- Email: [email protected]
Anatomic and physiologic measurements of the internal and external anal sphincters in normal females
Anatomic and Physiologic Measurements of the Internal and External Anal Sphincters in Normal Females DEE E. FEiVNER, MD, J. SCOTT KRIEGSHAUSER, MD, HENRY H. LEE, MD, ROBERT W. BEART, MD, AMY WEAVER, MS, AND JEFFREY L. CORNELLA, MD Objective: To assess the correlation between anal sphincter magnetic resonance imaging (MRI) measurements and manometric anorectal vectography pressures. Methods: Ten healthy, nulliparous women underwent anal sphincter MRI with examination of sagittal, axial oblique, and coronal planes. Anal manometry was performed with a radial eight-channel catheter. Customary functional measurements were recorded, including anterior and posterior sphincter length, squeeze length, length of the high-pressure zone, and maximal resting and squeeze pressures. The Spearman rank correlation coefficient was used to assess correlation. Results: The manometric squeeze length and the manometric length to maximum squeeze pressure were correlated negatively with the posterior sphincter length by MRI (P = .049 and .044, respectively). The manometric high-pressure zone squeeze length was correlated positively with the posterior sphincter length by MRI (P = .042). The mean * standard deviation (SD) posterior sphincter length was 27.3 + 6.0 mm. Anatomically, the cylindric shape of the anal sphincter is characterized by a gradual increase in muscle thickness cephalad. The external striated sphincter was much thicker posteriorly (24.7 & 4.6 mm) than anteriorly (6.6 f 1.7 mm) in the proximal or caudal third. The proximal internal smooth muscle sphincter was nearly equal in thickness anteriorly and posteriorly (9.0 f 1.4 mm and 9.6 f 1.7 mm, respectively). Although variation in the thickness of both the smooth and striated muscle was found, manometric pressures did not correlate with the muscle thickness along the sphincter. Conclusion: The length of the anal sphincter correlated positively with the functional information, as determined by manometry. An anal sphincter length of 3 cm is consistent, from an anatomic and functional view, in these ten normal women. (Obstet Gynecol 1998;91:369-74. 0 1998 by The American College of Obstetricians and Gynecologists.)
From the Section of Gynecologic the Mayo Foundation, Scottsdale,
VOL.
91, NO.
3, MARCH
1998
Surgery, Arizona.
Muyo
Clinic
Scottsdale
and
The anal sphincter is divided into internal and external components. The internal anal sphincter, or visceral component, is a continuation of the circular smooth muscle of the rectum. The internal sphincter, under autonomic control, provides approximately 85% of the maximum resting pressure recorded during manometric evaluation.’ Manometric determinations of the maximum resting pressures have implied an internal sphincter length of 3-5 cm in women.’ Anatomic depiction of the internal sphincter is a thinned, attenuated muscle with an anterior length of 2 cm and a posterior length of 3 cm.3 The somatic, or external, component of the anal sphincter is a striated skeletal muscle under voluntary control. It is composed of the funnel-shaped levator ani muscles and the external anal sphincter muscles. The external sphincter exerts the maximum recorded voluntary squeeze pressure during manometric evaluation. The external sphincter has been described as a cylinder encircling the anal canal, with a narrowing and shortening of the muscle anteriorly to a band 1 cm in diameter and 2 cm in length.4 Magnetic resonance imaging (MRI) is an effective and safe method for evaluating the rectal musculature.5,6 Aronson and Lee6 found the shape of the anal sphincters to be nearly cylindric, with an anterior component averaging 18.3 mm in thickness and 28.0 mm in length. Fifty-four percent of the anterior thickness was attributed to the internal sphincter.6 The substantial thickness of the anterior sphincter contrasts markedly with that in previous reports7 but is consistent with manometric profiles of the anal canal. Resting pressures are primarily the result of tonic contraction of the internal anal sphincter.’ The external sphincter is studied by measuring the maximum squeeze pressure exerted by a sustained voluntary contraction. The purpose of this study is to correlate the lengths and thicknesses of the internal
0029-7844/98/$19.00 PI1 SOO29-7844(97)00678-9
369
and external anal sphincters, as measured by MRI, with anorectal pressures measured by manometry. Whereas previous studies2,4’6 have evaluated anatomic or physiologic characteristics independently, this study compares directly the anatomy and functional measurements of normal women.
Materials and Methods The study was approved by the Mayo Clinic Institutional Review Board, and all subjects gave written consent. Ten healthy, nulliparous female volunteers, with a mean age of 28 years and a range of 22-34 years, participated in this study at the Mayo Clinic, Scottsdale, Arizona. All subjects were solicited through advertisement. Detailed medical, gynecologic, and colorectal histories were obtained before enrollment. Subjects were free of any previous anorectal or vaginal trauma or surgery. Exclusion criteria included pregnancy, a past pathologic process involving the rectum or anus, a history of gastrointestinal disease, fecal incontinence, obstipation, medications known to affect gastrointestinal motility, and physical findings showing evidence of pelvic neuropathy or pelvic mass. On the day of testing, each subject received a hypertonic sodium phosphate enema (C.B. Fleet, Inc., Lynchburg, VA) 1 hour before testing. A tampon was placed in the vagina for both MRI and manometric studies to improve imaging during MRI studies. Fifty percent of the patients had manometric studies before MRI and 50% had MRI before manometric testing. All manometric evaluations were performed by the same investigator, and all MRI and assessment, by a single radiologist. For manometric evaluations, subjects were in the left lateral decubitus position with knees and hips flexed. Anal pressure profiles were measured with a low compliant recording system Synectic PC polygraph version 5.0 (Synectics Medical AB, Stockholm, Sweden), and eight-channel flexible catheter 4.8 mm optical density with a spiral and radial configuration of eight ports beginning 3 cm from the distal tip. The catheter was perfused continuously with 25 PL of distilled water per minute. Zero-pressure calibration was done at the level of the anal orifice before insertion. The catheter was well lubricated, and no anesthesia was used. Two resting and squeeze pressure profiles were obtained using a continuous pull method with a rate of 1 mm per second. Proximal or cephalad measurements were begun at 7 cm above the anal verge, and middle and distal measurements were obtained as the catheter was withdrawn. Magnetic resonance imaging was performed with insertion of a HO pediatric catheter, the same diameter as the manometric probe, into the anorectal canal at the
370
Fenner et al
MRI/Anal
Sphincter
Measurements
exact location of the manometric catheter at its highest point during manometric evaluation. The MRI technique was modified from that of Aronson and Lee.’ Each subject was supine in the MRI unit, with a coupled dual 5-inch coil system positioned above and below the perineum. Scout images were obtained in the sagittal and coronal planes by using a spin echo (repetition time 400/echo time 20) pulse sequence with a 128 X 256 matrix, 24-cm field of view, and 1 number of excitations or acquisitions to ensure correct coil placement. These scout images were used to prescribe the sagittal spin echo sequence (repetition time 250/echo time ll), 256 X 256 matrix, 20-cm field of view, 2 number of excitations or acquisitions, and 3-mm-thick contiguous slices. From the midsagittal image, oblique axial images perpendicular to the plane of the anal canal were obtained using a spin-echo sequence (repetition time 250/echo time 15) with 256 X 256 matrix, 20-cm field of view, 2 number of excitations or acquisitions and 3-mm-thick contiguous slices through the anorectal musculature. Magnetic resonance imaging measurements were obtained with electronic calipers directly from the magnetic resonance images as follows. Measurements of the internal anal sphincter were obtained from the axial Tl weighted images and included the thickness anteriorly, posteriorly, right, and left. The mucosa and submucosa and the longitudinal muscle layer were included in these measurements. The anterior and posterior lengths of the external anal sphincter were obtained from the midline sagittal Tl weighted images. The thicknesses anteriorly and posteriorly were obtained at the proximal, middle, and distal portions of the external anal sphincter. The measurements included the entire wall thickness from the mucosa to the outer external anal sphincter, and external anal sphincter thickness was calculated by subtracting the measurements of the internal anal sphincter. Manometric pressure measurements and lengths were determined by the Synectic PC polygraph version 5.0, (Synectics Medical AB). The distributions of the MRI and manometric measurements were summarized by means 2 standard deviations (SD). Correlations between MRI and manometric measurements were assessed using the Spearman rank correlation coefficient. All calculated P values were two-tailed with a null hypothesis of no correlation. The lengths of the anterior and posterior sphincters, as determined by MRI, were compared with the manometric resting and squeeze lengths, the high-pressure zone resting and squeezing lengths, and the length to maximum pressure during resting and squeeze profiles. The thicknesses of the internal and external anal sphincters in the anterior, posterior, right, and left quadrants were compared with pressure readings along the sphincter for resting and squeeze profiles, respectively.
Obstetrics
& Gynecology
Length, 0
5
10
15
mm 20
25
30
35
I I I I I I I 1 Anterior
(22.6 &5.8 (range
14.8-36.2))
Distal Anal Canal
Figure 1. Magnetic resonance imaging scan in axial with labeled anatomic structures. AI = anterior inferior; right.
oblique plane L = left; R =
Results Figures 1 and 2 show the MRI scans from one woman in the axial oblique and midline sagittal planes. Scans show the sphincter to be cylindric in shape, with a mean (+- SD) anterior length of 22.6 t 5.8 mm and a mean (I SD) posterior length of 27.3 2 6.0 mm (Figure 3). The internal sphincter was nearly identical anteriorly and posteriorly in thickness, with a narrowing on the sides (Figure 4). Ant eriorly, the internal anal sphincter contributed greater than 50% of the total thickness of the sphincter. The external anal sphincter was thicker posteriorly than anteriorly throughout the entire length of the sphincter, with a marked thickening cephalad. Table 1 shows the standard measurements obtained
I
Posterior
Figure 3. Mean (2 anterior and posterior evaluation.
(27.3 56.0 (range
19.0-39.2))
standard deviation and range) length of the anal sphincters by magnetic resonance imaging
during manometric evaluation. There was a positive correlation (P = .042) between mean (i: SD) posterior sphincter length (2.7 2 0.6 cm) as measured by MRI, and mean (t SD) high-pressure zone length with squeeze (3.1 * 0.8 cm) recorded on manometry (Table 2). The high-pressure, or continence, zone, is defined as the zone recorded during manometry where the pressures exceed greater than 50% of the maximum pressure recorded.8 The squeeze and resting pressures tended to be highest in the midportion of the anal canal, with a decrease in pressures distally (Figures 5 and 6). There was no significant correlation between manometric pressures at rest, or with squeeze, and the muscle thickness of the internal or external sphincters at corresponding sites along the canal. When evaluating the sequence effect, there was no apparent difference between the group that had MRI, versus manometry, first.
Discussion
Figure 2. Magnetic resonance imaging scan in sagittal plane with labeled anatomic structures. P = proximal or cephalad; M = middle; D = distal or caudad.
VOL.
91, NO.
3, MARCH
1998
Magnetic resonance imaging of the pelvis, anal manometry, ultrasonography of the internal and external anal sphincters, and nerve studies to evaluate pudendal neuropathies are being used by gynecologists and colorectal surgeons to evaluate obstetric injuries and defecation disorders.’ Before evaluating and treating pathologic conditions, it is important to understand normal structural and functional anatomy and physiology. Our MRI scans found the mean (2 SD) anterior length to be 22.6 ? 5.8 mm and the mean (rt SD) posterior length, 27.3 ? 6.0 mm. The length of the internal sphincter has been thought to correspond with the length of the high-pressure zone obtained during
Fenner
et al
MRI/Anal
Sphincter Measurements
371
Figure 4. Average thickness (mm) of anal sphincters by magnetic resonance imaging evaluation. The internal sphincter is represented by davk shading, and the external sphincter, by light gray. L = left; R = ngnr.
Posterior
Proximal anal manometry. Our study confirmed this correlation and found a positive relationship between the posterior sphincter length measured by MRI and the length of the high-pressure zone with squeeze determined by anal manometry (P = .042). The high-pressure zone is an important factor in the maintenance of fecal continence. We found the mean (2 SD) length of the high-pressure zone to be 3.0 2 0.6 cm at rest and 3.1 i 0.8 cm with squeeze. These findings are consistent with studies by Nivatvongs et al,iO who found the high-pressure zone in women to be 3.7 cm in length. Both anatomic and physiologic tests demonstrate the length of the continence zone to be at least 3 cm. Surgeons performing operations such as repair of acute or chronic thirddegree sphincter lacerations may not recognize the need to extend the repair cephalad to this 1evel.l’ Our findings support an anatomic and functional repair that extends 3 cm in length cephalad to maintain this high-pressure, or continence, zone. During squeeze, the anal canal lengthens as the
Table
1. Summary
Mean
Length to maximum resting pressure (cm) Length to maximum squeeze pressure (cm) Resting length (cm) Squeeze length (cm) Maximum resting pressure (mmHg) Maximum squeeze pressure (mmHg) High-pressure zone resting length (cm) High-pressure zone squeeze length (cm) SD = standard
372
Fenner
SD
Minimum
Maximum
3.6
0.5
2.8
4.6
4.8
1.0
3.5
6.8
5.4 6.6 150.6
0.5 1.0 33.4
4.6 5.3 81.1
6.0 8.4 184.6
MRI Anterior
200.3
27.7
150.4
3.0
0.6
2.2
3.9
3.1
0.8
2.0
5.0
Posterior
MRl/Anal
2. Correlation Length and length length
257.9
length
Between Magnetic Resonance Imaging Manometry
Sphincter
Measurements
Spearman rank correlation coefficient
Manometry Length to maximum resting pressure Length to maximum squeeze pressure Resting length Squeeze length High-pressure zone resting length High-pressure zone squeeze length Length to maximum resting pressure Length to maximum squeeze pressure Resting length Squeeze length High-pressure zone resting length High-pressure zone squeeze length
MRI = magnetic resonance Two-tailed P > .05.
deviation.
et al
Table
Manometry Measurements
of Anal
Measurement
external sphincter contracts, and the puborectalis muscle shortens and elevates the anal canal. By anal manometry, the mean (t SD) resting length was 5.4 2 0.5 cm, and with squeeze, the canal lengthened to 6.6 2 1.0 cm. Whereas there was no significant correlation between the manometric length and the anatomic length at rest, there was a positive correlation between the posterior length as measured by MRI and the length of the high-pressure zone with squeeze (P = .042). In other words, the longer the posterior sphincter measured on MRI, the longer the high-pressure zone measured on manometry. A negative correlation was seen
imaging;
-0.13
(I’ = NS)
-0.33
(P = NS)
0.19 (P = NS) -0.32 (I’ = NS) 0.51 (P = NS) 0.49 (I’ = NS) 0.01 (P = NS) -0.64
(P = ,044)
0.25 (P = NS) -0.63 (P = .049) 0.51 (P = NS) 0.65 (P = .042)
NS = not significant.
Obstetrics 8 Gynecology
Figure 5. Average sures by manometry left; R = right.
resting (mmHg).
presL =
Posterior
Proximal
Distal when comparing the posterior length and the manometric length with maximum squeeze pressure and the total squeeze length (P = ,044 and .049, respectively). The shorter the posterior length of the sphincters on MRI, the more cephalad were the maximum pressures recorded with squeeze for each participant. Physiologically, squeeze efforts prevent leakage upon presentation of enteric contents to the proximal canal at a socially inopportune time. Squeeze pressure is not responsible for maintaining fecal continence from hour to hour.’ The length of the high-pressure zone and the pressure difference between rectum and anal canal aid continence. As enteric contents descend, individuals with a shorter sphincter length may maintain continence by a greater lengthening of the anal canal and high-pressure zone. They may have compensated for a shorter canal by developing increased strength of the puborectalis. Taylor and Beartl’ found that squeeze pressures in the proximal anal canal were highest
Figure 6. Average sures by manometry left; R = right.
squeeze (mmHg).
posteriorly and lowest anteriorly. Our study found the squeeze pressures in the proximal anal canal to be distributed equally anteriorly and posteriorly, and squeeze pressures were higher posteriorly than anteriorly in the distal anal canal. Posterior resting pressures have been shown to be highest along the posterior-proximal and posteriordistal segments.13 McHugh and Daimant14 showed that the anterior resting pressures in women were highest distally in the anal canal, but in men, they were highest proximally. Our study showed the anterior resting pressure to be highest in the midportion of the anal canal. We found no significant correlation between muscle thickness and pressures at corresponding sites along the canal. Sultan et a1,15 when comparing anal endosonography and manometry, also found no correlation between internal and external anal muscle thicknesses and manometric pressures. The lack of
presL =
Posterior
Proximal
VOL.
91, NO.
3, MARCH
1998
Fenner
et al
MRI/Anal
Sphincter
Measurements
373
correlation may be due to the fact that anal manometry is not only a measurement of pressure exerted by the sphincter muscles, but also a measure of static resistance and tissue turgor. In addition, the force exerted during squeeze is dependent on voluntary effort. Other considerations are the limited sample size in their investigation, in which there is only a 49% chance of detecting correlations beyond 2 0.60 SD that are significantly different from the null hypothesis of no correlation. There also is some controversy about the basic anal anatomy as seen on MRI.r6 In our study, using Tl weighted magnetic resonance images, the internal anal sphincter was not differentiated from the mucosa, submucosa, or longitudinal muscle layer. However, the thicknesses of these other structures are fairly constant, as has been shown on recent studies using T2 weighted images and on endosonography.i5-i7 Endosonography and T2 weighted images show better differentiation of the mucosa and submucosa from the internal sphincter, but generally less distinct separation of the external anal sphincter from the longitudinal muscle layer. Discrepancies between our measurements and those in other studies are accounted for by the inclusion or exclusion of adjacent layers. Sultan et ali5 included the longitudinal muscle layer with their external anal sphincter measurements. For our ten healthy patients, MRI and anal manometry complemented each other in evaluations of the important functional length of the anal sphincter musculature. Although the mechanism that maintains fecal continence is complex, with future use of techniques such as MRI and anal manometry along with neurophysiologic testing, we will begin to understand more fully this important neuromuscular unit.
References 1. Frenckwer B, Euler CV. Influence of pudendal block on the function of anal sphincters. Gut 1975;16:482-9. 2. Wunderlich M, Parks AG. Physiology and pathophysiology of the anal sphincters. Int Surg 1982;67:291-8. 3. Davey M. The riddle of the sphincters. Am Surg 1987;53:298-306.
374
Fenner et al
MRI/Anal
Sphincter Measurements
4. Oh C, Kark AE. Anatomy of the external anal sphincter. Br J Surg 1972;59:717-23. 5. McCarthy S, Vaguer0 E. Gynecologic anatomy with magnetic resonance imaging. Am J Obstet Gynecol 1986;155:255-9. 6. Aronson MI’, Lee RA. Anatomy of anal sphincters and related structures in continent women studied with magnetic resonance imaging. Obstet Gynecol 1990;76:846-51. 7. Wilde FR. The anal intermuscular septum. Br J Surg 1949;36:27985. 8. Coller JA. Clinical application of anorectal manometry. Gastroenterol Clin North Am 1987;16:17-33. 9. Pemberton JH. Surgery of the alimentary tract. Vol IV. 3rd ed. Philadelphia: WB Saunders Company, 1991. 10. Nivatvongs S, Stem HS, Fryd DS. The length of the anal canal. Dis Colon Rectum 1981;24:600-1. 11. Thompson JD, Rock JA. TeLinde’s operative gynecology. 7th ed. Philadelphia: WB Saunders Company, 1991. 12. Taylor BM, Beart RW. Longitudinal and radial variations of pressure in the human anal sphincter. Gastroenterology 1984;86: 693-7. 13. Coller SA. Clinical application of anorectal manometry. Gastroenterol Clin North America 1987;16:17-33. 14. McHugh SM, Diamant NE. Anal canal pressure profile: A reappraisal as determined by rapid pull through technique. Gut 1987;28:1234-41. 15. Sultan AH, Kamm MA, Hudson CN. Endosonography of the anal sphincters: Normal anatomy and comparison with manometry. Clin Radio1 1994;49:368-74. 16. Hussain SM, Stoker J, Lameris JS. Anal sphincter complex: Endoanal MR imaging of normal anatomy. Radiology 1995;197:671-7. 17. Schafer A, Enck I’, Furst G, Kahn T, Frieling T, Lubke HJ. Anatomy of the anal sphincters-Comparison of anal endosonography to magnetic resonance imaging. Dis Colon Rectum 1994;37:777-81.
Address reprint requests to: Jefwy L. Cornella, MD Mayo Clinic Scottsdale 13400 E. Shea Boulevard Scottsdale, AZ 85259 E-mail: [email protected]
Received July 7, 1997. Received in revised form October 20, 1997. Accepted October 29, 1997. Copyright 0 Gynecologists.
1998 by The American Published by Elsevier
College of Obstetricians Science Inc.
and
Obstetrics 6 Gynecology
From the Section of Gynecologic the Mayo Foundation, Scottsdale,
VOL.
91, NO.
3, MARCH
1998
Surgery, Arizona.
Muyo
Clinic
Scottsdale
and
The anal sphincter is divided into internal and external components. The internal anal sphincter, or visceral component, is a continuation of the circular smooth muscle of the rectum. The internal sphincter, under autonomic control, provides approximately 85% of the maximum resting pressure recorded during manometric evaluation.’ Manometric determinations of the maximum resting pressures have implied an internal sphincter length of 3-5 cm in women.’ Anatomic depiction of the internal sphincter is a thinned, attenuated muscle with an anterior length of 2 cm and a posterior length of 3 cm.3 The somatic, or external, component of the anal sphincter is a striated skeletal muscle under voluntary control. It is composed of the funnel-shaped levator ani muscles and the external anal sphincter muscles. The external sphincter exerts the maximum recorded voluntary squeeze pressure during manometric evaluation. The external sphincter has been described as a cylinder encircling the anal canal, with a narrowing and shortening of the muscle anteriorly to a band 1 cm in diameter and 2 cm in length.4 Magnetic resonance imaging (MRI) is an effective and safe method for evaluating the rectal musculature.5,6 Aronson and Lee6 found the shape of the anal sphincters to be nearly cylindric, with an anterior component averaging 18.3 mm in thickness and 28.0 mm in length. Fifty-four percent of the anterior thickness was attributed to the internal sphincter.6 The substantial thickness of the anterior sphincter contrasts markedly with that in previous reports7 but is consistent with manometric profiles of the anal canal. Resting pressures are primarily the result of tonic contraction of the internal anal sphincter.’ The external sphincter is studied by measuring the maximum squeeze pressure exerted by a sustained voluntary contraction. The purpose of this study is to correlate the lengths and thicknesses of the internal
0029-7844/98/$19.00 PI1 SOO29-7844(97)00678-9
369
and external anal sphincters, as measured by MRI, with anorectal pressures measured by manometry. Whereas previous studies2,4’6 have evaluated anatomic or physiologic characteristics independently, this study compares directly the anatomy and functional measurements of normal women.
Materials and Methods The study was approved by the Mayo Clinic Institutional Review Board, and all subjects gave written consent. Ten healthy, nulliparous female volunteers, with a mean age of 28 years and a range of 22-34 years, participated in this study at the Mayo Clinic, Scottsdale, Arizona. All subjects were solicited through advertisement. Detailed medical, gynecologic, and colorectal histories were obtained before enrollment. Subjects were free of any previous anorectal or vaginal trauma or surgery. Exclusion criteria included pregnancy, a past pathologic process involving the rectum or anus, a history of gastrointestinal disease, fecal incontinence, obstipation, medications known to affect gastrointestinal motility, and physical findings showing evidence of pelvic neuropathy or pelvic mass. On the day of testing, each subject received a hypertonic sodium phosphate enema (C.B. Fleet, Inc., Lynchburg, VA) 1 hour before testing. A tampon was placed in the vagina for both MRI and manometric studies to improve imaging during MRI studies. Fifty percent of the patients had manometric studies before MRI and 50% had MRI before manometric testing. All manometric evaluations were performed by the same investigator, and all MRI and assessment, by a single radiologist. For manometric evaluations, subjects were in the left lateral decubitus position with knees and hips flexed. Anal pressure profiles were measured with a low compliant recording system Synectic PC polygraph version 5.0 (Synectics Medical AB, Stockholm, Sweden), and eight-channel flexible catheter 4.8 mm optical density with a spiral and radial configuration of eight ports beginning 3 cm from the distal tip. The catheter was perfused continuously with 25 PL of distilled water per minute. Zero-pressure calibration was done at the level of the anal orifice before insertion. The catheter was well lubricated, and no anesthesia was used. Two resting and squeeze pressure profiles were obtained using a continuous pull method with a rate of 1 mm per second. Proximal or cephalad measurements were begun at 7 cm above the anal verge, and middle and distal measurements were obtained as the catheter was withdrawn. Magnetic resonance imaging was performed with insertion of a HO pediatric catheter, the same diameter as the manometric probe, into the anorectal canal at the
370
Fenner et al
MRI/Anal
Sphincter
Measurements
exact location of the manometric catheter at its highest point during manometric evaluation. The MRI technique was modified from that of Aronson and Lee.’ Each subject was supine in the MRI unit, with a coupled dual 5-inch coil system positioned above and below the perineum. Scout images were obtained in the sagittal and coronal planes by using a spin echo (repetition time 400/echo time 20) pulse sequence with a 128 X 256 matrix, 24-cm field of view, and 1 number of excitations or acquisitions to ensure correct coil placement. These scout images were used to prescribe the sagittal spin echo sequence (repetition time 250/echo time ll), 256 X 256 matrix, 20-cm field of view, 2 number of excitations or acquisitions, and 3-mm-thick contiguous slices. From the midsagittal image, oblique axial images perpendicular to the plane of the anal canal were obtained using a spin-echo sequence (repetition time 250/echo time 15) with 256 X 256 matrix, 20-cm field of view, 2 number of excitations or acquisitions and 3-mm-thick contiguous slices through the anorectal musculature. Magnetic resonance imaging measurements were obtained with electronic calipers directly from the magnetic resonance images as follows. Measurements of the internal anal sphincter were obtained from the axial Tl weighted images and included the thickness anteriorly, posteriorly, right, and left. The mucosa and submucosa and the longitudinal muscle layer were included in these measurements. The anterior and posterior lengths of the external anal sphincter were obtained from the midline sagittal Tl weighted images. The thicknesses anteriorly and posteriorly were obtained at the proximal, middle, and distal portions of the external anal sphincter. The measurements included the entire wall thickness from the mucosa to the outer external anal sphincter, and external anal sphincter thickness was calculated by subtracting the measurements of the internal anal sphincter. Manometric pressure measurements and lengths were determined by the Synectic PC polygraph version 5.0, (Synectics Medical AB). The distributions of the MRI and manometric measurements were summarized by means 2 standard deviations (SD). Correlations between MRI and manometric measurements were assessed using the Spearman rank correlation coefficient. All calculated P values were two-tailed with a null hypothesis of no correlation. The lengths of the anterior and posterior sphincters, as determined by MRI, were compared with the manometric resting and squeeze lengths, the high-pressure zone resting and squeezing lengths, and the length to maximum pressure during resting and squeeze profiles. The thicknesses of the internal and external anal sphincters in the anterior, posterior, right, and left quadrants were compared with pressure readings along the sphincter for resting and squeeze profiles, respectively.
Obstetrics
& Gynecology
Length, 0
5
10
15
mm 20
25
30
35
I I I I I I I 1 Anterior
(22.6 &5.8 (range
14.8-36.2))
Distal Anal Canal
Figure 1. Magnetic resonance imaging scan in axial with labeled anatomic structures. AI = anterior inferior; right.
oblique plane L = left; R =
Results Figures 1 and 2 show the MRI scans from one woman in the axial oblique and midline sagittal planes. Scans show the sphincter to be cylindric in shape, with a mean (+- SD) anterior length of 22.6 t 5.8 mm and a mean (I SD) posterior length of 27.3 2 6.0 mm (Figure 3). The internal sphincter was nearly identical anteriorly and posteriorly in thickness, with a narrowing on the sides (Figure 4). Ant eriorly, the internal anal sphincter contributed greater than 50% of the total thickness of the sphincter. The external anal sphincter was thicker posteriorly than anteriorly throughout the entire length of the sphincter, with a marked thickening cephalad. Table 1 shows the standard measurements obtained
I
Posterior
Figure 3. Mean (2 anterior and posterior evaluation.
(27.3 56.0 (range
19.0-39.2))
standard deviation and range) length of the anal sphincters by magnetic resonance imaging
during manometric evaluation. There was a positive correlation (P = .042) between mean (i: SD) posterior sphincter length (2.7 2 0.6 cm) as measured by MRI, and mean (t SD) high-pressure zone length with squeeze (3.1 * 0.8 cm) recorded on manometry (Table 2). The high-pressure, or continence, zone, is defined as the zone recorded during manometry where the pressures exceed greater than 50% of the maximum pressure recorded.8 The squeeze and resting pressures tended to be highest in the midportion of the anal canal, with a decrease in pressures distally (Figures 5 and 6). There was no significant correlation between manometric pressures at rest, or with squeeze, and the muscle thickness of the internal or external sphincters at corresponding sites along the canal. When evaluating the sequence effect, there was no apparent difference between the group that had MRI, versus manometry, first.
Discussion
Figure 2. Magnetic resonance imaging scan in sagittal plane with labeled anatomic structures. P = proximal or cephalad; M = middle; D = distal or caudad.
VOL.
91, NO.
3, MARCH
1998
Magnetic resonance imaging of the pelvis, anal manometry, ultrasonography of the internal and external anal sphincters, and nerve studies to evaluate pudendal neuropathies are being used by gynecologists and colorectal surgeons to evaluate obstetric injuries and defecation disorders.’ Before evaluating and treating pathologic conditions, it is important to understand normal structural and functional anatomy and physiology. Our MRI scans found the mean (2 SD) anterior length to be 22.6 ? 5.8 mm and the mean (rt SD) posterior length, 27.3 ? 6.0 mm. The length of the internal sphincter has been thought to correspond with the length of the high-pressure zone obtained during
Fenner
et al
MRI/Anal
Sphincter Measurements
371
Figure 4. Average thickness (mm) of anal sphincters by magnetic resonance imaging evaluation. The internal sphincter is represented by davk shading, and the external sphincter, by light gray. L = left; R = ngnr.
Posterior
Proximal anal manometry. Our study confirmed this correlation and found a positive relationship between the posterior sphincter length measured by MRI and the length of the high-pressure zone with squeeze determined by anal manometry (P = .042). The high-pressure zone is an important factor in the maintenance of fecal continence. We found the mean (2 SD) length of the high-pressure zone to be 3.0 2 0.6 cm at rest and 3.1 i 0.8 cm with squeeze. These findings are consistent with studies by Nivatvongs et al,iO who found the high-pressure zone in women to be 3.7 cm in length. Both anatomic and physiologic tests demonstrate the length of the continence zone to be at least 3 cm. Surgeons performing operations such as repair of acute or chronic thirddegree sphincter lacerations may not recognize the need to extend the repair cephalad to this 1evel.l’ Our findings support an anatomic and functional repair that extends 3 cm in length cephalad to maintain this high-pressure, or continence, zone. During squeeze, the anal canal lengthens as the
Table
1. Summary
Mean
Length to maximum resting pressure (cm) Length to maximum squeeze pressure (cm) Resting length (cm) Squeeze length (cm) Maximum resting pressure (mmHg) Maximum squeeze pressure (mmHg) High-pressure zone resting length (cm) High-pressure zone squeeze length (cm) SD = standard
372
Fenner
SD
Minimum
Maximum
3.6
0.5
2.8
4.6
4.8
1.0
3.5
6.8
5.4 6.6 150.6
0.5 1.0 33.4
4.6 5.3 81.1
6.0 8.4 184.6
MRI Anterior
200.3
27.7
150.4
3.0
0.6
2.2
3.9
3.1
0.8
2.0
5.0
Posterior
MRl/Anal
2. Correlation Length and length length
257.9
length
Between Magnetic Resonance Imaging Manometry
Sphincter
Measurements
Spearman rank correlation coefficient
Manometry Length to maximum resting pressure Length to maximum squeeze pressure Resting length Squeeze length High-pressure zone resting length High-pressure zone squeeze length Length to maximum resting pressure Length to maximum squeeze pressure Resting length Squeeze length High-pressure zone resting length High-pressure zone squeeze length
MRI = magnetic resonance Two-tailed P > .05.
deviation.
et al
Table
Manometry Measurements
of Anal
Measurement
external sphincter contracts, and the puborectalis muscle shortens and elevates the anal canal. By anal manometry, the mean (t SD) resting length was 5.4 2 0.5 cm, and with squeeze, the canal lengthened to 6.6 2 1.0 cm. Whereas there was no significant correlation between the manometric length and the anatomic length at rest, there was a positive correlation between the posterior length as measured by MRI and the length of the high-pressure zone with squeeze (P = .042). In other words, the longer the posterior sphincter measured on MRI, the longer the high-pressure zone measured on manometry. A negative correlation was seen
imaging;
-0.13
(I’ = NS)
-0.33
(P = NS)
0.19 (P = NS) -0.32 (I’ = NS) 0.51 (P = NS) 0.49 (I’ = NS) 0.01 (P = NS) -0.64
(P = ,044)
0.25 (P = NS) -0.63 (P = .049) 0.51 (P = NS) 0.65 (P = .042)
NS = not significant.
Obstetrics 8 Gynecology
Figure 5. Average sures by manometry left; R = right.
resting (mmHg).
presL =
Posterior
Proximal
Distal when comparing the posterior length and the manometric length with maximum squeeze pressure and the total squeeze length (P = ,044 and .049, respectively). The shorter the posterior length of the sphincters on MRI, the more cephalad were the maximum pressures recorded with squeeze for each participant. Physiologically, squeeze efforts prevent leakage upon presentation of enteric contents to the proximal canal at a socially inopportune time. Squeeze pressure is not responsible for maintaining fecal continence from hour to hour.’ The length of the high-pressure zone and the pressure difference between rectum and anal canal aid continence. As enteric contents descend, individuals with a shorter sphincter length may maintain continence by a greater lengthening of the anal canal and high-pressure zone. They may have compensated for a shorter canal by developing increased strength of the puborectalis. Taylor and Beartl’ found that squeeze pressures in the proximal anal canal were highest
Figure 6. Average sures by manometry left; R = right.
squeeze (mmHg).
posteriorly and lowest anteriorly. Our study found the squeeze pressures in the proximal anal canal to be distributed equally anteriorly and posteriorly, and squeeze pressures were higher posteriorly than anteriorly in the distal anal canal. Posterior resting pressures have been shown to be highest along the posterior-proximal and posteriordistal segments.13 McHugh and Daimant14 showed that the anterior resting pressures in women were highest distally in the anal canal, but in men, they were highest proximally. Our study showed the anterior resting pressure to be highest in the midportion of the anal canal. We found no significant correlation between muscle thickness and pressures at corresponding sites along the canal. Sultan et a1,15 when comparing anal endosonography and manometry, also found no correlation between internal and external anal muscle thicknesses and manometric pressures. The lack of
presL =
Posterior
Proximal
VOL.
91, NO.
3, MARCH
1998
Fenner
et al
MRI/Anal
Sphincter
Measurements
373
correlation may be due to the fact that anal manometry is not only a measurement of pressure exerted by the sphincter muscles, but also a measure of static resistance and tissue turgor. In addition, the force exerted during squeeze is dependent on voluntary effort. Other considerations are the limited sample size in their investigation, in which there is only a 49% chance of detecting correlations beyond 2 0.60 SD that are significantly different from the null hypothesis of no correlation. There also is some controversy about the basic anal anatomy as seen on MRI.r6 In our study, using Tl weighted magnetic resonance images, the internal anal sphincter was not differentiated from the mucosa, submucosa, or longitudinal muscle layer. However, the thicknesses of these other structures are fairly constant, as has been shown on recent studies using T2 weighted images and on endosonography.i5-i7 Endosonography and T2 weighted images show better differentiation of the mucosa and submucosa from the internal sphincter, but generally less distinct separation of the external anal sphincter from the longitudinal muscle layer. Discrepancies between our measurements and those in other studies are accounted for by the inclusion or exclusion of adjacent layers. Sultan et ali5 included the longitudinal muscle layer with their external anal sphincter measurements. For our ten healthy patients, MRI and anal manometry complemented each other in evaluations of the important functional length of the anal sphincter musculature. Although the mechanism that maintains fecal continence is complex, with future use of techniques such as MRI and anal manometry along with neurophysiologic testing, we will begin to understand more fully this important neuromuscular unit.
References 1. Frenckwer B, Euler CV. Influence of pudendal block on the function of anal sphincters. Gut 1975;16:482-9. 2. Wunderlich M, Parks AG. Physiology and pathophysiology of the anal sphincters. Int Surg 1982;67:291-8. 3. Davey M. The riddle of the sphincters. Am Surg 1987;53:298-306.
374
Fenner et al
MRI/Anal
Sphincter Measurements
4. Oh C, Kark AE. Anatomy of the external anal sphincter. Br J Surg 1972;59:717-23. 5. McCarthy S, Vaguer0 E. Gynecologic anatomy with magnetic resonance imaging. Am J Obstet Gynecol 1986;155:255-9. 6. Aronson MI’, Lee RA. Anatomy of anal sphincters and related structures in continent women studied with magnetic resonance imaging. Obstet Gynecol 1990;76:846-51. 7. Wilde FR. The anal intermuscular septum. Br J Surg 1949;36:27985. 8. Coller JA. Clinical application of anorectal manometry. Gastroenterol Clin North Am 1987;16:17-33. 9. Pemberton JH. Surgery of the alimentary tract. Vol IV. 3rd ed. Philadelphia: WB Saunders Company, 1991. 10. Nivatvongs S, Stem HS, Fryd DS. The length of the anal canal. Dis Colon Rectum 1981;24:600-1. 11. Thompson JD, Rock JA. TeLinde’s operative gynecology. 7th ed. Philadelphia: WB Saunders Company, 1991. 12. Taylor BM, Beart RW. Longitudinal and radial variations of pressure in the human anal sphincter. Gastroenterology 1984;86: 693-7. 13. Coller SA. Clinical application of anorectal manometry. Gastroenterol Clin North America 1987;16:17-33. 14. McHugh SM, Diamant NE. Anal canal pressure profile: A reappraisal as determined by rapid pull through technique. Gut 1987;28:1234-41. 15. Sultan AH, Kamm MA, Hudson CN. Endosonography of the anal sphincters: Normal anatomy and comparison with manometry. Clin Radio1 1994;49:368-74. 16. Hussain SM, Stoker J, Lameris JS. Anal sphincter complex: Endoanal MR imaging of normal anatomy. Radiology 1995;197:671-7. 17. Schafer A, Enck I’, Furst G, Kahn T, Frieling T, Lubke HJ. Anatomy of the anal sphincters-Comparison of anal endosonography to magnetic resonance imaging. Dis Colon Rectum 1994;37:777-81.
Address reprint requests to: Jefwy L. Cornella, MD Mayo Clinic Scottsdale 13400 E. Shea Boulevard Scottsdale, AZ 85259 E-mail: [email protected]
Received July 7, 1997. Received in revised form October 20, 1997. Accepted October 29, 1997. Copyright 0 Gynecologists.
1998 by The American Published by Elsevier
College of Obstetricians Science Inc.
and
Obstetrics 6 Gynecology