Management Other anorectal abnormalities are excluded by anorectal examination. Biopsy specimens ofsolitary rectal ulcers show characteristic fibromuscular obliteration of the lamina propria. is treated and patients are advised to avoid excessive straining at stool. Many agents, including topical steroid , have been used but none significantly affects the course of the disease. Surgery is BMJ: first published as 10.1136/bmj.305.6847.246 on 25 July 1992. Downloaded from usually unrewarding, but some doctors advocate rectopexy if there is evidence of abnormal descent. Symptoms tend to be intermittent and usually little more than a nuisance. Occasionally profuse haemorrhage may occur, which may necessitate urgent excision ofthe . White indurated base of a solitary rectal ulcer viewed through a proctoscope.

Proctalgia fugax Proctalgia fugax is a benign condition that usually affects young men with anxiety. It is often familial and is characterised by attacks ofpain in the rectum, perineum, or urethra, which rise to a crescendo of severe , lasting from a few minutes to halfan hour. It is not especially related to Treatment of proctalgia fugax defecation and often occurs at night, waking the patient up. The pain is * Explanation and reassurance spasmodic and responds to smooth muscle relaxants such as nitrates. * Treatment, after exclusion of other organic disease, is centred on * Antispasmodics explaining the symptoms and reassuring patients that there is no serious * Glyceryl trinitrate underlying pathology. Analgesics or glyceryl trinitrate may be tried if * Puborectalis stretch necessary, but treatment with drugs is not generally satisfactory. In severe cases the patient should be taught to insert a gloved and lubricated finger into the rectum and stretch the puborectalis which is in spasm, in much the same way as forcibly extending a gastrocnemius in spasm during an attack ofcramp. The photographs were produced by the department of medical illustration, Salford Health Authority, and the line drawings were prepared by Paul Somerset, medical illustration department, Wythenshawe Hospital. Mr N J Andrews is consultant surgeon, department ofgeneral surgery, Pilgrim Hospital, Boston, Lincolnshire, and Mr D J Jones is lecturer and honorary senior registrar, Hope Hospital, Salford.

The ABC ofColorectal Diseases has been edited by Mr D J Jones and Professor M H Irving, http://www.bmj.com/ department of , Hope Hospital, Salford. Current Issues in Cancer

Colorectal cancer on 25 September 2021 by guest. Protected copyright.

R H J Begent

This is the twelfth in a series of Each year 22 000 people die of cancer of the colon and These trends have stronger support in the United articles examining recent rectum in Britain, making this disease the second most States than in Britain, where there has been concern developmtents in cancer common cause ofdeath from cancer. Five year survival that quality of life will be diminished by introducing is less than 30%, and until recently surgery was the unproved investigation or treatment. Validated only treatment that reduced mortality. Moves are now methods of measurement of quality of life make it being made towards more screening, adjuvant therapy, possible to address such concerns.' This review and intensive follow up with resection or chemo- considers the case for a more prospective and active therapy for recurrence. Implementation of these approach to management of . The policies requires structured planning rather than various stages ofmanagement are listed in table I. reaction to crises. Department ofClinical Oncology, Royal Free TABLE I-Components ofmanagement ofcolorectal cancer Hospital School of Screening Medicine, London Component Requirements of comments SIGMOIDOSCOPY NW3 2PF The benefits of sigmoidoscopy in healthly popula- R H J Begent, Ronald Raven Case finding Public and professional awareness Screening Family history, genetic markers, knowledge tions are controversial. The proposal that sigmoido- professor of predisposing conditions scopy, preferably with a flexibile instrument, should Surgical resection Specialist colorectal surgeons Series edited by: Dr GM Adjuvant therapy Some indicationsbut more trials needed be done every three to five years in people over the age Mead. Follow up Role of intensive follow up uncertain of 50 is supported by the National Cancer Institute of Palliation Chemotherapy and radiotherapy useful; the United States, the American Cancer Society, and quality of life must be assessed BMJ 1992;305:246-9 the American College of Physicians. However, the

246 BMJ VOLUME 305 25 JULY 1992 King's Fund consensus statement,' the United States TABLE I -Current recommendations for adjuvant therapy preventive services task force, and the Canadian periodic health examination task force considered that Site and stage Recommended adjuvant therapy the case for screening was not strong enough. A case- Rectum: control study has since shown a 59% reduction in A None mortality from bowel cancer in people screened by B Radiotherapy and chemotherapy C Radiotherapy and chemotherapy sigmoidoscopy, with the reduction applying only to Colon: BMJ: first published as 10.1136/bmj.305.6847.246 on 25 July 1992. Downloaded from tumours arising in the screened part of the bowel.' A Nonie The potential benefit of screening comes largely B None from identifying and removing premalignant adeno- C Chemotherapy mas, though carcinomas may also be diagnosed. Evidence that sigmoidoscopy rather than colonoscopy patients after apparently complete resection of rectal is appropriate for initial screening comes from a study cancer and is more common when there is local in which the standardised incidence ratio of colonic infiltration into perirectal tissues or on to serosal cancer was 3-6 ifa single villous, tubulovillous, or large surfaces. Radiotherapy either before or after surgery (>1 cm) adenoma was found in the rectosigmoid by reduces this incidence. Preoperative treatment has rigid sigmoidoscopy.4 If the adenomas were multiple lower morbidity, but patients with early disease the standardised incidence ratio was 6-6. Thus a large, and little chance of recurrence may be treated un- tubulovillous, or villous adenoma in the rectosigmoid necessarily. Survival benefit from perioperative radio- seems to be predictive ofcarcinoma at remote proximal therapy has not been shown conclusively, but combined sites in the colon and such patients should have follow analysis of the relevant trials suggests a benefit of about up by colonoscopy. The case for routine screening of 10%. More data on survival are expected shortly from people aged over 50 by sigmoidoscopy should be re- mature analysis of the Medical Research Council's examined. second and third trials of radiotherapy, and the issue is being studied further in the United Kingdom Co- FAECAL OCCULT BLOOD ordinating Committee on Cancer Research AXIS trial Large randomised studies of the value of faecal (Adjuvant X ray and Infusion Study). occult blood testing in healthy populations are in progress, but definitive results in terms of mortality cannot be expected until 1995. Sensitivity and Chemotherapy specificity are limited but a higher proportion of good Chemotherapy given in addition to postoperative prognosis (Dukes's stage A) tumours is detected in radiotherapy prolongs survival in patients at high risk screened populations than in controls.5 of recurrence. In a randomised study patients with nodal disease or infiltration through the muscle layers HIGH RISK GROUPS received radiation alone or radiation with chemo- Family doctors, surgeons, and clinical oncologists therapy.9 5-Fluorouracil was given with radiotherapy have an important role in identifying risk by taking a and 5-fluorouracil with semustine before and after comprehensive family history and acting on it. Right radiotherapy. There was a significant reduction in the sided tumours, carcinoma occurring at a young age, local recurrence rate in the group receiving chemo- familial adenomatous polyposis, a carcinoma in a first therapy in addition to radiotherapy, which was sug- degree relative, and family history of ovarian, breast, or uterine carcinoma, particularly at a young age, are http://www.bmj.com/ pointers to a possible familial risk. Assessment of risk by a clinical geneticist is particularly valuable in such cases and the development of genetic markers-for example, in familial adenomatous polyposis-is increasing the precision of risk assessment. Colono- scopy should be done regularly in these people6 as well as in those with ulcerative or who have had

resection of a carcinoma or adenoma. on 25 September 2021 by guest. Protected copyright.

Intitial surgery Operative mortality and postoperative morbidity and survival are better, especially for rectal cancer, if surgery is done by a specialist colorectal surgeon.78 The King's Fund consensus statement on colorectal cancer considered that the evidence was strong enough to recommend that arrangements should be made in each British health district for a specialist emergency and elective service or for patients to be referred to a hospital that makes this provision.' This could be achieved if health authorities placing contracts for cancer care made it a required quality criterion.

Adjuvant therapy of rectal carcinoma Rectal and colonic carcinoma are considered separately because local recurrence ofrectal carcinoma in the pelvis gives radiotherapy a role in treatment of rectal carcinoma which does not apply to colonic tumours (table II). Computed tomograms showing (above) lung metastases before treat- RADIOTHERAPY ment and (below) after 12 weeks ofchemotherapy with S-fluorouracil Local recurrence occurs in about one third of and leucovorin

BMJ VOLUME 305 25 JULY 1992 247 gested to be the result of the radiosensitising effect of 5- Follow up after initial treatment fluorouracil given concurrently with radiotherapy. Five year survival has changed little for colonic and The overall death rate was reduced by 29%Yo with the rectal carcinoma in recent years. It is about 80'% for addition of chemotherapy. This study is the most Dukes's stage A, 50%Y, for stage B, and 30% for stage C. convincing to date, but before its publication a con- Very few patients presenting with distant metastases or sensus statement from the National Cancer Institute in unresectable disease live for five years.

the United States considered that there was strong As many as 20% ofpatients relapsing after apparently BMJ: first published as 10.1136/bmj.305.6847.246 on 25 July 1992. Downloaded from evidence in favour of chemotherapy for Dukes's stage curative resection can be cured by resection of hepatic B and C carcinoma of the rectum.' or pulmonarv metastases.' Results are best when there At present, inadequate information is available to is a solitary deposit. Intensive follow up can detect reconmmend what the chemotherapy regimen should recurrence before it produces symptoms. Monthly be. The trials needed to answer such questions require measurements of serum carcinoembryonic antigen large numbers of patients and prolonged follow up. concentration predict clinically evident recurrence Several trials of different aspects of adjuvant chemo- by an average of six months in about 60%/o of patients therapy and radiotherapy for rectal cancer are in and trigger other investigations or exploratory laparo- progress and the most constructive approach is to enter tomv.'9 Other serum tumour markers may also patients into these. In Britain the AXIS trial addresses contribute. Routine chest radiography, liver ultra- a relevant question in that patients are randomised sonography, and abdominal computed tomography to no chemotherapy or portal vein infusion of 5- and colonoscopy have also been used for early detec- fluorouracil for seven days after surgery. 5-Fluorouracil tion. Immunoscintigraphy can be helpful in locating given in this way has access to the systemic circulation metastases when serum carcinoembryonic antigen and could contribute to disease control in the pelvis concentrations rise.)" and elsewhere but is unlikely to have a radiosensitising As only solitary metastases will be curable and these effect, being given before radiotherapy. will eventually be evident by simple clinical follow up and still be resectable, some people have argued against intensive follow up. In a randomised trial Adjuvant therapy in colonic carcinoma supported by the Cancer Research Campaign, serum 5-Fluorouracil gives partial or complete response in carcinoembryonic antigen concentrations are measured 10- 15 % of patients with advanced disease and the regularly for three years. If there is a significant rise in argument for using it as adjuvant treatment after antigen concentration patients are randomised to apparently complete tumour resection is weak. Perhaps second look laparotomy and tumour resection unless surprisingly, meta-analysis oftrials ofadjuvant therapy there is a contraindication. In the control arm the including 5-fluorouracil showed a small (3 4%) but surgeon is not informed of the antigen concentration significant survival benefit for treatment." This has not and continues to follow up the patient in the normal in general been accepted as sufficient reason for using way. Over 1100 patients have been entered but a result intravenous 5-fluorouracil for adjuvant therapy. Many is not yet available. trials combining 5-fluorouracil with other agents have given disappointing results.'2 Developments incorporating 5-fluorouracil seem Chemotherapy for advanced disease more interesting. Infusion of 5-fluorouracil into the Most tumour recurrences detected on follow up are portal vein for seven days after resection ofthe primary not resectable, but could early diagnosis be beneficial ttimour gave impressive results in some trials.'" Other in their management? The low response rate to 5- http://www.bmj.com/ trials have shown no benefit from this approach and it fluorouracil has rightly limited its use to palliation, is being further investigated in the AXIS trial. The trial but the higher response rates now achieved with has the advantage that the treatment is completed combination therapy justify re-examination of this during admission to hospital for primary surgery. issue. A study in the Nordic countries randomised 183 Patients with all Dukes's stages are eligible but manv patients with advanced but asymptomatic colorectal surgeons prefer to omit patients with Dukes's stage cancer for treatment either immediately or only when A tumours because of their. good survival without symptoms developed.' Median survival was five adjuvant therapy. months longer (p<0-01) and the symptom free period on 25 September 2021 by guest. Protected copyright. The combination of 5-fluorouracil and levamisole as six months longer (p<0001) in patients treated with adjuvant therapy has been shown to produce a signifi- methotrexate, leucovorin, and 5-fluorouracil from the cant survival advantage over levamisole alone and over detection of their metastases rather than when no treatment in patients with Dukes's stage C colonic symptoms developed. In a subset of patients in whom carcinoma. '-' The death rate was reduced by a third quality of life was assessed, those in the early treatment compared with controls. For patients with Dukes's arm also fared better. Ifreproducible, this study argues stage B2 disease (tumour affecting pericolic fat or the for careful follow up including measurement of serum serosa) a significant advantage in survival was not carcinoembryonic antigen concentration to give chemo- shown, but there were relatively few patients in this therapy as soon as recurrence is detected. arm and longer follow up is needed before the case for adjuvant therapy in stage B disease can be judged. Levamisole, which was initially used for its anti- Palliation helminthic properties, also has immunomodulatory Radiotherapy is effective in palliation of symptoms actions. Objectors to the adoption ofthis regimen point produced by localised tumours, particularly in the out that there was no 5-fluorouracil only arm in the pelvis. Chemotherapy also has an place in palliation.' major trial, although there was in a smaller study," and Response rates with 5-fluorouracil alone are low (10- that the place of levamisole as a single agent is not 15%). The effect is dose related and infusion over secure. a few days increases response rates." The effect of 5-Fluorouracil modulated by leucovQrin has about 5-fluorouracil is modulated by administration with double the response rate of 5-fluorouracil alone in leucovorin, which prolongs the inhibition of thymidy- patients with advanced disease. It is now being studied late svnthase activitv and hence inhibits DNA svnthesis in randomised adjuvant therapy trials and results can through stabilisation of the ternary complex of 5,10- be expected in the next few years. Until optimal methvlenetetrahvdrofolic acid with fluorodeoxyuridine chemotherapy regimens are established the best option monophosphate (a metabolite of 5-fluorouracil) and is to enter patients into relevant clinical trials. thymidylate synthase. Response rates are increased to

248 BMJ VOLUME 305 25 JULY 1992 20-40% and reliefofsymptoms is reported in up to 75% Conclusion of patients. Increased survival compared with that Surgery followed by only simple palliation for after 5-fluorouracil alone has been reported. Meta- symptoms caused by advanced disease is no longer analysis has not confirmed this report, although sufficient in management of colorectal cancer. Early response rates were increased from 11% to 23% diagnosis through case finding and screening of high (p< 10 7) with 5-fluorouracil and leucovorin compared risk groups can lead to treatment when prognosis is

with 5-fluorouracil alone.24 The relatively low overall relatively good. Adjuvant chemotherapy and radio- BMJ: first published as 10.1136/bmj.305.6847.246 on 25 July 1992. Downloaded from response rate and crossover of patients from one therapy have a place but the best way ofapplying them regimen to the other may conceal a survival benefit. is not clear and it is important that patients are entered When metastases are confined to the liver 5-fluoro- into relevant clinical trials. Early chemotherapy for deoxyuridine infused through the hepatic artery recurrence seems to improve survival, and ifconfirmed produces higher response rates than intravenously this would justify careful follow up after initial surgery. administered 5-fluorouracil.2S However, increased Particular attention should be paid to measuring survival has not been shown, probably because of quality of life in clinical trials and in routine practice. metastases outside the liver which were too small to Teams including the relevant specialists are likely to be detect when treatment was begun. A current trial of most effective in implementing these measures in an this approach is further investigating survival and efficient and cost effective way. quality of life. These chemotherapy regimens are generally well I thank Dr J C Thompson for her help in preparing this tolerated but if there is no convincing survival benefit article. their use has to be justified by improved quality oflife.' Data are sparse, but improved quality of life has been I Byrne M. Cancer chemmotherapv and qtialit ol life.BIM1992;304: 1523-4. 2 Seventh King's Fund Forum. Canicer of/iieii colon and retiuni consensus sialtenie1. reported with one regimen of 5-fluorouracil and London: King's Fund, 199(. leucovorin compared with 5-fluorouracil alone.7' More 3 Selbv JV, Friedman GI), Qucenshury CP1, Wciss NS. A case contirol stludy of screening sigmoidoscopy and mortality Irom colorectal cancer. N 1ngI17 studies using validated quality of life assessments are Med 1992;326:653-5. necessary before the place of palliative chemotherapy 4 Aitken WS, Morson BC, Cuzick J. Long terin risk of colorectal canicer after excision Of rectosigmoid adenomas. N lEnglJ M1d 1992;326:658-62. can be properly assessed. 5 Hardcastlc JD), Farrands PA, Balfour 'W. Controlled trial of fiecal octcllt blood testing in the detection of colorectal cantcer. Lancc 1983;ii: 1-4. 6 Houlston RS, Murdav V, Harocopos C, Williams CB, Slack J. Scrceniilng anld genetic cotinselling for relatises of partirets with colorectal cancer in a family Future prospects cancer clinic. BAIJ 1990;301:18-25. The outlook for people with colorectal cancer has 7 McArdle CS, Hole ). Impact oIf variabilitv amonlg surgeonis ott postoperatRis morbidity and mortality and ultimate survival. BJM7 1991;302:15011-5. improved slightly, and several new approaches being 8 Darby CR, Berry AR, Mortensen N. Management variability itn surgery for investigated in clinical trials may have a further colorectal emergencies. Br] Surg 1992;79:2(6- 1(). impact. 9 Krook JE, Moertcl CG, Gunderson Ll, Wieand LL, Collins R'', cart RW,ci al. Effective surgical adjuvant therapy of high risk rcctal cancer. N lngl7 Med 1991;324:709-45. SURGERY 10 Office of Medical Applications of Rcsearch. NIH consenssis cotilfrccic. Adjuvant therapy for patients with colott and rectal cancer. 7.1A14 The frequency of local recurrence may be reduced 1990;264: 1444-50. 11 Buyse M, Zeleniuch-Jacquotte A, Chambers TC. Ad'usaint therapy of by radioimmunoguided surgery. Antibodies reacting colorectal cancer. Why we still doit't know. 7AIMA. 1988;259:3571-8. with colorectal cancer antigens and radiolabelled with 12 Schnall SF, MacDonald JS. Adjuvant therapy itt colorectal cancer. .Senin iodine-125 are given intravenously before surgery and Oncol 1991;18:560-70. a hand held gamma detecting probe is used to locate the 13 Taylor 1, Machin D, Mullee M. A randimised controlled trial ol adjuvant portal vein cytotoxic perfusion in tolorectal canccr. Br 7 Surg 1985;72: http://www.bmj.com/ tumour at operation. Residual tumour, not detectable 359-63. 14 Moertel CG, Fleming FR, MiacDonald JS. Levamisolc and fliloroltracil in by conventional means, can be located in the tumour adjuvant therapy for resectedtcolon carcinoma. N EngI7 etid 19901;322: bed or at the margin ofthe resected specimen.26 Occult 352-8. 15 Moertel CG, Fleming 1, Macdonald JS, Haller 1), Lauric I.Thc itstcrgrotLp metastases have also been detected. study of fluorouracil (5-FU) plus lcvamisole (LEV) and lesamisutlealonc as adjusant therapy for stagc C colon cancer. A final report (abstrlct). DRUG RESISTANCE Proceedings of the American Association oJ (,lini/cal Oncologvy 1992;Mia\y:457. 16 Windle R, Bell P'RF, Shaw D). Five year results of a randomised trial of The best regimens containing 5-fluorouracil give adjuvant 5-fluorouracil and levamisolc in colorectal canccr. Ir 7 Surg 1987;74:569-72. only 20-40% response rates, and drug resistance 17 Arnaud JP, Bvse Mi, Nordlinger B. Adjuvant therapy ol poor prognosis colon on 25 September 2021 by guest. Protected copyright. eventually develops in almost all patients. The effects cancer with levamisole: results of an EORRIC double blind randomised of 5-fluorouracil may be enhanced modulation with clinical trial. fIr Surg 1989;76:284-9. by 18 August DA, Ottow RT, Sugarbaker P'H. Clinical perspectives of huiman drugs other than leucovorin. Interferon alfa is an colorectal cancer mctastases. Cancer Metastasi.s Rev 1984;3:303-24. example and is being studied in combination with 5- 19 Begent RHJ, Rustin GJS. Tumour markcrs: from carcinoembryonic antigcn to products of hybridoma technology. Cancer Surzn 1989;8:1118-2 1. fluorouracil and leucovorin in a randomised trial. 20 Begent RHJ, Keep PA, Searle F, Green AJ, Mitchcll HDC, Joncs BE, 't al. Uridine and dipyridamole are other modulators Radioimmunolocalisation and selection for surgery in rccurrent ctlorcctal cancer. Brj Surg 1986;73:64-7. of pyrimidine metabolism whose roles are being in- 21 Glimelius B for the Nordic Gastroltitestinal TIumour Adjuvant Therapy vestigated. Group. Expectaiscy of primary chemotherapy itn paients with advantced asymptomatic colorectal cancer: a randomised trial. Fur .7 Can-cer 19991; Suppl 2:S82. BIOLOGICAL THERAPY 22 Lokich JJ, Ahigrcn J1), Gullo JJ, IPhillips JA, Fryer JG. A prospective randomised comparison of continuous inifision fluorouracil with a coul- New approaches based on our improved under- ventional schedule in metastatic colorcctal carcinoma. A mid-Atlantic standing of the molecular basis of neoplasia show oncology prrogram study. ] Clin Oncol 1989;7:425-32. 23 Poon MA, O'Connell MJ, Moertcl CG, Wicand HS, Cullinan SA, tIscrsotn promise. Clinical trials of antibody targeted therapy LK, et al. Bitchemical modulation ol fluorouracil: evidence of signiificant are in progress in colorectal cancer. Antibodies improsement of sursival and quality of lifC in paticnts with advanced targeting colorectal cancer have been used to activate colorectal carcinoma. ] Clin Oncol 1989;7: 14107-18. 24 Piedbois 1P Buysc M, Rustum Y, Machotcr I), Erlichman C, Carlson RWX, ( natural immune effector mechanisms, to target al. Miodullation of 5-fluorouracil by leucosvorin in patients with adxvaisced therapeutic radionuclides or toxins, and to target an colorectal caticcr: a meta-analysis (abstract). 1roiceudings o/ ih/i Anerican Society of Clinical Oncologv 1992 May: 1126. enzyme which will activate a prodrug at the tumour 25 Chang AE, Schneider 11), Sugarbaker l'H, Simpson C, Cilnane Ni, Stlitibcrg site. A vaccine composed ofa human antibody directed M. A prospective randomised trial of regiontal versus systemic contintouIs 5- fluoroxyItridinc chermotherapy in the treatment oi colorectal liver mciastascs. against the idiotype of an antitumour antibody Ann Surg 1987;206:685-93. mimics the tumour associated antigen but is immuno- 26 Blair SI), Theodrou NA, Begent RHJ, )awson 'M, Salmon NI, Riggs l, ti al. Comparisois of anti-foetal microsillus and anti-CLA antibodies its genic, breaking natural tolerance to the tumour peroperativc radioimmunolocalisation of colorectal cailccr. Ir 7 Cancer antigen. 1991;61:891-4.

BMJ VOLUME 305 25 JULY 1992 249