Prognosis in Ulcerative Colitis

Inflammatory Bowel Disease

0025-7125/90 $0.00

+ .20

Prognosis in Ulcerative Colitis

Olle Brostrom, MD, PhD* "Prognosis applies to the act or art of foretelling the course and the termination of a disease; the term usually implies a correct diagnosis and knowledge of how the disease will affect the patient as it runs its course and how it will end." WEBSTER·S NEW DICTIONARY OF SYNONYMS"

There are several reasons why knowledge of the prognosis of a disease such as ulcerative colitis is clinically important. It enables the physician to convey a proper and balanced picture to the patient and his or her relatives in order to help them cope with future problems; moreover, it helps the physician emphasize the relevant facts to the individual patient. For instance, the colon cancer risk in ulcerative colitis and its management needs more attention in a patient with a total colitis whereas the patient with proctitis can be reassured of the low risk in his or her case. Knowledge of factors influencing prognosis may also lead to improved treatment and indications for future research. The concept of prognosis can be seen as outlined in Figure 1. The "natural history of disease" as it occurs is modified by various factors. These factors will affect "the quality of life" and potentially life expectancy and cause of death. The aim of this review is to examine the various aspects involved in determining the prognosis. Our knowledge of the natural history in ulcerative colitis is briefly considered and followed by a review of studies on the mortality of this condition. The importance and impact of the factors influencing prognosis are then discussed.

NATURAL HISTORY OF ULCERATIVE COLITIS Ulcerative colitis is a disease with a wide clinical spectrum affecting all age groups, both sexes alike, with a predominance in the young and middle aged. It is characterized by an insidious onset of eventually bloody *Assistant Professor, Karolinska Institute; and Staff, Section of Gastroenterology, Second Department of Medicine, Sodersjukhuset, Stockholm, Sweden

Medical Clinics of North America-Vol. 74, No. 1, January 1990

201

202

OLLE BROSTR0M

diarrhea in a majority of patients and a more acute onset including the fulminant form in a minority. The continued course of disease is either of a chronic intermittent type with recurrent attacks with great individual variation or, more rarely, a chronic continuous type with constant symptoms. Factors affecting recurrence of attacks are not completely known but may include infectionsparticularly upper respiratory-emotional stress, and drugs. 25, 33 A small group is seen as having only one attack and no recurrence of symptoms over many years. This group probably contains misdiagnosed patients with specific causes such as various infectious agents. Furthermore, a small number of patients will have their diagnosis changed to Crohn's colitis during the course of the disease, whereas the opposite reversion of diagnosis is rare. Finally, some patients remain with features of both forms of colitis. A majority of the patients have a distal disease at the first attack whereas the rest have a more proximal extent including total involvement. Followup studies reveal that up to one fifth of patients with distal disease progress to have increased extent during the disease course. Important factors as family history and extraintestinal manifestations are considered elsewhere in this volume. The colon cancer risk will be discussed further on in this article.

PROGNOSIS: POPULATION STUDIES There are many studies available on the prognosis in ulcerative colitis reflecting the experience from Europe, United States, and Israel over the past 50 years. 7 , 9, 10, 39 In this review, an emphasis is put on the results from five recent population studies where groups of un selected patients from defined geographic areas in the United Kingdom, United States, and MODIFYING FACTORS Treatment: Surgical/Medical Pregnancy

~

NATURAL HISTORY Age at Onset Severity of First Attack Course of Disease Complications: Acute attack Liver disease Colon cancer

~

QUALITY OF LIFE Psychosocial Adjustment Postoperative

LIFE

~ EXPECTANCY

I

BIR~:-:O:-:-N:-:S-=E=T-------D--E...JATH OF ULCERATIVE COLITIS

Figure 1. Concept of prognosis,

PROGNOSIS IN ULCERATIVE COLITIS

203

Scandinavia have been employed. 4, 17,38,40,44 Although there are considerable differences in presentation and analysis, the inception cohorts allow a reasonable base for comparison. The work from Oxford by Edwards and Truelove 9 is included since it constitutes one of the first major studies on prognosis in ulcerative colitis. In order to make comparisons, feasible results from the studies are seen in Tables 1 to 4, Numbers have been derived from figures and tables in the original papers and should be considered as approximate, Aspects on Methodology The problems and pitfalls in conducting studies of this kind have been discussed in detail by Sackett and Whelan,37 and the actuarial analysis method by Devroede. 8 The ideal study would be a prospective, covering a long period of time with a large number of patients available for detailed analysis. However, studies over long time periods usually have to be retrospective, The main advantage is the possibility of including a large number of patients with the possibility of detecting trends-that is, survival patterns-not possible in short-term studies. The main disadvantages include problems of proper definitions, incomplete case records, and old methods of diagnosis. Moreover, the results of long-term studies, in particular on prognosis, may reflect the outcome of different influenCing factors such as new drugs and surgical methods during the study period, This may make it difficult to deduct the results for application to present clinical problems, Therefore, a difference in size and length of studies as that discussed below may emphasize different aspects of prognosis. Study Populations (Table 1) Except for the Oxford study, all groups include an unselected colitis population from defined geographic areas representing urban (Uppsala, Stockholm, Copenhagen) as well as mixed (Northeastern Scotland, Rochester, MN) catchment areas. 4, 9, 17, 38, 40, 44 They all include a reference population for analysis. The study populations vary from 182 to 1274 patients and the study periods vary from 10 to 44 years. Except for the Copenhagen study, the studies are all retrospective, 17 Patients At diagnosis the distribution of patients with proctitis, left-sided, and extensive/total colitis varied conSiderably. The Copenhagen and Scottish groups included only 16 per cent with total colitis whereas Uppsala, Rochester, and Stockholm had 28 to 36 per cent.4, 17, 38, 40, 44 Apart from the possibility of true differences in disease spectrum, the differences probably represent varying success in including patients with distal disease. These differences are important when considering the cumulative survival probability. Cumulative Survival Probability The excess mortality compared with the expected is most pronounced in Uppsala when all patients are considered. ll The difference at 5 years is 9 per cent increasing to 23 per cent at 25 years' duration of disease. The

t-:l

~

783 182

1274

1985 1987

1987

Copenhagen, Denmark Rochester, Minnesota, USA Stockholm, Sweden

537 41 24

1955-1979

23 47 74

1960-1978 1935-1979

1938-1962 1945-1974 1967-1976

titis

66

37

41

43 21 15

*Actual cumulative survival, no expected survival estimated. tOnly men over 40 years. NS = not stated.

UK

220

624

1963-1964 1976 1983

Oxford, VK Uppsala, Sweden Northeastern Scotland,

PATIENTS TlME PERIOD

PUBLISHED

STUDY REGION

36

16 33

36 28 16

total 22 11

10

25 11

(years)

15

7 4

4

2

6

15

4% at 2 yearst No difference at 20 years

No difference at 10 years

15 9

5

23

25

20 18 (all men) 11 15 (all women)

32 15

20

3

9

90 20

NS NS 6

68 25 16

5

7

65 15 16

10

45 25

15 (years)

10

12

45' 27

20

13

25

COLITIS, PAT. Ol"LY VS EXPECTED (PER CENT)

!eftsided

ALL PATIENTS VS EXPECTED (PER CENT)

DiAGNOSIS (PER CENT)

proc-

DIFFERENCE IN CUMULATIVE SURVIVAL TOTAL

DIFFERENCE IN CUMULATIVE SURVIVAL

EXTENT OF DISEASE AT

Table 1. Cumulative Survival in Patients with Ulcerative Colitis

PROGNOSIS IN ULCERATIVE COLITIS

205

trend is similar in Stockholm, but less pronounced, plU1icularly among women. 4 Thus the excess mortality seems to increase with duration of disease. In Copenhagen only men over the age of 40 and during the first and second year after diagnosis had an excess mortality of 2.1 and 1.5 per cent. 17 Neither the Scottish nor the Rochester study showed any difference with the expected when all patients were considered. 40. 44 When patients with total colitis at follow-up are analyzed separately, the trends of increased excess mortality is emphasized in Stockholm and Uppsala and it also becomes evident in Scotland. 4. 38. 40 The three studies further show an accentuated excess mortality during the first years after diagnosis. This is, in part, probably a reflection of the outcome of the first attack discussed below. It is worth noting that the trends in cumulative survival probability are similar to the Oxford figures but less pronounced. In Oxford, the survival probability for total colitis patients was extremely poor with only 45 per cent surviving 15 years' duration. This series included patients from the era before sterojd treatment was introduced. Age-related Mortality In Oxford and Scotland, a marked increase in mortality rate during the first attack and relapse attacks was noted in patients aged over 60 and 70 years. 9. 40 In Stockholm there was an increasing excess mortality among patients over 45 years old (84 deaths observed versus 36 expected).4 Longterm follow-up in Uppsala showed no such excess mortality in the elderly but rather among the age groups 15 to 44 years where all deaths were related to the colitis. 38 The overall impression remains that the risk of dying increases with old age in patients with ulcerative colitis. The magnitude of the excess mortality is difficult to compare in different studies since some refer to first attack cases only. Moreover, the cause of death is not only colitis-related as will be discussed below. Mortality and Time Periods In Oxford there was an almost threefold improvement in mortality rate (24 per cent 1938 to 1952 versus 9 per cent 1953 to 1962) after the introduction of steroid treatment. 9 In Stockholm, where steroid treatment was available from the beginning, an improved survival was also noted when 1955 to 1969 was compared with 1970 to 1979 (observed/expected ratio, 2.8 and 1. 75; P
206

OLLE BROSTROM

lowest in Italy and Israel. 43 The mortality was closely correlated to the incidence and it was suggested that this reflected the mortality rate rather than differences in the outcomes of treatment. The sharpest decline in mortality during the 1970s and early 1980s was seen in those countries with a high mortality rate. There are no obvious explanations for this observation. Newcombe et aF8 also observed the high mortality in Northwestern Europe and a low mortality in the Southern hemisphere: Australia, New Zealand, and South Africa. They also observed a general decline in mortality that did not differ greatly among different racial groups in the United States, and South Africa. Little data are available from other parts of the world but D'Oliveira et al6 noted a high mortality in Venezuela, 16.4 per million population per year, compared with 7.4 in England and Wales. Even here a sharp decline was observed over time. Although there seems to be a varied mortality rate among ulcerative colitis patients in different parts of the world, a general time trend of a decline in the rate is evident. CLINICAL FEATURES The First Attack and Its Outcome In Oxford the severity and extent of the first attack significantly influenced the prognosis (Table 2).9 The majority of the attacks were mild and carried a low mortality; however, the 11 per cent of the patients with a severe attack had a 26 per cent mortality. The Scottish study confirmed the distribution of the severity of attacks seen in Oxford. 9, 40 In Uppsala the severe attacks exclusively affected those patients with total colitis and none with left-sided or distal disease. 38 The three studies show quite similar outcomes of the first attack where 3 to 6.9 per cent die, 86 to 95 per cent go into remission or improve, and 0.8 to 5.5 per cent underwent colectomy. The slightly better results from Scotland is probably a reflection of the unselected patient group, as well as the more recent time period studied, implying an improved treatment. Course of Disease The further course of disease is not uniformly described in the studies but permits some comparison. In Oxford a majority of patients (64 per cent) ran a chronic intermittent and 7 per cent chronic continuous course, whereas 8 per cent died in the first attack offulminant disease. 9 Surprisingly, as many as 18 per cent had only one attack which may reflect some inability at the time of diagnosis to find various forms of infectious colitis possible to diagnose today. In Uppsala the figures are quite similar except for a lower rate of single attack cases. 38 The same trends, although not presented in detail, are shown from Rochester. 44 The Copenhagen figures are particularly interesting since the patients have been monitored prospectively allowing for repeated evaluations by the same observers.17 After 6 to 7 years' duration, a steady pattern is reached where approximately 55 per cent of patients ran an inactive, 35

0...1

to <:>

Mortality

'At 10 years. NS = not stated.

Rochester, Minnesota, USA

Northeastern Scotland, UK

Severe 6 23

NS

Moderate 26 2.9

NS

7

44

NS

Moderate 23 27

Moderate 23 2.4

Severe 26

11 26

Copenhagen, Denmark

Distrihution Total Left Proctitis Mortality

Mortality

Severe

Stockholm, Sweden

Uppsala, Sweden

Oxford, UK

FIRST ATTACK (PER CENT)

Mild 68 1.4

Mild 50 72 26 0

Mild 64 0.4

NS

Died Complete remission Unchanged Colectomy

NS

NS

Died Complete remission Improved Unchanged Colectomy

Died Complete remission Improved Unchanged Colectomy

3 87 8 3

6.2 46.9 33.8 7.6 5.5

4.5 85.9 8.2 0.6 0.8

OUTCOME OF FIRST ATTACK (PERCENT) COURSE OF DISEASE (PER CENT)

Transient Intermittent Unremitting

NS

At 6 years' duration Inactive course Intermittent course Chronic continuous Other

NS

Acute fulminant Chronic intermittent Chronic continuous One attack Other

Acute fulminant Chronic intermittent Chronic continuous One attack Other

Table 2. Clinical Features in Ulcerative Colitis

27.5 65.4 5.5

55 35 8 2

9.2 74 7.8 6.4 2.6

8 64 7 18 3

l00~29

Distal Left Total

100~30'

Distal Left Total

NS

NS

Distal Left Total 100 --> 19 --> 15 100 --> 18

Distal Left Total 100 --> 8--> 2 100 --> 14

EXTENT (PERCENT)

SPREAD OF

208

OLLE BROSTR0M

per cent an intermittent, and 8 per cent a chronic continuous course. The large proportion of patients with inactive disease could be due to the fact that many patients only have distal or left-sided disease and would run a more benign course. This explanation, however, would be in contrast with the Oxford figures where patients with left-sided and distal disease had approximately the same proportions of different clinical activity as those with total colitis. Spread of Extent of Disease

It is difficult to record the change in extent of disease, since there are differences in frequency, quality, and interpretation of barium enemas. The dynamics of progress and regress of mucosal changes seen at colonoscopy (recently described) adds to the problem of evaluating extent. 30 The Oxford and Uppsala series show a spread of distal disease to left-sided in 8 to 19 per cent of cases and further spread of 2 to 15 per cent to total colitis. 9 , 38 In patients with initial left-sided disease, 14 to 18 per cent of patients progress to total colitis. In Scotland, an increase from distal to more extensive disease (not specified) occurred at 5 and 10 years' duration in 12 and 30 per cent of the patients40 ; this underlines the time factor as important. We lack proper data today whether this trend of spread of disease continues or levels off with time. This would have interesting implications to treatment practice. Will sulfasalazine prophylaxis influence the spread of disease and, thus, change the disease pattern? Mortality Pattern, Cause of Deaths In Table 3, the number of patients dead at follow-up are listed in the different series. In Uppsala and Stockholm 40 to 50 per cent of the deaths were attributable to ulcerative colitis whereas only 16 per cent were in Rochester. 4, 38, 44 This low figure may have been influenced by the fact that also probable cases of ulcerative colitis were included. In Uppsala and Stockholm the causes of deaths unrelated to colitis Table 3. Death Among Ulcerative Colitis Patients NOT RELATED TO ULCERATIVE COLITIS

RELATED TO ULCERATIVE COLITIS

COLON CANCER DEATHS

Oxford, UK 123/624 (19.7) Uppsala, Sweden 62/220 (28.1) Northeastern NS Scotland, UK 671783 (8.6) Copenhagen, Denmark

NS 30 NS

NS 32 NS

17 5 NS

NS

NS

NS

Rochester, Minnesota, USA Stockholm, Sweden

NO. OF DEATHS AT FOLLOW-UP (PERCENT)

(PER CENT OF ALL DEATHS)

(14) (8)

37/182

(20.3)

31

6

3

(8)

109/1274

(8.6)

41

68

6

(5.5)

NS = not stated.

PHOGNOSIS IN ULCEHATIVE COLITIS

209

were analyzed and no difference in mortality pattern with that of the general population was found. 4. 38 In Stockholm, however, an excess mortality compared with the expected was found in the group of patients whose deaths were unrelated to colitis. 4 This was interpreted by the authors as a combined effect of increasing age and ulcerative colitis, causing an additive effect on the normal disease spectrum. Except for Uppsala where a similar but not pronounced tendency was found, this has not been reported in other unselected series, but merits further studies. In contrast, Cyde et aI, 14 in a careful analysis of mortality in a selected series of ulcerative colitis patients from the Birmingham area, found a deficit of deaths from diseases of the circulatory and respiratory systems. In view of recent data on smoking habits in ulcerative colitis patients, it is tempting to speculate if this could have affected the results in this study. In fact, in a subsequent case control study, the authors found that ulcerative colitis patients smoked less and had lower systolic and diastolic blood pressures-particularly among those with extensive disease and after panproctocolectomy.13 The number of deaths attributable to colon cancer varied between 5.5 and 14 per cent of all deaths in the different series. These figures indicate the size of the group that potentially could be influenced by cancer prophylactic measures such as surgery and surveillance. The prognosis of colitis cancer patients has been analyzed by Cyde et al. 15 In a series of 35 colitis cancers from the Birmingham area, they found that the 5-year survival of 33.5 per cent was not different from that of the expected survival in idiopathic colon cancers. The colon cancer incidence in different studies is discussed below. Complications Only potentially fatal local and systemic complications affecting prognosis, except colon cancer, are discussed here. The frequency from some of the series discussed previously are listed in Table 4. In Oxford and Uppsala, perforation was found at the same rate carrying a high mortality. 9. 38 Interestingly, several of these cases were diagnosed at postmortem examinations. Dilatation in itself and particularly combined with perforation also carried a high mortality whereas massive hemorrhage per se may be less dangerous-although four patients in Stockholm had bleeding cited as a major cause of death. 4 In the clinical setting, several serious complications interact reflecting the difficulty of identifying one single responsible factor. This is probably pertinent when considering pulmonary embolism responsible for three deaths in the Stockholm series. 4 Liver disease is a complex problem to analyze in ulcerative colitis since diagnostic procedures are not used uniformly in all patients with signs of liver disease. Incidence of liver disease therefore has to be viewed with caution. In Stockholm, three patients died of liver failure. 4 In a separate detailed analysis from this series, 150 of 1274 patients had hepatobiliary disease diagnosed (unpublished data). They included 11 with pericholangitis, two primary sclerosing cholangitis, two primary biliary cirrhosis, three neoplasms, and five with chronic active hepatitis. At follow-up, a majority of patients showed no progress of disease indicating that most patients run a benign course. It seems that in an unselected colitis population, severe

~ .....

o

3.2 3.2 0.4

*Only deaths recorded. NS = not stated.

Oxford Uppsala Stockholm

Per cent

7

20

n

PERF.

(15) (5) (5)

(n = dead) 1.6 2.2 0.3

Per cent 5

10

n

DILAT.

(4)

(3)

(NS)

(n = dead) 3.4 3.6 0.3

Per cent

8

21

n

(3) (4)

(NS)

(n = dead)

MASSIVE HEMORRHAGE

Table 4. Complications in Ulcerative Colitis (Except Colon Cancer)

1.6 4.1 0.2

10 9

n

(3)

(NS) (NS)

(n = dead)

PULMONARY EMBOLISM

Per cent

PROGNOSIS IN ULCERATIVE COLITIS

211

liver disease is found infrequently. However, when specific entities are analyzed the prognosis seems worse. In a long-term follow-up of 45 patients with primary sclerosing cholangitis, Aadland et aP estimated a reduced life expectancy of about 30 years. A small excess mortality in hepatobiliary carcinoma appeared in Birmingham. 14 Colorectal Cancer Risk An increased colon cancer incidence in ulcerative colitis has been known for many years. 3 This has prompted a search for a clinical approach such as cancer surveillance, and is discussed elsewhere in this volume. Estimating colon cancer incidence in ulcerative colitis highlights the requirement of using un selected patient populations in order to obtain adequate incidence data. 37 Data from some studies where emphasis is on defined areas and populations are summarized in Table 5. Despite relatively large cohorts and long follow-up, small numbers of cancer cases remain for analysis. Subsequently a joint study of three cohorts from West Midlands, Oxford, and Stockholm has been done showing a modest but remarkably similar increase in incidence in the three areas (Fig. 2).16 Further analysis showed there was also a slight excess risk of cancer in left-sided colitis, as has been suggested by Greenstein et al. l l However, reservations on the diagnostic precision in defining left-sided colitis were noted in the joint study. Except for duration and extent of disease, young age at onset has been suggested as a risk factor for cancer development; however, no conclusive answer has been given yet. In this joint study, however, it seemed as if the age of the patient at cancer diagnosis was more important than the actual duration of disease. The maximum risk seemed to be around 50 years. Supportive evidence from other trials is clearly needed before any changes in the basis of screening should be considered. In conclusion, it appears that complications in ulcerative colitis affecting prognosis are those connected with the acute attack, liver disease, and colon cancer. The prognosis in the acute attack has been improved greatly over the last decades as discussed below. The impact on the incidence of colon cancer with measures such as cancer surveillance has yet to be shown although some evidence of impact has appeared. 19 A high colectomy rate has been claimed to account for a low incidence of colon cancer in Copenhagen. 17 It seems clear that the colon cancer risk is the single most important risk factor affecting long-term prognosis where it accounts for 5 to 14 per cent of all deaths. Pregnancy Pregnancy in women with ulcerative colitis raises important questions by patients and physicians alike. What is the risk to the fetus? Will the disease itself deteriorate? Is the medical therapy safe for the fetus/infant? In a survey by Willoughby and Truelove,50 the overall outcome of pregnancies in 147 women was similar to that of the general population. Those patients with active disease had a somewhat lower chance of producing a normal live baby. In a case-controlled retrospective analysis, Porter and

tQ ..... tQ

*18 years' duration. NS = not stated.

Copenhagen Rochester Uppsala West Midlands Oxford Stockholm

71783 3/179 7/220 38/823

PATIENTS

nffotal n

6 NS NS NS

A t Diagnosis of Colitis

7

31

NS 2

At Follow-up

TOTAL OR EXTEl"SIVE COLITIS

o

2.5

o

5 0.8 1 5.5 0.7

10 1.1 1 8.4 3.4

15

1.4* 2.5 11.7 7.2

20

(Years' duration)

CUMULATIVE INCIDENCE

Table 5. Incidence of Colorectal Cancer in Ulcerative Colitis

11.6

15.3

(all patients) (all patients) (extensive colitis) (extensive colitis)

25

213

PROGNOSIS IN ULCERATIVE COLITIS

lOO

Cumulative proportion with cancer

10

( 'I.)

Birmingham _____

··0·:

o

10

20

Oxford

0····0

Sweden

.- --.

30

Years from onset of U C

40

Figure 2. Extensive colitis. Cumulative proportion of patients with colorectal cancers by referral center (log scale). (From Cyde SN, Prior P, Allan RN, et al: Colorectal cancer in ulcerative colitis: A cohort study of primary referrals from three centers. Cut 29:213, 1988; with permission.)

Stirrat32 confirmed the overall good outcome as well as Jiirnerot 21 in a review of 1155 pregnancies from 18 literature reports. In the Oxford study, 30 per cent of patients with inactive disease at the onset of pregnancy relapsed and 14 per cent did so in the first trimester. 50 Fifty-two per cent of those with active disease at conception showed no change or worsened whereas in the puerperium the figure was only 7 per cent. Treatment with sulfasalazine did not seem to affect the outcome of pregnancy. A small excess of congenital abnormalities was considered likely due to clinical activity of the disease itself. 50 Although sulfasalazine enters breast milk and theoretically could cause kernicterus, there is no evidence of such a risk in a healthy normal child. 20 In a survey of 132 ulcerative colitis patients, Mogadam et aP6 found no increase in fetal complications compared with the general population. In conclusion, it seems justified not to discourage pregnancy in patients with ulcerative colitis but, if possible, advise that conception be delayed until the disease is inactive. Treatment should be as in the nonpregnant state and continue in the postpartum period with no substantial risk to the mother, fetus, or the newborn infant. Treatment and Prognosis Surgical and medical treatments have greatly affected the prognosis in ulcerative colitis, particularly in those with a severe course of disease. This probably accounts for most of the improvement during the last decades.

214

OLLE BROSTROM

Table 6 summarizes the outcomes of surgery in some large studies. They differ in selection of patients but all show a higher mortality in acute cases (5.3 to 30 per cent) than in elective cases (1.6 to 6 per cent). In Copenhagen and Northeastern Scotland, an estimate of the cumulative colectomy rate shows that surgery is employed most frequently during the first years of disease, 18 and 8 per cent, respectively, at 5 and 30 years and 11 per cent at 10 years. 17. 40 This means that after approximately 5 years the risk of having surgery is constant. The probability of surgery is naturally higher in those with severe and total colitis, 35 to 50 per cent in Scotland. 40 The postoperative complications are high. In Stockholm, there were 56 per cent major complications in emergency operations and 25 per cent in elective operations (unpublished data). This is in accordance with other studies. 2, 22, 35 Old age also affects the outcome of surgery with an increased mortality. 2, 4, 14, 36 There has been a time trend of decreased postoperative mortality shown in other studies. 2 This improvement is probably a result of intensive preoperative treatment as outlined by Truelove and Jewell,45 but also influenced by a better understanding of the importance of fluid and electrolyte balance, the prophylactic antibiotic treatment, the antithrombotic treatment, improvement in anesthesiology with less lung complications, and the introduction of intensive care units. The surgical procedure itself could influence the outcome, Lee and Truelove 23 found that subtotal colectomy carried a higher mortality in

The Arabian nights 2

inflamed rectum. In a Danish study with the same findings, it was suggested that the excess mortality was caused by a tendency to treat more severely ill patients with subtotal colectomy. 22 Recent experience with new surgical technique with the pelvic pouch and ileoanal anastomosis shows low or no mortality in several series. 49 Pouchitis and sometimes stricture formation have been problems and the long-term functional results are not yet completely clear. The influence of medical treatment on prognosis is mainly due to effects on the acute attack Table 6. Outcome of Surgery for Ulcerative Colitis in Present and Previously Published Series

AUTHORS

Lennard-Jones et al (1960) Brooke et al (1960) Daly et al (1968) Ritchie (1972) Ritchie (1974) Koudahl et al (1976) van Heerden et al (1978) Albrechtsen et al (1981) Goligher (1984) Hawley (1988) Stockholm (1989)

NO, OF

POSTOPERATIVE MORTALITY

PATIENTS

(PER CENT)

TIME PERIOD

Acute

Elective

Acute

Elective

Overall

(1951-1959) (1955-1962) (1947-1966) (1955-1969) (1967-1972) (1961-1974) (1961-1975) (1969-1978) (1955-1979) (1963-1986) (1955-1984)

26 62 149 118 131 35 47 132 184 178 185

113 142 293 128 140 66 312 158 320

30 16 24 15 46 23 25 5,3 14 7,4 9,2

4.4 6.0 3,4 1.6 2.1 6,0 1.9 2,5 2,8

9,3 9,3 10 8,1 24 12 8,9 3,7 6,7

1.7

4,5

301

PROGNOSIS IN ULCERATIVE COLITIS

215

and prophylaxis with sulfasalazine. The prophylaxis has been suggested as a possible explanation to a fall in patient years with severe attacks observed in a multinational survey.41 However, problems oflow compliance has been noted by van Hees et al. 47 Fortunately, serious side effects of medical treatment potentially affecting prognosis are rare. They include rare cases of steroid-induced psychiatric disorders. Recently, steroid-related osteonecrosis in inflammatory bowel disease has been described. 46 Serious side effects with sulfasalazine include unusual cases of agranulocytosis, pneumonitis, and massive hepatic necrosis. 34 Psychosocial Prognosis The patients have to adapt to the disease itself and, when operated, to the particular procedure such as incontinent or continent ileostomy, ileorectal anastomosis, or pelvic ileal reservoir. Already in the Oxford study it was noted that nearly 90 per cent of the patients were living a normal (69 per cent) or essentially normal (19 per cent) life. 9 In Copenhagen, 90 per cent of the patients were fully capable of work except for the first few years after diagnosis. 17 In a separate analysis with controls, the patients were similar in marital status, frequency of severe family or sexual problems, leisure activities, physical and earning capacity, incidence of mental disorders, and drug and alcohol abuse. IS The patients belonged to higher socioeconomic groups than controls which they did even before diagnosis. Adaptation to conventional ileostomy was generally good in a large questionnaire survey of 1803 patients by Morowitz et aP7 although one fourth required ileostomy revision and 230 had reoperation for bowel obstruction. Physical and emotional health were considered good or excellent in almost 80 per cent of 279 patients from the Cleveland Clinic. 24 A study of sexual problems among married ileostomists in England5 showed that 12 per cent ascribed marital tension, unhappiness, or even separation to the presence of the stoma. After rectal excision, almost one third of the men and women reported sexual dysfunction. The need for proper counseling was stressed since only 7 per cent felt they received helpful advice from any source. In other studies with continent ileostomy, women reported a clearly enhanced sexual function. 29 Fifty-five patients from St. Marks converting from an incontinent ileostomy to a pelvic reservoir preferred this procedure in 87 per cent. 31 While most felt there was no significant disadvantage with the reservoir, 20 per cent regarded the long convalescent period and 18 per cent the requirement for catheterization as drawbacks.

SUMMARY AND CONCLUSIONS In unselected prognosis is good appreciable excess after diagnosis and

patient populations with ulcerative colitis the overall and has improved over the years. There is still an mortality, however, particularly during the first years it tends to increase with duration of disease. Patients

216

OLLE BROSTROM

with severe attacks, total colitis, and high age at diagnosis are particularly at risk. The disease runs an inactive or intermittent course in the majority of patients, although up to one fifth of the patients have a progress of the original extent of the colitis. Worldwide, there has been a time trend of decreased mortality possibly affecting younger patients in particular. Complications of the acute attack with or without surgery, liver disease, and colon cancer account for the major part of the colitis-related deaths whereas the mortality pattern in other respects does not differ significantly from that of the general population. The colon cancer incidence seems lower than previously reported but still accounts for approximately one tenth of all deaths. If this figure can be improved with cancer surveillance and prophylactic colectomy seems probable but remains to be shown. Pregnancy, if planned, should be encouraged when the patient is in remission although the disease or its standard treatment does not seem to dangerously affect the patient, fetus, or the newborn infant. Surgical and medical treatment probably accounts for most of the improvement in prognosis seen over the years. The postoperative mortality has been reduced, especially in series where new surgical procedures have been used. A high frequency of major postoperative complications still remains a challenge for improvement. The medical intensive treatment of the acute attack has contributed to the improved prognosis. If compliance is good, the sulfasalazine prophylaxis may be one of the explanations to the change into a milder disease pattern that has been observed recently. Finally, and most important, a majority of patients sustain a normal life with full working capacity. Those who have surgery adapt well, particularly when a continence-saving procedure is used. The sexual function follows the improvement although the patient's need for support and counseling should not be underestimated.

REFERENCES 1. Aadland E, Schrumpf E, Fausa 0, et al: Primary sclerosing cholangitis: A long term follow-up study. Scand J Gastroenterol 22:655-664, 1987 2. Albrechtsen D, Bergan A, Nygaard K, et al: Urgent surgery for ulcerative colitis: Early colectomy in 132 patients. World J Surg 5:607-615, 1981 3. Bargen JA: Chronic ulcerative colitis associated with malignant disease. Arch Surg 17:561576, 1928 4. Brostr6m 0, Monsen U, Nordenvall B, et al: Prognosis and mortality of ulcerative colitis in Stockholm county, 1955-1979. Scand J Gastroenterol 22:907-913, 1987 5. Burnham WR, Lennard-Jones JE, Brooke BN: Sexual problems among married ileostomists. Survey conducted by the Ileostomy Association of Great Britain and Ireland. Gut 18:673-677, 1977 6. D'Oliveira R, Mayberry JF, Newcombe RG, et al: International comparison of mortality from inHammatory bowel disease in the Latin-speaking countries Venezuela, Italy and France. Digestion, 29:239-241, 1984 7. Demling L, Hegeman G, Classen M, et al: The prognosis of ulcerative colitis. German Med Monthly 14:209-214, 1969

PROGNOSIS IN ULCERATIVE COLITIS

217

8. Devroede G, Taylor WF: On calculating cancer risk and survival of ulcerative colitis in patients with the life table method. Gastroenterology 71:505-509, 1976 9. Edwards FC, Truelove SC: The course and prognosis of ulcerative colitis, Part 1-4. Br Med J 4:299--315, 1963 & 5:1-22, 1964 10. Gilat T, Ribak J, Benaroy J, et al: Ulcerative colitis in the Jewish population of Tel Aviv Jafo. Gastroenterology 66:335-342, 1974 11. Greenstein AJ, Sachar DB, Smith H, et al: Cancer in universal and leftsided ulcerative colitis: Factors determining risk. Gastroenterology 77:290-294, 1979 12. Greenstein AJ, Sachar B, Gibas A, et al: Outcome of toxic dilatation in ulcerative and Crohn's colitis. J Clin Gastroenterol 7:137-144, 1985 13. Gyde SN, Prior P, Alexander S, et al: Ulcerative colitis: Why is the mortality from cardiovascular disease reduced? Q J Med (new series L Ill) 211:351-357, 1984 14. Gyde SN, Prior P, Dew MJ, et al: Mortality in ulcerative colitis. Gastroenterology 83:3643, 1982 15. Gyde SN, Prior P, Thompson H, et al: Survival of patients with colorectal cancer complicating ulcerative colitis. Gut 25:228-231, 1984 16. Gyde SN, Prior P, Allan RN, et al: Colorectal cancer in ulcerative colitis: A cohort study of primary referrals from three centers. Gut 29:206-217, 1988 17. Hendriksen C, Kreiner S, Binder V: Long term prognosis in ulcerative colitis-based on results from a regional patient group from the county of Copenhagen. Gut 25: 158-163, 1985 18. Hendriksen C, Binder V: Social prognosis in patients with ulcerative colitis. Br Med J 581-583, 1980 19. Jones HW, Grogono J, Hoare AM: Surveillance in ulcerative colitis: Burdens and benefit. Gut 29:325-331, 1988 20. Jarnerot G, Into-Malmberg M: Sulfasalazine treatment during breastfeeding. Scand J Gastroenterol 14:869-871, 1979 21. Jarnerot G: Fertility, sterility, and pregnancy in chronic inflammatory bowel disease, review. Scand J Gastroenterol 17:1-4, 1982 22. Koudahl G, Kristensen M: Postoperative mortality and complications after colectomy for ulcerative colitis. Scand J Gastroenterol 11(suppl 37):117-122, 1976 23. Lee CGE, Truelove SC: Proctocolectomy for ulcerative colitis. World J Surg 4:195-201, 1980 24. Mcleod RS, Lavery JC, Leatherman JR, et al: Factors affecting quality of life with a conventional ileostomy. World J Surg 10:474-480, 1986 25. Mee AS, Jewell DP: Factors inducing relapse in inflammatory bowel disease. Br Med J 2:801-802, 1978 26. Mogadam M, Dobbins WO Ill, Korelitz BJ, et al: Pregnancy in inflammatory bowel disease: Effect of sulfasalazine and corticosteroids on fetal outcome. Gastroenterology 80:72-76, 1981 27. Morowitz DA, Kirsner JB: Ileostomy in ulcerative colitis. Am J Surg 141:370-375, 1981 28. Newcombe RG, Mayberry JF, Rhodes J: An international study of mortality from inflammatory bowel disease. Digestion 24:73-78, 1982 29. Nilsson LO, Kock NG, Kylberg F, et al: Sexual adjustment in ileostomy patients before and after conversion to continent ileostomy. Dis Col Rect 24:287-290, 1981 30. Niv Y, Bat L, Ron E, et al: Change in the extent of colonic involvement in ulcerative colitis: A colonoscopic study. Am J Gastroenterol 82:1046-1051, 1987 31. Pezim ME, Nicholls RJ: Quality of life after restorative proctocolectomy with pelvic ileal reservoir. Br J Surg 72:31-33, 1985 32. Porter RJ, Stirrat GM: The effects of inflammatory bowel disease on pregnancy: A casecontrolled retrospective analysis. Br J Obstet Gynaecol 93:1124-1131, 1986 33. Rampton DS, McNeil NI, Sarner M: Analgesic ingestion and other factors preceding relapse in ulcerative colitis. Gut 24:187-189, 1983 34. Ribe J, Benkov KJ, Thung SN, et al: Fatal massive hepatic necrosis: A probable hypersensitivity reaction to sulfasalazine. Am J Gastroenterol 81:205-208, 1986 35. Ritchie JK: Ulcerative colitis treated by ileostomy and excisional surgery. Br J Surg 59:345-351, 1972 36. Ritchie JK: Results of surgery for inflammatory bowel disease: A further survey of one hospital region. Br Med J 1:264-268, 1974

218

OLLE BROSTR0M

37. Sackett DL, Whelan G: Cancer risk in ulcerative colitis. Scientific requirements for the study of prognosis. Gastroenterology 78:1632-1635, 1980 38. Samuelsson S-M: Ulcerative colitis in the County of Uppsala. Clinical, epidemiological and sociomedical aspects [dissertation). Acta Universitatis Upsalienis, 1976 39. Sedlack RE, Nobrega FT, Kurland L, et al: Inflammatory colon disease in Rochester, Minnesota, 1935-1964. Gastroenterology 62:935-941, 1972 40. SincIair TS, Bount PW, Ashley N, et al: Non-specific proctocolitis in Northeastern Scotland: A community study. Gastroenterology 85:1-11, 1983 41. SoftIey A, Clamp SE, Watkinson G, et al: The natural history of inflammatory bowel disease: Has there been a change in the last 20 years? Scand J Gastroenterol 23(suppl 144):20-23, 1988 42. Sonnenberg A: Mortality from Crohn's disease and ulcerative colitis in England-Wales and the US from 1950 to 1983. Dis Colon Rectum 29:624-629, 1986 43. Sonnenberg A: Geographic variation in the incidence of and mortality from inflammatory bowel disease. Dis Colon Rectum 29:854-861, 1986 44. Stonnington CM, Philips SF, Zinsmeister AR, et al: Prognosis of chronic ulcerative colitis in a community. Gut 28:1261-1266, 1987 45. Truelove SC, Jewel! DP: Intensive intravenous regimen for severe attacks of ulcerative colitis. Lancet 1:1067-1070, 1974 46. Vakil N, Sparberg M: Steroid-related osteonecrosis in inflammatory bowel disease. Gastroenterology 96:62-67, 1989 47. van Hees PAM, van Tongeren JHM: Compliance to therapy in patients on a maintenance dose of sulfasalazine. J Clin Gastroenterol 4:333-336, 1982 48. Webster's New Dictionary of Synonyms. Springfield, G & C Merriam Co, 1973, p 238 49. Williams NS, Johnston D: The current status of mucosal proctectomy and ileo-anal anastomosis in the surgical treatment of ulcerative colitis and adenomatous polyposis. Br J Surg 72:159-168, 1985 50. Willoughby CP, Truelove SC: Ulcerative colitis and pregnancy. Gut 21:469-474, 1980

Address reprint requests to Ol!e Brostrom, MD, PhD Karolinska Institute Section of Gastroenterology Second Department of Medicine Sodersjukhuset Box 3800 S-10064, Stockholm Sweden