Core Topics in General & Emergency Surgery: Companion to Specialist Surgical Practice (49 page)
Read Core Topics in General & Emergency Surgery: Companion to Specialist Surgical Practice Online
Authors: Simon Paterson-Brown MBBS MPhil MS FRCS
Acute colonic pseudo-obstruction (ACPO) is the term used to describe the syndrome in which patients present with symptoms and signs of large-bowel obstruction but in whom no mechanical cause can be demonstrated at contrast radiology. In more than 80% of patients with ACPO, an underlying precipitating condition exists, of which at least 50 have been described.
27
The most common of these associated conditions are metabolic disorders, trauma and cardiorespiratory disease (
Box 10.2
). The term ‘Ogilvie's syndrome’ is loosely used in the literature as a synonym for ACPO, although it was actually used first to describe the pseudo-obstruction associated with retroperitoneal malignant infiltration. The mortality rate of ACPO is high, partly as a result of the underlying disorders but also related to failure to recognise the condition, leading to inappropriate operation. The true incidence is hard to ascertain since a number of unrecognised cases are likely to resolve spontaneously and diagnostic criteria are variable. However, it has been estimated that some 200 deaths per annum in the UK may result from ACPO.
28
Box 10.2
Predisposing conditions in acute colonic pseudo-obstruction
Chest infection
Myocardial infarction
Cerebrovascular accident
Renal failure
Puerperium
Retroperitoneal malignancy (Ogilvie's syndrome)
Orthopaedic trauma
Myxoedema
Electrolyte disturbance
The state of colonic motility at any point in time is determined by a balance of the inhibitory influence of the sympathetic nervous supply and the stimulatory effect of the parasympathetic system. It has been suggested that ‘neuropraxia’ of the sacral parasympathetic nerves may be a factor in the aetiology of ACPO, leading to a failure of propulsion in the left colon. This would also explain the ‘cut-off’ between dilated and collapsed bowel, which is located on the left side of the large bowel in 82% of patients.
29
Many of the conditions commonly associated with ACPO, such as sepsis, are likely to result from sympathetic overactivity.
The clinical features of ACPO are almost identical to those of mechanical large-bowel obstruction, making differentiation on clinical grounds alone almost impossible. In a review of 400 patients it was noted that the clinical features of ACPO were abdominal pain (83%), constipation (51%), diarrhoea (41%), fever (37%) and abdominal distension (100%).
29
On examination, the abdomen is generally very distended and tympanitic, but tenderness is often less than expected. The majority of patients will already have had operative procedures or have been hospitalised for some time because of some other disorder, and serum electrolytes are often abnormal.
Chapter 5
)
Plain radiographs of the abdomen in ACPO typically show gross distension of the large bowel with cut-off at the splenic flexure, rectosigmoid junction or, less commonly, the hepatic flexure (
Fig. 10.7
). Gas–fluid levels are less commonly seen on the plain radiograph in patients with ACPO compared with those presenting with mechanical obstruction.
30
It has been suggested that a prone lateral view of the rectum may be useful in making the diagnosis since gaseous filling of the rectum will tend to exclude mechanical obstruction. The caecal diameter should be measured on sequential abdominal radiographs since it is believed that the risk of caecal rupture increases greatly with increasing caecal diameter.
30
Figure 10.7
Plain abdominal radiograph of a patient with acute colonic pseudo-obstruction.
There is overwhelming support for the use of a contrast enema in all patients with suspected ACPO in order to establish the diagnosis, since the differentiation from mechanical obstruction can be extremely difficult. This is well illustrated in a study reported by Koruth et al., who performed a contrast enema on 91 patients with suspected large-bowel obstruction.
31
Of the 79 patients who were thought clinically to have mechanical obstruction, the diagnosis was confirmed in 50. There was free flow of contrast to the caecum in the remaining 29 patients. Of these 29, 11 had non-obstructing colonic pathology such as diverticular disease and ulcerative colitis and 18 patients had pseudo-obstruction. Of the 12 patients who were thought to have pseudo-obstruction before the water-soluble contrast enema, two were shown to have carcinoma of the colon.
The initial management of ACPO is non-operative and the underlying cause is treated if possible. Any medications that cause gut stasis should be discontinued, particularly analgesics. A nasogastric tube is routinely inserted to prevent swallowed air from entering the intestine and the use of enemas and flatus tubes is said to be of value in the treatment of early colonic pseudo-obstruction; in a number of patients even the water-soluble contrast enema used to establish the diagnosis may have a useful therapeutic effect. In most patients, the condition will resolve without intervention. One study found that it took an average of 6.5 days for complete resolution to take place in a group of 26 patients treated medically.
30
Progress should be checked by serial examination of the abdomen and by abdominal radiographs.
It is only when the risk of perforation increases substantially that more active intervention becomes necessary. The risk of perforation is approximately 3% and it has been shown that there is a correlation between perforation and the duration of distension.
32
The mean duration of distension was 6 days in the group of patients who went on to perforate compared with a mean duration of only 2 days in the group that did not progress to perforation.
In a randomised, double-blind, controlled trial of neostigmine only, 10 of 11 patients who were treated with intravenous neostigmine had prompt passage of flatus or stool, with reduced abdominal distension, compared with none of 10 patients who received placebo injection.
33
These results were mirrored in a trial of 28 patients, with rapid resolution in 26. Time to pass flatus varied from 30 seconds to 10 minutes. In the two patients who failed to resolve, one was found to have a sigmoid cancer and the other died of multi-organ failure.
34
There is a risk of bradycardia with cholinergic agonists, and it has been suggested that patients with cardiac instability should not be treated with neostigmine. Interestingly, there is anecdotal evidence that the concomitant administration of glycopyrrolate with neostigmine seems to offset the risk of bradycardia and may be considered in patients with cardiac instability.
35
Epidural anaesthesia blocks sympathetic outflow, and improvement has been observed in a number of patients with ACPO who have had this form of treatment.
36
The use of colonoscopy to decompress the colon in ACPO has become well established and it is successful in 73–90% of patients.
37
The procedure can be difficult and tedious, requiring a skilled colonoscopist, and air insufflation must be kept to a minimum. Frequent small-volume irrigation is required to ensure good visibility in the colon and maintain the patency of the colonoscope suction channel. A further advantage of colonoscopy is that necrotic patches can be identified on the colonic mucosa, allowing pre-emptive surgical treatment before perforation supervenes. The risk of perforation of the colon during colonoscopy for this condition has been estimated at around 3%,
38
and other complications are very unusual. It should be emphasised, however, that radiographs taken after successful clinical response often fail to show complete resolution of caecal distension, and one disadvantage of colonoscopic treatment is the tendency for the condition to recur. The overall rate of recurrence following initial colonoscopic decompression varies from 15% to 29%.
39
There is some difference of opinion about the best method of management of recurrent ACPO, but the safety and efficacy of repeat colonic decompression has now been reported.
27
,
37
,
39
A potential means of avoiding recurrence is intubation of the caecum with a long intestinal tube passed alongside the colonoscope.
40
The indications for surgery include the following:
Caecal distension – the extent of distension varies between authors, from 9 to 12 cm,
41
the threshold rising with increasing availability of medical therapy.
Continuing caecal distension beyond 48–72 hours despite maximum medical therapy.
Pain over the right iliac fossa, i.e. the caecum.
Pneumoperitoneum.
There are doctors who recommend percutaneous caecostomies,
42
where a tube is inserted into the caecum using radiological guidance for the purpose of decompression. A trephine caecostomy may be performed under local anaesthesia. To avoid contamination, the caecum can be sutured to the incised external or internal oblique muscles and only opened when the peritoneal cavity is sealed off. Only when perforation of the caecum is suspected should a full laparotomy be performed. If a perforation or necrosis of the caecum has already occurred, a full laparotomy is necessary and a right hemicolectomy is the treatment of choice. When resection of the right colon is required, it is probably safest to bring out an ileostomy and mucous fistula and reanastomose the two ends of bowel at a later date. Primary anastomosis may be feasible if contamination of the peritoneal cavity is not a feature and the remaining colon looks healthy.
Approximately 85% of colonic emergencies are due to colon cancer and most of these occur in the elderly patient.
43
However, only 8–29% of colorectal cancer patients present with intestinal obstruction.
44
Approximately half of splenic flexure tumours present with obstruction, compared to 25% of those in the left colon, 6% of rectosigmoid lesions and 8–30% of right-sided carcinomas.
45
Both obstruction and perforation occur together in approximately 1% of all colon cancers, but in patients who have an obstruction caused by cancer, 12–19% will have a perforation.
44
The perforation may either be at the site of the tumour or in the caecum, caused by back pressure from the distal obstructing lesion.
The influence of obstruction on prognosis is controversial. Some studies suggest that the apparent adverse effect of obstruction on prognosis is a consequence of the stage of the disease rather than obstruction itself, as 27% have liver metastasis at the time of operation.
46
Other reports, however, suggest that obstruction is an independent predictor of poor prognosis.
47
Symptoms associated with large-bowel obstruction frequently reflect the site of the tumour. In right-sided obstruction, particularly at the level of the ileocaecal valve, the onset of colicky central abdominal pain may be quite sudden and vomiting a relatively early feature. If the obstruction is at the rectosigmoid junction, there may be a history of a change in bowel habit and of rectal bleeding, with vomiting uncommon.
On examination, abdominal distension is the most notable feature. Peritoneal irritation suggests that perforation is either imminent or may have already occurred. Palpation of an irregular liver edge suggests that liver metastasis may be present and a palpable mass on rectal examination obviously suggests a carcinoma of the rectum.
Plain abdominal radiography will usually provide the diagnosis of large-bowel obstruction. The pattern of gas distribution in both the small and large bowel will depend on the site of obstruction and also on whether the ileocaecal valve is competent. However, as already mentioned, differentiation from ACPO requires a contrast enema.
A water-soluble contrast enema (
Fig. 10.8
) should be carried out in all patients with suspected large-bowel obstruction without evidence of perforation.
31
Figure 10.8
Water-soluble contrast enema demonstrating complete obstruction in the proximal sigmoid colon.
A water-soluble contrast enema will exclude other conditions such as volvulus or pseudo-obstruction and, in addition, may go some way to cleansing the colon distal to the obstructing lesion. Sigmoidoscopy or colonoscopy can be useful, particularly if the suspected obstructing lesion is in the distal colon. In addition, either technique can be used to exclude synchronous carcinoma or adenoma below the level of obstruction. CT with intravenous and water-soluble rectal contrast is increasingly used in the emergency setting and this has the added advantage of providing information on the spread of disease preoperatively (
Fig. 10.9
).
Figure 10.9
CT scan of a patient with a perforated caecum (perforation indicated by arrow). The scan also revealed an obstructing splenic flexure tumour and extensive liver metastases.
The morbidity and mortality rate associated with emergency procedures for obstruction of the colon is at least twice that for elective surgery. This has encouraged surgeons to develop methods such as the use of expandable intraluminal stents. These can be used to palliate patients or to convert emergency operations into elective procedures. Several colonic stents are currently available and can be placed using radiological or endoscopic assistance. Various models can be deployed over guidewires or through an endoscope.
A review of 54 case series that included 1198 patients
48
found that in 791 palliative patients, technical success was possible in 93% and clinical success (defined as decompression in less than 48 hours) in 91%. Stents also provided successful conversion to elective procedures in 92% of 407 patients. Intraluminal stenting was more successful in shorter, distal lesions and colonic primaries. Technical failure usually results from an inability to pass the guidewire or stent catheter across the lesion. Reported complications include perforation in 3.8%, stent migration in 12% and re-obstruction in 7%.
When surgery is undertaken, patients with right-sided obstruction should be positioned flat on the operating table. Those with left-sided large-bowel obstruction are placed in the lithotomy/Trendelenburg position to allow access to the anus during the procedure for purposes of irrigation of the rectal stump or anal insertion of a surgical stapling instrument. It also allows the surgical assistant the option of standing between the patient's legs. The abdomen is opened through a midline incision. If the bowel is tense, it should be decompressed: first, to improve visualisation of the rest of the abdomen and, second, to prevent spillage of faecal content. The large bowel can be decompressed by inserting a 16 G intravenous catheter or 18 F nasogastric tube obliquely through the colonic wall, following which suction is applied. This is often enough to make it possible to handle the large bowel without fear of rupture as the distension is mainly gas. After localisation of the primary tumour, synchronous tumours should be excluded. The presence of direct spread to adjacent structures should also be assessed, in addition to any peritoneal seedlings and the presence of liver secondaries. Based on these observations, a decision can be made as to whether the operation is potentially curative or palliative.
When there is a prospect of curative resection, standard techniques of radical cancer therapy should be employed, including wide excision of the lesion en bloc with the appropriate blood vessels and mesentery. If the lesion is adherent to other structures, an attempt should be made to resect the affected part en bloc. High cure rates are possible with locally advanced tumours if a radical resection is performed and clear resection margins obtained. The presence of liver or peritoneal metastases does not preclude resection of the primary carcinoma and gastrointestinal continuity should be restored if at all possible.
If there is a closed loop obstruction because of competence of the ileocaecal valve, the caecum and right colon may be very tense. Decompression can be achieved as a preliminary to resection by making a small enterotomy in the terminal ileum and passing a Foley catheter through the ileocaecal valve into the caecum. This technique is particularly useful in situations where there is splitting of the taenia on the caecum, indicating impending rupture. The range of operations available for treatment of right-sided tumours causing obstruction includes right hemicolectomy with primary anastomosis, right hemicolectomy with exteriorisation of both ends of the large bowel, and ileo-transverse colon bypass. There is general agreement that a right hemicolectomy with primary anastomosis is the treatment of choice in most patients; however, this procedure is by no means free of complications. One report noted an operative mortality rate of 17% in 195 patients who had emergency right hemicolectomy with primary anastomosis for obstructing colonic carcinoma.
49
In addition, a leak rate of 10% was noted in 179 patients who had a right hemicolectomy and primary anastomosis for obstruction. This compares with a leak rate of 6% in 579 patients with right colon cancer who did not have an obstruction.
49
Other studies have shown similar mortality rates, and many of the deaths resulted from anastomotic failure. This suggests that instead of subjecting all patients with obstruction to right hemicolectomy with primary anastomosis, it may be wiser to use a policy of selection, subjecting patients with good risk status to primary anastomosis and managing patients with risk factors for anastomotic failure by resection and exteriorisation of the bowel ends.
The anastomotic technique used will depend on the surgeon's preference. If the obstructed bowel is very thickened and oedematous, care should be taken with the use of stapling instruments since there is a tendency for the instruments to cut through oedematous bowel. On these occasions a sutured anastomosis is preferable. Only on the relatively rare occasion when locally advanced disease is unresectable should the patient be subjected to an ileo-transverse colon bypass procedure. There is almost no place for caecostomy in the current management of right-sided large-bowel obstruction; trephine ileostomy, which can usually be achieved under local anaesthesia, affords better palliation in very sick patients not fit for operation under general anaesthesia.
Most surgeons would advocate an extended right hemicolectomy for patients with transverse colon carcinoma, and decompression of the colon may be necessary to facilitate mobilisation. For the patient who has a large carcinoma obstructing the transverse colon, achieving clearance may be difficult because of involvement of the transverse mesocolon and adjacent organs. The splenic flexure is mobilised to assist with a primary anastomosis between ileum and upper descending colon. In the sick patient who already has a perforated caecum, resection and an ileostomy may be appropriate.
For obstructing left-sided colonic tumours, the two most frequent options are resection and either primary anastomosis, or colostomy and rectal closure (Hartmann's procedure). A three-stage procedure (diverting colostomy, resection and colostomy closure) is now rarely used, and only in critically ill patients. Two-stage procedures became popular during the 1970s and remain the procedure of choice for most surgeons, with an overall mortality around 10%.
44
However, the main disadvantage of this approach, apart from the fact that a second operation is required for reconstruction, is that approximately 40% of patients will be left with a permanent stoma.
The second stage of the two-stage operation may be difficult because of adhesions, although it is facilitated if the rectal stump has been divided above the peritoneal reflection. The timing of the second stage is also important and most surgeons will wait 2–6 months. The second stage of the two-stage procedure is increasingly being performed using laparoscopic techniques. If successful the hospital stay is reduced, but conversion to open surgery may be necessary.
Primary resection and anastomosis is associated with a shorter hospital stay, reduced mortality and morbidity, and the avoidance of a stoma. It is increasingly becoming the operation of choice for colorectal surgeons, with the exact procedure depending on a number of factors. Subtotal colectomy followed by ileosigmoid or ileorectal anastomosis has been reported to have a low operative mortality of 3–11% and a hospital stay of around 15–20 days.
50
Segmental colectomy (left hemicolectomy, sigmoid colectomy or anterior resection of the rectosigmoid) with on-table irrigation followed by primary anastomosis has a reported operative mortality rate around 10%, anastomotic leakage around 4% and hospital stay of approximately 20 days.
51
A retrospective study of 243 patients who underwent emergency operation for obstructing colorectal cancers showed that primary resection and anastomosis for left-sided malignant obstruction either by segmental resection with on-table lavage or by subtotal colectomy was not more hazardous than primary anastomosis for right-sided obstruction.
52
Single-stage resection with primary anastomosis was possible in 197 patients. Of the 101 patients with left-sided obstruction, segmental resection with on-table colonic lavage was performed in 75 patients and subtotal colectomy in 26. There were no differences in the mortality or leakage rates between patients with right-sided and left-sided lesions (mortality 7.3% vs. 8.9%; leakage 5.2% vs. 6.9%, respectively).
More recent papers have described segmental resection and primary anastomosis without any attempt to clean the bowel. One study reported only one leak and one postoperative death in 58 consecutive patients with left-sided malignant colonic obstruction who underwent bowel decompression without irrigation, followed by resection and primary colocolic anastomosis.
53
The patients in this study had a mean age of 63 (range 54–89) years, the leak occurring in a 61-year-old with a sigmoid carcinoma and the death in an 80-year-old due to myocardial infarction. None of the carcinomas described were rectosigmoid or rectal. A further study of left-sided obstruction in which 40 of 60 patients had carcinoma compared one-stage resection with Hartmann's procedure.
54
There was no significant difference in outcome or time taken to complete the operation, with the only death being in the Hartmann's group. Again, none of the tumours described was more distal than the sigmoid colon.
Circumstances usually dictate the operative choice. In the elderly, segmental resection may be preferable to a more extensive resection, both in terms of operative morbidity and postoperative problems of incontinence. Subtotal colectomy is clearly the procedure of choice when there are synchronous tumours in the colon. When subtotal colectomy with ileorectal anastomosis is selected as the operation of choice, the whole colon is mobilised and the rectum washed out as for elective surgery. After resection of the colon, an ileorectal or ileosigmoid anastomosis is performed using either a sutured or stapled technique. The first randomised trial comparing subtotal versus segmental resection was reported in 1995.
55
This study involved 91 eligible patients recruited by 18 consultant surgeons in 12 centres; 47 were randomised to subtotal colectomy and 44 to on-table irrigation and segmental colectomy. There was no significant difference in operative mortality, hospital stay, anastomotic leakage or wound sepsis between the two groups. There was, however, a significantly higher permanent stoma rate in the subtotal colectomy group compared with the segmental colectomy group (7 vs. 1). The high permanent stoma rate in the subtotal colectomy group was partly accounted for by four patients who were randomised to subtotal colectomy but who underwent Hartmann's procedure because this was thought clinically more appropriate at the time of surgery by the operating surgeon. Two additional patients had the anastomosis taken down at a later date and a stoma formed. At follow-up 4 months after the operation, there was a significantly greater number of patients who had three or more bowel movements a day after subtotal colectomy than after segmental resection (14 of 35 vs. 4 of 35). One patient had 12 bowel movements per day after subtotal colectomy. Nearly one-third of patients randomised to subtotal colectomy had night-time bowel movements during the first few months after operation. In contrast, less than 10% of those who had segmental resection had this problem.
Segmental resection of the colon, with or without intraoperative irrigation, is the preferred treatment for left-sided malignant colonic obstruction.
55
,
56
Although the SCOTIA study
55
addressed the immediate and early results after these two procedures, it did not investigate the long-term implications of either procedure. It has been argued that there are advantages in performing a subtotal colectomy rather than segmental resection because synchronous tumours will be removed along with the obstructing lesion and, since the length of colon left is small, there should be less risk of developing a metachronous tumour. Clearly the operating surgeon needs to take all these factors into account when making the final decision on which procedure is most suitable for the individual patient.