POSTGRAD.   MED.   J. (I96I), 37, 783

                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
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
                                                                                    ..       ..       ..           83        45% 78%
suggest and is inlong-term recurrenceto is
         that the
                                                                    Pollock, I958     .      ..       ..                      ?  44%
higher              the region of 60      80% (Van                  Leeds
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 ',
      cases were indexed
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
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 ..
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 .
                                                                                          ..  ..
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
entirely on clinical were and the following
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
   (2)diarrhoea Episodes of minor to interfere with
                not severe enough
                                     abdominal pain     they are classified under Crohn's original ter-
                                                        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 '
                                                                                           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
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
    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.-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
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
    (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
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

CASES            46
                CASES         45
                                                                             120 CASES
                           CASES         36

                                                             69o NON
                    __\                   94/

                              3 /       FAI L E

                                         <$     \
                                                             6901 REC
                            ^§            ^
                                                                                          FAIR 260/,
 0-I YRS.         -2YRs.   0-5 YRS.     O-IO1YRS.
FIG. 2.-Failure rate of conservative treatment from   48
disease the greater the chance of recurrence. In
                                                              31!/o REC                   POOR 18/o
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

                                                                              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
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-
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
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.
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.

To top