POSTGRAD. MED. J. (I96I), 37, 783 THE PROGNOSIS OF CROHN'S DISEASE OF THE SMALL INTESTINE J. M. DAVIS, M.CHIR., F.R.C.S. Surgeon to the Whittington Hospital, London, N. 19 REGIONAL ileitis has only been widely recognized TABLE I since the classical description by Crohn, Ginsberg ESTIMATES OF THE RECURRENCE RATE OF CROHN'S and Oppenheimer in I932. It is a rare disease- DISEASE FROM THE MOST RECENT LITERATURE the average general surgeon or physician will treat Over- Io- only one or two new cases each year. Consequently - all Year been very few large series have yet allow astudied for Author No. Recur- Recur- Cases rence rence periods of timethe enough to A most conclusive long Rate Rate assessment of prognosis. important aspect of the prognosis, the recurrence rate, has Crohn, 1958 Mt. Sinai, N. Y. .. .. .. 348 30% been very variably reported (Table I). Early Van Patter and others, 1954 .. series suggested that a recurrence followed surgery 297 64% in about 15%/ of patients (Garlock and Crohn, Mayo Clinic Marshall and Mathieson, 1955 233 30% ? 1945), butshowed series with longer periods of later Lahey Clinic Jackson, 1958 . .. 126 follow-upand that that this figure was an under- recurrence occurred in about Mass. General .. 55% ? estimate Gump and Lepore, 960 .. 30%0 of patients (Crohn, 1958). 10Other reports 97 40% 43% Presbyterian, N. Y. containing patients followed for rate15 much to years Cooke, 1955 Birmingham .. .. .. 83 45% 78% suggest and is inlong-term recurrenceto is that the Pollock, I958 . .. .. ? 44% higher the region of 60 80% (Van Leeds 7 Patter, Bargen, Dockerty, Feldman, Mayo and Present series .. .. .. . 20 31% ? Waugh, I954; Cooke, I955). of In this paper a further series patients with TABLE 2 Crohn's disease of the small intestine is reported to FINAL DIAGNOSIS OF 263 CASES INDEXED AS CROHN'S provide additional this series and thethe prognosis. information about DISEASE The results from reports from Cases included in present series: Cases the literature are critically reviewed, the factors Chronic terminal ileitis.. .. .. .. 30 which may predispose to recurrence are examined Chronic segmental ileitis .. .. .. 13 and an attempt is made to explain some of the Chronic segmental jejunitis .. .. .. 6 factors which may have led to such wide variations Chronic ileo-jejunitis .. .... .. 2 in the estimates of the recurrence rate. I51 Present Series Cases excluded from present series: A retrospective survey has been made of all Acute terminal ileitis .. .. .. .. 40 Acute segmental ileitis . .. .. .. 8 cases of Crohn's disease listed in the diagnostic Acute segmental jejunitis . .. .. 4 indices of seven non-teaching hospitals and St. Ileo-colitis .. .. .. .. . 23 Mark's Hospital for Diseases of the Rectum during Segmental colitis . .. .. the years I949 to I958. With the exception of St. Caecitis . . . . . . . . .. . . 10 Miscellaneous ........ .. 8 Mark's, these are all general hospitals serving an average cross section of the as 'Crohn'sAltogether community. 112 263 112 cases have been excluded eitherdisease ', but cases were indexed because the diagnosis was inconclusive or because it fell These I51 cases of Crohn's disease of the small into one of the controversial categories listed intestine have been followed up not only to deter- in Table 2. This left 151 cases of Crohn's disease mine the incidence of recurrence, but also to of the small intestine, which form the main basis determine the general health of the patients regard- of this study. less of the estimated activity of the intestinal 784 POSTGRADUATE MEDICAL JOURNAL December I961 lesion. Each patient has been examined by the TABLE 3 writer whenever possible, but when this has not MAIN SYMPTOMS AND MODES OF PRESENTATION FROM proved practical assessment has or from a by a been made I51 CASES OF CROHN'S DISEASE OF THE SMALL INTESTINE the patient questionnaire fromown doctor. Estimations report from the patient's of the Symptoms haemoglobin and the erythrocyte sedimentation Abdominal pain .. 95 rate have been performed routinely on all the Diarrhea.. .... . 85 Weight loss . 45 patients examined, but X-rays of the gastro- intestinal tract have only been performed when Pyrexia 3.. ... Fistula in ano .. .. 31 2 indicated on clinical grounds. Altogether I41 of Amenorrhoea .. .. 9 (of women) Melana .... . 3 151 patients have been traced: I6 patients have Delayed puberty .. 3 died, 82 patients have been assessed by examina- tion and 43 patients have been assessed by Modes of Presentation questionnaire. of recurrent or active disease has The diagnosis Abdominal mass ... .. 44 Intestinal obstruction .. .. 17 been made by consideration of the clinical, radio- 'Acute appendicitis' . .... 10 logical and pathological evidencesections have the together with Pelvic mass .. .. Ileo-rectosigmoid fistula . .. .. .. 10 7 Histological operative findings. the been reviewed if only Peritonitis .. ..... . 3 original report was incon- Incidental discovery .. .. 3 clusive. In patients with negative or inconclusive Ileo-vesical fistula .. .. .. 2 X-rays the diagnosis of recurrence was made Other .. .. .. .. .. 4 grounds entirely on clinical were and the following accepted symptoms and signs Abdominal as evidence of (3) Incidence of New Cases and Relation to the recurrent disease: pain; diarrhoea Population. An average of 12 new cases a year more than four times a day (provided this was not have occurred over a Io-year period. This in- the only symptom); palpable mass; internal or cidence is almost certainly too low because the external abdominal fistulae; loss of weight; efficiency of the diagnostic indices varies from elevated ESR. Four patients had radiological but hospital to hospital. If the three most reliable no other evidence of recurrent disease and they indices, from hospitals serving a known population, have not been classified as having a recurrence. only are considered, then the probable incidence Assessment of the general health of each patient of Crohn's disease of the small intestine is in the on simple clinical grounds was much more straight- region of one new case a year in each Ioo,ooo of forward than the diagnosis of recurrent disease the population. and the following classification has been used: (4) The Site of the Pathological Lesion. In 130 Good. Symptom free (excluding diarrhoea (I) four times a patients the lesion affected the distal ileum, ex- up to day). tending right down to the ileo-caecal valve, and and (2)diarrhoea Episodes of minor to interfere with Fair. not severe enough abdominal pain they are classified under Crohn's original ter- work. minology as chronic terminal ileitis. Within this regularPoor. Severe group nine patients had ' skip ' lesions which were (3) work. symptoms. Unable to perform not in continuity with the main disease process and regular affected either an isolated segment of upper ileum or of jejunum. In 1 patients with chronic ter- Main Clinical Features minal ileitis the disease extended to the cacum, (I) Age and Sex. In the present series the maximum incidence occurred in the third and but to no other more distal part of the large in- testine. Patients with more extensive large bowel fourth decades; the oldest patient was 72 years involvement have been classified as ileo-colitis and and the youngest was 14 years. The average age are excluded from the present series. at the onset of symptoms was 32 years; this is In the remaining 2 patients the disease affected slightly higher than thebecause of age exclusion average of most the proximal ileum or jejunum and they are classi- reported series, probably of acute the fied as chronic segmental ileitis or jejunitis. of a large number of cases ileitis, many of which occur in the younger age groups. The Follow-up Findings incidence in males and females was almost equal (i) Mortality. There were i6 deaths from 141 (females 5 %). and traced cases of Crohn's disease of the small in- Signs. The of (2) Symptoms and the modesincidence of the and signs testine. Three deaths were due to unrelated inter- symptoms presentation current disease and the remaining I3 deaths were were very similar to those in most other series and directly attributable to Crohn's disease, giving a they are shown in Table 3. mortality of 9.8%. Six of these I3 deaths occurred December 1961 DAVIS: The Prognosis of Crohn's Disease of the Small Intestine 785 at a late stage and were caused by recurrent or TABLE 4 active disease; the remaining seven deaths fol- PERIOD OF FOLLOW-UP FROM ONSET OF SYMPTOMS OF lowed surgical treatment of the primary lesion. I4I CASES OF CROHN'S DISEASE OF THE SMALL INTESTINE Most of the operative deaths occurred in com- Years cases-in two patients there was peri- plicateddue to from o-I I-2 2-3 3-4 4-5 5-10 10-20 Over Total tonitis perforation of the primary lesion and Onset 20 three patients died of sepsis following a secondary Cases I4 15 I8 4 7 43 I6 4 I41 resection after a previous unsuccessful short- circuiting operation. Only two deaths occurred following a primary resection disease is (2.7%). in 75 patients significant numbers of cases have not been fol- lowed for the longer periods of time. In addition, The death rate of Crohn's surprisingly the statistical difficulties of interpreting a diminish- low for a disease of such high morbidity. Most reportedThis low death rate is mortality of 5 the series have an overall to almost insurmountable. ing series arebeen made to estimate the However, an io%. supported by the attempt has approximate Registrar io General's figures (I959)-during long-term rate with by' comparing an 'annual ' results a pseudo cumulative ' recur- previousnumber of deaths due to and Wales the years in England recurrence average Crohn's disease rence rate (Fig. i). The 'annual' rate is the a year was only 60, or I per 750,000 population. number of recurrences that have been traced up If the annual incidence of new cases is, as sug- to the period under consideration expressed as a above, per Ioo,ooo population, and gesteddisease isi not on the of the then the ratio of percentage of the total number of cases followed if the increase, to the same period. The figures for this series at new cases to deaths each year is 7.5: i. These two, five and io years are I6%, 17% and 36% figures suggest that only onedisease in sevendies person or certainly an respectively. Thisthe true is almostrate because incidence eight who it.develops Crohn's actually underestimate of long-term recurrent because of many of the short-term patients with (2) Duration of Follow-up. One hundred and disease have not yet been followed to the later from periods forty-one patients have been traced time of onset years. The ' pseudo cumulative' recurrence rate of six months to 29 years from the includes these short-term recurrences and is an of the disease (Table 4). Many cases were first attempt to relate them to the total numbers of indexed in the period under review (1949-58), later periods. It has been patients followed to the the number of recurrences although theyand several of these treated many had originally been calculated by expressing years earlier, long-term cases followed up to and before the period under con- only reappeared orbecause they had developed sideration as a percentage of the sum of the same recurrent disease other complications. figure of recurrences added to the number ofwhole non- (3) Recurrent Disease following Surgery. One recurrences who have been followed for the hundred and twenty-seven patients were eventually length of the same period. A true cumulative treated by definitive surgery; there were seven recurrence rate for a diminishing series is im- operative deaths, leaving I20 cases available for possible because as time progresses the number of recurrences outnumber the total of traced cases study of the recurrence rate. Of these 120 patients, and the rate exceeds Ioo%. 37 have developed recurrent disease, giving an overall recurrence rate of 31%, unrelated to the It must be stressed that the 'pseudo cumula- time of follow-up. Almost half of the total recur- tive' recurrence rate is an artificial figure which rences developed symptoms within two years of overestimates the true recurrence rate. Neverthe- operation and the remaining recurrences occurred less, it allows a useful comparison with the sporadically during the of the years (Table 5). ensuing 'annual' recurrence rate, because in a perfect An accurate estimate long-term recurrence series, with every case followed for the maximum rate is not possible in the present series because period of time, these two rates would coincide. TABLE 5 FOLLOW-UP PERIODS OF 120 PATIENTS TREATED BY DEFINITIVE SURGERY AND TIME OF ONSET OF RECURRENT DISEASE Years from Over Operation 0-I 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-1o 10o-5 15-20 20 Total Cases .. .. 13 16 17 9 17 11 9 5 3 6 6 5 3 120 Onset recurrence.. 11 7 3 3 I 2 3 2 I 0 2 0 2 37 Note.D(31%) Note.-Deaths following initia 1operation are excluded. 786 POSTGRADUATE MEDICAL JOURNAL December I961 Consequently, if these two estimates are plotted on a graph, the truth must lie somewhere between 120 them, and for this series the figures suggest that the true recurrence rate is in the region of 25% at five years and 45% at io years (Fig. I). n 80 ANNUAL (4) Active Diseasesfollowing Conservative Treat- ment. Although 48 of I41 traced cases were first treated conservatively, the majority eventually U came to surgery, leaving only 14 patients treated 40 entirely without operation (laparotomy without definite surgery is counted as conservative treat- ment). Of these 14 patients treated conservatively, two died of active disease and two were severely incapacitated by the disease; the remaining io 120 patients remained reasonably well, nine of them having been followed up for more than four years. Conservative treatment is considered to have PSEUDO failed, either because surgery had become neces- so80 - CUMULATIVE sary or because the condition of the patient had deteriorated, in 54% of patients at two years and 94% at Io years (Fig. 2). The overall medical U salvage rate of about Io% is similar to that of most 40 series, although this estimate is possibly now on the low side because in recent years the indications for surgical treatment have probably become more stringent. (5) Clinical State Regardless of Active or Recur- rent Disease. The general health of each patient has been assessed as good, fair or poor by con- 880 sideration of the simple criteria described earlier. This assessment has been made at the end of the follow-up period except for patients who died of g60 unrelated intercurrent disease when assessment was made from the period preceding the terminal J40 M 40 C0 illness. The results of this assessment of the clinical state of the whole series of 141 patients are: ^*^ /--. good 5I%, fair 27%, poor 13%, died 9%. It is W 20 T surprising that only just over one-fifth of the patients have either been classified as poor or have died as a direct result of the disease. Comparison of the overall recurrence rate and the clinical state of the 120 patients treated by 1 2 3 4 5 6 7 8 9 10 definitive surgery shows that, whereas 31% of YEARS SINCE OPERATION patients developed recurrent disease, only i8% of FIG. i.-Two methods of expressing recurrence rate patients were incapacitated or had died of the from I20 patients after surgery. Recurrences are shown in black. See text for explanation of the disease (Fig. 3). This difference is explained partly because several patients with recurrent disease terms ' annual' and ' pseudo-cumulative'. suffered from relatively few symptoms and also because a small number of patients had apparently very important fact that 70 to 80% of patients been cured by a second operation. This dis- remain surprisingly well for long periods of time. crepancy between the estimates of the recurrence Factors which may Predispose to Recurrence rate and the clinical state has also been noted by Cooke (I955), who reports a surprisingly low The following factors which may predispose to rate of invalidism in his series despite an overall recurrent disease have been examined: recurrence rate of 45% and Io-year recurrence (i) Extent of the Intestinal Lesion. This is prob- rate of 78%. In recent years so much attention ably the only important single factor which may be has been focused on the problem of the recurrence of help in predicting the likelihood of recurrence rate that there has been a tendency to overlook the in any individual case-the more extensive the December 1961 DAVIS: The Prognosis of Crohn's Disease of the Small Intestine 787 48 CASES 46 CASES 45 120 CASES CASES 36 CASES 566 GOOD 69o NON __\ 94/ 73 3 / FAI L E <$ \ 6901 REC FAILED 391o 54% FAILED ^§ ^ FAILED FAIR 260/, 0-I YRS. -2YRs. 0-5 YRS. O-IO1YRS. FIG. 2.-Failure rate of conservative treatment from 48 patients. disease the greater the chance of recurrence. In 31!/o REC POOR 18/o OR the present series the extent of the lesion was recorded accurately in only 74 patients; the __DIED average length of the diseased intestine in the FIG. 3.-Comparison of the overall recurrence rate with patientswas idid not develop a with who recurrence of the .the clinical state from 20zo patients after surgery. disease in., as compared 19 in. in the patients who did develop a recurrence. Van Patter defunctioning short-circuiting operation. It is and others (i954) have reported similar findings. most important to distinguish between the two 'Skip ' lesions do not seem to carry a particu- forms of short-circuiting procedure, the essential larly bad prognosis. From the nine patients in which a 'skip ' lesion occurred in the present difference being that in the defunctioning type the bowel is completely divided and the distal end is series only two developed recurrent disease; seven oversewn to leave a blind loop, thus excluding the of the ' skip ' areas were resected en bloc with the diseased area from the main intestinal stream main lesion and two were resected separately. One (Fig. 4). In the present series just over two-thirds of the recurrences occurred at the site of separate of the patients were treated by primary resection resection. and the remainder were treated by a short- Cooke (1955) has drawn attention to the high circuiting operation; only 11 patients were treated incidence of recurrence in the patients who have by a defunctioning short-circuiting operation as a been found to have steatorrhcea before operation. primary procedure. He suggests that steatorrhcea indicates an extensive Accurate comparison of the incidence of recur- intestinal lesion and consequently a greater risk of rence after the three types of operation is not recurrence. In the present series facal fat estima- possible in the present series because the numbers tions have only occasionally been performed prior of the sub-groups are too small to be significant, to operation and therefore there is no further but the trend strongly supports the conclusion of information about what is probably a most useful the Mayo Clinic (Van Patter and others, I954) prognostic guide. that the incidence of recurrence bears no relation (2) Choice of Operation. There are three main to the type of operation performed. However, operations which can be performed for Crohn's the morbidity following short-circuiting operations disease of the small intestine-resection, a non- was very high, particularly after the non-defunc- defunctioning short-circuiting operation and a tioning procedures; in the present series almost 788 POSTGRADUATE MEDICAL JOURNAL December 1961 o FIG. 4.-Two types of short-cir- cuiting operation-non-defunc- .'r. .L ".· ·.- ' ., tioning without exclusion and ..' defunctioning with exclusion. ", · . ··· ··.: :L .. , o-·L .-WtTH0JT "EXLUSION l TH ECLUSIO $SHOR 'EXCLUON- WTHOUT CIRCUIT SHORT CIRCUIT WITH EXCLUSION one-third (eight of 26 cases) treated by this opera- in obstructed cases-with modern instruments for tion subsequently required resection, not for intestinal decompression at the time of operation recurrent disease, but for persistent local or sys- (Savage, I960) this operation should now rarely temic symptoms. The morbidity after defunction- be necessary. ing operationstreated quite so high, but two of the was not (3) Extent of Excision. Van Patter and others 1 patients by this method developed (1954) found no correlation between the recurrence mechanical loop complications; one patient de- rate and the length of normal intestine proximal veloped intussusception ofa the blind loop and to the diseased area in the specimens of patients treated by resection. In the present series too few another patient developed closed-loop obstruc- tion. Garlock and Crohn (1945), who are strong specimens have been measured accurately to protagonists of theit defunctioning short-circuiting decide whether a radical excision reduces the risk of recurrence. This is a point of great importance operation because carries a low mortality, admit that from 90 patients treated by this method 25 to the surgeon-on the one hand, a very extensive for persisting (27%) eventually required resectiondisease. operation is no guarantee against recurrence and site of the symptoms at the originalresection is now without may in itself cause malabsorption-on the other It is concluded that hand, the pathological process sometimes proves doubt the treatment of choice. The main original to be more extensive than has been apparent at objection to resection, that it carried an operative the time of operation and recurrence is almost mortality in the region of I4% (Crohn, I958), inevitable if resection is performed through has been rendered invalid by improved operative diseased intestine. In the present series five technique and improved pre- and post-operative patients were found to have disease extending care. Even in the present series, drawn mainly right up to the line of section and four developed from non-teaching hospitals, there were only two symptoms of recurrent disease within six months. operative deaths following primary with resection in A reasonable compromise has been suggested by 75 uncomplicated patients (patients perfora- Crohn (1958), who advocates section of the bowel tion of the ileum have been excluded). The I2 in. proximal to the apparent upper limit of the majority of specialized centres now confirm anwhen even disease. lower mortality. However, there are times (4) Duration of Symptoms Prior to Surgery. resection may be considered inadvisable, either Pollock (1958) suggests that the risk of recurrence because of the patient's general condition or is greater in patients who have a long history prior because of dense adhesions or gross thickening of to operation. He gives a Io-year recurrence rate the mesentery and, under these circumstances, a of 75% from eight cases of 'chronic' Crohn's defunctioning a short-circuiting operation is non- con- disease (history over two years), as compared with sidered to be reasonable alternative. The a recurrence rate of 20% from 10 cases of ' sub- defunctioning short-circuiting operation has been acute' Crohn's disease (history one month to two very justly condemned by Lewisohn (1938) and years). These figures are really too small to be it should be reserved only as a-temporary expedient significant and they conflict with Crohn's view December 1961 DAVIS: The Prognosis of Crohn's Disease of the Small Intestine 789 that the greatest percentage of cures comes from of resolution or recrudescence of the primary lesion the patients in which operation has been post- after one of the short-circuiting operations has not poned for about two years. In fact, Crohn (1958) been classified as recurrent disease. The only form advocates delaying operation for about this period of recurrence recognized after this type of opera- of time ' until the inflammatory process has cooled tion has been a new lesion in the ileum proximal off and healing resolution has begun'. In the to the anastomosis or a spread of the disease to the present series there was no significant correlation large intestine. between the recurrence rate and the length of the The definition of' recurrence ' in the literature pre-operative history-46% of recurrences had a is very variably defined and its meaning is not history of more than two years compared with always restricted to patients who have received 41% of non-recurrences. In practice the timing of the operation is usually some form of definitive surgery, but is also used to include failures of conservative treatment and ex- dictated by the patients' symptoms and few in- acerbations of the disease in patients who have only dividual patients pass through a series of predict- received a laparotomy or an appendicectomy. Used able phases. However, exacerbations of the disease, in this sense ' recurrence ' loses its true meaning of probably caused by episodes of secondary acute 'recurrence following definitive surgery' and inflammation, will usually resolve if treated con- gives artificially high figures. servatively. It is probably wise to avoid operating (2) Selection of Cases. One of the main prob- during these episodes because the acute inflamma- lems in analysing any series of Crohn's disease is tion may obscure the extent of the underlying to decide whether or not to include two large granulomatous process and so make it difficult for categories of patients-acute ileitis and Crohn's the proximal extent of the lesion to be defined. disease of the large intestine. Inclusion of the (5) Age and Sex. In the present series slightly former will influence estimates of the prognosis more females developed recurrence than males. favourably and inclusion of the latter will have the Females, 21 cases; males, i 6 cases. This difference converse effect. is not significant. Acute ileitis in the present series has been ex- The age distribution of the patients who de- cluded if the history has not been longer than veloped recurrent disease shows no marked dif- seven days. The relationship of acute ileitis to ference from that of the whole series. The older Crohn's disease has been controversial, but the age groups were not relatively immune, as has weight of evidence now strongly suggests that if been suggested by Jackson (1958), as seven of the the history is short this condition only very rarely recurrences were over the age of 40 at the time of progresses to form the chronic granulomatous onset of the disease. lesion. There may well be two forms of acute Factors which Influence Estimates of the ileitis, one being a separate disease entity with a Recurrence Rate good prognosis and the other being genuine Crohn's disease, which has given only minor The estimate of the recurrence rate for any given symptoms until an episode of secondary acute in- series of Crohn's disease is influenced by variations fection has become superimposed on the under- of definition, selection and interpretation. The lying chronic granulomatous lesion. In retrospect data from any series can be manipulated in several the main distinguishing feature between these two different ways to give divergent results. If, for forms of acute ileitis is the length of the history. example, the patients from the present series who had been followed for less than two years had been Fifty-two patients with a short history of acute ileitis have been excluded from the present series eliminated, and if the patients with symptomless because 42 of this group have been traced and only radiological ' recurrences' had been accepted as clinical recurrences, then the overall recurrence one has developed symptoms suggesting a chronic rate would have risen from 31 to 43%. Con- granulomatous lesion (Table 6). In comparison, 23 patients of acute ileitis with a history of more sequently, it is very important to examine carefully all the factors concerned before accepting the than a week have been traced and are included in the present series because 17 have progressed to recurrence figures for any given series. These give definite signs of a chronic granulomatous various factors are discussed in turn with reference lesion. Several recent authorities have adopted a to the present series and to other reported series, similar policy: Pollock (I958) distinguishes acute where relevant. Crohn's disease with a history of less than one (i) Definition of 'Recurrence '. In the present series ' recurrence' is used only to denote a new month and Gump and Lepore (i960) have ex- cluded a large group of cases presenting with chronic granulomatous lesion developing in symptoms suggestive of acute appendicitis. patients treated either by resection or by one of Crohn's disease of the large intestine constitutes the two types of short-circuiting operation. Failure the second main problem of selection. Twenty- 790 POSTGRADUATE MEDICAL JOURNAL December 1961 TABLE 6 after resection or exclusion of the terminal ileum, THE EFFECT OF THE LENGTH OF THE HISTORY ON THE and it is probably caused by the loss of an im- PROGNOSIS OF 75 CASES OF ACUTE ILEITIS portant area of water-absorbing intestinal mucosa. Acute Developed In this series diarrhrea alone has not been accepted History Ileitis Traced Died Chronic as evidence of recurrent disease. Cases Lesions The second difficulty is the problem of the Less than patient who is free of symptoms, but who has 7 days.. 52 42 2 i (2%) radiological changes suggestive of recurrence. More than Dyson, Hodes and Rhoades (I954) first drew attention to this dilemma and suggested that minor 7 days.. 23 23 o 17 (73%) abnormalities of the intestinal mucosa and dila- tion of the intestine might be associated merely three patients with ileo-colitis (histological Crohn's with deformity in the region of the anastomosis, disease of the terminal ileum associated with dif- but Van Patter and others (1954) think that X-ray fuse or segmental lesions of the large intestine) signs may precede a clinical recurrence. In the have been excluded from the present series (Table present series there have been five symptomless 2). The main reason for their exclusion is that the patients with doubtful X-ray findings and they Crohn's diagnosis of been disease ininthese patients has have not been classified as recurrence. Cooke often only established retrospect after (1955) partly attributed his high recurrence rate to many years of various forms of management and the inclusion of several symptomless radiological multiple piecemeal operations. Recurrence after recurrences. surgery for ileo-colitis appears common, but many (4) Duration of the Period of Follow-up. In the failures appear to have been due to limited resec- present series almost half of the patients who tions through diseased tissue and it would only recurred after surgery developed the symptoms of confuse the picture of the overall prognosis to recurrent disease within two years of operation. include this group together with Crohn's disease After two years the risk of recurrence diminishes, of the small intestine. However, four cases of but it is always present-one patient developed ileo-colitis have been included in the present series the symptoms of recurrence 29 years following because the ileal lesion preceded the large gut the initial operation. Consequently the length of lesion by one or more years: in these patients the the period of follow-up is a most important factor colitis was clearly a complication of the small gut in assessing the overall recurrence rate. Low lesion and they have been classified as having recurrence figures from the early reports were un- recurrent disease. Many of the reported series doubtedly mainly due to the small number of long- contain cases that would fall into the category of term cases. Even now very few series, including ileo-colitis and thus increase the recurrence figures. the present one, contain enough long-term cases (3) Criteria of Recurrent Disease. Histological to form a conclusive assessment of the long-term proof of recurrent disease was only obtained in recurrence rate. about half of the cases of the present series. In The high overall recurrence rate of 64% re- the remaining cases the diagnosis of recurrence ported by the Mayo Clinic (Van Patter and others, was made on clinical and radiological grounds and 1954) at first sight appears to be due to the large was usually straightforward, but in a few cases this number of long-term cases in this series. However, diagnosis was very difficult and sometimes im- examination of the data shows that an unusually possible to establish with certainty. This small high proportion of recurrences, almost three- proportion of doubtful cases may be interpreted quarters, developed symptoms within two years of by one observer as recurrent disease and as non- operation. Therefore this high recurrence rate recurrent disease by another, depending on the cannot be explained only by the large number of criteria accepted. Although only a small number long-term cases and there must be some other of cases are involved, they may be enough to reason to account for the unusually high incidence influence the recurrence figures in a small series by of early recurrences. The most likely explanation o1 to I5%. This lack of an absolute standard of is that the criteria accepted for the diagnosis of what constitutes a clinical recurrence probably recurrent disease by the Mayo Clinic has been accounts for the main differences of opinion about different from that of most other series. the incidence of recurrent disease. (5) Statistical interpretation. There are several There are two main difficulties. The first is the difficulties in assessing the long-term recurrence problem of the patient with diarrhoea but with no other symptoms or confirmation of recurrent rate of Crohn's disease not only because the total numbers of cases in most series are small by disease. Whilst it is true that diarrhoea is usually statistical standards, but also because the numbers the first symptom of recurrence, it is very common of cases followed inevitably decrease with time and December I961 DAVIS: The Prognosis of Crohn's Disease of the Small Intestine 791 the results from a diminishing series can be inter- ishes to 40% by 15 years. Whilst agreeing with pretedthe several different ways.the has beenseries in It seen their general hypothesis that the recurrence rate from follow-up results of present over io years does not approach Ioo%, it is diffi- that there are several methods of expressing the cult to accept this anomalous result. There can long-term recurrence rate and that any estimate of obviously be nowith absolute diminution of the true the figures at 0o years can be no more than recurrence rate the progression of time and an intelligent guess. The main reasons are that this surprising finding is probably best regarded as this series is not only overloaded with an un- a statistical curiosity. representative selection of long-term cases, but it Summary also contains a high proportion of short-term cases, many of which have recurred but have not yet A series of I41 cases of Crohn's disease of the been followed to the later years. Most series con- small intestine have been followed for varying tain the same statistical difficulties and most long- periods. The main conclusions are: term figures should be viewed tentatively. (I) The overall recurrence rate after surgery is It is noteworthy that in the two largest series 31% (12o cases). (Crohn, I958; Van there has been no from the Patter and others (2) The failure rate with conservative manage- Mayo Clinic, I954) attempt to ment is 94% (48 cases). give definite figures for the long-term recurrence (3) Primary resection is the operation of choice. rate, presumably because of the statistical difficul- ties. Cooke (I955) has tried to calculate the long- (4) Consideration of the recurrence figures alone gives an inaccurate picture of the overall term recurrence rate from the Mayo Clinic data and estimates a 68% and 8o% recurrence rate at prognosis. Io and 15 years for this series. It is difficult to (5) Estimates of the long-term recurrence rate determine how these estimates have been made, are fraught with difficulties of definition, selection because the necessary data are not available in the and interpretation and they should be viewed very Mayo Clinic report to reach these conclusions. These high figures may reflect a ' pseudo cumula- critically. tive ' result, but they cannot reflect the proportion I am very grateful to the physicians and surgeons of of recurrences present in patients followed for io the following hospitals for allowing me to follow up and 15 years. Cooke (I955) has estimated his own their patients: Addenbrooke's, Cambridge; Ashford, Middlesex; Central Middlesex; Chase Farm, Middle- long-term results 'on the same basis ' as 78% at io years. sex; Edgware General, Middlesex; St. Mark's, London; West Middlesex; Whittington, London. Gump and Lepore (I960) in a series of 97 cases I am also very grateful for financial assistance from followed for o or more years report a' cumulative ' the Department of Gastroenterology, Central Middlesex recurrence rate of 43% at Io years, which dimin- Hospital, and the Research Department, St. Mark's Hospital. REFERENCES COOKE, W. T. (1955): Nutritional and Metabolic Factors in the Etiology and Treatment of Regional Ileitis, Ann. roy. Coll. Surg. Eng., 17, 137. CROHN, B. B., GINSBURG, L., and OPPENHEIMER, G. D. (1932): Regional Ileitis: A Pathologic and Clinical Entity, J. Amer. med. Ass., 99, 1323. - (958): ' Regional Ileitis'. New York: Grune and Stratton. DYSON, W. L., HODES, P. J., and RHOADES, J. E. (1954): Late Results in the Surgical Treatment of Regional Enteritis, Penn. med. J., 57, 443. GARLOCK, J. H., and CROHN, B. B. (1945): An Appraisal of the Result of Surgery in the Treatment of Regional Ileitis, J. Amer. med. Ass., 127, 205. GUMP, F., and LEPORE, M. L. (I960): Prognosis in Acute and Chronic Regional Enteritis, Gastroenterology, 39, 694. JACKSON, B. B. (1958): Chronic Regional Enteritis, Ann. Surg., 148, 8i. LEWISOHN, R. (1938): Segmental Enteritis, Surg. Gynec. Obstet., 66, 215. MARSHALL, S. F., and MATHIESON, W. L. (1955): Surgical Treatment of Chronic Regional Enteritis, Lahey Clin. Bull., 9, 66. POLLOCK, A. V. (1958): Crohn's Disease, Brit. J. Surg., 197, 193. REGISTRAR GENERAL'S STATISTICAL REVIEW OF ENGLAND AND WALES FOR THE YEAR I959: Part I, 29. SAVAGE, P. J. (1960): The Management of Acute Intestinal Obstruction, Brit. J. Surg., 47, 643. VAN PATTER, W. N., BARGEN, J. A., DOCKERTY, M. B., FELDMAN, W. H., MAYO, C. W., and WAUGH, J. M. (1954): Regional Enteritis, Gastroenterology, 26, 347.
Pages to are hidden for
"OF THE SMALL INTESTINE"Please download to view full document