bladder exstrophy by tensile testing. J.S. SUSSMAN ... with the intact symphysis; the best load-to-failure (#2 ... load-to-failure, stiçess and energy-to-failure. The.
British Journal of Urology (1997), 79, 979–984
A comparison of methods of repairing the symphysis pubis in bladder exstrophy by tensile testing J .S . SU SS M AN, P. D. SPO NS E L L ER , J .P. GE AR HAR T*, A. D.C. VAL DE VIT , J . K IE R- YOR K and E .Y. S. CHA O Department of Orthopaedic Surgery and *Urology, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA
Objective To compare the ecacy of several fixation techniques in the reconstruction of diastasis of the symphysis pubis in bladder exstrophy. Materials and methods The symphyses of 32 pelves removed from piglets about 1 month old were disrupted and repaired using one of eight methods. After repair, each pelvis was tested biomechanically for load-to-failure, stiness and energy-to-failure. The various repair techniques were compared with one another and to a group of six pelves tested intact. Results Four of the methods tested, including a #2 nylon suture placed through bone in a horizontal mattress arrangement, several loops of #2 nylon suture tied around the pubes, Mersilene tape tied around the
Introduction Bladder exstrophy, a rare condition aecting roughly 1 in 30 000 births [1], involves a failure of inferior midline mesenchymal migration and fusion, resulting in both soft tissue and bony deformities. The urinary bladder and urethra are open on the abdominal wall, there is a wide diastasis of the symphysis pubis and the external genitalia are dysmorphic. Exstrophy involves a spectrum of bony pelvic anomalies. Diastases as wide as 12 cm have been reported [2]. Studies of computed tomograms of the pelves of patients with exstrophy have found a mean diastasis of 5.9 cm, with external rotation and shortening of the pubic rami, external rotation of the posterior pelvis and retroversion of the acetabulae [3,4]. The goals of surgical repair are to close the bladder, urethra, and abdominal wall, to achieve continence, to preserve renal function, and to construct functional and cosmetically acceptable external genitalia. However, urological closure of the bladder is sometimes not possible [5] and the failure rate of corrective surgery is high. Most commonly, the cause of failure is a dehiscence of the midline closure caused by tension exerted by the uncorrected pelvis. To minimize the failure rate, a pelvic osteotomy is performed in most cases to normalize the pelvic ring and Accepted for publication 4 February 1997 © 1997 British Journal of Urology
pubes, and Mitek G-II suture anchors placed into the superior pubic rami, showed the highest stiness and load-to-failure. All methods were very weak compared with the intact symphysis; the best load-to-failure (#2 nylon horizontal mattress suture) was less than half of that for intact bone, and the best stiness (mersilene tape) was less than one-third that for the intact symphysis. Conclusion The repairs varied greatly in the variables tested and those which are most promising merit further investigation to assess methods of improving their performance. Keywords Bone, osteotomy, pelvis, bladder exstrophy, repair, suture
allow for approximation of the symphysis. This reduces tension on the abdominal wall, decreasing stress on the repaired bladder and abdominal wall, and is followed by a lower rate of wound dehiscence after urological repair [6]. The urological results achieved in patients with exstrophy correspond closely to the degree to which the pubic diastasis is closed [7]. Repairing the pelvic ring also allows reconstruction of the levators ani and puborectalis and for placement of the bladder neck and urethra within the pelvic ring. This results in a greater rate of urinary continence [8]. However, even with osteotomy the long-term failure rate is substantial. A significant degree of recurrent diastasis of the symphysis, which can undermine the urological repair over time, is seen in most patients [9–12] and a better method for fixing the symphysis pubis is needed (Fig. 1). Wires passed through both obturator foramena and twisted together anteriorly have tended eventually to cut through the bone [12–15]. A large suture through the bone has often been used [5,8,11,16] but has also cut through the bone [16] and progressive diastasis usually occurs. Other methods have included external fixation [17,18], wires wrapped around each superior pubic ramus and then twisted together [7], fascia lata looped and tied through the obturator foramena [14] or sutured to the superior rami [10], or a combination or wires, bolts, and plates [2]. Several of these methods are depicted in Fig. 2. However, none has 979
980
J.S . SU SS M AN et al.
a
Fig. 1. Recurrence of diastasis caused by the wire suture cutting through in a 40-year-old patient who had posterior osteotomies and wire fixation.
b
been very successful in maintaining approximation of the symphysis. The properties of the immature bone, ligament and cartilage of the symphysis in newborns and infants are poorly understood, and it is not clear what method of fixation can best hold these structures together. In addition, any fixation method must allow for mobility [19]. Little relevant experience is available from traumatology, as ruptures of the symphysis in paediatric patients rarely require fixation. To date, no study has compared proposed methods of fixation. We sought to do so by using dierent methods to fix the symphysis pubis in immature animal pelves and then comparing the biomechanical properties and investigating the failure modes for each repair. We examined whether a repair through bone or around it was stronger, and which material was most eective. Several recently developed materials were tested for their applicability.
Materials and methods Forty piglets (body mass 3.5–5.9 kg) were weighed and dissected; the pelvic bones were removed, wrapped in towels soaked in 0.9% saline solution, placed into plastic bags, and frozen at −20°C. Photographs and radiographs were taken of one specimen so that the bony and cartilaginous structure of the pubic symphysis could be analysed. These included anteroposterior views of the pelvis and anteroposterior views of the pubes and symphysis after removal of the iliac and ischial bones. The specimens were subsequently thawed by placing the plastic bag into water at room temperature and then prepared for testing. Anteroposterior diameters were
c Fig. 2. Composite illustration of methods of fixation described previously [7,10,14] (reproduced with permission). a, Wire loop around two pubes. b, Twisted wire loop around superior rami. c, Fascia lata sutured to superior rami.
measured from the posterosuperior edge of the symphysis to the antero-inferior edge of S3 vertebra. Pelves in Group 1 were kept intact to be tested as controls. On pelves in Groups 2 through 9, the symphysis pubis was cut with a scalpel and repaired using one of eight dierent techniques, with five replicates except where indicated. Group 2 The two pubic bones were sutured together using one #2 nylon suture passed through two holes in the bone on each side of the symphysis in a horizontal mattress arrangement; this is a commonly used surgical method. Group 3 A loop of #2 nylon was passed anteriorly to posteriorly through each obturator foramen, and then brought © 1997 British Journal of Urology 79, 979–984
ME TH ODS OF RE PA IRI NG TH E SY MPH IS I S P UBIS I N BL ADD E R E XS TROP HY
anteriorly through the midline. The two free ends were passed medially through each loop and tied anteriorly to the contralateral free ends. Group 4 Two loops of #2 nylon were placed on each pubic bone as in Group 3, and tied anteriorly. Group 5 Loops of Securestrand sutures, a braided nonabsorbable suture, were placed on each pubic bone by a method identical to that in Group 3. Group 6 (n=3) Mitek G-II suture anchors (Mitek Surgical Products, Norwood, MA, USA) were placed in each superior pubic ramus, as far laterally as possible, and tied using #2 nylon. The anchors are essentially small bolts, with wings which catch deep to cortical bone. The morphology of the pig pelvis necessitated that these were placed close to the triradiate cartilage; this would not be necessary in the human. Group 7 (n=2) Mitek G-II Mini suture anchors were placed and tied in the same manner as in Group 6; these are similar to the G-II anchors, but smaller. Group 8 (n=3) A strip of Mersilene tape (Ethicon, Somerville, NJ, USA) was passed through both obturator foramena and tied anteriorly. Mersilene is 5 mm woven polyester fibre tape.
Fig. 3. Diagram of the fixation methods. © 1997 British Journal of Urology 79, 979–984
981
Group 9 (n=4) A strip of tendon was harvested from the hindfoot flexors and sutured across the superior pubic rami. This was intended to simulate a repair technique described by Scherz et al. [10] who harvested a section of ilio-tibial tract and grafted it across the superior rami in a similar way. The repair techniques are depicted in Fig. 3; after repair, each pelvis was again wrapped in towels soaked in 0.9% saline, placed in a plastic bag and frozen at −20°C. For testing, each pelvis was thawed in water at room temperature and then placed into a specially designed jig for tensile testing on an MTS machine (MTS Systems Inc., Minneapolis, MN, USA). The pelvis was positioned on its lateral side and large portions of the two ilia were removed. Most of one hemipelvis was embedded in a block of Woods metal, which was then axed to the lower part of the MTS machine. A loop of webbing (breaking strength about 26 kN) was passed through the obturator foramen of the other hemipelvis and each end of the loop was passed around a metal crossbar attached to the MTS machine actuator. Because of the poor ischial bone quality, this was considered the best way to simulate the physiological load on the symphysis, which consists of a distractive force strongest at the inferior pubic symphysis [19,20]. Figure 4 shows a pelvis mounted for testing. After alignment and placement into the jig, the ligaments of the sacro-iliac joints were cut to remove any stabilizing eect these joints would exert on the pelvis and to thus ensure that only the symphyseal repair was being tested. For pelves repaired with suture anchors or with a tendon sutured across the superior rami, the ilium was removed before testing. As the load was being applied to these pelves inferior to the repair, they would
982
J.S . SU SS M AN et al. a 600 **
P < 0.02
Failure load (% BW)
500 400 300 200 **
**
**
100 0
**
1
2
3
4
5 Group
6
7
8
9
b 0.8
Fig. 4. The loading method; the pelvis is positioned with the cranial direction to the left, caudal to the right. The right hemipelvis is embedded in Woods metal. Webbing is placed around the left ischium and secured to the MTS tester.
Results Results for load-to-failure, stiness and energy-to-failure are presented in Fig. 5. No repair method was able to withstand a higher load-to-failure than the horizontal mattress nylon suture, which failed at a mean (sd) load of 206% (55) of body weight. Three other repair options (Groups 4, 6, and 8) were not significantly weaker: Groups 3, 5, 7 and 9 all failed at significantly weaker loads. The loads were normalized by body weight to correct for the heavier piglets being slightly older and having stronger bone. The intact pelves withstood a mean (sd) load of 450% (114) of body weight before failure. Group 8, the pelves repaired with mersilene tape, proved stiest when load was applied, with a mean (sd) stiness of 0.162 mm−1 (0.015), again normalized by body weight. Group 6, pelves fixed with Mitek G-II suture anchors and #2 nylon, were not significantly less sti. All other repair methods, including the horizontal mattress suture, were less sti. Intact pelves showed a mean (sd ) stiness of 0.607 mm−1 (0.096).
P < 0.001
Stiffness (/mm)
0.6
0.4
0.2 *
*
*
*
* 0.0
1
2
3
* *
4
5 Group
6
7
8
9
*
*
8
9
c 0.05 *
Energy (% BW mm)
have tended to pivot around a point of contact between ilium and sacrum. Once the ilium was removed, a small initial load brought the loading point in line with the repair. Once each specimen was aligned and the necessary cuts made, tensile force was applied until failure. Mean load-to-failure, stiness and energy-to-failure were calculated for each repair technique, and unpaired t-tests and analysis of variance used to compare the results. In addition, the failure mode for each specimen was recorded.
*
P < 0.05
0.04 0.03 0.02
*
*
*
0.01 0.00
1
2
3
4
5 Group
6
7
Fig. 5. Graph of the, a, load-to-failure, b, stiness and, c, energy absorbed to failure for the various repairs. P values represent significance versus, a, group 2, b, group 8 and c, group 4. 1, Intact pubis. 2, No. 2 Nylon mattress. 3, No. 2, Nylon loop. 4, No. 2 Nylon 2 loops. 5, Secure strand. 6, Mitek suture anchor. 7, Mitek mini anchor. 8, Mersilene tape. 9, Tendon.
© 1997 British Journal of Urology 79, 979–984
ME TH ODS OF RE PA IRI NG TH E SY MPH IS I S P UBIS I N BL ADD E R E XS TROP HY
Group 4, pelves repaired with two loops of #2 nylon, had the greatest mean (sd ) energy-to-failure of 0.0275%.mm (0.0054) of body weight. Group 2, the pelves repaired with the horizontal mattress suture, did not have a significantly lower energy-to-failure and neither did intact pelves. All other repair methods failed at lower energy. Most failure modes involved fracture of the bone. The intact pelves in group 1 failed by fracture of both pubic rami of one hemipelvis. In group 2, the horizontal mattress suture showed substantial stretch or subcatastrophic failure and then tore out of the bone at failure. Repair of pelves in groups 3, 4, and 5 all involved a suture material looped around the pubis and through itself. When load was applied, these loops tightened and gradually tore through the pubis until failure was complete (Fig. 6). Pelves in groups 6 and 7 failed when the suture anchors pulled free of the bone (Fig. 7). Yield to that point was a result of stretching of the suture material and partial tearing of the anchor through the bone. Pelves in group 8 failed by fracture of one pubis at the site of contact with the mersilene tape. This was the stiest repair method, and little diastasis was seen before failure. Pelves in group 9 showed no failure of bone, but the nylon sutures tore through the tendinous graft. This occurred gradually, which explains the low stiness as well as the low failure load.
Discussion Examining the radiographs and the bone, we concluded that the piglet model was valid. The porcine pelvis, like that of most quadrupeds, is elongated craniocaudally and is a substantially dierent shape than the human pelvis. Most of the strength in the human symphysis
Fig. 6. Failure of the nylon loop; the suture cut through the bone/ cartilage of the pubis. © 1997 British Journal of Urology 79, 979–984
983
Fig. 7. Failure of the Mitek suture anchor; the anchor pulled out of the pubis on one side.
probably lies in the inferior (arcuate) ligament. However, this study tested only the ability of the pubic bones to withstand load after dierent methods of repair. Like those of the human infant, the pubic bones of the piglet were only partially ossified, with a large amount of cartilage still present at the symphysis. Thus, the failure modes of the pelves tested would probably reflect those seen in humans. Piglets are roughly 1.1 kg at birth and reach 50–100 kg when fully grown, with about half of their growth during the first year. The specimens in this study were several weeks old and thus their bones were still immature. The few samples in this study limited the significance of any dierences in the results, but some conclusions are clear. The intact pelves failed by fracture of the pubic rami rather than symphyseal disruption. The softness of the bone dictated that the failure of repaired pelves involved bone at contact points and not the repair material. Those methods with greatest load-to-failure all had larger areas of contact between bone and repair material than the weaker methods. Maximizing the distribution of load should thus be a primary consideration in determining what repair method is best suited for surgical use. The only repair method in which the repair material failed was the tendon sutured across both superior pubic rami. However, this method failed at such a low load that it provides little insight. Stiness was highest for the mersilene tape and the G-II suture anchors; repairs made only with suture material were all less sti. It is possible that progressive failure was occurring with these specimens as the suture began to tear through the weakest sections of bone, and became catastrophic only when the bone was completely broken at one contact point. The three repair methods which withstood a load
984
J.S . SU SS M AN et al.
closest to that in group 2 could easily be augmented. Group 4 pelves were repaired with multiple nylon loops through the obturator foramen and more nylon loops could be used. Group 6 pelves were repaired with two G-II suture anchors, one in each superior ramus. Anchors could be added to each inferior ramus and sutures tied in dierent patterns to distribute the load. Group 8 pelves were repaired with one strip each of mersilene tape; several could be used to maximize the area of contact between the bone and repair material. These ideas merit investigation in a future study. All the repair methods were very weak compared to the intact symphysis. The best load-to-failure (#2 nylon horizontal mattress suture) was less than half of that for intact bone and the best stiness (Mersilene tape) was less than one-third that for intact symphysis. Further investigation is warranted to find methods which can better approximate the strength of intact symphysis. Finally, this study did not address behaviour in the biological environment. The material chosen must not be excessively prominent under the thin subcutaneous bridge and must also withstand repetitive low-stress cyclic loading rather than single catastrophic loading. Further research is indicated to devise a repair which is more physiological, stronger and more durable.
8
9
10
11
12
13
14 15
16 17
References 1 Epidemiology of bladder exstrophy and epispadias: a communication from the International Clearinghouse for Birth Defects Monitoring Systems. Teratology 1987; 36: 221–7 2 Perovic S, Brdar R, Scepanovic, D. Bladder exstrophy and anterior pelvic osteotomy. Br J Urol 1992; 70: 678–82 3 Sponseller PD, Bisson LJ, Gearhart JP, Jes RD, Magid D, Fishman E. The anatomy of the pelvis in the exstrophy complex. J Bone Joint Surg Am 1995; 77: 177–89 4 McKenna PH, Khoury AE, McLorie GA, Churchill BM, Babyn PB, Wedge JH. Iliac osteotomy: a model to compare the options in bladder and cloacal exstrophy reconstruction. J Urol 1994; 151: 182–6 5 Jes RD, Guice SL, Oesch I. The factors in successful exstrophy closure. J Urol 1982; 127: 974–6 6 Rickham PP, Stauer UG. Exstrophy of the bladder progress of management during the last 25 years. Prog Pediatr Surg 1984; 17: 169–88 7 Aadalen RJ, O’Phelan EH, Chisholm TC, McParland FA, Sweetser TH. Exstrophy of the bladder: long-term results of
18
19
20
bilateral posterior iliac osteotomies and two-stage anatomic repair. Clin Orthop 1980; 151: 193–200 Sponseller PD, Gearhart JP, Jes RD. Anterior innominate osteotomies for failure or late closure of bladder exstrophy. J Urol 1991; 146: 137–40 Sutherland D, Pike L, Kaufman K, Mowery C, Kaplan G, Romanus B. Hip function and gait in patients treated for bladder exstrophy. J Pediatr Orthop 1994; 14: 709–14 Scherz HC, Kaplan GW, Sutherland DH, Packer MG. Fascia lata and early spica casting as adjuncts in closure of bladder exstrophy. J Urol 1990; 144: 550–3 Schmidt AH, Keenen TL, Tank ES, Bird CB, Beals RK. Pelvic osteotomy for bladder exstrophy. J Pediatr Orthop 1993; 13: 214–9 Gokcora IH, Yazar T. Bilateral transverse iliac osteotomy in the correction of neonatal bladder extrophies. Int Surg 1989; 74: 123–5 Montagnani CA. Innominate osteotomy in reconstructive surgery for exstrophy of the bladder. J Pediatr Surg 1967; 2: 583–9 O’Phelan EH. Iliac osteotomy in exstrophy of the bladder. J Bone Joint Surg Am 1963; 45: 1409–21 Horoszowski H, Israeli A, Heim M, Jonash P, Farine I. A new orthopedic fixation method in the treatment of bladder extrophy. Clin Orthop 1982; 165: 200–3 Mollard P. Bladder reconstruction in exstrophy. J Urol 1980; 124: 525–9 Mollard P, Mouriquand PDE, Buttin X. Urinary continence after reconstruction of classical bladder exstrophy (73 cases). Br J Urol 1994; 73: 298–302 Slongo T, Jakob RP. The small AO external fixator in paediatric orthopaedics and trauma. Injury 1994; 25 Suppl 4: S-D77–84 Varga E, Hearn T, Powell J, Tile M. Eects of method of internal fixation of symphyseal disruptions on stability of the pelvic ring. Injury 1995; 26: 75–80 Gamble JG, Simmons SC, Freedman M. The symphysis pubis: anatomic and pathologic considerations. Clin Orthop 1986; 203: 261–72
Authors J.S. Sussman, BA. P.D. Sponseller, MD. J.P. Gearhart, MD. A.D.C. Valdevit, MS. J. Kier-York, RN. E.Y.S. Chao, PhD. Correspondence: Dr P.D. Sponseller, Department of Orthopaedic Surgery, Johns Hopkins University School of Medicine, 601 N. Caroline Street, Baltimore, Maryland 21287–0882, USA.
© 1997 British Journal of Urology 79, 979–984
