Postgrad Med J 1997; 73: 225 - 229 (© The Fellowship of Postgraduate Medicine, 1997

Original articles

Prognosis and management of Crohn's disease inthe over-55 age group

RS Walmsley, CD Gillen, RN Allan

Queen ElizabethHospital, Edgbaston,Birmingham, B152TH, UKRS WalmsleyCD GillenRN Allan

Correspondence to Dr RSWalmsley, c/o Dr Allan

Accepted 10 May 1996

SummaryThe course, prognosis and management of62 patients with Crohn's disease aged 55years or over at diagnosis has been re-viewed. The distal ileus was the commonestsite of disease in the older patient, wherethe characteristic presentation was acuteafter initially mild symptoms. Early localresection was often required, particularlywhere there was diagnostic doubt or suspi-cion of caecal malignancy. Recurrencerates were much lower in the older patientthan after resection in younger patients.Medical treatment played a minor role inthe management ofpatients with distal ilealdisease, in partbecause stricture formationwas present at diagnosis and the acutenature of symptoms at presentation led toearly surgical treatment. Colonic Crohn'sdisease was usually confined to the distal orleft side of the colon and initially could bedifficult to distinguish from diverticulardisease. Extensive colonic Crohn's diseasewas rare. The apparently limited diseasewas not necessarily associated with a goodprognosis, since disease at this site some-times progressed rapidly, necessitatingurgent surgical resection.Medical treatment (corticosteriod ther-

apy, with or without azathioprine) wasusually effective initially for treatment ofsymptomatic colonic Crohn's disease, butsustained remission was rare. Those pa-tients with persistent symptoms were re-stored to good health with surgicaltreatment but at a price, in that nearly halfeventually required a permanent stoma.

Keywords: Crohn's disease

Crohn's disease typically presents in youngadult life, but also occurs in the older patient.Indeed, some reports suggest a bimodaldistribution with a second peak occurringbetween the ages of 50 and 70.' The increasingincidence of both Crohn's disease and theproportion of elderly people in the population,enhances the significance and importance ofCrohn's disease in this age group. Early studiesof prognosis suggested that Crohn's diseasemight be more aggressive in the older patient,2but since the recognition of Crohn's colitis in1960,' and its distinction from ischaemiccolitis in 1963,4 more recent evaluation ofCrohn's disease in the older patient hassuggested a relatively benign course,5'6

although the total number of patients reviewedin these studies was small.We have undertaken an analysis of the

clinical presentation, course, and prognosis ina series of 62 consecutive patients aged 55years or over at diagnosis of their Crohn'sdisease between 1970 and October 1993.

Methods

The case records were reviewed of 62 patients(30 male) who were at least 55 years of age atpresentation, identified from among a series ofmore than 850 Crohn's disease patients underlong-term review. Only patients diagnosedfrom 1970 onwards were included in thisstudy, when the distinction between ischaemicand Crohn's colitis was fully appreciated, andalso because during this period the patients hadreceived consistent joint medical and surgicalcare.

Results

The mean age at onset, diagnosis, duration offollow-up and mortality are summarised intable 1. The male to female ratio was 1:1.2.The distribution of disease at presentation issummarised in table 2. Patients with terminal

Table 1 Mature onset Crohn's disease:patient characteristics (n=62)

Median Mean Range

Age at diagnosis (years) 67.1 68.35 56.9-92Age of onset of 66.6 67 56-82symptoms (years)

Interval onset to 6 16.3 0-186diagnosis (months)

Follow-up (years) 10 8 0.1-20Age at death (years) 76 75 63-95

Table 2 Mature onset Crohn's disease: dis-tribution of disease (n=62)

Number ofpatients (%)

Colonic disease (all sites) 38 (61)Total or extensive colonic disease 4 (6)Distal colonic disease 22 (35)Other colonic disease 3 (5)Left-sided colonic disease 8 (13)Terminal ileum + right colon 24 (39)Peri-anal disease alone 1 (1)

226 Walmsley, Gillen, Allan

+/- right colonic involvement at presentationconstituted 39% of the total. In an earlier studyof young patients aged 16 or less presenting toour unit from 1935 to 1980, 56% had terminalileal involvement.7 Total or extensive colonicinvolvement, defined as disease affecting atleast four of either the right, transverse,descending, sigmoid colon or rectum, oc-curred in only four of our patients. Distaldisease is confined to the rectum +/- thesigmoid colon. Left-sided colonic disease in-cludes the descending colon, sigmoid andrectum. Nineteen patients with colonic diseasealso had peri-anal disease (50%). Only onepatient with distal ileal disease also had peri-anal disease.The prevalence of coexistent disorders is

summarised in box 1. Diverticular disease wascommon (45%), particularly in patients withcolonic Crohn's disease, as noted previously,8although the prevalence is similar to that in thegeneral population of this age.9 The othercommon problem was thrombo-embolic dis-ease in 10% of patients, even after those withdisorders predisposing to thrombophlebitissuch as carcinoma, had been excluded. Thisis consistent with the recently observed in-crease in incidence compared with the generalpopulation. 10

In common with recent reports of theassociation of smoking with Crohn's disease,'128 (45%) of the total group were currentsmokers at diagnosis with a positive smokinghistory at some time in nearly two thirds(63%), higher than the general populationprevalence of around 50%.

There were 25 deaths in total during follow-up, 14 male, seven disease-related. Table 3summarises the standardised mortality ratiosusing sex- and age-specific mortality rates forEngland and Wales in 1990, assuming aPoisson distribution. There was no significantdifference for either sex.

Mature onset Crohn's disease:coexisting conditions (n=62)

* diverticular disease 28* deep vein thrombosis/pulmonary embolus 6* peptic ulceration 6* gall stones 5* seronegative arthritis 3* sacroileitis 2* osteoporosis 2* pyoderma gangrenosum 1* erythema nodosum 1

Box 1

Table 3 Mature onset Crohn's disease:mortality

Standard 95%mortality confidence

Observed Expected ratio intervals

Male 14 10.2 1.37 0.75-2.3Female 11 7.4 1.48 0.74-2.66

DISTAL ILEAL DISEASE +/- ASCENDINGCOLONThe diagnosis was based on radiological andhistological criteria in 12 (50%), on radiologi-cal grounds alone in five cases, and histologyalone in a further six. One patient had changeson barium enema examination consistent withthe diagnosis of Crohn's disease which wasconfirmed at laparotomy with characteristichistological changes in the lymph nodes.

PresentationMost patients presented with multiple symp-toms including recurrent abdominal pain andweight loss,10 ofwhom four also had diarrhoeaand two a palpable mass. Three patientspresented with weight loss, diarrhoea and a massbut no pain. Presentation with isolated symp-toms was uncommon and included diarrhoea(4), pain (2), weight loss (1), and symptomaticanaemia (1). Acute presentations includedperforation (2) and acute obstruction (1).

Medical managementBecause many patients came to resection soonafter presentation, medical treatment was onlyused on a few patients pre-operatively. Onepatient with a narrow segment of terminalileum was given prednisolone for six weekswith no improvement in symptoms and subse-quently underwent resection. Four patientswith long narrow strictures associated withrecurrent colicky abdominal pain received nopre-operative medical therapy. Four patientsrequired emergency laparotomy and in eightcases there was diagnostic doubt. One patientwith a long terminal ileal stricture at presenta-tion received medical therapy for nine yearsbefore undergoing resection. One patient withmultiple entero-enteric fistulae around theterminal ileum received anti-diarrhoeals for14.5 years before laparotomy revealed a dis-seminated carcinoid tumour. Five patientsreceived medical treatment only and did notrequire surgical intervention. Two required acourse of systemic steroids for symptomaticrecurrent disease following resection.

Surgical managementTwenty patients had a laparotomy, four as anemergency and all but one underwent resec-tion. The commonest indication for surgerywas recurrent subacute obstructive symptoms(47%), but the need to exclude malignancywas also an important indication for laparot-omy (37%). Seventeen came to resectionwithin three months of presentation. In theother three patients the interval was one, nine,and 14.5 years, respectively.

Recurrent disease was uncommon (sixpatients), but when it did occur it was usuallyshortly after resection (in four, less than oneyear), the others suffering symptomatic recur-rence at 2.5 and 12 years. However only onepatient requiring a second resection threemonths after his first.

MortalityDuring the period of follow-up there were eightdeaths; ischaemic heart disease (2), stroke (1),

Crohn's disease in the over-55s 227

disseminated carcinoma of the prostate (1),carcinoma of the oesophagus (1). One deathwas due to disseminated malignancy from aprimary carcinoma of the caecum found atinitial laparotomy in an elderly woman whohad had symptoms attributable to her Crohn'sdisease for more than 15 years. The exactcause of death was not fully established in twocases.One 68-year-old woman developed severe

post-operative hyponatraemia with markeddeterioration in her mental faculties such thatshe required rehousing in sheltered accommo-dation. Another woman with entero-entericfistulae for 14.5 years underwent ileocaecalresection for relief of obstructive symptomsand was found to have a disseminated carci-noid tumour at this site.

Current statusFifteen of the surviving patients are well, eightrequiring anti-diarrhoeal medication. Two arereceiving vitamin B 12. One man has developedmyelofibrosis and polycythaemia. The womanwith a carcinoid tumour is terminally ill with alarge pelvic mass.

COLONIC DISEASEOnly five of the 38 patients with colonicinvolvement were secondary referrals to theunit.

PresentationDiarrhoea was the commonest presentingcomplaint occurring in 26 patients, accompa-nied by blood in the stool (13), or weight loss(11), peri-anal symptoms (4) and vaginaldischarge (1). Five had altered bowel habitbut no diarrhoea, of whom four had blood intheir stools and two weight loss. Otherpresentations included, abdominal pain alone(2), pain with weight loss (1), peri-analdiscomfort and incontinence (1), and onewoman who presented with a right ileac fossamass and colicky pain due to terminal ilealdisease as well as bloody stools due toproctosigmoiditis.

DiagnosisThe diagnosis was established on histologicaland radiological grounds except for five basedon histological evidence alone, while two hadcharacteristic barium studies.

25-22 W At presentation

20 - 19 | Final extent

15-

10 897

5 4 43

0Left-sided Other

Distal Total/extensiveFigure 1 Distribution of colonic Crohn's disease

The initial and final distribution of disease issummarised in figure 1. In most patients therewas no change in the site or extent of thedisease once the diagnosis had been estab-lished, despite a mean follow-up of 10 years.Four patients with initial disease apparentlyconfined to the distal or left colon, and onewith patchy disease developed extensive colo-nic Crohn's disease.

Medical treatmentMore than half of this group received corticos-teroids (55%), with the addition of azathiopr-ine in five. While initially there was usually agood response to therapy, only six patients hada complete or sustained remission. Surgicalresection was needed in the majority to restoregood health. Rectal corticosteroids alone wereeffective in three, and 5-acetylsalicylic acidpreparations successfully treated five others.

Surgical treatmentMore than half had definitive colonic resection(excluding drainage procedures for sepsis orlocal peri-anal surgery). The median time fromdiagnosis to the first resection was 1.3 years(mean 1.6, range 0 to 6.6). The indications aresummarised in box 2. The initial surgery andfinal outcome is summarised in figure 2. Eightpatients required a second resection and tworequired a third. Local resection with sigmoidcolectomy was often undertaken in patientswith left-sided disease, but half of themeventually came to panproctocolectomy be-cause of symptomatic recurrence. Ileostomyalone was sometimes used in the severely illand those loath to accept panproctocolectomy,but the benefit was usually short-lived.The median interval between the first and

second resections was 0.6 years (mean 0.6,range 0.08 to 1.3).

MortalityThere were 17 deaths in this group during theperiod of follow-up. The unrelated deaths weredue to ischaemic heart disease (3), pneumonia(3), pulmonary emboli (2, associated withcarcinoma of the prostate and carcinoma ofthe bronchus), disseminated carcinoma of thethyroid (1) and septicaemia following emer-gency panproctocolectomy for perforation ofischaemic bowel (1). The seven disease-relateddeaths are summarised in table 4. Two of thefour post-operative deaths occurred amongpatients who were moribund at presentation.Two patients developed adenocarcinoma of

Indications for surgery in patientswith colonic disease (n=21)

* failed medical therapy 14* perforation 2* obstruction 2* entero-vesical fistula 1* severe peri-anal disease 1* ? malignancy 1

Box 2

228 Walmsley, Gilen, Allan

COLONIC DISEASE (38)

Surgical treatment (21) Conservative treatment (17)

Colectomy + Colectomy + Loop Sigmoid Sigmoid colectomy Sigmoid colectomywith IRA end ileostomy colectomy + colostomy + loop ileostomy

(5) ileostomy (1) (4) (5) (1) (2)

Neoterminal ileal Proctectomy Anastomotic Proctectomyresection (1) (1) resection (1) (1)

PPC(3) PPC(3) PPC(1) PPC1) PPC(1)

Figure 2 Treatment of colonic Crohn's disease. IRA=ileorectal anastomosis; PPC=panproctocolectomy; figures in parentheses are numbers ofpatients in each group

the large intestine, both unexpected findings. ADuke's A tumour was found at electivepanproctocolectomy in a 67-year-old man withleft-sided disease of five years duration. A 70-year-old man with extensive colonic disease foreight years had a polyp removed endoscopi-cally which histologically was a carcinoma in-situ.

Current statusOf the 21 patients alive when last reviewed, 13are on no medication at all, three take oral 5-acetylsalicylic acid preparations, including onewho also takes rectal steroid preparations, threereceive anti-diarrhoeal agents, two oral pre-dnisolone and one 61-year-old woman whoinitially had proctosigmoiditis is on a reducingcourse of steroids prescribed for an exacerba-tion of her disease which now involves thetransverse colon. A 73-year-old woman whotakes 2.5 mg of prednisolone daily to controlher diarrhoea also has myelodysplasia, requir-ing frequent blood transfusions.

Discussion

It is often difficult to establish the diagnosis ofCrohn's disease in the older patient becausethe disease is uncommon and the initialsymptoms are often mild or atypical. Themean interval between onset of symptomsand establishing the diagnosis in this series

was just over 16 months (range up to 15 years).The explanation for the delay in diagnosis wasusually that the symptoms were initially mildand non-specific and only became severeenough immediately before presentation towarrant medical attention.We found an equal male to female ratio in

patients with Crohn's disease in the olderpopulation and could not confirm earlierreports of a female predominance,"12-'4 per-haps because only small numbers of patientswere studied previously.

DISTAL ILEAL DISEASEAfter initially mild, often atypical, symptoms,patients with distal ileal disease then presentwith focal, clear-cut symptoms; usually sub-acute intestinal obstruction from underlyingstricture formation. A small but significantnumber of patients present acutely. Five of the24 patients (21.7%) with distal ileal diseasepresented with a right iliac fossa mass; thispercentage was higher than that in an earlierreview of patients of all ages with distal ilealCrohn's disease presenting to this unit between1970 and 88 (which includes some patients inthis present series), where only 12 of 139patients (8.6%) had such a finding.'5 Lapar-otomy may be needed when the diagnosis isuncertain, particularly if the radiological find-ings are difficult to interpret or the clinicalpresentation is unusual.

Table 4 Crohn's disease related deaths (colonic)Age at On systemic steroidsdeath (years) Sex Cause of death at time of death

82 F Septicaemia following elective panproctocolectomy No72 M Septicaemia following emergency colectomy for obstruction Yes63 M Septicaemia from septic arthritis complicated by short bowel Yes

syndrome73 M Septicaemia following colectomy for perforation Yes67 M Bronchopneumonia and renal failure following elective No

panproctocolectomy81 M Septicaemia following perineal gangrene arising from a perineal No

abscess85 F Small bowel volvulus 11 years after an elective panproctoco- No

lectomy, plus an incidenal carcinoma bronchus

Crohn's disease in the over-55s 229

Sunnary points

Distal ileal disease* is the most common site* delay in diagnosis means characteristically

acute presentation* recurrent disease is uncommon and rarely

requires further resection

Colonic disease* usually distal* medical management is often difficult* operative intervention in 55% of cases, ofwhom over half require panproctocolectomy

Box 3

The initial resection rate (79%) is similar tothat reported in earlier reviews of 45 patientsfrom this unit (77%)5 and consistent withoperative rates for all ages, both here (79% of139 cases)15 and in the Cleveland Clinic series(91.5% of 246 patients).16 Recurrent distal ilealCrohn's disease requiring resection is rare inthe older patient. Only one of the 19 patientstreated by resection, with a mean follow-up ofeight years, required a second resection com-pared to 33 out of 110 patients in an earlierstudy of patients of all ages treated inBirmingham with a similar mean follow-up of10 years.15

COLONIC DISEASEThis study confirms that colonic Crohn'sdisease in the older patient is usually confinedeither to the rectum or the sigmoid colon,while total or extensive colitis is uncommon.However, it cannot be assumed that theapparently limited disease is associated with agood prognosis or benign outcome. Severeexacerbations of apparently distal disease needintensive medical treatment and even soresponse is often poor, perhaps because of an

additional ischaemic element in the affectedcolon.We found that local resection of colonic

disease in the older patient often only broughttemporary improvement and early recurrentdisease in the residual colon was common, sothat an initial local resection was often followedby panproctocolectomy. Perhaps in somepatients this might have been undertakeninitially. Emergency surgical treatment in theolder patient was associated with a pooroutcome so that, wherever possible, electivesurgery is preferable and likely to minimisemorbidity. It follows that better surgical resultscould be achieved at the price of some patientsaccepting a permanent stoma.

This study suggests that colonic Crohn'sdisease is not the benign disease that recentreports have suggested.5'6'"2 While the patientshave predominantly distal colonic disease, theoverall operative rate in this study of 55% issimilar to that seen in 360 patients of all agespresenting to our unit between 1944 and 1986.Andrews et all7 showed an operative interven-tion rate of 62% at 10 years and a population-based study from Denmark found a rate of55% after a mean follow-up of 10 years.'8 Overhalf of those who underwent resection even-tually required a panproctocolectomy; this issimilar to 41% of 270 surgically treatedpatients of all ages with colonic Crohn'sdisease'2 and seven of 15 patients describedby Tchirkow from the Cleveland Clinic.'4 Thefact that only three of the 10 panproctocolec-tomies were performed as the initial operationsuggests that a more radical approach might beappropriate when faced with persistent severesymptoms, even when the disease is confinedto the distal colon. While the high morbiditymay in part reflect the inherent biases of ahospital-based series, there were only fivesecondary referral patients included in thecurrent study.

1 Rose JDR, Roberts GM, Williams G, Mayberry JP, RhodesJ. Cardiff Crohn's disease jubilee: the incidence over 50years. Gut 1988; 29: 346- 50.

2 de Dombal FT, Burton IL, Clamp SE, et al. Short termcourse and prognosis of Crohn's disease. Gut 1974; 15:435-43.

3 Lockhart-Mummery HE, Morson BC. Crohn's disease(regional enteritis) of the large intestine and its distinctionfrom ulcerative colitis. Gut 1960; 1: 87-105.

4 Boley SJ, Schwartz S, Lash J, et al. Reversible vascularocclusion of the colon. Surg Gynecol Obstet 1963; 116: 53-60.

5 Fabricius PJ, Gyde SN, Shoulder P, et al. Crohn's disease inthe elderly. Gut 1985; 26: 461-5.

6 Harper PC, McAuliffe TL, Beeken WL. Crohn's disease inthe elderly. A statistical comparison with younger patientsmatched for sex and duration of disease. Arch Intern Med1986; 146: 753-6.

7 Puntis J, McNeish AS, Allan RN. Long term prognosis ofCrohn's disease with onset in childhood and adolescence.Gut 1984; 25: 329-36.

8 Ritchie JK, Lennard-Jones JE. Crohn's disease in the distallarge bowel. Gastroenterology 1976; 11: 433-6.

9 Hughes LE. Post-mortem survey of diverticular disease ofthe colon. Gut 1969; 10: 336-51.

10 Webberley MJ, Hart MT, Melikian V. Thromboembolismin inflammatory bowel disease; role of platelets. Gut 1993;34: 247-51.

11 Tobin MV, Logan FA, Langman MJS, McDonnell RB,Gilmore IT. Cigarette smoking and inflammatory boweldisease. Gastroenterology 1987; 93: 316-21.

12 Softley A, Myren J, Clamp SE, Bouchier IAD, WatkinsonG, de Dombal FT. Inflammatory bowel disease in theelderly patient. Scand Jf Gastroenterol 1988; 23 (suppl 144):31: 27-30.

13 Lee FI, Giaffer M. Crohn's disease of late onset inBlackpool. Postgrad Med J 1987; 63: 471.

14 Tchirkow G, Lavery IC, Fazio VW. Crohn's disease in theelderly. Dis Colon Rectum 1983; 26: 177-81.

15 Andrews HA, Keighley MRB, Alexander-Williams J, AllanRN. Strategy for management of distal ileal Crohn's disease.BrJ Surg 1991; 78: 679 - 82.

16 Farmer RG, Whelan G, Fazio VW. Long-term follow up ofpatients with Crohn's disease. Relationship between theclinical pattern and prognosis. Gastroenterology 1985; 88:1818-25.

17 Andrews HA, Lewis P, Allan RN. Prognosis after surgeryfor colonic Crohn's disease. Br J Surg 1989; 76: 1184- 90.

18 Binder V, Hendriksen C, Kreiner S. Prognosis in Crohn'sdisease based on the results from a regional patient groupfrom the County of Copenhagen. Gut 1985; 26: 146- 50.