- Email: [email protected]
Osteitis pubis after Marshall-Marchetti-Krantz urethropexy: A pubic osteomyelitis
Osteitis pubis after Marshall-Marchetti-Krantz urethropexy: A pubic osteomyelitis Dorothy N. Kammerer-Doak, MD, Jeffrey L. Cornella, MD, Javier F. Magrina, MD, C. Robert Stanhope, MD, and Jerry Smilack, MD Scottsdale, Arizona, and Rochester, Minnesota OBJECTIVE: Our purpose was to review cases of osteitis pubis encountered at our institution after MarshallMarchetti-Krantz retropubic urethropexy. STUDY DESIGN: The charts of patients diagnosed with osteitis pubis subsequent to Marshall-MarchettiKrantz retropubic urethropexy from 1980 to 1994 were reviewed. RESULTS: Fifteen cases of osteitis pubis were diagnosed after 2030 Marshall-Marchetti-Krantz procedures (0.74%). Onset of symptoms related to osteitis pubis began a mean of 69.8 days postoperatively (range 10 to 459 days). Although initial plain films of the symphysis pubis were normal in 7 (54%), radiographic abnormality was eventually demonstrated in all a mean of 25.7 weeks after surgery (range 4 to 78 weeks). A variety of conservative treatments resulted in symptomatic relief in 47%. Seven of the remaining patients underwent operative therapy with partial or complete relief noted in all. Subsequent bone cultures were positive in 5 (71%). At follow-up a mean of 58 months after the Marshall-Marchetti-Krantz procedure complete resolution of symptoms was noted in 33% and continued pain or ambulatory difficulty in the remainder. There was no relationship between postoperative urinary tract infections, postoperative complications, presenting sign of fever, elevated leukocyte count or sedimentation rate, and subsequent operative intervention (P > .05). CONCLUSIONS: Osteitis pubis after urogynecologic surgery is an uncommon event requiring aggressive surgical and antibiotic therapy. When bone cultures are performed, a microbial cause may be demonstrated in as many as 71% of patients. (Am J Obstet Gynecol 1998;179:586-90.)
Key words: Osteitis pubis, Marshall-Marchetti-Krantz retropubic urethropexy, osteomyelitis
Osteitis pubis has been described as a noninfectious, self-limited inflammatory condition of the symphysis pubis associated with urologic and gynecologic surgical procedures, trauma, connective tissue disorders, and pregnancy.1-5 Since its first description in the English literature by Beer4 in 1924 after a suprapubic prostatectomy, the etiology, pathophysiologic features, and treatment of this uncommon condition of the symphysis pubis has remained obscure. The purpose of this descriptive study is to review the experience of patients diagnosed with osteitis pubis after Marshall-Marchetti-Krantz retropubic urethropexy at Mayo Clinic Rochester and Scottsdale over the past 14 years. Presenting signs and symptoms, modalities used for diagnosis, effectiveness of various treatments, and disease course are reviewed.
From the Department of Operative Gynecology, Mayo Clinic. Received for publication July 16, 1997; revised February 4, 1998; accepted February 11, 1998. Reprint requests: Dorothy N. Kammerer-Doak, MD, Department of Obstetrics and Gynecology, University of New Mexico Hospital, 2211 Lomas Blvd NE, ACC-4, Albuquerque, NM 87131. Copyright © 1998 by Mosby, Inc. 0002-9378/98 $5.00 + 0 6/1/89509
586
Material and methods All female patients with the diagnosis of osteitis pubis from 1980 to 1994 were identified by a computer-generated search. After review of these charts, 15 women were found to have developed this complication subsequent to Marshall-Marchetti-Krantz retropubic urethropexy. The charts of these patients were reviewed for patient characteristics, concomitant surgical procedures, type of suture used, perioperative complications, presenting signs and symptoms of osteitis pubis, diagnostic modalities, treatment, and disease course. Fisher’s exact test was used to examine the relationship between perioperative variables, presenting symptoms, and ultimate operative treatment. Statistical analysis was performed with use of EpiInfo version 6 (Centers for Disease Control and Prevention, Atlanta, Ga) with P < .05 considered statistically significant. Results Over this 14-year time span, 15 cases of osteitis pubis were diagnosed after 2030 Marshall-Marchetti-Krantz nonlaparoscopic procedures, a known occurrence of 0.74%. The mean age and weight of the patients developing this complication was 62.6 years (SD 7.2) and 70.7
Kammerer-Doak et al 587
Volume 179, Number 3, Part 1 Am J Obstet Gynecol
Fig 1. Plain radiograph of pelvis with symphyseal sclerosis, lysis, and widening of joint spaces.
kg (SD 11.93), respectively; 80% were postmenopausal. None of the patients had a documented urinary tract infection preoperatively. Retropubic drains were used in 4 (27%) patients. Permanent suture, Ethibond (Ethicon, Cincinnati, Ohio), was used in 10 (67%) cases and delayed absorbable in 2. In the remaining 3 patients the type of suture was not specified. A total of 2 to 4 sutures were placed periurethrally and brought through symphyseal cartilage. Intentional cystotomy for precise placement of sutures was performed in 80%, and 1 patient had an inadvertent cystotomy. Forty-one percent of patients received prophylactic antibiotics with a secondgeneration cephalosporin. No patient received perioperative heparin for deep venous thrombosis prophylaxis. Many patients underwent additional surgical procedures (Table I). Postoperative complications were encountered in 4 (27%) patients (Table II). Postoperative urinary tract infections were caused by a variety of organisms, including Escherichia coli, Klebsiella, Staphylococcus aureus, enterococci, Pseudomonas, and other coliforms in 7 (47%). Three of these patients had multiple postoperative urinary tract infections with more than 1 organism. Suprapubic or transurethral catheters were used in all patients for bladder drainage a mean of 15.8 days (SD 9.7). One patient had prolonged urinary retention necessitating self-intermittent catheterization. The diagnosis of osteitis pubis was based on classic symptoms (suprapubic pain, difficulty, and pain with ambulation) and abnormal radiographic findings. The mean time of onset of symptoms from the date of surgery was 69.8 days (range 10 to 459 days). If the single patient who had osteitis >1 year after the Marshall-MarchettiKrantz procedure is excluded, the mean time was 42 days (SD 35.5). Fever (temperature >38°C) was noted in 27%, leukocytosis (>10 000/µL) without left shift in 20%, and an elevated sedimentation rate (>20 mm/h) in 67%. Plain anteroposterior radiograph of the symphysis pubis
Fig 2. Computed tomographic scan demonstrating soft tissue mass (arrow).
Table I. Concurrent procedures
Hysterectomy Appendectomy Moschcowitz culdoplasty Paravaginal defect repair Abdominal sacral colpopexy Posterior repair
No.
%
5 1 2 2 2 2
33 7 13 13 13 13
was used as the diagnostic modality in 13 (87%), whereas bone scan and computerized tomography were diagnostic of osteitis pubis in the other 2 patients. Pelvic radiographic findings included pubic bone sclerosis, widening of the joint spaces, and rarefaction (Fig 1) and were detected a mean of 25.7 weeks after the MarshallMarchetti-Krantz procedure (range 4 to 78 weeks). Initial radiographs obtained at a mean of 13 weeks (SD 24.7) from surgery and 14 weeks (SD 16.7) from onset of symptoms were normal in 7 of 13. Subsequent x-ray films obtained a mean of 12.3 weeks (SD 17) after the initial radiograph demonstrated changes consistent with osteitis pubis in all patients. Computed tomographic scan was performed in 7 patients, with abnormalities noted in 6 (86%), including 2 with soft tissue masses suggestive of abscess or sequestrum (Fig 2). Bone or indium-labeled white blood cell scan was positive for increased symphyseal uptake in 12 of 13 patients (Fig 3). Fine needle aspirate, performed in 6 patients, was positive for E coli in 1. All the patients underwent a variety of conservative treatments (Table III), resulting in partial or complete symptomatic relief in 47% after a mean of 18.7 weeks
588 Kammerer-Doak et al
September 1998 Am J Obstet Gynecol
Fig 3. Radionuclide bone scan showing increased uptake in region of symphysis pubis (large arrow).
Table II. Postoperative complications No. Wound seroma Wound infection Pyelonephritis Urinary tract infection None
2 1 1 7 10
% 13 7 7 47 73
Table III. Conservative treatment Oral steroids Injected steroids Oral antibiotics Intravenous antibiotics Nonsteroidal anti-inflammatory drugs Walker
33% 7% 67% 20% 80% 13%
(range 4 to 30 weeks) from initiation of therapy. Many patients received more than 1 treatment modality. Three patients received intravenous antibiotic therapy, 2 whom were rehospitalized for postoperative fevers before the diagnosis of osteitis pubis. One patient with mild symptoms had no relief after conservative treatment only. Seven of the remaining women underwent operative therapy a mean of 20.4 weeks after the Marshall-Marchetti-Krantz procedure consisting of symphyseal debridement (86%), removal of permanent suture (57%), symphyseal wedge resection and partial cystectomy (14%), and open needle biopsy (14%). Subsequent bone cultures were positive in a total of 5 (71%), including 4 of the 5 with negative fine needle aspirate cultures and the 1 with positive fine needle
aspirate culture. The identified organisms included enterococci (n = 1), lactobacilli (n = 1), E coli (n = 1), and staphylococci (n = 2). These patients were treated with 3 to 8 weeks of oral or intravenous antibiotics appropriate for the organism identified on culture. Partial or complete relief was noted in all patients a mean of 11.9 weeks (SD 17.8) after operative intervention and antibiotic therapy. Incontinence recurred in 1 patient several months after operative treatment consisting of open needle biopsy and was treated with a repeat Marshall-Marchetti-Krantz procedure 1 year later without sequelae. There was no relationship between concurrent operative procedures, postoperative urinary tract infection, any postoperative complication, presenting sign of fever, elevated leukocyte count or sedimentation rate, and subsequent operative intervention (P > .05). The mean length of follow-up after the MarshallMarchetti-Krantz procedure was 58 months (range 1.5 to 142 months). Complete resolution of symptoms was noted in 33% of patients. Improvement in pubic pain and normal ambulation was reported in 47% and difficulty with walking but no pubic pain in 7%. Comment The diagnosis of osteitis pubis is based on typical symptoms of suprapubic discomfort, difficulty with ambulation and wide-based, waddling gait, and radiographic changes of irregular bony margins and rarefaction and widening of the symphyseal joint spaces.1-4, 6-8 Historically, its clinical course resolved over several months to several years.2, 3, 7, 9 Various treatments, both operative and medical, have been used with considerable variation of effectiveness. Because early reports suggested
Volume 179, Number 3, Part 1 Am J Obstet Gynecol
that osteitis pubis was not responsive to antibiotics, the disease was described as noninfectious, although an association was noted with abscesses, draining sinuses, and wound and urinary tract infections.1, 3, 7, 10, 11 A minority of patients had bone cultures, many which grew organisms considered minimally virulent, leading to the mistaken conclusion that postoperative osteitis pubis was noninfectious.3, 8, 9 Interestingly, Lavalle and Hamm8 in 1951 performed open bone biopsy and cultures on 3 patients with osteitis pubis after retropubic prostatectomy, from which Pseudomonas was isolated. A probable noninfectious, inflammatory type of osteitis pubis is noted to occur with trauma, athletic activity, pregnancy, and rheumatic disorders.3, 5, 6, 12 Typical symptoms and radiographic findings of osteitis pubis are noted in these patients. Erosion and sclerosis of the symphysis pubis have been noted as incidental findings in asymptomatic patients, making it necessary to distinguish between osteitis pubis as a symptom-producing disease and incidental radiographic findings in the absence of disease.3, 6 Treatment for osteitis pubis associated with these conditions with nonsteroidal anti-inflammatory agents, physical therapy, and steroids can be successful, although some patients may require operative treatment with symphyseal wedge resection.6, 7, 9, 13, 14 The etiology of osteitis pubis has been unclear. The 3 main hypotheses for the cause of postoperative osteitis pubis include trauma, impaired vascular circulation, and infection.1-4, 6, 9, 11, 14 Beer4 was the first to postulate that osteitis pubis was the result of intraoperative trauma to the symphysis and its muscular attachments either from surgical instruments or retractors. However, attempts at producing osteitis pubis through surgical trauma to the symphysis pubis have been unsuccessful in laboratory animals.15, 16 Additionally, even with the MarshallMarchetti-Krantz procedure, in which sutures are placed directly into the periosteum or cartilage of the symphysis pubis, osteitis pubis is uncommon, occurring only in 1% to 2.5%.17 Finally, subsequent to symphysiotomies and pubiotomies, osteitis and osteomyelitis is rare.9 Impaired circulation resulting from damage to the venous plexus supplying the symphysis leading to obstruction, venous dilation, thrombosis, and bony demineralization has also been proposed as a causative factor in osteitis.2, 3, 9 With use of phlebography, impaired venous drainage from the pubic bone to the pelvic veins has been demonstrated.9 Additionally, response to treatment with heparinization in patients with osteitis pubis has been reported.18, 19 However, this response to heparinization is not consistent, and perioperative heparin prophylaxis does not prevent the disease from developing.19 Symptoms and findings of osteitis pubis, including fever, leukocytosis, and bone biopsy specimens, can be identical to those of osteomyelitis, supporting an infectious etiology.3, 7, 20, 21 The differentiating factors between these 2
Kammerer-Doak et al 589
entities are stated to be radiographic findings of bilateral, symmetric symphyseal involvement and absence of sequestra and other evidence of progressive, bony destruction and the self-limited nature of osteitis pubis compared with osteomyelitis.3, 7, 20 Radiographic changes of osteitis pubis can be identical to those of osteomyelitis.3, 22 Osteomyelitis has been reported to resolve spontaneously without antibiotics or debridement.8, 10, 11 Although early reports noted treatment of osteitis pubis with antibiotics to be unsuccessful,1-3, 7, 8, 11 these failures may represent inadequate and inappropriate antibiotic treatment because osteomyelitis may require 4 to 6 weeks of intravenous antibiotic therapy.7, 10, 11, 20-23 Because blood flow to the symphysis pubis can be compromised with untreated osteomyelitis, surgical debridement and antibiotic therapy are usually necessary.24 Bone cultures from patients with osteitis pubis do not always yield bacterial growth.2, 3, 6-8, 14 However, even when bone cultures are negative, a diagnosis of osteomyelitis can be made on the basis of radiographic and pathologic findings.7, 20, 22 A review noted that in 50% of patients with osteomyelitis no bacteria were isolated.7 Thus postoperative osteitis pubis shares many common signs, symptoms, and radiographic, microbial, and pathologic findings with osteomyelitis, indicating a infectious etiology. Recent publications, mostly within the urologic literature, have noted a high number of positive bone cultures in patients with postoperative osteitis pubis, most commonly aerobic gram-negative organisms, especially E coli, Pseudomonas, and Proteus, as well as streptococcal and staphylococcal species.7, 10, 11, 20-23 Anaerobic bacteria have also been isolated from patients with postoperative osteitis pubis.23, 25 Our study has the largest numbers of patients with osteitis pubis after gynecologic surgery published in the last decade. Pubic osteomyelitis, on the basis of positive bone cultures, was noted in 71% of those with cultures obtained with open surgical procedures. The occurrence of osteitis pubis noted in this study is 0.74%, lower than the average 2.5% noted in a review of the literature including 2700 Marshall-Marchetti-Krantz procedures.17 The range of incidence noted in that publication was 0% to 10%, with the higher number probably reflecting a clinical diagnosis of osteitis pubis without radiographic confirmation.17 Because our study is retrospective, it is possible that some cases of osteitis pubis were not identified by computer search, leading to a falsely low occurrence. We believe a significant degree of this is unlikely, given the morbidity of osteomyelitis. Postoperative osteitis pubis was an obscure disease, with multiple theories of pathogenesis and inconsistent results with a wide variety of treatment modalities. Many of the initial series were reported before the antibiotics available today and without diagnostic radiographic modalities such as magnetic resonance imaging and bone and computed tomographic scans.1-4, 8, 13, 15 The
590 Kammerer-Doak et al
recommended management of osteitis pubis developing subsequent to pelvic surgery is based on opinions derived from this dated literature, especially the dictum that postoperative osteitis pubis is not an infection.11 However, as noted in our series, the clinical, radiologic, and microbiologic data are indicative of osteomyelitis. Such clinically significant infections are best treated with bactericidal antibiotics that achieve high serum and tissue concentrations, with or without surgical debridement, accompanied by analgesics and ambulatory aids. Osteitis pubis, an inflammatory and noninfectious condition, may occur after pelvic surgery, but it is probably more commonly noted in association with pregnancy, trauma, athletic activity, and rheumatic disorders. We concur with the belief of Sexton et al11 that an inflammatory process involving the symphysis pubis after uterine or bladder suspension surgery probably represents infection and that an aggressive diagnostic approach to establish a microbial etiology be strongly considered so that a rational treatment plan can then be undertaken.
September 1998 Am J Obstet Gynecol
8. 9. 10. 11.
12.
13. 14.
15. 16. 17. 18. 19.
REFERENCES
1. Cibert J. Post-operative osteitis pubis: causes and treatment. Br J Urol 1952;24:213-5. 2. O’Leary JA. Osteitis pubis following vesicourethral suspension. Obstet Gynecol 1964;24:73-6. 3. Coventry MB, Mitchell WC. Osteitis pubis: observations based on a study of 45 patients. JAMA 1961;178:898-905. 4. Beer E. Periostitis of the symphysis and descending rami of the pubes following suprapubic operations. Int J Med Surg 1924;37:224-5. 5. Sequeira W. Diseases of the symphysis pubis. Semin Arthritis Rheum 1986;16:11-21. 6. Fricker PA, Taunton JE, Ammann W. Osteitis pubis in athletes: infection, inflammation or injury? Sports Med 1991;12:266-79. 7. Rosenthall RE, Spickard WA, Markham RD, Rhamy RK.
20. 21. 22.
23. 24.
25.
Osteomyelitis of the symphysis pubis: a separate disease from osteitis pubis. J Bone Joint Surg 1982;64:123-8. Lavalle LL, Harem FC. Osteitis pubis: its etiology and pathology. J Urol 1951;66:418-22. Steinbach HL, Petrakis NL, Gilfillan RS, Smith DR. The pathogenesis of osteitis pubis. J Urol 1955;74:840-6. Bouza E, Winston DJ, Hewitt WL. Infectious osteitis pubis. Urology 1978;12:663-9. Sexton DJ, Heskestad L, Lambeth WR, McCallum, Levin LS, Corey GR. Postoperative pubic osteomyelitis misdiagnosed as osteitis pubis: report of four cases and a review. Clin Infect Dis 1993;17:695-700. Gamble JG, Simmons SC, Freedman M. The symphysis pubis: anatomic and pathologic considerations. Clin Orth Related Res 1986;203:261-72. Schnute WJ. Osteitis pubis. Clin Orthoped 1961;20:187-92. Grace JN, Sim FH, Shives TC, Coventry MB. Wedge resection of the symphysis pubis for the treatment of osteitis pubis. J Bone Joint Surg 1989;71:358-64. Wheeler WK. Periostitis pubes following suprapubic cystostomy. J Urol 1941;45:467-75. Beneventi FA, Spellman R. Unsuccessful attempts to produce osteitis pubis in dogs. Urology 1953;69:405-6. Mainprize TC, Drutz HP. The Marshall-Marchetti-Krantz procedure: a critical review. Obstet Gynecol Surv 1988;43:724-9. Merimsky E, Canetti R, Firstater M. Osteitis pubis: treatment by heparinisation. Br J Urol 1981;53:154-6. Nissenkorn I, Servadio C, Lubin E. The treatment of osteitis pubis with heparin. J Urol 1981;125:528-9. Burns JR, Gregory JG. Osteomyelitis of the pubic symphysis after urologic surgery. J Urol 1977;118:803-5. del Busto R, Quinn EL, Fisher EJ, Madhavan T. Osteomyelitis of the pubis: report of seven cases. JAMA 1982;248:1498-500. Lupovitch A, Elie JC, Wysocki R. Diagnosis of acute bacterial osteomyelitis of the pubis by means of fine needle aspiration. Acta Cytol 1989;33:649-51. Gilbert DN, Azorr M, Gore R, Hofeldt R. The bacterial causation of postoperative osteitis pubis. Surg Gynecol Obstet 1975;141:195-8. Hoyme UB, Tamimi HK, Eschenbach DA, Ramsey PG, Figge DC. Osteomyelitis pubis after radical gynecologic operation. Obstet Gynecol 1984;63 Suppl: 47S-53S. Jenkins FH, Raft MJ, Florman LD, Day TG, Templeton WC. Pubic osteomyelitis due to anaerobic Bacteria. Arch Intern Med 1984;144:842-3.
Key words: Osteitis pubis, Marshall-Marchetti-Krantz retropubic urethropexy, osteomyelitis
Osteitis pubis has been described as a noninfectious, self-limited inflammatory condition of the symphysis pubis associated with urologic and gynecologic surgical procedures, trauma, connective tissue disorders, and pregnancy.1-5 Since its first description in the English literature by Beer4 in 1924 after a suprapubic prostatectomy, the etiology, pathophysiologic features, and treatment of this uncommon condition of the symphysis pubis has remained obscure. The purpose of this descriptive study is to review the experience of patients diagnosed with osteitis pubis after Marshall-Marchetti-Krantz retropubic urethropexy at Mayo Clinic Rochester and Scottsdale over the past 14 years. Presenting signs and symptoms, modalities used for diagnosis, effectiveness of various treatments, and disease course are reviewed.
From the Department of Operative Gynecology, Mayo Clinic. Received for publication July 16, 1997; revised February 4, 1998; accepted February 11, 1998. Reprint requests: Dorothy N. Kammerer-Doak, MD, Department of Obstetrics and Gynecology, University of New Mexico Hospital, 2211 Lomas Blvd NE, ACC-4, Albuquerque, NM 87131. Copyright © 1998 by Mosby, Inc. 0002-9378/98 $5.00 + 0 6/1/89509
586
Material and methods All female patients with the diagnosis of osteitis pubis from 1980 to 1994 were identified by a computer-generated search. After review of these charts, 15 women were found to have developed this complication subsequent to Marshall-Marchetti-Krantz retropubic urethropexy. The charts of these patients were reviewed for patient characteristics, concomitant surgical procedures, type of suture used, perioperative complications, presenting signs and symptoms of osteitis pubis, diagnostic modalities, treatment, and disease course. Fisher’s exact test was used to examine the relationship between perioperative variables, presenting symptoms, and ultimate operative treatment. Statistical analysis was performed with use of EpiInfo version 6 (Centers for Disease Control and Prevention, Atlanta, Ga) with P < .05 considered statistically significant. Results Over this 14-year time span, 15 cases of osteitis pubis were diagnosed after 2030 Marshall-Marchetti-Krantz nonlaparoscopic procedures, a known occurrence of 0.74%. The mean age and weight of the patients developing this complication was 62.6 years (SD 7.2) and 70.7
Kammerer-Doak et al 587
Volume 179, Number 3, Part 1 Am J Obstet Gynecol
Fig 1. Plain radiograph of pelvis with symphyseal sclerosis, lysis, and widening of joint spaces.
kg (SD 11.93), respectively; 80% were postmenopausal. None of the patients had a documented urinary tract infection preoperatively. Retropubic drains were used in 4 (27%) patients. Permanent suture, Ethibond (Ethicon, Cincinnati, Ohio), was used in 10 (67%) cases and delayed absorbable in 2. In the remaining 3 patients the type of suture was not specified. A total of 2 to 4 sutures were placed periurethrally and brought through symphyseal cartilage. Intentional cystotomy for precise placement of sutures was performed in 80%, and 1 patient had an inadvertent cystotomy. Forty-one percent of patients received prophylactic antibiotics with a secondgeneration cephalosporin. No patient received perioperative heparin for deep venous thrombosis prophylaxis. Many patients underwent additional surgical procedures (Table I). Postoperative complications were encountered in 4 (27%) patients (Table II). Postoperative urinary tract infections were caused by a variety of organisms, including Escherichia coli, Klebsiella, Staphylococcus aureus, enterococci, Pseudomonas, and other coliforms in 7 (47%). Three of these patients had multiple postoperative urinary tract infections with more than 1 organism. Suprapubic or transurethral catheters were used in all patients for bladder drainage a mean of 15.8 days (SD 9.7). One patient had prolonged urinary retention necessitating self-intermittent catheterization. The diagnosis of osteitis pubis was based on classic symptoms (suprapubic pain, difficulty, and pain with ambulation) and abnormal radiographic findings. The mean time of onset of symptoms from the date of surgery was 69.8 days (range 10 to 459 days). If the single patient who had osteitis >1 year after the Marshall-MarchettiKrantz procedure is excluded, the mean time was 42 days (SD 35.5). Fever (temperature >38°C) was noted in 27%, leukocytosis (>10 000/µL) without left shift in 20%, and an elevated sedimentation rate (>20 mm/h) in 67%. Plain anteroposterior radiograph of the symphysis pubis
Fig 2. Computed tomographic scan demonstrating soft tissue mass (arrow).
Table I. Concurrent procedures
Hysterectomy Appendectomy Moschcowitz culdoplasty Paravaginal defect repair Abdominal sacral colpopexy Posterior repair
No.
%
5 1 2 2 2 2
33 7 13 13 13 13
was used as the diagnostic modality in 13 (87%), whereas bone scan and computerized tomography were diagnostic of osteitis pubis in the other 2 patients. Pelvic radiographic findings included pubic bone sclerosis, widening of the joint spaces, and rarefaction (Fig 1) and were detected a mean of 25.7 weeks after the MarshallMarchetti-Krantz procedure (range 4 to 78 weeks). Initial radiographs obtained at a mean of 13 weeks (SD 24.7) from surgery and 14 weeks (SD 16.7) from onset of symptoms were normal in 7 of 13. Subsequent x-ray films obtained a mean of 12.3 weeks (SD 17) after the initial radiograph demonstrated changes consistent with osteitis pubis in all patients. Computed tomographic scan was performed in 7 patients, with abnormalities noted in 6 (86%), including 2 with soft tissue masses suggestive of abscess or sequestrum (Fig 2). Bone or indium-labeled white blood cell scan was positive for increased symphyseal uptake in 12 of 13 patients (Fig 3). Fine needle aspirate, performed in 6 patients, was positive for E coli in 1. All the patients underwent a variety of conservative treatments (Table III), resulting in partial or complete symptomatic relief in 47% after a mean of 18.7 weeks
588 Kammerer-Doak et al
September 1998 Am J Obstet Gynecol
Fig 3. Radionuclide bone scan showing increased uptake in region of symphysis pubis (large arrow).
Table II. Postoperative complications No. Wound seroma Wound infection Pyelonephritis Urinary tract infection None
2 1 1 7 10
% 13 7 7 47 73
Table III. Conservative treatment Oral steroids Injected steroids Oral antibiotics Intravenous antibiotics Nonsteroidal anti-inflammatory drugs Walker
33% 7% 67% 20% 80% 13%
(range 4 to 30 weeks) from initiation of therapy. Many patients received more than 1 treatment modality. Three patients received intravenous antibiotic therapy, 2 whom were rehospitalized for postoperative fevers before the diagnosis of osteitis pubis. One patient with mild symptoms had no relief after conservative treatment only. Seven of the remaining women underwent operative therapy a mean of 20.4 weeks after the Marshall-Marchetti-Krantz procedure consisting of symphyseal debridement (86%), removal of permanent suture (57%), symphyseal wedge resection and partial cystectomy (14%), and open needle biopsy (14%). Subsequent bone cultures were positive in a total of 5 (71%), including 4 of the 5 with negative fine needle aspirate cultures and the 1 with positive fine needle
aspirate culture. The identified organisms included enterococci (n = 1), lactobacilli (n = 1), E coli (n = 1), and staphylococci (n = 2). These patients were treated with 3 to 8 weeks of oral or intravenous antibiotics appropriate for the organism identified on culture. Partial or complete relief was noted in all patients a mean of 11.9 weeks (SD 17.8) after operative intervention and antibiotic therapy. Incontinence recurred in 1 patient several months after operative treatment consisting of open needle biopsy and was treated with a repeat Marshall-Marchetti-Krantz procedure 1 year later without sequelae. There was no relationship between concurrent operative procedures, postoperative urinary tract infection, any postoperative complication, presenting sign of fever, elevated leukocyte count or sedimentation rate, and subsequent operative intervention (P > .05). The mean length of follow-up after the MarshallMarchetti-Krantz procedure was 58 months (range 1.5 to 142 months). Complete resolution of symptoms was noted in 33% of patients. Improvement in pubic pain and normal ambulation was reported in 47% and difficulty with walking but no pubic pain in 7%. Comment The diagnosis of osteitis pubis is based on typical symptoms of suprapubic discomfort, difficulty with ambulation and wide-based, waddling gait, and radiographic changes of irregular bony margins and rarefaction and widening of the symphyseal joint spaces.1-4, 6-8 Historically, its clinical course resolved over several months to several years.2, 3, 7, 9 Various treatments, both operative and medical, have been used with considerable variation of effectiveness. Because early reports suggested
Volume 179, Number 3, Part 1 Am J Obstet Gynecol
that osteitis pubis was not responsive to antibiotics, the disease was described as noninfectious, although an association was noted with abscesses, draining sinuses, and wound and urinary tract infections.1, 3, 7, 10, 11 A minority of patients had bone cultures, many which grew organisms considered minimally virulent, leading to the mistaken conclusion that postoperative osteitis pubis was noninfectious.3, 8, 9 Interestingly, Lavalle and Hamm8 in 1951 performed open bone biopsy and cultures on 3 patients with osteitis pubis after retropubic prostatectomy, from which Pseudomonas was isolated. A probable noninfectious, inflammatory type of osteitis pubis is noted to occur with trauma, athletic activity, pregnancy, and rheumatic disorders.3, 5, 6, 12 Typical symptoms and radiographic findings of osteitis pubis are noted in these patients. Erosion and sclerosis of the symphysis pubis have been noted as incidental findings in asymptomatic patients, making it necessary to distinguish between osteitis pubis as a symptom-producing disease and incidental radiographic findings in the absence of disease.3, 6 Treatment for osteitis pubis associated with these conditions with nonsteroidal anti-inflammatory agents, physical therapy, and steroids can be successful, although some patients may require operative treatment with symphyseal wedge resection.6, 7, 9, 13, 14 The etiology of osteitis pubis has been unclear. The 3 main hypotheses for the cause of postoperative osteitis pubis include trauma, impaired vascular circulation, and infection.1-4, 6, 9, 11, 14 Beer4 was the first to postulate that osteitis pubis was the result of intraoperative trauma to the symphysis and its muscular attachments either from surgical instruments or retractors. However, attempts at producing osteitis pubis through surgical trauma to the symphysis pubis have been unsuccessful in laboratory animals.15, 16 Additionally, even with the MarshallMarchetti-Krantz procedure, in which sutures are placed directly into the periosteum or cartilage of the symphysis pubis, osteitis pubis is uncommon, occurring only in 1% to 2.5%.17 Finally, subsequent to symphysiotomies and pubiotomies, osteitis and osteomyelitis is rare.9 Impaired circulation resulting from damage to the venous plexus supplying the symphysis leading to obstruction, venous dilation, thrombosis, and bony demineralization has also been proposed as a causative factor in osteitis.2, 3, 9 With use of phlebography, impaired venous drainage from the pubic bone to the pelvic veins has been demonstrated.9 Additionally, response to treatment with heparinization in patients with osteitis pubis has been reported.18, 19 However, this response to heparinization is not consistent, and perioperative heparin prophylaxis does not prevent the disease from developing.19 Symptoms and findings of osteitis pubis, including fever, leukocytosis, and bone biopsy specimens, can be identical to those of osteomyelitis, supporting an infectious etiology.3, 7, 20, 21 The differentiating factors between these 2
Kammerer-Doak et al 589
entities are stated to be radiographic findings of bilateral, symmetric symphyseal involvement and absence of sequestra and other evidence of progressive, bony destruction and the self-limited nature of osteitis pubis compared with osteomyelitis.3, 7, 20 Radiographic changes of osteitis pubis can be identical to those of osteomyelitis.3, 22 Osteomyelitis has been reported to resolve spontaneously without antibiotics or debridement.8, 10, 11 Although early reports noted treatment of osteitis pubis with antibiotics to be unsuccessful,1-3, 7, 8, 11 these failures may represent inadequate and inappropriate antibiotic treatment because osteomyelitis may require 4 to 6 weeks of intravenous antibiotic therapy.7, 10, 11, 20-23 Because blood flow to the symphysis pubis can be compromised with untreated osteomyelitis, surgical debridement and antibiotic therapy are usually necessary.24 Bone cultures from patients with osteitis pubis do not always yield bacterial growth.2, 3, 6-8, 14 However, even when bone cultures are negative, a diagnosis of osteomyelitis can be made on the basis of radiographic and pathologic findings.7, 20, 22 A review noted that in 50% of patients with osteomyelitis no bacteria were isolated.7 Thus postoperative osteitis pubis shares many common signs, symptoms, and radiographic, microbial, and pathologic findings with osteomyelitis, indicating a infectious etiology. Recent publications, mostly within the urologic literature, have noted a high number of positive bone cultures in patients with postoperative osteitis pubis, most commonly aerobic gram-negative organisms, especially E coli, Pseudomonas, and Proteus, as well as streptococcal and staphylococcal species.7, 10, 11, 20-23 Anaerobic bacteria have also been isolated from patients with postoperative osteitis pubis.23, 25 Our study has the largest numbers of patients with osteitis pubis after gynecologic surgery published in the last decade. Pubic osteomyelitis, on the basis of positive bone cultures, was noted in 71% of those with cultures obtained with open surgical procedures. The occurrence of osteitis pubis noted in this study is 0.74%, lower than the average 2.5% noted in a review of the literature including 2700 Marshall-Marchetti-Krantz procedures.17 The range of incidence noted in that publication was 0% to 10%, with the higher number probably reflecting a clinical diagnosis of osteitis pubis without radiographic confirmation.17 Because our study is retrospective, it is possible that some cases of osteitis pubis were not identified by computer search, leading to a falsely low occurrence. We believe a significant degree of this is unlikely, given the morbidity of osteomyelitis. Postoperative osteitis pubis was an obscure disease, with multiple theories of pathogenesis and inconsistent results with a wide variety of treatment modalities. Many of the initial series were reported before the antibiotics available today and without diagnostic radiographic modalities such as magnetic resonance imaging and bone and computed tomographic scans.1-4, 8, 13, 15 The
590 Kammerer-Doak et al
recommended management of osteitis pubis developing subsequent to pelvic surgery is based on opinions derived from this dated literature, especially the dictum that postoperative osteitis pubis is not an infection.11 However, as noted in our series, the clinical, radiologic, and microbiologic data are indicative of osteomyelitis. Such clinically significant infections are best treated with bactericidal antibiotics that achieve high serum and tissue concentrations, with or without surgical debridement, accompanied by analgesics and ambulatory aids. Osteitis pubis, an inflammatory and noninfectious condition, may occur after pelvic surgery, but it is probably more commonly noted in association with pregnancy, trauma, athletic activity, and rheumatic disorders. We concur with the belief of Sexton et al11 that an inflammatory process involving the symphysis pubis after uterine or bladder suspension surgery probably represents infection and that an aggressive diagnostic approach to establish a microbial etiology be strongly considered so that a rational treatment plan can then be undertaken.
September 1998 Am J Obstet Gynecol
8. 9. 10. 11.
12.
13. 14.
15. 16. 17. 18. 19.
REFERENCES
1. Cibert J. Post-operative osteitis pubis: causes and treatment. Br J Urol 1952;24:213-5. 2. O’Leary JA. Osteitis pubis following vesicourethral suspension. Obstet Gynecol 1964;24:73-6. 3. Coventry MB, Mitchell WC. Osteitis pubis: observations based on a study of 45 patients. JAMA 1961;178:898-905. 4. Beer E. Periostitis of the symphysis and descending rami of the pubes following suprapubic operations. Int J Med Surg 1924;37:224-5. 5. Sequeira W. Diseases of the symphysis pubis. Semin Arthritis Rheum 1986;16:11-21. 6. Fricker PA, Taunton JE, Ammann W. Osteitis pubis in athletes: infection, inflammation or injury? Sports Med 1991;12:266-79. 7. Rosenthall RE, Spickard WA, Markham RD, Rhamy RK.
20. 21. 22.
23. 24.
25.
Osteomyelitis of the symphysis pubis: a separate disease from osteitis pubis. J Bone Joint Surg 1982;64:123-8. Lavalle LL, Harem FC. Osteitis pubis: its etiology and pathology. J Urol 1951;66:418-22. Steinbach HL, Petrakis NL, Gilfillan RS, Smith DR. The pathogenesis of osteitis pubis. J Urol 1955;74:840-6. Bouza E, Winston DJ, Hewitt WL. Infectious osteitis pubis. Urology 1978;12:663-9. Sexton DJ, Heskestad L, Lambeth WR, McCallum, Levin LS, Corey GR. Postoperative pubic osteomyelitis misdiagnosed as osteitis pubis: report of four cases and a review. Clin Infect Dis 1993;17:695-700. Gamble JG, Simmons SC, Freedman M. The symphysis pubis: anatomic and pathologic considerations. Clin Orth Related Res 1986;203:261-72. Schnute WJ. Osteitis pubis. Clin Orthoped 1961;20:187-92. Grace JN, Sim FH, Shives TC, Coventry MB. Wedge resection of the symphysis pubis for the treatment of osteitis pubis. J Bone Joint Surg 1989;71:358-64. Wheeler WK. Periostitis pubes following suprapubic cystostomy. J Urol 1941;45:467-75. Beneventi FA, Spellman R. Unsuccessful attempts to produce osteitis pubis in dogs. Urology 1953;69:405-6. Mainprize TC, Drutz HP. The Marshall-Marchetti-Krantz procedure: a critical review. Obstet Gynecol Surv 1988;43:724-9. Merimsky E, Canetti R, Firstater M. Osteitis pubis: treatment by heparinisation. Br J Urol 1981;53:154-6. Nissenkorn I, Servadio C, Lubin E. The treatment of osteitis pubis with heparin. J Urol 1981;125:528-9. Burns JR, Gregory JG. Osteomyelitis of the pubic symphysis after urologic surgery. J Urol 1977;118:803-5. del Busto R, Quinn EL, Fisher EJ, Madhavan T. Osteomyelitis of the pubis: report of seven cases. JAMA 1982;248:1498-500. Lupovitch A, Elie JC, Wysocki R. Diagnosis of acute bacterial osteomyelitis of the pubis by means of fine needle aspiration. Acta Cytol 1989;33:649-51. Gilbert DN, Azorr M, Gore R, Hofeldt R. The bacterial causation of postoperative osteitis pubis. Surg Gynecol Obstet 1975;141:195-8. Hoyme UB, Tamimi HK, Eschenbach DA, Ramsey PG, Figge DC. Osteomyelitis pubis after radical gynecologic operation. Obstet Gynecol 1984;63 Suppl: 47S-53S. Jenkins FH, Raft MJ, Florman LD, Day TG, Templeton WC. Pubic osteomyelitis due to anaerobic Bacteria. Arch Intern Med 1984;144:842-3.