of ulcerative colitis.
12 male and 18 female cases. Oesophageal motility While on the subject of aetiology, we have
studies were carried out in 14 and revealed aperistal- reported two cases of transitional cell carcinoma of
sis of either the entire, or a large segment, of the the rectum in male homosexuals (Leach and Ellis,
oesophagus. Motility distal to the oesophagus was 1981). One of the patients was aged 39 years, the
not studied as extensively but two of three patients other 36 years. Both were anoreceptive and both
evaluated for gastric motility were normal and one had had gonococcal proctitis in the past. We were
copyright.
of six investigated for small bowel motility was able to find five other reported cases and postulate
normal as were two of the four patients evaluated that there might be a connection between chronic
for colonic motility. Pathological examination of all infection or inflammatory disease of the rectum and
or part of the gastrointestinal tract was carried out carcinoma in these patients. We looked for raised
in 26 patients and abnormalities were detected in antibodies to herpes simplex in our patients but they
less than half of these; these included hypertrophy, were not present, neither were inclusion bodies seen
villous atrophy and degeneration of ganglion cells. on electronmicroscopy in the second case.
Only 48% of cases demonstrated clinical improve- From the point of view of prognosis, clinicians
ment and eight patients ultimately died. Treatment have become used, over the years, to the value of the
has been disappointing, although resection or by- Dukes classification of large bowel cancers (A, con-
pass may lead to improvement in certain cases with fined to the mucosa, B invading the muscle wall and
well localized involvement. There are no specific C with lymph node involvement). From St Mark's
histological criteria for the diagnosis of this condi- Hospital, Talbot and his colleagues (1981) now in-
tion. A wide number of therapeutic agents have been troduce a new important dimension, which is
tried, all without long-term success. whether or not there is invasion of extramural veins
in the resected specimen. The prognosis under such
Large bowel circumstances is particularly poor, with only 15
Carcinoma of the large bowel is the second survivors out of 91 patients at 5 years. It was inter-
commonest cause of deaths from cancer in the esting that in the study of a large series of patients
United Kingdom and this reflects the situation only intramural venous invasion was found in
throughout the Western World. It is not surprising, Dukes A cases and such invasion was not correlated
therefore, that this topic should feature prominently with metastatic spread. The combination of the
in current surgical publications. An important study presence or absence of extramural venous invasion
by Segal and his colleagues (1981) from Barag- together with the classical Dukes classification un-
wanath Hospital, Johannesburg, stresses the interest- doubtedly refines prognostication for the patient's
ing difference in incidence between black and white future.
races. Although there is a 13 per 100 000 incidence From St Mark's Hospital, also, comes a further
of colorectal cancer among the white population, interesting paper on prognosis of carcinoma in
this falls to 0-8% among the blacks. Adenomatous ulcerative colitis (Ritchie, Hawley and Lennard-
polyps were found in only 8 % of resected specimens Jones, 1981). They compare 67 patients with carcin-
in black patients compared to 33% in whites and, oma complicating ulcerative colitis treated betweenPostgrad Med J: first published as 10.1136/pgmj.58.680.325 on 1 June 1982. Downloaded from http://pmj.bmj.com/ on October 27, 2021 by guest. Protected by
332 H. Ellis
1947 and 1980 with 4817 patients without colitis the control group of 67 patients compared with only
seen over the same period. Although there was a 7 in the perfusion group of 63 patients. Liver meta-
higher proportion of inoperable and high grade stases were present in 13 control and two perfusion
tumours in the colitic group, the prognosis was patients at time of death.
found to be very similar in patients with and without Invasion of adjacent structures at the time of
colitis. initial laparotomy was found adversely to effect
Undoubtedly the most important technical ad- prognosis by Wood and his colleagues (1981) who
vance in colorectal surgery in recent years is the consider that this should be employed, together with
EEA stapling instrument which enables anastomoses Dukes classification and the differentiation of the
to be made low down in the pelvis after rectal ex- tumour in staging of colorectal growths. However,
cision. There are now many publications reporting Kelley and his colleagues (1981) found that although
satisfactory results and undoubtedly patients who the operative mortality was higher in such cases, the
might otherwise have been submitted to abdomino- 5 year survival was not remarkably different from
perineal excision of the rectum can now have the uncomplicated cases. Blamey and his colleagues
anal sphincter preserved (Heald and Leicester, 1981; (1981) specifically considered ovarian involvement
Rothenberger and Goldberg, 1981). Whether or not in carcinoma of the colon and rectum. The 5 year
we shall see an epidemic of anastomotic recurrences survival of women without ovarian involvement
over the next few years due to over-enthusiasm at after curative resection was 72 % but this fell to 50%
preserving the sphincter at the expense of adequate when ovarian involvement was present and the
excision remains to be seen. patient was submitted to curative excision of the
There is undoubtedly a need to carry out careful lesion. Certainly radical surgery for advanced local
assessment of the stapling machine against standard carcinoma of the colon can be surprisingly reward-
hand-sewing of the anatomosis. Beart and Kelly ing (Ellis, 1981).
(1981) report a careful study from the Mayo Clinic
in which patients with carcinoma of the rectum at Recurrent and metastatic disease
least 5 cm from the dentate line were randomized What of the patients who develop local recurrences
copyright.
between a two layer suture anastomosis and a stap- or metastatic disease? The presence of liver meta-
ling anastomosis. There were 35 patients in each stases, of course, is virtually a death sentence.
group. The stapling machine was found to be Bengtsson et al. (1981) from Lund note that patients
quicker, especially in a technically difficult opera- with 25 % of the liver occupied by metastases lived
tion. Postoperative complications were the same, on average for only 6 months, this fell to 5 when
but use of the stapling machine was more likely to up to 75 % of the liver was involved and when the
be associated with rectal tears and anastomotic liver was still more extensively implicated the
defects. These authors estimate that the stapling average survival was only 3 months. The longest
machine will save some 12% of rectums which are survival of an untreated patient with liver metastases
at present removed. was 36 months. Recurrence at the anastomosis,
In an effort to improve prognosis, following again, is often extremely serious but curative
resection, we have been involved at Westminster as resection is still possible. Vassilopoulos and his
part of a multicentre trial using adjuvant razoxane colleagues (1981) operated on 30 patients for such
after colorectal cancer resection (Gilbert et al., 1981). recurrent disease, all referred from other institutions.
Of patients in Dukes group B and C randomized There was one postoperative death. Fifteen under-
between controls (49) and treated (47) cases the went a curative second resection with a median
recurrence rate in the first six months was 20 % and survival rate of 59 months and a 5-year survival of
28% respectively in the Dukes B and C controls 49%. Three of these patients are alive and well at
compared with 4% and 9% in the corresponding 96, 91 and 72 months.
razoxane treated patients. The adjuvant group had Pihl and his colleagues (1981) in Melbourne
received 125 mg razoxane twice daily for five con- review 1315 patients with large bowel cancer treated
secutive days every week indefinitely. Importantly, by Sir Edward Hughes by potentially curative resec-
the razoxane treated patients experienced no signi- tion between 1950 and 1978. Thirty-five (27 %) sub-
ficant toxicity apart from a readily reversible mild sequently presented with a recurrent tumour at the
leukopenia in 52 %, while gastrointestinal symptoms site of the anastomosis, 9 following resection of
necessitated stopping the drug in only 4 patients. colonic tumours and 26 resection of primary
Taylor (1981) records his interesting randomized tumours of the rectum. Fifteen of these patients
prospective clinical trial of adjuvant portal vein per- underwent further surgical resection, 14 with curative
fusion with 5-fluorouracil (5FU) by cannulation of intention and one by means of palliative resection in
the obliterated umbilical vein. At the latest follow the presence of liver metastases. The remaining 20
up, 23 patients had died with recurrent disease in had such widespread local or distant organ dissemi-Postgrad Med J: first published as 10.1136/pgmj.58.680.325 on 1 June 1982. Downloaded from http://pmj.bmj.com/ on October 27, 2021 by guest. Protected by
Review of general surgery 1981 333
nation that no further surgical treatment was con- patients submitted to splenectomy for congenital
sidered to be indicated. The 14 patients who had spherocytosis, one in 81 patients with idiopathic
attempted curative resection had a median cancer- thrombocytopenic purpura, 2 in 11 patients with
specific survival time of41 months whereas the median myelofibrosis and no fewer than 6 in 16 patients
survival time of the remainder was 8 5 months. with secondary thrombocytopenic purpura. The rise
Anastomotic recurrence must be differentiated in the platelet count after splenectomy might be
from the development of a second metachronous assumed to carry an increased risk of postoperative
tumour, which itself may be eminently resectable. venous thrombosis. However, MacPherson notes
Welch (1981), reviewing this situation, gives a 2-8% only two confirmed instances of portal thrombosis
incidence of metachronous growths in a review of the in 151 patients with an anatomically normal portal
patientsat the Massachusetts General Hospital. Of 63 bed but there were four early and four late throm-
patients with metachronous cancers, three developed boses in 92 patients with abnormal portal drainage.
a third, and one a fourth, metachronous tumour. However, even with radioactive scanning, he has
Of course, there is great interest in the role of been unable to show an increased risk of deep
carcinoembryonic antigen (CEA) estimations in the venous thrombosis.
detection of recurrent cancer of the colon at an early Recent years have seen increasing concern about
enough stage to make further resection possible. the risk of infection following splenectomy, particu-
Wanebo (1981) reviews four reported series of larly in children. King and Schumacker (1952), in a
patients from Colombus Ohio, Rothwell Park, now classical paper, reported meningitis in 4 infants,
Boston and the Sloane-Kettering Institute. Positive one of whom died, and an undiagnosed fatal pyrexia
explorations were reported in between 78 and 94% in a fifth, within 3 years of splenectomy for congeni-
of patients submitted to a 'second look' procedure tal spherocytosis. Other reports soon confirmed that
on account of a rise in the CEA level. Resectable severe infection and deaths in children due to ful-
disease was found in between 7 and 72% of these minating sepsis was indeed a hazard after splenec-
patients. The true effect on patient survival requires tomy for haematological disorders, particularly in
a carefully controlled prospective trial comparing in- younger children. Further reports then appeared
copyright.
terventions done as a result of a rise in CEAlevels with reporting overwhelming sepsis following splenec-
the results of treatment based on clinical follow up. tomy for trauma in children and then there were
reports of deaths in adults under the same circum-
Early detection stances. Of 36 published cases of post-splenectomy
Farrands, Griffiths and Britton (1981) have docu- sepsis, Dickerman (1981) reports that 16 have been in
mented an interesting experiment in population adults and 11 of these were fatal. The organism in
screening in Frome, Somerset. A group of 8925 these adult cases was usually the pneumococcus, and
people over the age of 40 years were invited to have one was due to Haemophilus influenzae. One case
Haemoccult testing for occult blood. Of these, 2439 developed 31 years and one other 25 years after
accepted. There were 39 false positive tests which splenectomy. The importance of the spleen in infec-
became negative after dietary restriction. Patients tion is due to its production of immunoglobulin
with piles and other anal conditions had these treated (IgM) antibodies against circulating bacterial anti-
and the test was repeated 6 weeks later. A total of 8 gens. The spleen is the sole source of tuftsin, a
adenomas and 4 cancers were detected and one false specific cell-bound leucophilic gamma globulin frac-
negative (a rectal carcinoma) has occurred so far. tion that is essential for maximal stimulation of
phagocytic activity of neutrophils. The spleen is
Splenectomy and its dangers also important in the regulation of both T and B
The commonest indication for splenectomy is lymphocytes (Sherman, 1981).
trauma (and this includes accidental injury to the The characteristic clinical picture of post-splen-
spleen in the course of an upper abdominal opera- ectomy sepsis is of sudden onset with nausea, vomit-
tion, for example vagotomy or hiatus hernia ing, headache and confusion leading to coma. The
repair). There are a wide variety of other indications infection is usually fulminant and the mortality rate
for the operation, particularly congenital sphero- exceeds 50 %. The infecting organism is the pneumo-
cytosis and idiopathic thrombocytopenic purpura. coccus in over 50 % of cases.
These days, more and more splenectomies are being Although the incidence of overwhelming infection
performed in staging laparotomies for Hodgkin's following splenectomy for trauma is apparently low,
disease. The procedure is a major one, often per- the consequences are devastating. A number of
formed on seriously ill patients. This is reflected in measures have therefore been investigated in an
the hospital mortality, which bears a direct relation effort to reduce or eliminate this complication,
to the gravity of the primary disease. Thus Mac- particularly in children. They include non-operative
Pherson (1981) reports one hospital death in 60 management, surgical repair of the injured spleen,Postgrad Med J: first published as 10.1136/pgmj.58.680.325 on 1 June 1982. Downloaded from http://pmj.bmj.com/ on October 27, 2021 by guest. Protected by
334 H. Ellis
partial splenectomy, autotransplantation of portions Francke and Neu (1981) advise that all splenec-
of the spleen, prophylactic pneumococcal vaccina- tomized patients should be vaccinated with pneumo-
tion and antibiotic prophylaxis. coccal vaccine and this should be given, if at all
Careful conservative treatment with radioisotope possible, in patients who are to undergo elective
scanning has been used in selected patients with splenectomy. Those patients with low initial levels
minor splenic injury. Sherman (1981) details the of antibodies to pneumococci or those with a high
various conservative operations, including simple risk of exposure, those on chemotherapy for lympho-
suture and omental patch of minor lacerations, the reticular or other malignant disorders, children
use of topical haemostatic agents, partial splenec- under 5 years of age, and possibly all patients in the
tomy and occasionally ligation of the splenic artery first 2 or 3 years after splenectomy should receive
together with suture of a capsular tear. Experimental prophylactic penicillin therapy (penicillin V or
support for the concept of preservation of at least a amoxicillin). Erythromycin can be used in patients
part of the spleen is given by Greco and Alvarez allergic to penicillin. Any 'flu-like' illness after
(1981). They performed up to an 80% splenectomy splenectomy should initiate the taking of cultures
in one-month-old rats and showed that this residuum and treatment with an anti-pneumococcal, anti-
protected the animals against pneumococcal bacter- Haemophilus influenzae antibiotic, since early treat-
aemia. ment may reduce morbidity and mortality pending
Oakes and Charters (1981) give an interesting culture results.
account of their attempts at conservative treatment Schwartz (1981) is more conservative in his
in 24 cases of splenic trauma aged between 6 and 71 advice-polyvalent pneumococcal vaccine (Pneumo-
years. Two of these were due to knife wounds, vax) and oral penicillin until puberty for children but
7 iatrogenic during surgery and 15 were due to blunt no active treatment for adults, where risk of infection
trauma. In four cases it was impossible to do any- is very low. He considers that there is need to compare
thing other than remove the spleen but in 20 a the rate of sepsis in adult patients undergoing splen-
repair was attempted. In one of these cases disrup- ectomy for haematological disease treated with
tion occurred on the sixth day and an emergency vaccine and antibiotics with that in those with
copyright.
splenectomy was carried out, but in the others suc- identical disease in whom the spleen is not re-
cess was obtained. These authors estimate that the moved.
risk of post-splenectomy sepsis is 0-58% after In a fascinating paper from Rochester, New York,
trauma compared with an 0-1% risk of a septic Patel and his colleagues (1981) give an account of
death in the general population. four examples of autotransplantation of slices of
Pachter, Hofstetter and Spencer (1981) review 27 traumatized spleen in four patients for preservation
patients with splenic trauma, 18 of whom were 15 of splenic function. In each, thinly sliced (3 mm)
years of age or older. Three required splenectomy segments of spleen, roughly 20 g, were placed in an
because of complete destruction or avulsion but in omental pouch. All patients survived without com-
the other 24 the spleen was rescued by debridement, plications. Postoperative studies showed that, at 4
partial splenectomy or suture repair with no un- weeks, Howell-Jolly bodies and target cells had
toward consequences. An interesting series of 68 disappeared from the blood film, platelet counts
consecutive splenic injuries in children up to the age returned to normal range and initially low IgM
of 16 is presented by King and his colleagues (1981). levels increased to normal. Scans at 8 weeks con-
Twenty-two (32%) required splenectomy, and the firmed the presence of functioning splenic tissue. In
only two deaths in the series occurred in this group subsequent discussion on the paper, the authors
as a result of associated cranial trauma. Parenchymal reported that a total of 10 patients had now under-
repair was performed on 16 occasions (24%) and gone such splenic implants, 7 at splenectomy for
non-operative treatment in 30 of the children (44 %). traumatic rupture, one in whom the spleen had been
One of the parenchymal repairs had recurrent removed during distal pancreatectomy, another in
haemorrhage which necessitated splenectomy. Other whom the spleen had been removed during staging
reports of a successful conservative approach have laparotomy for Hodgkin's disease after the patholo-
been published this year. Cooney (1981), reviewing gist had shown that the sections were free from
31 children, noted that 20 were managed conser- disease and another in whom the patient had suf-
vatively with one requiring later surgery, 5 had fered spontaneous rupture during the course of in-
splenic repair and only 6 needed immediate splenec- fectious mononucleosis. Obviously this procedure
tomy. Giuliano and Lim (1981) found that 33 splenic will require serious evaluation.*
injuries could be managed by conservative surgery *Editor-(A detailed review of the risks of splenectomy and
without complications in a series of 92 adult cases; conservative approaches to the management of splenic
the rest were submitted to immediate splenec- trauma has recently been published (Werbin, N. & Lodher,
tomy. K. (1982) Postgraduate Medical Journal, 58, 65)).Postgrad Med J: first published as 10.1136/pgmj.58.680.325 on 1 June 1982. Downloaded from http://pmj.bmj.com/ on October 27, 2021 by guest. Protected by
Review of general surgery 1981 335
Breast highly encouraging (Ellis, 1981). Witte (1981) con-
Publications on carcinoma of the breast continue siders that mastectomy is only required if the prim-
to proliferate, with particular interest in the conserva- ary tumour is large and states 'removal of a female
tive treatment of breast cancer, the role of adjuvant breast, a source of considerable physical and
chemotherapy and the value of hormone receptors in emotional trauma, is seldom necessary'. Martinez
prognosis. and Goffinet (1981), from Stanford University, give
For many decades, radical mastectomy was details of their iridium implantation technique which
regarded as the only form of treatment for operable gives a highly localized irradiation, with excellent
breast cancers-indeed, to suggest any other treat- local control. Lavigne and Minet (1981), from Liege
ment modality was considered sacrilegious. There is in Belgium note that the 5 year survival of 114
now excellent evidence that comparable results can patients submitted to local excision plus radio-
be achieved by more conservative measures involv- therapy is the same as following more aggressive and
ing far less mutilation of the patient. Turner and his more mutilating treatment.
colleagues (1981), for example, report 534 patients Many surgeons have considered that the proof
in Manchester randomized between radical and that conservative surgical treatment followed by
more conservative mastectomy. The trial took place radiotherapy is comparable to conventional treat-
between 1969 and 1976 and showed no difference in ment must require a controlled randomized study
total survival, local recurrence, distant metastases but there have been obvious difficulties in a trial of
or disease free period. In many centres, simple this nature, particularly in the randomization of
mastectomy is being combined with prosthetic im- patients into a radical surgery and conservative
plantation or reconstruction, either at the time of group. However, an important paper from Veronesi
mastectomy or later (Gant and Vasconez, 1981). and his colleagues (1981) from the National Cancer
Although the cosmetic result is often not particu- Institute in Milan now presents such a study. This
larly pleasing, many patients are satisfied by the was carried out from 1973 to 1980 and comprised
result. Some surgeons try to conserve the nipple and 701 patients with breast cancer measuring less than
areola in these procedures in order to improve the 2 cm in diameter and with no palpable axillary
copyright.
cosmetic results. However, Andersen, Gram and nodes, who were randomized between the classical
Pallesen (1981) studied 80 consecutive mastectomy radical mastectomy and local excision of the tumour
specimens and found that in no fewer than 35 with axillary dissection and radiotherapy. There
(43-8%) of these histological examination revealed were 349 patients in the radical and 352 in the con-
nipple or areolar involvement. Only six had obvious servative group. The two groups were comparable
clinical changes in this area. in age distribution, size and site of the primary tum-
Some surgeons have gone to the other extreme of our, menopausal status and frequency of axillary
advising local excision of the tumour alone. How- deposits. There were three local recurrences in the
ever, multi-centric lesions in clinical breast cancer radical mastectomy group and one in the conserva-
have been found in up to 41% of cases (depending tive group. Actuarial curves showed no difference
on the care which the pathologist takes in searching between the two groups in disease-free or overall
for small foci of cancer in distant parts of the breast) survival. The authors stress that the results cannot
(Leading Article, 1981). For example, Westman- be considered final, since the longest follow up period
Naeser and his colleagues (1981), in Sweden, is 7j years, but it appears unlikely that the two
reviewed 173 mastectomy specimens. Fifteen of these groups will show different results with respect to
showed macroscopic multifocal lesions, 26 demon- local recurrence after a longer follow up period since
strated microscopic invasive foci elsewhere in the most local or regional recurrences occur within 3
breast and 11 showed epithelial proliferation with years of treatment. Another unsolved problem is the
severe atypia or cancer, a total of 30 % of multifocal risk of late carcinogenesis induced by radiation
lesions. These were not related to the size of the therapy in the very long term.
tumour or the presence of axillary metastases.
However, there is now intense interest throughout Prognosis and adjuvant therapy
the Western world in the role of minimal surgery One of the most important thrusts in clinical
(local excision of the tumour itself) combined with research in breast cancer involves attempts at defin-
intensive radiotherapy. Published results from centres ing those patients whose prognosis is grave and in
in France, Canada and the U.S.A. give survival whom adjuvant therapy might be indicated, and'also
rates which are identical with those of radical mastec- to try and define those patients with disseminated
tomy, combined with a high degree of local control disease who are likely to respond to hormonal or
and excellent cosmetic and functional results. At cytotoxic therapy.
Westminster we have gone over to this technique The anomaly of the patient with an apparently
with enthusiasm and our immediate results are 'early' breast cancer who yet rapidly disseminatesYou can also read
