Ren#{233}H. Kruyt,
Normal Dynamic
MD
#{149} Johannes
B. V. M. Delemarre,
Anorectum: MR Imaging
U
1
From
the
This
27, 1990; Address
RSNA, See also ,c
revision reprint
Hospital, requested
of Diagnostic October
requests to R.H.K. 1991 the article by Yang et al (pp
25-33)
were
analyzed
be
used
as a baseline
for
the
variation large
in
these measuredefecography
(2). of the anorectum, magresonance (MR) imaging has used up till now only in the was
for
In evaluation netic
been case
of the
perforate
congenital
anus
(3,4),
anomaly
neoplasms,
of im-
and
inflammatory processes (5-9). In this issue of Radiology, Yang et al (10) meport on the evaluation of pelvic floor descent with MR imaging, especially in pelvic prolapse. We studied the normal anatomic relationships of the anorectum in the sagittal plane duning rest, penineal contraction, and straining in 10 asymptomatic mdividuals. The purpose of the present investigation is to provide normal values
(R.H.K.,
10, 2300
4; revision
can
observer ments
Radiology
Rijnsburgerweg
subjects
study of patients with complaints melated to defecation, such as obstipation, rectal incontinence, mucus and blood in stools, pain, unsuccessful defecation, and feeling of incomplete evacuation after defecation. There was a wide mange of measurements of the anorectal angle and position of the anorectal junction, and the inter-
179:159-163
Departments
University
NTIL
and interobsenvem variation in the measurement of the dynamics and position of the anorectal junction and anomectal angle was determined (2).
ies, 757.1214
Leiden
Doornbos,
in this
J.D.,
RC Leiden,
received
November
issue.
PhD
J.
#{149} Huib
Vogel,
MD
Anatomy’
asymptomatic
Index terms: Anus, 757.929 #{149} Anus, MR studies, 757.1214 #{149} Defecography, 757.1299 #{149} Magnetic resonance (MR), comparative studies, 757.1214 #{149} Rectum, 757.929 #{149} Rectum, MR stud-
1991;
#{149} Joost
now, the imaging modality of choice for dynamic analysis of the anorectum has been defecognaphy, a radiographic study of the anorectum in lateral projection after opacification of the perineum, anorectum, and vagina with barium sulfate contrast medium, first described in 1952 (1). In 1989, a study was published in which defecograms of
In this study, the anatomy of the anorectum in relation to the surrounding structures and the anorectal angle were analyzed with magnetic resonance (MR) imaging at rest, during perineal contraction, and during straining in 10 asymptomatic subjects. The intraand interobserver and intraand interpatient variations in the measurements of the anorectal angle, position of the anorectal junction, and position of the plica of Kohlrausch in the rectum were established at rest, during perineal contraction, and during straining. The values for the anorectal angle and position of the anorectal junction obtained with MR imaging were compared with standard radiographic defecography findings. It was shown that MR imaging has the potential for measuring these parameters in a more precise and more patient-friendly way than defecography. Unlike dynamic defecography, MR imaging is able to depict the mobility of the posterior rectal wall. A descent of over 20 mm from rest to straining should be considered pathologic. This finding might play a role in patient selection for operation.
Radiology
MD
H.J.V.)
and
The
Netherlands. 12; accepted
Surgery
(J.B.V.M.D.), Received
November
July
for
parameters
of the
static
and
dy-
namic anatomy of the anorectum-to be used as a baseline for the study abnormal anonectal function-and compare
tional
the
technique
with
of to
conven-
defecognaphy.
PATIENTS
AND
METHODS
Ten patients (five women and five men aged 28-42 years [mean, 33 years]) without bowel dysfunction (such as rectal blood loss, discharge of mucus, constipation, incontinence, or sequelae of pelvic surgery)
were
analyzed
with
a Philips
0.5-T imager (Gyroscan; Philips Medical Systems, Best, The Netherlands). The MR imaging was performed without preparations like introduction of contrast material, diet, or enema or any other manipulation. As a scout view, a single transverse section
of
10 mm
was
obtained
at the
1ev-
el of the sacrococcygeal junction to determine the midsagittal position with a spinecho pulse sequence (repetition time msec/echo time msec 400/30). Five adjacent sagittal sections 10 mm thick were obtained in the anorectum at rest with a pulse sequence of 350/30 (acquisition time, 300 seconds). In the optimal planedemonstrating the plica of Kohlrausch (plica transversales recti) in the rectal wall, the sacrococcygeal junction, and the rectal opening-images were obtained twice at rest, twice during perineal contraction (squeezing), and twice during straining (Valsalva maneuver). This was done with a single-section gradient-echo sequence (100/14; flip angle, 60#{176}) in an acquisition time of 20.4 seconds per image. With a 1.5-T MR imager, higherquality images can be obtained in a considerably shorter acquisition time (10). A spatial resolution of approximately 1.5 mm was obtained with a field of view of 300 mm and an acquisition matrix of 204 256 (Figs 1, 2). Image acquisition was performed twice in every stage, with an interval of 10 minutes between the two series of images obtained at rest, during perineal contraction, and during straining. The patients were examined while in the prone position, enabling air to collect
x
19.
Abbreviation:
SD
standard
deviation.
159
in the rectum and creating an excellent natural contrast medium for MR imaging (11). The images were obtained during suspended respiration, and the patients were asked to void before the examination to prevent compression of the recturn by a full bladder. The anorectal junction is defined as the intersection point of the line along the posterior border of the distal part of the rectum and the central axis of the anal canal. The angle between these lines is defined as the anorectal angle in concordance with conventional defecography (2,12-16). The position and mobility of the anorectal junction are related to the symphysiosacral baseline (Figs lb. 2b). Measurement of the anorectal angle and position of the anorectal junction was performed blinded. The measurements for each parameter were evaluated twice by each of two observers (R.H.K. and J.B.V.M.D.), with an interval of 1 week between evaluations, to analyze the interand intraobserver variations and compare them with the findings at conventional defecography (2,17). As a parameter for posterior rectal wall mobility,
the
distance
of the
dorsal
part
with
conventional
radiography (17). The
during barium small sample size
determination
of sex-
and
enema did not
lateral
study allow
age-related
normal values. To obtain an impression of whether the static and dynamic anatomy of the anorectum in the sitting position (in which defecography is performed) and in the prone position (in which MR imaging is performed) is equivalent, we examined five patients who also underwent conventional defecography in the same session in the prone position. The anorectal angle and the position of the anorectal junction at rest, during penineal contraction, and during straining were compared for the sitting and prone positions.
The interand intnaobservem variations of the parameters as measured with MR imaging are listed in Tables 1 and 2. The values are much smaller those obtained The difference
with defecogmafor the anomectal
angle at rest is illustrated in Figure 3 and compared with the intemobservem variation at defecogmaphy, as analyzed by Goei et al (2). Although not shown, graphs of the other parameters showed similar results (2). In this study, the mean difference between observers 1 and 2 was 4. 1 #{176}(maximum, 14#{176}) for the anomectal angle. For the position of the anorectal junction, the mean difference was 2.1 mm
160
Radiology
#{149}
I
.
a.
coccyx
b.
Figure 2. (a) MR image of the same volunteer as in Figure 1 during straining, obtained with the same technique. A anterior, P posterior. (b) Schematic representation of a. Note increase in anorectal angle (ARA) and descent of the anorectal junction and plica of Kohlnausch (PK) in relation to the baseline when compared with Figure lb. Also, the coccyx demonstrates a considerable descent during straining.
Table 1 Interobserver Variation in the Measurement 10 Asymptomatic Subjects
of Parameters
with
Anorectal Angle
Mean
difference
Mean Mean
difference difference
Note.-ARJ the dorsal plica
RESULTS
than phy.
1. (a) MR image obtained in the sagittal plane with gradient-echo pulse sequence (100/ 14; flip angle, 60#{176}; acquisition time, 20.4 seconds) with the patient at rest. A anterior, P = posterior. (b) Schematic representation of a. ARA = anorectal angle, BASELINE = junction line between cranial symphysis pubis and distal sacrum, PK = plica of Kohlrausch.
of
the plica of Kohlrausch in the rectal wall from the symphysiosacral baseline was measured. These values were also determined twice by the two observers. The distance between the internal nectal wall and the sacral bone was measured in concordance
b.
Figure
=
± SD at rest ± SD during ± SD during
perineal straining
3.5#{176} ± 4.5#{176} 4.9#{176} ± 6.6#{176} 4.0#{176} ± 4.9#{176}
contraction
distance of the anorectal junction from of Kohlrausch from the symphysiosacral
(maximum, 9 mm), and for the position of the plica of Kohlrausch the mean difference was 1 .4 mm (maximum, 7 mm). The mean intmaobserver variation for observers 1 and 2 was 3.1#{176} (maximum, 11#{176}) for the anorectal angle. For the position of the anonectal junction, the mean intraobserver variation was 1.9 mm (maximum, 10 mm), and for the position of the plica of Kohlmausch the mean intmaobserven variation was 08 mm (maximum, 4 mm). The mean values and standard deviations (SDs) for the anorectal angle and distance of the
the symphysiosacral baseline.
MR Imaging ARJ (mm)
PK (mm)
2.6 ± 3.8 1.9 ± 2.7 1.9 ± 2.4 baseline,
In
1.5 ± 1.7 1.5 ± 2.4 1.3 ± 2.7 PK
=
distance
of
anorectal junction and dorsal plica of Kohlrausch from the symphysiosacmal baseline at rest, during penineal contraction, and during straining and the interpatient variations are listed in Table 3. The mean differences and SDs between the measurements in every stage (the intrapatient variation) are listed in Table 4. The distance of the coccyx from the baseline changes between nest and straining; a descent of up to 26 mm was measured. The rectosacral space was between 2 and 6 mm (mean, 3.5 mm).
April
1991
Table
Intraobserver 10 Asymptomatic
Variation in the Subjects
Measurement
o f Parameters
with
Anorectal Angle Mean Mean Mean
difference difference difference
Note.-ARJ the dorsal plica
Table
=
± SD at rest ± SD during ± SD during
perineal straining
from
junction from the symphysiosacral
MR
I maging
ARJ (mm)
the symphysiosacral baseline.
in
PK (mm)
1.9 ± 2.3 2.1 ± 2.7 1.8 ± 3.5
3.4#{176} ± 4.2#{176} 3.5#{176} ± 2.4#{176} 2.5#{176} ± 2.7#{176}
contraction
of the anorectal
distance
of Kohlrausch
baseline,
0.9 ± 1.2 0.8
±
1.2
0.7 ± 1.0 PK
=
distance
of
Variation
Asymptomatlc
in the
Measurement
of Parameters
with
during during
-
Note.-ARJ the dorsal plica
obs.1
Im aging
MR
in 10
Subjects Anorectal Angle
Atrest±SD (At rest (At rest
-
=
perineal straining)
contraction) ± SD
distance of the anorectal junction from of Kohlrausch from the symphysiosacral
ARAobs.2
ARJ (mm)
109#{176}±8#{176} -10#{176}± 8#{176} 29#{176} ± 21’
± SD
cograms. dorsal during sumed.
A mean elevation plica of Kohlmausch perineal contraction The spatial resolution
of the of 1.1 mm was meaof the
MR imager was approximately mm, therefore the true elevation the plica was not detected. The mean descent on straining was
3
Interpatient
ARA
from the pubosacral baseline at rest, during penineal contraction, and duming straining, could be identified on MR images in 10 asymptomatic subjects (Figs 1, 2). This anatomic landmark is generally not visible on defe-
2
PK (mm)
18±6 -4 ± 5 19 ± 11
the symphysiosacral baseline.
baseline,
-3
PK
=
0±4 ± 3 5 ± 5
distance
of
(DEGREES)
1.5 of 4.8
mm (range, -1-15 mm). This small group justifies an estimation of normal values as mean ± 3 SD. This means a normal value for the distance of the dorsal plica of Kohlrausch from the symphysiosacral baseline of 0 ± 16 mm and, for the descent during straining, of 4.8 ± 15 mm. The intrapatient variation (Table 4) and the interand intraobserver variations in the measurement of the position of the plica of Kohlmausch (Tables 1, 2) were limited. Pathologic mobility of the rectum plays a role in internal intussusception of the rectum and possibly in the event of anterior wall prolapse of the rectum in mectocele (18,19). Pathologic mobility of the rectum is therefore considered a reason for operation in carefully selected patients (12,13,15, 19,20,21-25). This study provides two parameters for rectal mobility: (a) the mobility of the anomectal junction ventrally and (b) the mobility of the
posterior
rectal
wall
at the
origin
of
the plica of Kohlmausch dorsally. In posterior mectopexy, the mobility of the posterior rectal wall is reduced. Therefore, we expect this measurement to have the potential to help in 60
80
100
120
ARAobs.1+
ARAobs.2 (DEGREES)
2
Figure
3.
Difference
at rest, as measured ing (#{149}).(Reprinted,
1 60
140
versus mean for the values for anorectal angle by observers 1 (obs.l) and 2 (obs.2) at defecography with permission, from reference 2.)
(ARA) (0)
with the patient and at MR imag-
deciding whether to perform posterior mectopexy, an operation with the aim of preventing excessive descent of the rectum during straining by fixation of the posterior rectal wall in the area of the plica of Kohlmausch to the periosteum of the sacral bone.
Anorectal In the five defecography studies, measurements of the anorectal angle and of the position and mobility of the anorectal junction were made at rest, during perineal contraction, and during straining in the sitting and prone positions. The differences in these measurements between the sitting and prone positions were as follows: for the anorectal junction, 1.0 ± 4.5 at rest, -0.4 ± 2.7 during contraction, and 0 ± 2.7 during straining; for the anorectal angle, 1.0 ± 3.9 at rest, -1.0 ± 6.4 during contraction, and
Volume
179
#{149} Number
1
0.4 ± 6.0 during straining. therefore no substantial the values between the prone positions.
There was difference in sitting and
Angle
The mobility of the posterior rectal wall in the craniocaudal direction, determined by measuring the distance of the anatomic landmark the
Determination of the anomectal angle plays a mole in analysis of patients with rectal incontinence. If this angle is above 130#{176} at rest, rectal incontinence is likely to be due to a distumbance in muscle activity on the part of the pubomectal sling (16,20). The mean values of the anomectal angle on the MR studies (Table 3) are in the same mange as those found in defecography of asymptomatic subjects (2,13,19), but the intempatient vamia-
dorsal
tion
DISCUSSION Plica
of Kohlrausch
part
of the
plica
of Kohlrausch
is less
in MR
imaging.
The
Radiology
mean
161
#{149}
interobservem variation was 4.10 (maximum, 14#{176}) which is far less than in defecogmaphy (mean, 11.2#{176}; maximum, 33#{176}) (Fig 3). This might explain the smaller intempatient vanation in MR imaging. This small group justifies an estimation of normal values as mean ± 3 SD or 109#{176} ± 24#{176} at rest, 99#{176} ± 24#{176} during perineal contraction, and 138#{176} ± 63#{176} during straining. In contrast with defecography (2), the observers agreed on increase or decrease of the anomectal angle in every patient during penineal contraction and straining in the
MR
studies.
The
observer
variations
of the measurements are mainly due to variations in drawing the tangent in the curved caudal inner rectal wall. The rectal opening is rather consistently identified, because of the perianal fat in the sagittal plane (Figs 1, 2). The difference between the measurements in the same stage in the same patients (Table 4) is memarkable. The mean overall difference is 7.6#{176},with a maximum of 33#{176} during straining. These values are considerably higher than the intraand interobserver variations (Tables 1, 2). To our knowledge, these intrapatient variations have never before been determined for defecographic studies.
Anorectal
Junction
Determination of the position and mobility of the anorectal junction plays a role in the diagnosis of the spastic pelvic floor syndrome, frequently encountered in the solitary rectal ulcer syndrome (12,18) and in the penineal descent syndrome (18). From dynamic defecogmaphy studies, it is known that during straining the descent of the anomectal junction is less than 2 cm in normal subjects. If this junction demonstrates a descent as fan as 3 on 4 cm below the pubococ-
cygeal
line,
this
supports
the
diagno-
sis of perineal descent syndrome (16,18,20). In our study, a line could be drawn between the superior border of the pubis and the sacrococcygeal junction, and these landmarks were clearly detectable on all of the images of each subject. We prefer to relate the position of the anomectal junction to this symphysiosacral baseline instead of the symphysio-
coccygeal
baseline
(2), because
the
coccyx itself demonstrates a descent at MR imaging of up to 26 mm from rest to pemineal contraction (Figs 1, 2). The mean distances and SDs of the anomectal junction from the baseline in the presented MR studies are
162
Radiology
#{149}
Table 4 Intrapatient Asymptomatic
Variation in the Measurement Subjects
of Parameters
with
MR Imaging
Anorectal Angle Mean
difference
Mean
difference
Mean
difference
± SD at rest ± SD during ± SD during
perineal straining
ARJ (mm)
6.2#{176} ± 8.1 0 4.5#{176} ± 6.0#{176} 12.2#{176} ± 15.7#{176}
contraction
in 10 PK (mm)
39 45 4.1 ± 5.8 4.6 ± 5.9
1.9 ± 2.4 3.2 ± 3.5 4.6 ± 5.9
Note.-ARJ = distance of the anorectal junction from the symphysiosacral baseline, PK = distance of the dorsal plica of Kohlrausch from the symphysiosacral baseline. There were two series of MR examinations at rest, during perineal contraction, and during straining in every subject. with an interval of 10 minutes between them.
listed in Table 3. This small group justifies an estimation of normal values as mean ± 3 SD. This was 18 ± 18 mm at rest. During penineal contraction, the mean elevation was 4.6 mm (maximum, 14 mm; minimum, -6 mm). During straining, the mean descent was 18.6 mm (maximum, 32 mm; minimum, -3 mm). These values are in concordance with those of the study of Yang et al (10) published in this issue. In contrast to the analysis of defecogmaphic studies (2), the observers were equivocal on the dimection of the motion in the MR studies of all subjects; the mean interobserver variation was 2.1 mm, with a maximum of 9 mm. The mean interobserver variation of this parameter in defecogmaphy is approximately 15 mm (17). The intrapatient variation for this parameter was limited (Table 4). To our knowledge, this variation has never been determined for defecognaphy.
Rectosacral
Space
Determination of the width of the mectosacnal space is important in the diagnosis of malignancies, inflammatory processes, lipoma, and myelomeningocele (17). This space was invariably cleanly visible on our MR studies and ranged from 2 to 6 mm. The upper limit in a group of 100 subjects from the general population was 16 mm, as measured with lateral radiography during barium enema examination (17).
earlier, the advantages of defecogmaphy were (a) the physiologic position of the patient, sitting on a modified toilet chair, and (b) the ability to visualize the act of defecation itself, enabling visualization of rectal prolapse and internal rectal intussusception. The advantages of MR imaging over defecography were as follows:
(a) MR
MR
seems
Comparing cognaphy
MR in the
of no relevance. imaging manner
imaging
with defedescribed
no ill biologic
ef-
of the
anorectum
with
References 1.
Wallden
deep
L.
Defecation
rectogenital
1952;(suppl
2.
3.
4.
in
Acta
cases
Chir
of
Scand
165):1-121.
subjects.
Radiology
1989;
141. Mezzacappa PM, Price MR and CT demonstration
173:137-
Haller JO, et al. of levator sling in congenital anorectal anomalies. J Comput Assist Tomogr 1987; 11:273-275. Sato Y, Pringle KC, Bergman RA, et al. Congenital
5.
block
pouch.
Goei R, van Engelshoven J, Schouten H, Baeten C, Stassen C. Anorectal function: defecographic measurement in asymptomatic
Comparing values for anorectal angle and position of the anomectal junction obtained with defecogmaphy for the sitting and prone positions mevealed no substantial difference. Thus, the difference in position of the patient during defecogmaphy and
has
the patient in the prone position, as gas collects in the rectum, creating an excellent natural contrast medium (11). (c) The intemobservem variation for the position of the anorectal junction and value of the anomectal angle at rest, during perineal contraction, and during straining at MR imaging is far less than that at defecography (Fig 3) (2). (d) Movements of the posterior rectal wall at the level of the plica of Kohlrausch can be analyzed with MR imaging. We believe that MR imaging is a method well suited for anomectal measurements, yielding better-defined baseline values than defecography. U
CONCLUSIONS
MR study
imaging
fects, to our knowledge, whereas irradiation during lateral pelvic fluoroscopy and radiography is considerable, with a gonadal dose of up to 12 mSv (20). (b) Opacification of the vagina and rectum is not necessary for
AP,
anorectal
anomalies:
MR
aging. Radiology 1987; 168:157-162. de Lange EE, Fechner RE, Wanebo Suspected
recurrent
rectosigmoid
ma after
abdominoperineal
imaging
and
histopathologic
diology
1989;
170:323-328.
im-
HJ. carcino-
resecton:
MR
findings.
Ra-
April
1991
6.
7.
Fishman-Javitt MC, Lovecchio JL, Javors B, et al. The value of MRI in evaluating perirectal and pelvic disease. Magn Reson Imaging 1987; 5:371-380. Bryan
PJ,
Butler
HE,
LiPuma
JP.
13.
14.
Magnet-
ic
8. 9.
10.
11.
12.
resonance imaging of the pelvis. Radiol Clin North Am 1984; 22:897-915. Heiken JP, Lee JKT. MR imaging of the pelvis. Radiology 1988; 166:11-16. Zilch HG, Baumgartl FW. MR anatomy des weiblichen beckens. Rontgenpraxis 1986; 39:407-414. Yang A, Mostwin JL, Rosenshein NB, Zerhouni EA. Pelvic floor descent in wornen: dynamic evaluation with fast MR imaging and cinematic display. Radiology 1991; 179:25-33. Bryan PJ, Butler HE, LiPuma JP. NMR scanning of the pelvis: initial experience with a 0.3 T system. AJR 1983; 141:11111118. Goei R, Baeten C, Janevski B, Engelshoven J. The solitary rectal ulcer syndrome: diagnosis with defecography. AJR 1987; 149:933-936.
Volume
179
Number
#{149}
1
1 5.
16.
17.
18.
19. 20.
Ch#{243}i DL, Ekberg 0. Functional analysis of anorectal junction. ROFO 1988; 148:5053. Mahieu P, Pringot J, Bodart P. Defecography. I. Description of a new procedure and results in normal patients. Gastrointest Radiol 1984; 9:247-251. Pescatori M, Ravo B. Diagnoitic anorectal functional studies. Surg Clin North Am 1988; 68:1231-1248. Mahieu P, Pningot J, Bodart P. Defecography. II. Contribution to the diagnosis of defecation disorders. Gastrointest Radiol 1984; 9:253-261. Chnispin AR, Fry 1K. The presacral space shown by barium enema. Br J Radiol 1963; 36:319-322. Goei R, Baeten C, Arends JW. Solitary rectal ulcer syndrome: findings at barium enema study and defecography. Radiology 1988; 168:303-306. Ekberg 0, Nylander G, Fork FT. Defecography. Radiology 1985; 155:45-48. Broden B, Snellman B. Procidentia of the rectum studied with cineradiography. Dis Colon Rectum 1968; 11:330-347.
21.
Hoffman
MJ,
Kodner
intussusception
22.
23.
25.
rectum:
Internal
diagnosis
and surgical management. Dis Colon Recturn 1984; 27:435-441. Kruyt RH. Dynamische defaecografie. In: Maag-, darmen leverziekten 1989. 1st ed. Aalsmeer,
The
loren, Berman
1990; 90-94. IR, Manning
Wright
K.
agnosis
and
prolapse.
24.
IJ, Fry RD.
of the
Netherlands:
DH,
Anatomic Dis
Dudeley-
specificity
treatment Colon
Mur-Kostver-
in
of internal Rectum
1985;
826. Johanson bances
C, Ihre T, Ahibach SO. in the defecation mechanism
special
reference
to intussusception
the
di-
rectal 11:816-
Disturwith of the
rectum (internal procidentia). Dis Colon Rectum 1985; 12:920-924. Kruyt RH, Delemarne JBVM, Gooszen HG, Vogel HJ. On selecting patients for postenor rectopexy. Br J Sung 1990; 10:11831184.
Radiology
#{149} 163