A new look at pelvic relaxation A.
CULLEN
JAMES
RICHARDSON, B.
NANCY
L.
.4tlantn,
LYON, WILLIAMS,
M.D
M.D. B.S.,
R.N.
Georgia
The concept is presented that most cystoceles andlor urethroceles result from insolated defects in the connective tissue supports of the anterior quadrant of the pelvis. Four areas in which defects have been found to occur are identified. Sixty patients are presented who were found to have isolated defects in the endopelvic fascia at the lateral sidewall of the pelvis with significant cystourethroceles and stress urinary incontinence. The surgical treatment ronsisted only of a direct approach to and closure of the isolated defect. The operative results at 3 to 48 months ulere excellent in 91.7 per cent, improved in 5 per cent, and failed in 3.3 per cent. Discussion is offered of the possibility of the study of the pelvic,poor from the viewpoint of a mechanical engineer. (AM. J. OBSTET. GYNECOL. 126: 568, 1976.)
PE L v I c relaxation associated with urethrocele and/or cystocele has been usually ascribed to a generalized relaxation or attenuation and stretching of the connective tissue supports of the anterior quadrant of the pelvis.‘-3 Several years ago the authors began to feel that generalized stretching or attenuation of the anterior quadrant fascia was the exception, if it occurred at all. With careful physical examination, isolated areas of weakness could be detected in either the pubocervical fascia or pubourethral ligaments of patients with cystocele and/or urethrocele. This suggested an isolated defect in one of these structures. The defects w-ere found to occur in four different areas. It appeared that all patients with cystocele and/or urethrocele had one or more of the four defects. This observation led the authors to decide that most cystoceles and/or urethroceles were the result of isolated defects or breaks in the pelvic connective tissue supports. The pubocervical fascia was described as the principal static support of the bladder and urethra by Curtis, Anson, and McKay4 in 1939. They discussed the relationship of this supporting layer to cystocele and urein 1951, called attention to the throcele. Krantz,’ From
the Department of Gynecology and Obstetrics, Emory School of Medicine at Craulford W. Long Hospital.
Lhziversity Mewwrial
Presented at the Thirty-eighth Annual Meeting of the South Atlantic Association of Obstetricians and Gynecologists, Hamilton, Bermuda, January 24-29, 1976. Reprint requests: Dr. A. Cullen Richardson, 710 Peachtree St., N.E., Atlanta, Georgia ?0308.
pubourethral ligaments which attach the distal urethra to the overlying symphysis. The areas in the anterior quadrant of the pelvis in which defects were identified and the order of the frequency in which they were observed in this patient population was as follows. I. Lateral defect. A defect is found in the fascia laterally at or near its attachment to the levator insertion in the lower margin of the superior pubic ramus. It usually results in a mild to moderate cystourethrocele, a loss of the urethrovesical angle, and significant stress urinary incontinence. It can be unilateral or bilateral. 2. Transverse okfect. A transverse separation occurs in the pubocervical fascia from its insertion into the pericervical ring of connective tissue. It usually results in a large cystocele in which the bladder herniates beneath the mucosa of the anterior vaginal fornix. The urethra remains well supported with a good urethrovesical angle. There is rarely, if ever, stress urinary incontinence. If this defect is repaired improperly, however, the patient can develop severe stress incontinence postoperatively. 3. Midline defect. An anteroposterior separation of the fascia occurs between the vagina and overlying bladder and/or urethra. Depending on the length of the defect, it results in a cystocele and/or urethrocele. If the defect is beneath the vesical neck, there will be a loss of the urethrovesical angle and stress urinary incontinence. This defect is the one most easily repaired and excellent results can be expected with the Kelly-Kennedy type procedures.
Pelvrc relaxation
J. Puhoztr~thml ligzmrrzt rkfect. The defect results from a loss of the integrity of the pubourethral ligaments. It is the least common of the four defects. Kepair can be extremely difficult as the defect allows the urethra to telescope externally beneath the symphysis. above an essentially intact anterior vaginal wall. The outward protrusion straightens the urethrovesical angle M ith trouhlesomc and persistant stress urinary incontinence. The major clinical problem initially was identification of the lateral defect as distinct from the midline defect. ‘1‘0 make this clistinction, a tongue blade was placed along the vaginal axis in either side of the vagina. .i‘he tongue blades were brought up to the lower- margin of the symphysis. If there was still a cystonrethrocele when the patient strained, the defect was considered to be midline. If there was no bulge, the defect uxs considered to be lateral. If the lateral defect was unilateral. then support on that side alone oblitereated the bulge \\hen the patient strained. In addition. if the examiner palpated carefully through the anterior vaginal wall along the posterior surface of the s>-mphvsis and superior pubic ramus, the lateral weakness could be detected. This finding was subtle, but definite and recognizable. A combination of two or more of the above defects within the anterior quadrant of the pelvis was rarely seen. A posterior pelvic quadrant defect (enterocele) and/or ;t central pelvic fascial defect with uterine prolapse were lkequently associated with one of the anterior quadrant defects. It was decided that a reasonably satisfactory validation of the localized nature of the defects would be the clinical results obtained if the surgical correction were limited to the area of the defect. With this concept in mind, the authors abandoned all general plication operative tee hniques and those procedures which created abnormal anatomy. An attempt was made in surgery to approach each defect directly and to close the defec,t with the restoration of normal anatomy. The most common anterior quadrant defect which had been demonstrated was the lateral one in which the fascia separated from its lateral attachment at the levator muscle insertion. The results of the surgical treatmem in this group of patients are reported at this time.
Material Between January, 1971. and October, 1975, the authors operated on 93 patients (all from their private practice) with symptomatic anterior quadrant pelvic relaxation. Of these, 62 were deemed to have lateral defects which resulted in a cystourethrocele with
569
significant stress urinary incontincnc-cs. I)ue to cxtrnsive scarring from previous surgerv. CIVOpatients were treated with other operative procedures and thus eliminated from the study group. Of the remaining 60 with lateral defects. 15 were felt to have biklreral and 45 unilateral defects. Of the 45 unilateral detects, five were on the left side and 40 were on the +$t side. Sis of the 60 patients had had a previou, amerior vaginal repair but they did not have extensiv(L SC;Irring.
Met hod It was decided that all patients with lateral defects should be treated with some type of. retropuhic procedure. After a few minor variations during the earlv months of the study, the following proccclure became the routine treatment for cystocele andior urethrocele associated with a lateral defect. Through a PGnnensteil incision, the retroupubic space is entered in the midline and dissection is carried laterally (III the affected side(s), separating the bladder enrirely trcml the pubic ramus anteriorly and from the urethra and anterior one third of the vagina below. The dissrc tion is carried cautiously downward until the upper 2 to 3 tm. of the medial attachment of the levator muscle cotne into view. At this time, defect in the lateral expanse of the pubocervical f‘ascia is detectable. The medial margin of the separation has usually retracted almost to, if not to, the lateral superior sulcus of the vagina. With a fingel in the vagina, the superior lateral sulcus ot the vagina is elevated and sutures are placed through the fascia adjacent to the superior lateral vaginal sulc~is and then into the semitendinous insertion of the I(.vator muscle into the superior pubic ramus. Foun to six such suturt~~ are placed. Care is taken to avoid the nlumerous small veins in the area and also to avoid the periosteum of the pubic ramus. If hemostasis is satisfactorv, no drains are inserted and the abdomen is clr~rd, Because the procedure is carried out at a distance from the vesic-al neck. no appreciable edema or inflammarorv reaction develops around the urethrovesical jumtion. and neither suprapubic nor transurethral drainage of the bladder is necessarv.
Results The follow-up in this group varies from 3 to 4X months, with an average of 20 months tor all patients. Fifty-eight of the 60 patients have had satisfactor! results. Of these, five state that thev have \tress urinar) incontinence with coughing on rare occasions hut not of such a degree to be considered by theln a problem. Three, although greatly improved, still tcel that their control is not quite as saCsfactory as they would like it to be. There have been only two operative failures. and
570
Richardson,
Lyon,
and Williams
both were apparent at 6 weeks after surgery. One was an error in the preoperative assessment of the patient, for she returned with a cystourethrocele (midline defect) in spite of good lateral support on each side. She was subsequently cured with a Kelly urethroplasty. The second failure was in a patient who had had a previous extensive repair. A Moskowitz procedure for an enterocele was done at the same time as the repair of the lateral pelvic wall defect. At 4 weeks after surgery, she felt “something pull” while coughing and both the enterocele and cystocele recurred. The follow-up of 3 to 48 months gives an over-all functional result of satisfactory to excellent in 9 1.7 per cent of the patients, improved in 5.0 per cent and failed in 3.3 per cent. This compares favorably with many published reports. ‘-’ The anatomic results at this time are the same as the functional results. An additional gratifying finding has been the shortened postoperative hospital stay. The postoperative stay averaged 4.9 days for all patients despite the fact that 45 patients (75 per cent) had additional major intra-abdominal procedures, such as hysterectomy and/or enterocele repair. During the early part of the series, suprapubic cystotomy tubes were inserted in six cases. All of the suprapubic tubes were removed by the third day and none of these patients required catheterization. In the remaining 54 patients who did not have suprapubic drainage, only four (7 per cent) required catheterization. Three required catheterization one to three times. Only one patient required an indwelling catheter for 24 hours, and she had the only recorded urinary tract infection in the series. No serious complications occurred. There have been no instances of postoperative urgency incontinence or osteitis pubis.
Comment The final evaluation of the results of this report requires additional follow-up. It is reassuring, however, that the two failures and three satisfactory, but not excellent results, were apparent immediately. There had been no change in the functional or anatomic results in any patient after 6 weeks postoperative evaluation. In the future it is possible that some patients will develop defects in other locations. A preponderance of right lateral over left lateral defects was observed (40 on the right and five on the left). A possible explanation could be that the sigmoid colon and the mesosigmoid provide additional protection to the pelvic floor on the left. Only six of the above 60 patients had had some type of previous anterior vaginal surgery. One of the fail-
ures was in this group. This operative approach n~av not be as useful if it is performed as a secondarv procedure as compared with its cftectiveness as 11 primary procedure. Once the anatomy has been clistorted by an anterior vaginal plication, it is difficult to determine the location of the original primary defect. Furthermore, even if’ a lateral defect is evident, the previous plication may have so shortened the supports in the midline as to make it surgicallv impossible to close the defect laterally. This report is intended to document the concept ot localized defects in the pubocervical fascia as the cause of most anterior pelvic quadrant relaxation and to emphasize the importance of this concept in the selection of the primary operative procedure, which should be based on an assessment of the anatomic defect iii each individual patient. The surgical procedure described above is easily performed and has served the authors well. The results obtained were very satisfactory. It is readily admitted, hovvever. that t.here are other procedures that will accomplish the same desired result, which is the correction of the isolated defect. As in the case of anterior quadrant defects, it is felt that relaxation in the posterior pelvic quadrant and in the central pelvis also frequently results from localized breaks in the continuity of the supporting structures. A specific surgical approach to these areas as well as the other anterior quadrant defects has likewise been practiced since January, 1971. These results will be the subject of a later report. The nature and etiology of the localized defects of the endopelvic fascia remain undetermined. It seems likely that at least some of these defects are actual tears in the fascia. On gross examination at the time of surgery, the appearance of the defect is often exactly as one might expect of an actual tear. How and when such tears would occur is not as yet apparent. Occasionally, a patient can recall the exact time when she felt “something pull” during physical exertion. Subsequently, she noted soreness in her lower abdomen and a few weeks later began to experience pelvic pressure and stress urinary incontinence. Frequently, however, the onset was totall\; insidious. A surgical repair of the pelvic floor is, in fact, an exercise in mechanical engineering. An engineer* is now participating with the authors in their continuing study of this problem. He has suggested that the pelvic floor should be approached as one might study the capacity of a suspension bridge. Such a viewpoint leads one to ask questions about (1) the structural compo*Hyland Engineering Technology,
Y. L. Chen, Ph.D., Assistant Science and Mechanics, Atlanta, Georgia.
Professor, School Georgia Institute
of of
Volume 126 Number5
nents, (2) the architectural configuration, and (3) the potential loads to be applied. As to the structural components, one would want answers tl) the following questions: What structures in the pelvis actually have the capacity to bear weight? What is the tensile strength of the tissue components of each structure? How do they behave under stress? What are the differences in the capacities of these structures to withstand dynamic or impact stress and static or chronic stress? How does this capacity change with age, nutrition, and/or hormonal status? How does the weight-bearing capacity of these structures correlate with the relative amount of collagen, elastin, smooth muscle, and ground substance when seen with light microscopy? Are there qualitative or chemical changes in one or more of these components? Kegarding its structural configuration, one would ask: What is the architecture and configuration of the
Pelvic
relaxation
571
supporting tissues? Is the contour of the pelvic floor such as to create high stress points, anal does this change with parity, obestiy, or age? Is the distribution of supporting structures such that there art’ inherentlv weak and strong areas? When considering the potential load IO be applied the questions would be: What is the pressure applied to the pelvic floor, and what factors i&luence the variations in pressure? Is the load transmitted invariably as hydrostatic pressure, or is thele a “hall-insling” effect? Does the pelvic floor sometimes give way purely as a result of “overloading”? It so, what can be done to decrease the load? These and many other questions neetl to be answered before the best surgical treatment of pelvic relaxation can be determined and befor-tx intelligent consideration can be given to prevention of this problem.
REFERENCES 1. Danforth. D. N., editor: Textbook of Obstetrics and Gynecology, ed. 2, New York, 1966, Harper & Row, Inc., p. 828. 2. Brewer, J. I.: Textbook of Gynecology, ed. 3, Baltimore, 196 1, The Williams & Wilkins Company, pp. 12 1 and 404. 3. Jeffcoate, T. N. A.: Principles of Gynecology, ed. 2, Washington, 1962, Butterworth & Company, Ltd., p. 725. 4. Curtis, A. H., Anson, B. I., and McKay, C. B.: Surg. Gynecol. Obstet. 68: 161, 1939.
Discussion DR.MARKPENTECOST,JR., Atlanta, Georgia(Officia1 guest). Dr. Richardson has presented an intriguing and interesting preliminary report on anterior quadrant pelvic relaxation with many new ideas. The authors’ goal has been to search out and identify specific anatomic defects in pelvic relaxation producing stress urinary incontinence and to perform a procedure to repair the specific anatomic defect and thereby restore normal anatomy, which is in contrast to the general plication procedures. The underlying premise is that there must be a specific defect or tear as opposed to a genearlized defect such as stretching or attenuation of fascia and ligaments. The report of Francis’ that stress urinary incontinence develops during pregnancy rather than at the time of delivery and improve post partum would suggest that attenuation and stretching do play a role in pelvic relaxation. From their findings the authors have identified four areas in which defects have been found to occur and have arranged them in order of their frequency. Surprisingly, the most common is the lateral defect. I would have thought the midline defect more common, as the Kelly type midline plication produces good results in a large percentage of cases.2 If this finding is
5. Krantz. K. E.: AM. J. OBSTET. GYNECOL. 691: 374, 1951. 6. Haydon, G. B., and Yonkilis, M. H.: Obstet. Gynecol. 35: 898, 1970. 7. Ball, T. J.. Knapp, R. C., Nathanson, B., and Lagasse. L. D.: AM.J. OBSTET. GYNECOL.~~:~~?'. 1966. 8. Burch,J. C.: AM.J. OBSTET. GYNECOL. 100: 764, 1968. 9. Green, T. H.: In Reid, D. E., and Christian. C. D., editors: Controversy in Obstetrics and Gynecology II, Philadelphia, 1974, W. B. Saunders Company. p. 3!+5.
substantiated, it is very important. If. the defect is lateral to where we are repairing, we hake only taken a tuck and not really corrected the situation and would expect this to gradually fail. We are in a sense splicing the rope at the wrong place. Since so much of the investigation of pelvic relaxation with stress urinary incontinence has used the urethrocystogram, it would seem very helpful to obtain this type of study both preoperatively and postoperatively to compare and correlate this study with other work. One might speculate that the authors’ midline defect might produce the Type 1 incontinence described by Green3 and that the lateral defect might be one source of Green’s Type II incontinence. Another source of Type II incontinence might be the pubourethral ligament defect described. If this speculation is correct, then the authors’ double tongue blade urethrovesicle elevation test would be a simple, inexpensive clinical niethod ot help in deciding on the type of’ corrective procedure to use. Also rhr authors make a strong point of restoring normal anatomy. These x-ray studies would be helpful in proving or disproving this point. I ask the question: does the authors’ unilateral procedure pull the urethra to that side? The authors do not seem to be stressing their opera-
CULLEN
JAMES
RICHARDSON, B.
NANCY
L.
.4tlantn,
LYON, WILLIAMS,
M.D
M.D. B.S.,
R.N.
Georgia
The concept is presented that most cystoceles andlor urethroceles result from insolated defects in the connective tissue supports of the anterior quadrant of the pelvis. Four areas in which defects have been found to occur are identified. Sixty patients are presented who were found to have isolated defects in the endopelvic fascia at the lateral sidewall of the pelvis with significant cystourethroceles and stress urinary incontinence. The surgical treatment ronsisted only of a direct approach to and closure of the isolated defect. The operative results at 3 to 48 months ulere excellent in 91.7 per cent, improved in 5 per cent, and failed in 3.3 per cent. Discussion is offered of the possibility of the study of the pelvic,poor from the viewpoint of a mechanical engineer. (AM. J. OBSTET. GYNECOL. 126: 568, 1976.)
PE L v I c relaxation associated with urethrocele and/or cystocele has been usually ascribed to a generalized relaxation or attenuation and stretching of the connective tissue supports of the anterior quadrant of the pelvis.‘-3 Several years ago the authors began to feel that generalized stretching or attenuation of the anterior quadrant fascia was the exception, if it occurred at all. With careful physical examination, isolated areas of weakness could be detected in either the pubocervical fascia or pubourethral ligaments of patients with cystocele and/or urethrocele. This suggested an isolated defect in one of these structures. The defects w-ere found to occur in four different areas. It appeared that all patients with cystocele and/or urethrocele had one or more of the four defects. This observation led the authors to decide that most cystoceles and/or urethroceles were the result of isolated defects or breaks in the pelvic connective tissue supports. The pubocervical fascia was described as the principal static support of the bladder and urethra by Curtis, Anson, and McKay4 in 1939. They discussed the relationship of this supporting layer to cystocele and urein 1951, called attention to the throcele. Krantz,’ From
the Department of Gynecology and Obstetrics, Emory School of Medicine at Craulford W. Long Hospital.
Lhziversity Mewwrial
Presented at the Thirty-eighth Annual Meeting of the South Atlantic Association of Obstetricians and Gynecologists, Hamilton, Bermuda, January 24-29, 1976. Reprint requests: Dr. A. Cullen Richardson, 710 Peachtree St., N.E., Atlanta, Georgia ?0308.
pubourethral ligaments which attach the distal urethra to the overlying symphysis. The areas in the anterior quadrant of the pelvis in which defects were identified and the order of the frequency in which they were observed in this patient population was as follows. I. Lateral defect. A defect is found in the fascia laterally at or near its attachment to the levator insertion in the lower margin of the superior pubic ramus. It usually results in a mild to moderate cystourethrocele, a loss of the urethrovesical angle, and significant stress urinary incontinence. It can be unilateral or bilateral. 2. Transverse okfect. A transverse separation occurs in the pubocervical fascia from its insertion into the pericervical ring of connective tissue. It usually results in a large cystocele in which the bladder herniates beneath the mucosa of the anterior vaginal fornix. The urethra remains well supported with a good urethrovesical angle. There is rarely, if ever, stress urinary incontinence. If this defect is repaired improperly, however, the patient can develop severe stress incontinence postoperatively. 3. Midline defect. An anteroposterior separation of the fascia occurs between the vagina and overlying bladder and/or urethra. Depending on the length of the defect, it results in a cystocele and/or urethrocele. If the defect is beneath the vesical neck, there will be a loss of the urethrovesical angle and stress urinary incontinence. This defect is the one most easily repaired and excellent results can be expected with the Kelly-Kennedy type procedures.
Pelvrc relaxation
J. Puhoztr~thml ligzmrrzt rkfect. The defect results from a loss of the integrity of the pubourethral ligaments. It is the least common of the four defects. Kepair can be extremely difficult as the defect allows the urethra to telescope externally beneath the symphysis. above an essentially intact anterior vaginal wall. The outward protrusion straightens the urethrovesical angle M ith trouhlesomc and persistant stress urinary incontinence. The major clinical problem initially was identification of the lateral defect as distinct from the midline defect. ‘1‘0 make this clistinction, a tongue blade was placed along the vaginal axis in either side of the vagina. .i‘he tongue blades were brought up to the lower- margin of the symphysis. If there was still a cystonrethrocele when the patient strained, the defect was considered to be midline. If there was no bulge, the defect uxs considered to be lateral. If the lateral defect was unilateral. then support on that side alone oblitereated the bulge \\hen the patient strained. In addition. if the examiner palpated carefully through the anterior vaginal wall along the posterior surface of the s>-mphvsis and superior pubic ramus, the lateral weakness could be detected. This finding was subtle, but definite and recognizable. A combination of two or more of the above defects within the anterior quadrant of the pelvis was rarely seen. A posterior pelvic quadrant defect (enterocele) and/or ;t central pelvic fascial defect with uterine prolapse were lkequently associated with one of the anterior quadrant defects. It was decided that a reasonably satisfactory validation of the localized nature of the defects would be the clinical results obtained if the surgical correction were limited to the area of the defect. With this concept in mind, the authors abandoned all general plication operative tee hniques and those procedures which created abnormal anatomy. An attempt was made in surgery to approach each defect directly and to close the defec,t with the restoration of normal anatomy. The most common anterior quadrant defect which had been demonstrated was the lateral one in which the fascia separated from its lateral attachment at the levator muscle insertion. The results of the surgical treatmem in this group of patients are reported at this time.
Material Between January, 1971. and October, 1975, the authors operated on 93 patients (all from their private practice) with symptomatic anterior quadrant pelvic relaxation. Of these, 62 were deemed to have lateral defects which resulted in a cystourethrocele with
569
significant stress urinary incontincnc-cs. I)ue to cxtrnsive scarring from previous surgerv. CIVOpatients were treated with other operative procedures and thus eliminated from the study group. Of the remaining 60 with lateral defects. 15 were felt to have biklreral and 45 unilateral defects. Of the 45 unilateral detects, five were on the left side and 40 were on the +$t side. Sis of the 60 patients had had a previou, amerior vaginal repair but they did not have extensiv(L SC;Irring.
Met hod It was decided that all patients with lateral defects should be treated with some type of. retropuhic procedure. After a few minor variations during the earlv months of the study, the following proccclure became the routine treatment for cystocele andior urethrocele associated with a lateral defect. Through a PGnnensteil incision, the retroupubic space is entered in the midline and dissection is carried laterally (III the affected side(s), separating the bladder enrirely trcml the pubic ramus anteriorly and from the urethra and anterior one third of the vagina below. The dissrc tion is carried cautiously downward until the upper 2 to 3 tm. of the medial attachment of the levator muscle cotne into view. At this time, defect in the lateral expanse of the pubocervical f‘ascia is detectable. The medial margin of the separation has usually retracted almost to, if not to, the lateral superior sulcus of the vagina. With a fingel in the vagina, the superior lateral sulcus ot the vagina is elevated and sutures are placed through the fascia adjacent to the superior lateral vaginal sulc~is and then into the semitendinous insertion of the I(.vator muscle into the superior pubic ramus. Foun to six such suturt~~ are placed. Care is taken to avoid the nlumerous small veins in the area and also to avoid the periosteum of the pubic ramus. If hemostasis is satisfactorv, no drains are inserted and the abdomen is clr~rd, Because the procedure is carried out at a distance from the vesic-al neck. no appreciable edema or inflammarorv reaction develops around the urethrovesical jumtion. and neither suprapubic nor transurethral drainage of the bladder is necessarv.
Results The follow-up in this group varies from 3 to 4X months, with an average of 20 months tor all patients. Fifty-eight of the 60 patients have had satisfactor! results. Of these, five state that thev have \tress urinar) incontinence with coughing on rare occasions hut not of such a degree to be considered by theln a problem. Three, although greatly improved, still tcel that their control is not quite as saCsfactory as they would like it to be. There have been only two operative failures. and
570
Richardson,
Lyon,
and Williams
both were apparent at 6 weeks after surgery. One was an error in the preoperative assessment of the patient, for she returned with a cystourethrocele (midline defect) in spite of good lateral support on each side. She was subsequently cured with a Kelly urethroplasty. The second failure was in a patient who had had a previous extensive repair. A Moskowitz procedure for an enterocele was done at the same time as the repair of the lateral pelvic wall defect. At 4 weeks after surgery, she felt “something pull” while coughing and both the enterocele and cystocele recurred. The follow-up of 3 to 48 months gives an over-all functional result of satisfactory to excellent in 9 1.7 per cent of the patients, improved in 5.0 per cent and failed in 3.3 per cent. This compares favorably with many published reports. ‘-’ The anatomic results at this time are the same as the functional results. An additional gratifying finding has been the shortened postoperative hospital stay. The postoperative stay averaged 4.9 days for all patients despite the fact that 45 patients (75 per cent) had additional major intra-abdominal procedures, such as hysterectomy and/or enterocele repair. During the early part of the series, suprapubic cystotomy tubes were inserted in six cases. All of the suprapubic tubes were removed by the third day and none of these patients required catheterization. In the remaining 54 patients who did not have suprapubic drainage, only four (7 per cent) required catheterization. Three required catheterization one to three times. Only one patient required an indwelling catheter for 24 hours, and she had the only recorded urinary tract infection in the series. No serious complications occurred. There have been no instances of postoperative urgency incontinence or osteitis pubis.
Comment The final evaluation of the results of this report requires additional follow-up. It is reassuring, however, that the two failures and three satisfactory, but not excellent results, were apparent immediately. There had been no change in the functional or anatomic results in any patient after 6 weeks postoperative evaluation. In the future it is possible that some patients will develop defects in other locations. A preponderance of right lateral over left lateral defects was observed (40 on the right and five on the left). A possible explanation could be that the sigmoid colon and the mesosigmoid provide additional protection to the pelvic floor on the left. Only six of the above 60 patients had had some type of previous anterior vaginal surgery. One of the fail-
ures was in this group. This operative approach n~av not be as useful if it is performed as a secondarv procedure as compared with its cftectiveness as 11 primary procedure. Once the anatomy has been clistorted by an anterior vaginal plication, it is difficult to determine the location of the original primary defect. Furthermore, even if’ a lateral defect is evident, the previous plication may have so shortened the supports in the midline as to make it surgicallv impossible to close the defect laterally. This report is intended to document the concept ot localized defects in the pubocervical fascia as the cause of most anterior pelvic quadrant relaxation and to emphasize the importance of this concept in the selection of the primary operative procedure, which should be based on an assessment of the anatomic defect iii each individual patient. The surgical procedure described above is easily performed and has served the authors well. The results obtained were very satisfactory. It is readily admitted, hovvever. that t.here are other procedures that will accomplish the same desired result, which is the correction of the isolated defect. As in the case of anterior quadrant defects, it is felt that relaxation in the posterior pelvic quadrant and in the central pelvis also frequently results from localized breaks in the continuity of the supporting structures. A specific surgical approach to these areas as well as the other anterior quadrant defects has likewise been practiced since January, 1971. These results will be the subject of a later report. The nature and etiology of the localized defects of the endopelvic fascia remain undetermined. It seems likely that at least some of these defects are actual tears in the fascia. On gross examination at the time of surgery, the appearance of the defect is often exactly as one might expect of an actual tear. How and when such tears would occur is not as yet apparent. Occasionally, a patient can recall the exact time when she felt “something pull” during physical exertion. Subsequently, she noted soreness in her lower abdomen and a few weeks later began to experience pelvic pressure and stress urinary incontinence. Frequently, however, the onset was totall\; insidious. A surgical repair of the pelvic floor is, in fact, an exercise in mechanical engineering. An engineer* is now participating with the authors in their continuing study of this problem. He has suggested that the pelvic floor should be approached as one might study the capacity of a suspension bridge. Such a viewpoint leads one to ask questions about (1) the structural compo*Hyland Engineering Technology,
Y. L. Chen, Ph.D., Assistant Science and Mechanics, Atlanta, Georgia.
Professor, School Georgia Institute
of of
Volume 126 Number5
nents, (2) the architectural configuration, and (3) the potential loads to be applied. As to the structural components, one would want answers tl) the following questions: What structures in the pelvis actually have the capacity to bear weight? What is the tensile strength of the tissue components of each structure? How do they behave under stress? What are the differences in the capacities of these structures to withstand dynamic or impact stress and static or chronic stress? How does this capacity change with age, nutrition, and/or hormonal status? How does the weight-bearing capacity of these structures correlate with the relative amount of collagen, elastin, smooth muscle, and ground substance when seen with light microscopy? Are there qualitative or chemical changes in one or more of these components? Kegarding its structural configuration, one would ask: What is the architecture and configuration of the
Pelvic
relaxation
571
supporting tissues? Is the contour of the pelvic floor such as to create high stress points, anal does this change with parity, obestiy, or age? Is the distribution of supporting structures such that there art’ inherentlv weak and strong areas? When considering the potential load IO be applied the questions would be: What is the pressure applied to the pelvic floor, and what factors i&luence the variations in pressure? Is the load transmitted invariably as hydrostatic pressure, or is thele a “hall-insling” effect? Does the pelvic floor sometimes give way purely as a result of “overloading”? It so, what can be done to decrease the load? These and many other questions neetl to be answered before the best surgical treatment of pelvic relaxation can be determined and befor-tx intelligent consideration can be given to prevention of this problem.
REFERENCES 1. Danforth. D. N., editor: Textbook of Obstetrics and Gynecology, ed. 2, New York, 1966, Harper & Row, Inc., p. 828. 2. Brewer, J. I.: Textbook of Gynecology, ed. 3, Baltimore, 196 1, The Williams & Wilkins Company, pp. 12 1 and 404. 3. Jeffcoate, T. N. A.: Principles of Gynecology, ed. 2, Washington, 1962, Butterworth & Company, Ltd., p. 725. 4. Curtis, A. H., Anson, B. I., and McKay, C. B.: Surg. Gynecol. Obstet. 68: 161, 1939.
Discussion DR.MARKPENTECOST,JR., Atlanta, Georgia(Officia1 guest). Dr. Richardson has presented an intriguing and interesting preliminary report on anterior quadrant pelvic relaxation with many new ideas. The authors’ goal has been to search out and identify specific anatomic defects in pelvic relaxation producing stress urinary incontinence and to perform a procedure to repair the specific anatomic defect and thereby restore normal anatomy, which is in contrast to the general plication procedures. The underlying premise is that there must be a specific defect or tear as opposed to a genearlized defect such as stretching or attenuation of fascia and ligaments. The report of Francis’ that stress urinary incontinence develops during pregnancy rather than at the time of delivery and improve post partum would suggest that attenuation and stretching do play a role in pelvic relaxation. From their findings the authors have identified four areas in which defects have been found to occur and have arranged them in order of their frequency. Surprisingly, the most common is the lateral defect. I would have thought the midline defect more common, as the Kelly type midline plication produces good results in a large percentage of cases.2 If this finding is
5. Krantz. K. E.: AM. J. OBSTET. GYNECOL. 691: 374, 1951. 6. Haydon, G. B., and Yonkilis, M. H.: Obstet. Gynecol. 35: 898, 1970. 7. Ball, T. J.. Knapp, R. C., Nathanson, B., and Lagasse. L. D.: AM.J. OBSTET. GYNECOL.~~:~~?'. 1966. 8. Burch,J. C.: AM.J. OBSTET. GYNECOL. 100: 764, 1968. 9. Green, T. H.: In Reid, D. E., and Christian. C. D., editors: Controversy in Obstetrics and Gynecology II, Philadelphia, 1974, W. B. Saunders Company. p. 3!+5.
substantiated, it is very important. If. the defect is lateral to where we are repairing, we hake only taken a tuck and not really corrected the situation and would expect this to gradually fail. We are in a sense splicing the rope at the wrong place. Since so much of the investigation of pelvic relaxation with stress urinary incontinence has used the urethrocystogram, it would seem very helpful to obtain this type of study both preoperatively and postoperatively to compare and correlate this study with other work. One might speculate that the authors’ midline defect might produce the Type 1 incontinence described by Green3 and that the lateral defect might be one source of Green’s Type II incontinence. Another source of Type II incontinence might be the pubourethral ligament defect described. If this speculation is correct, then the authors’ double tongue blade urethrovesicle elevation test would be a simple, inexpensive clinical niethod ot help in deciding on the type of’ corrective procedure to use. Also rhr authors make a strong point of restoring normal anatomy. These x-ray studies would be helpful in proving or disproving this point. I ask the question: does the authors’ unilateral procedure pull the urethra to that side? The authors do not seem to be stressing their opera-
