Department of Obstetrics and Gynecology, Miller School of Medicine, University of Miami, Miami, FL, USA

Ultrasound Obstet Gynecol 2007; 29: 65 69 ublished online 14 December 2006 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/uog.3890 Comparison of CT- or ultrasound-guided with concomitant intravenous vs. intravenous alone in the management of tubo-ovarian abscesses N. GOHARKHAY, U. VERMA and F. MAGGIOROTTO Department of Obstetrics and Gynecology, Miller School of Medicine, University of Miami, Miami, FL, USA KEYWORDS: computed tomography; ; pelvic abscess; tubo-ovarian abscess; ultrasound-guided ABSTRACT Objective The purpose of this study was to compare the outcome of treatment of tubo-ovarian abscesses by imaging-guided and vs. intravenous alone. Methods A retrospective chart review of all patients hospitalized with a diagnosis of tubo-ovarian abscess was performed. atients were categorized into two groups. The first group consisted of subjects treated with intravenous alone. atients in the second group had primary image-guided with concomitant intravenous. Treatment failures in the primary group underwent salvage when feasible. The primary outcome of interest was complete response. Secondary outcomes included need for additional treatment, duration of resolution of fever, total length of hospital stay, and complication rates. We also evaluated the effectiveness of secondary in patients who failed primary antibiotic therapy alone. Results A total of 58 patients were included in the study. Fifty patients were treated primarily with intravenous ; eight patients had primary, which was guided by ultrasound in all cases. Complete response was noted in 29 (58%) patients treated with alone. All eight (100%) patients in the primary group responded to treatment. Of the 21 treatment failures with primary, two underwent surgery and 19 (90.5%) had salvage with either ultrasound or computed tomographic guidance; 18 of 19 salvage s led to complete recovery. Subjects in the primary group required shorter hospital stays and showed more rapid resolution of fever. No significant morbidity was noted as a consequence of procedures. A higher failure rate for secondary was noted in older patients, those with larger tubo-ovarian abscesses, and those with a history of pelvic inflammatory disease. Conclusion Drainage of tubo-ovarian abscesses with concomitant intravenous is an effective and safe treatment for the primary or secondary treatment of tubo-ovarian abscesses. Copyright 2006 ISUOG. ublished by John Wiley & Sons, Ltd. INTRODUCTION Tubo-ovarian abscesses (TOAs) are classically treated with intravenous followed by oral over a prolonged period 1. The response rate to antimicrobial treatment alone is about 70%. Failure of treatment is characterized by increase in the size of the TOA, persistently elevated temperature or signs of peritonitis 2.There is a high recurrence rate for TOAs after treatment with alone, requiring repeat hospital admissions in many cases. Surgical intervention becomes necessary in about 25% of all patients 2. Options for surgical procedures include laparoscopy or laparotomy with of abscess, unilateral or bilateral adnexectomy, or total hysterectomy with bilateral salpingo-ophorectomy as definitive treatment. The presence of infection and adhesions as well as friable inflammatory tissue makes the surgical approach difficult. An alternative method for the treatment of TOAs is under direct guidance by computed tomography Correspondence to: Assistant rofessor U. Verma, Department of Obstetrics and Gynecology (D-50), Miller School of Medicine, University of Miami,.O. Box 016960, Miami, FL 33101, USA (e-mail: uverma@med.miami.edu) Accepted: 25 September 2006 Copyright 2006 ISUOG. ublished by John Wiley & Sons, Ltd. ORIGINAL AER

66 Goharkhay et al. (CT) or ultrasound, with or without antibiotic injection into the abscess cavity. This study was performed to compare the outcome of conservative management of TOAs using intravenous alone vs. under CT or ultrasound guidance with concomitant use of intravenous. MATERIALS AND METHODS We performed a retrospective analysis of all patients admitted to Jackson Memorial Hospital, University of Miami Medical Center, between April 1 1999 and September 30 2001 with a diagnosis of tubo-ovarian abscess. The cases were identified through a search of our computerized database, based on diagnostic coding. Clinical information was obtained from the electronic clinical database as well as through review of paper charts on all patients. Additional data were obtained by review of computerized reports available through the Division of Ultrasound, Department of Obstetrics and Gynecology. The primary outcome was the rate of complete response to the primary therapy without need for other interventions (such as surgery or secondary in the primary antibiotic group). Our secondary outcomes were the need for additional procedures, duration of resolution of fever, total length of hospital stay, and procedurerelated complications, such as bladder or bowel injury and sepsis. We further evaluated the effectiveness of secondary in patients who failed primary antibiotic therapy alone. The antibiotic regimen in all patients consisted of intravenous gentamicin and clindamycin as well as ampicillin (if not penicillin-allergic). A patient was assigned to the primary group when she underwent the procedure within 24 h of admission and no indication for such a procedure was evident from her medical records. The decision whether primary was performed was based on the attending physician s clinical judgment. The regimen for was the same for either primary or secondary (salvage). Direct guidance by either ultrasound or CT was utilized. In general, the ultrasound-guided procedures were performed by an obstetrician-gynecologist, whereas CT-guided s were done by an interventional radiologist. Both procedures aimed at draining all accessible cystic areas through a needle by the transvaginal or transabdominal approach. When using ultrasound guidance, a 900 mg dose of clindamycin suspended in 30 ml of saline was injected into the abscess at the end of the procedure. Statistical analyses were performed using SSS software for Windows (SSS Inc., Chicago, IL, USA). Statistical significance was determined by using an alpha level of 0.05 and two-sided tests. The Kolmogorov Smirnov procedure was used to evaluate the normal distribution of the variables. Categorical variables were compared with χ 2 or Fisher s exact tests; ANOVA was used to compare continuous variables if normally distributed. In case of non-normally distributed variables, Mann Whitney U and Wilcoxon tests were applied. RESULTS We identified a total of 58 subjects admitted to our institution with the diagnosis of tubo-ovarian abscess. The demographic data on all subjects together with clinical and initial diagnostic findings are shown in Table 1. Of the patients in this study, 50 subjects (86.2%) were treated primarily with and eight (13.8%) underwent primary. All primary s were performed under ultrasound guidance. Figure 1 shows ultrasound images of a TOA taken before and after imaging-guided. There were no significant differences in demographic features, clinical findings, or imaging characteristics of TOAs between the two study groups (Table 1). All patients with an intrauterine device (IUD) underwent primary antibiotic treatment, which likely reflects a bias on the side of the providers. The average total volume of the TOAs was calculated by the ellipsoid volume formula (volume = height width length 0.52). Of the 50 patients who received primary antibiotic therapy, 29 (58%) had a complete response and required no further treatment. All eight patients who underwent imaging-guided with concomitant experienced complete resolution of symptoms without Table 1 Demographics of patients and tubo-ovarian abscess (TOA) characteristics by study group All patients (n (%)) 58 (100) 50 (86.2) 8 (13.8) Age (years, mean (range)) 31.8 (16 61) 31.7 (16 61) 32.5 (18 44) 0.834 arity (median (range)) 1 (0 5) 1 (0 5) 0.5 (0 4) 0.782 History of ID (n (%)) 7 (12.1) 6 (12.0) 1 (12.6) 0.968 IUD present (n (%)) 16 (27.6) 16 (32.0) 0 (0.0) 0.098 HIV-positive (n (%)) 5 (8.6) 5 (10.0) 0 (0.0) 0.298 Unilateral TOA (n (%)) 37 (63.8) 33 (66.0) 4 (50.0) 0.443 Bilateral TOA (n (%)) 21 (36.2) 17 (34.0) 4 (50.0) 0.443 volume (ml, mean (range)) 145.7 (8.0 442.4) 143.8 (3 167) 157.1 (21 160) 0.732 IUD, intrauterine device; ID, pelvic inflammatory disease.

Management of tubo-ovarian abscesses 67 Table 2 Treatment outcome by study group All patients (n) 50 8 Treatment failure 21 (36.2) 21 (42.0) 0 (0.0) 0.019 (n (%)) Surgery needed (n (%)) 3 (5.2) 3 (6.0) 0 (0.0) 0.882 Hospital stay (days, median (range)) 7 (4 16) 7 (4 16) 4.5 (4 8) < 0.001 58 subjects 50 primary 8 primary 29 resolved 2 TAH+ BSO 19 salvage All resolved 18 resolved 1 salpingectomy Figure 2 Flow chart of total patient population. BSO, bilateral salpingo-ophorectomy; TAH, total abdominal hysterectomy. Figure 1 Ultrasound images of a tubo-ovarian abscess before (a) and after (b) abscess. The complex has an approximate diameter of 11.47 cm prior to. It measures 4.47 3.18 cm after the procedure; 250 ml of pus were drained. any further interventions (Table 2). No major side effects were observed in any of the cases treated with abscess. The duration of hospital stay was significantly shorter for the primary group as compared to the primary group (Table 2). The 21 treatment failures in the primary group received further therapy as follows. Two subjects showed persistent severe symptoms requiring total abdominal hysterectomy with bilateral salpingoophorectomy. Nineteen patients (90.5% of treatment failures in group) were treated by salvage under imaging guidance (11 ultrasound-guided, eight CT-guided). In 18 of 19 patients (94.7%), salvage led to complete clinical response. One patient (5.3%) remained febrile and required a unilateral salpingo-ophorectomy, after which she achieved remission. Figure 2 represents a flow chart of all patients analyzed in our study. Several characteristics were significantly associated with treatment failure in our study population (Table 3). Nonresponders were on average older (36.4 ± 9.4 vs.28.0 ± Table 3 Characteristics of patients with failure to primary Antibiotic non-responders Antibiotic responders All patients on (n) 50 22 28 Age (mean (range)) 31.7 (16 61) 36.4 (20 61) 28.0 (16 45) 0.002 History of ID (n (%)) 6 (12.0) 5 (22.7) 1 (3.6) 0.075 IUD present (n (%)) 16 (32.0) 8 (36.4) 8 (28.6) 0.761 HIV-positive (n (%)) 5 (10.0) 4 (18.2) 1 (3.6) 0.174 Bilateral TOA (n (%)) 17 (34.0) 6 (27.3) 11 (39.3) 0.548 volume (ml, mean (range)) 143.8 (8 442.4) 200.3 (13.4 442.4) 99.5 (8.0 229.9) < 0.001 IUD, intrauterine device; ID, pelvic inflammatory disease; TOA, tubo-ovarian abscess.

68 Goharkhay et al. Table 4 Duration until fever resolution by study group atients with fever (n) 41 34 7 Fever lasting < 24 h(n (%)) 6 (14.6) 2 (5.9) 4 (57.1) 24 47.9 h (n (%)) 10 (24.4) 9 (26.5) 1 (14.3) 48 71.9 h (n (%)) 21 (51.2) 19 (55.9) 2 (28.6) > 72 h(n (%)) 4 (9.8) 4 (11.8) 0 (0.0) 9.4 years (mean ± SEM), < 0.002). Furthermore, failure to respond to antibiotic therapy corresponded with total TOA volume (99 ± 58 vs. 200 ± 123 ml (mean ± SEM), < 0.001). In contrast, overall failure rates were similar for unilateral vs. bilateral lesions (40.5% vs. 33.9%, < 0.81). Women with a history of pelvic inflammatory disease were more likely to fail primary antibiotic therapy (3.6% of responders vs. 22.7% of non-responders), although this relationship was not statistically significant. At the time of presentation, 34 subjects (68.0%) in the primary antibiotic group and seven (87.5%) in the primary group were febrile ( < 0.001). We analyzed the duration of febrile illness in these patients by study group (Table 4). Overall, fever resolution was achieved faster in the primary group ( = 0.016). DISCUSSION One-third of women hospitalized with pelvic inflammatory disease have tubo-ovarian abscesses 3.Therearean estimated 100 000 annual hospitalizations for TOA in the USA 4. The classical treatment of TOA in the past was hysterectomy and bilateral salpingo-ophorectomy, which, although it results in complete cure, has significant morbidity, leads to infertility, and can cause early menopause. In order to prevent such adverse effects in these mostly young patients, a more conservative approach is desirable. Success rates of 67 75% have been reported with prolonged use of intravenous alone 5,6. In patients with large adnexal masses, the success with intravenous antibiotic is low. It has been reported that patients with abscesses 10 cm required surgery more often (60%), while of those who had abscesses smaller than 5 cm, only 20% required surgery 7.8. More recently, minimally invasive approaches to abscess, such as laparoscopic and percutaneous methods, have become available 9. Drainage of abscesses with concomitant use of is another alternative. Using either ultrasound or CT guidance, several smaller and a large study have been reported, including a small randomized study 10 16. Success rates for primary have been in the range of 86 100%, with the largest study by Gjelland et al. reporting 93.4% complete response 14. These reported outcomes are in accordance with the findings of the current investigation. Our study also demonstrates an excellent outcome for imaging-guided in cases of failed primary antibiotic treatment, with a success rate of 94.7%. Aboulgahr et al. reported improvement of symptoms in all patients after aspiration of abscesses and patients complete pain relief within 3 days of aspiration, which is similar to our experience 13. Gjelland et al. and Aboulgahr et al. reported faster fever resolution with abscess, which we also found in this study 13,14. In our population, more than 70% of febrile patients undergoing primary were afebrile by 48 h. We report a trend towards a shorter hospital stay in the group, in accordance with previous data 15,16. Larger TOAs failed medical management in our study to a larger proportion. An estimated total volume of 150 ml may be a relative indication for vs. expectant management if imaging-guided is not routinely available. We administered a dose of clindamycin into the abscess at the time of ultrasound-guided. Aboulgahr and his group have used cefotaxime in a similar setting 13. Injection of antibiotic into the abscess cavity after aspiration is simple and can be performed before removing the needle. This may improve treatment response and result in faster control of the infection. We did not find any significant morbidity associated with procedures, even in febrile patients with severe symptoms and potential for the presence of bacteremia. All evidence indicates that in experienced hands needle-guided is a safe procedure with minimal risk of complications. Drainage of abscesses is a highly successful technique, with minimal procedure-related complications despite the presence of infection 17,18. It may lead to significant cost savings, as suggested by our finding of reduced hospital stay. This mode of treatment is currently under-utilized in clinical practice. Our case series is the largest reported from the USA and confirms the results of the recent Norwegian study 14 as well as the relatively small randomized prospective study by erez-medina et al. 15. Limitations of the current study are the retrospective nature of data collection as well as the smaller sample size in the primary group. A larger prospective randomized trial to precisely quantify the benefit of primary and salvage of TOAs is encouraged. Based on the available data, primary of TOAs appears safe and effective as a first-line treatment and should be recommended wherever available. Furthermore, imaging-guided as secondary therapy is useful in most patients as an alternative to invasive surgical procedures. ACKNOWLEDGMENTS This study was supported by the Department of Obstetrics and Gynecology, University of Miami School of Medicine.

Management of tubo-ovarian abscesses 69 REFERENCES 1. McNeeley SG, Hendrix SL, Mazzoni MM, Kmak DC, Ransom SB. Medically sound, cost-effective treatment for pelvic inflammatory disease and tuboovarian abscess. Am J Obstet Gynecol 1998; 178: 1272 1278. 2. Mirhashemi R, Schoell WM, Estape R, Angioli R, Averette HE. Trends in the management of pelvic abscesses. J Am Coll Surg 1999; 188: 567 572. 3. Willson JR, Black JR III. Ovarian abscess. Am J Obstet Gynecol 1964; 90: 34 43. 4. Landers DV, Sweet RL. Current trends in the diagnosis and treatment of tuboovarian abscess. Am J Obstet Gynecol 1985; 151: 1098 1110. 5. Landers DV, Sweet RL. Tubo-ovarian abscess: contemporary approach to management. Rev Infect Dis 1983; 5: 876 884. 6. Sweet RL, Schachter J, Landers DV, Ohm-Smith M, Robbie MO. Treatment of hospitalized patients with acute pelvic inflammatory disease: comparison of cefotetan plus doxycycline and cefoxitin plus doxycycline. Am J Obstet Gynecol 1988; 158: 736 741. 7. Reed SD, Landers DV, Sweet RL. Antibiotic treatment of tuboovarian abscess: comparison of broad-spectrum betalactam agents versus clindamycin-containing regimens. Am J Obstet Gynecol 1991; 164: 1556 1561. 8. Ginsburg DS, Stern JL, Hamod KA, Genadry R, Spence MR. Tubo-ovarian abscess: a retrospective review. Am J Obstet Gynecol 1980; 138: 1055 1058. 9. Wiesenfeld HC, Sweet RL. rogress in the management of tuboovarian abscesses. Clin Obstet Gynecol 1993; 36: 433 444. 10. van Sonnenberg E, D Agostino HB, Casola G, Goodacre BW, Sanchez RB, Taylor B. US-guided transvaginal of pelvic abscesses and fluid collections. Radiology 1991; 181: 53 56. 11. icker RH, McLennan AC, Robertson RD, orter RN. Conservative management of pelvic abscess in recurrent pelvic inflammatory disease. Ultrasound Obstet Gynecol 1991; 1: 60 62. 12. Teisala K, Heinonen K, unnonen R. Transvaginal ultrasound in the diagnosis and treatment of tubo-ovarian abscess. Br J Obstet Gynaecol 1990; 97: 178 180. 13. Aboulghar MA, Mansour RT, Serour GI. Ultrasonographically guided transvaginal aspiration of tuboovarian abscesses and pyosalpinges: an optional treatment for acute pelvic inflammatory disease. Am J Obstet Gynecol 1995; 172: 1501 1503. 14. Gjelland K, Ekerhovd E, Granberg S. Transvaginal ultrasoundguided aspiration for treatment of tubo-ovarian abscess: a study of 302 cases. Am J Obstet Gynecol 2005; 193: 1323 1330. 15. erez-medina T, Huertas MA, Bajo JM. Early ultrasoundguided transvaginal of tubo-ovarian abscesses: a randomized study. Ultrasound Obstet Gynecol 1996; 7: 435 438. 16. Caspi B, Zalel Y, Or Y, Bar DY, Appelman Z, Katz Z. Sonographically guided aspiration: an alternative therapy for tuboovarian abscess. Ultrasound Obstet Gynecol 1996; 7: 439 442. 17. Fabiszewski NL, Sumkin JH, Johns CM. Contemporary radiologic percutaneous abscess in the pelvis. Clin Obstet Gynecol 1993; 36: 445 456. 18. Corsi J, Johnson SC, Gonik B, Hendrix SL, McNeeley SG Jr, Diamond M. Transvaginal ultrasound-guided aspiration of pelvic abscesses. Infect Dis Obstet Gynecol 1999; 7: 216 221.