Uterine Perforation During 3-Dimensional Image-GuidedBrachytherapy in Patients With Cervical Cancer

Baskent University Experience

Cem Onal, MD,* Ozan Cem Guler, MD,* Yemliha Dolek, MSci,* and Gurcan Erbay, MDÞ

Objectives: This study aimed to determine the incidence and characteristics of uterineperforation at our department using 3-dimensional computed tomography (CT)Ybasedbrachytherapy (BRT). The characteristics of the patients presenting with perforation andimpact of the perforation on the treatment course were also analyzed.Materials and Methods: The clinical and radiologic data of 200 patients with biopsy-proven cervical cancer treated using CT-based BRT were retrospectively evaluated. Allpatients had undergone tandem-based intracavitary BRT, and 67 patients had undergonemagnetic resonance imaging (MRI) before BRT.Results: Of the 200 patients, 17 (8.5%) had uterine perforation. Of the 626 applicationswith CT images, 30 (4.8%) resulted in uterine perforation. The median age of patients withperforation was higher (68 years; range, 44Y89 years) than that of the patients withoutperforation (59 years; range, 21Y87 years), and the mean (SD) tumor size at diagnosis waslarger in patients with perforation (7.0 [1.5] cm) than in patients without perforation (5.0[1.5] cm). The most frequent perforation site was the posterior uterine wall (8 patients),followed by the fundus (5 patients) and anterior wall (4 patients). Of the 7 patients with aretroverted uterus, 4 had uterine perforation during BRT. In 67 patients with MRI deliveredbefore BRT, only 3 (4%) had uterine perforation, and 2 of the 3 patients with uterineperforation had a retroverted uterus. However, of the 133 patients with no MRI evaluationbefore BRT, 14 (11%) had uterine perforation. No life-threatening complications or signs ofintraperitoneal tumor cell seeding were observed.Conclusions: Older age, larger tumors, a retroverted uterus, and a stenotic cervical os wereall predisposing factors for uterine perforation during BRT. Preoperative MRI is a feasibleand safe method to decrease the risks of uterine perforation and could be used preoperativelyin centers where intraoperative ultrasonography is not used in routine practice.

Key Words: Cervical cancer, Brachytherapy, Uterine perforation, Magnetic resonance imaging

Received September 25, 2013, and in revised form October 25, 2013.Accepted for publication October 27, 2013.

(Int J Gynecol Cancer 2014;24: 346Y351)

The definitive treatment of choice for locally advanced orinoperable cervical cancer is external beam radiother-

apy (ERT), concomitant chemotherapy, and brachytherapy(BRT).1,2 Brachytherapy, an essential component in thetreatment of cervical carcinoma, has stimulated enthusiasmfor high-doseYrate (HDR) delivery in recent years, offeringoutpatient treatment, easy radiation protection, low cost, andimproved tumor dose distribution. In intracavitary (IC) BRTof cervical cancer, an applicator (a uterine tandem and vaginal

ORIGINAL STUDY

346 International Journal of Gynecological Cancer & Volume 24, Number 2, February 2014

Departments of *Radiation Oncology, Baskent University Faculty ofMedicine, Adana; and †Radiology, Baskent University Faculty ofMedicine, Ankara, Turkey.Address correspondence and reprint requests to Cem Onal, MD,

Department of Radiation Oncology, Adana Research andTreatment Centre, Baskent University Faculty of Medicine,01120 Adana, Turkey. E-mail: hcemonal@hotmail.com.

The authors declare no conflict of interest.Copyright * 2014 by IGCS and ESGOISSN: 1048-891XDOI: 10.1097/IGC.0000000000000048

Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.

ovoid or a ring-tandem applicator) is placed in the uterinecavity and vaginal fornices under sedation or anesthesia.

Intraoperative complications of IC BRT include vaginallacerations and penetration of the tandem into the uterine wallcausing perforation of the uterus and of other pelvic organswith the applicator. The incidence of uterine perforationduring BRT application ranges from 1.75% to 13.7%,3Y5 andthe most common sites of perforation are the uterine fundusand posterior wall of the fundus.3,5,6 The predisposing factorsfor uterine perforation are advanced age (more than 60 years),anatomic distortions of the cervix, advanced diseases causingcervical stenosis, postirradiation fibrosis, and previous conebiopsy.3,7 Other predisposing factors include extremely ananteverted or a retroverted uterus.4,6

Although most cases of uterine perforation resolvespontaneously without sequelae after conservative treatment,infection, hemorrhage, or peritoneal tumor seeding may oc-cur.5 Another important potential consequence of uterineperforation is the uncertainty in the prescribed dose distri-bution around the incorrectly inserted applicator, which maycause underdosage of the target volume, compromising thelocal control probability.8 Therefore, accurate positioning ofthe applicator is critical for delivering appropriate doses ofirradiation to the target volume while keeping the doses to thesurrounding organs at risk below their tolerance limits. Toachieve appropriate application, intraoperative ultrasound(IUS) facilitates ideal tandem placement and decreases therisk of uterine perforation, thereby diminishing an underap-preciated source of toxicity while optimizing diseasecontrol.9Y11 However, a recent survey of BRT specialists in-dicates that only 56% have used IUS at some point, and thosephysicians using IUS rarely use it for every insertion (median,42% of insertions).12

The main objective of the present study was to deter-mine the incidence and characteristics of uterine perforation atour department in the era of 3-dimensional (3D) computedtomography (CT)Ybased BRT. We evaluated the characteris-tics of the patients presenting with perforation and reviewedtheir management and the impact of the perforation on thetreatment course.

MATERIALS AND METHODS

PatientsThe clinical and radiologic data of 238 patients with

biopsy-proven cervical cancer were retrospectively evaluated.All patients had undergone tandem-based IC BRT as acomponent of definitive treatment between February 2008and June 2013. Thirty-eight patients were excluded becauseof 2-dimensionalYbased IC. The characteristics of the 200patients who had undergone 3D-based BRT are summarizedin Table 1.

Brachytherapy ApplicationPatients without distant metastasis were treated with a

combination of 3D conformal ERT with concurrent weeklydose of 40 mg/m2 cisplatin and HDR BRT, as previouslydescribed.13 A total of 50.4 Gy ERT (1.8 Gy per fraction,

daily,Monday throughFriday)was delivered using 18-megavoltsphotons. External beam radiotherapy was planned with a 4-field box technique using a treatment planning system(Eclipse; Varian Medical Systems, Palo Alto, CA). Three-dimensional BRT planning was performed using 7 Gy perfraction prescribed to the target minimum that was adminis-tered in 4 fractions. Brachytherapy was performed using aremote afterloading HDR unit with a radioactive iridium-192source (VariSource; Varian Medical Systems). The IC BRTprocedure was initiated at the end of ERT.

From 2007 to mid-2008, conventional treatment plan-ning was performed via fluoroscopy using orthogonal images.Since mid-2008, routine CT imaging has been performed withall HDR BRT procedures for treatment planning purposes. Inaddition, after 2009, magnetic resonance imaging (MRI) wasadministered after completion of ERT and before BRT ap-plication to assess treatment response and facilitate the BRTapplication.

Computed tomographyYcompatible Fletcher-Suit ap-plicators or ring-tandem applicators were used during IC BRTand consisted of uterine tandems with various angles (15, 30,and 45 degrees) and various lengths (2, 4, and 6 cm). Inaddition, a pair of ovoids with various diameters (20, 25, and30 mm) in Fletcher-Suit applicators or a ring-tandem appli-cator was used. A thorough gynecologic examination wascarried out by an experienced radiation oncologist (C.O.)before each insertion. The application technique and appli-cators were individually adapted to each patient at eachinsertion based on the clinical and radiologic findings.

TABLE 1. Patient characteristics

Characteristics No. Patients %

Median age, y 60 (21Y89)Median tumor size, cm 5.2 (2Y13)FIGO stage

IB2 21 11IIA 14 7IIB 110 55IIIA 18 9IIIB 29 14IVA 8 4

Histologic diagnosisSquamous cell carcinoma 178 89Adenocarcinoma 22 11

Applicator typeFletcher Suit 93 47Ring tandem 107 53

Uterus positionAnteverted 193 96Retroverted 7 4

MRI evaluationPresent 67 34Absent 133 66

International Journal of Gynecological Cancer & Volume 24, Number 2, February 2014 Uterine Perforation in Cervical Cancer

* 2014 IGCS and ESGO 347

Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.

In patients who had undergone an MRI before BRT, the ap-plication and treatment planning were performed according tothe treatment response.

The insertion was carried out with sedation and anal-gesia and was placed in the lithotomy position. Patients werenot given specific instructions for rectal preparation, but theywere encouraged to empty their bowels before a simulationprocedure and before the subsequent IC BRT procedure.Appropriate anterior and posterior vaginal packing was usedto fix the applicator position and to displace the bladder andrectum away from the vaginal applicators. After the IC ap-plication, the applicator was fixed with a universal applicatorclamping device (Varian), which was positioned underneaththe patient.

After the insertion, the patientswere transported to theCTscanner. The paradigm of our BRT dose prescription andtreatment planning practice is represented by the image-guidedadaptive BRT of cervix cancer described elsewhere.13,14

Toxicity was graded according to the Radiation Ther-apy Oncology Group/European Organization for Researchand Treatment of Cancer common toxicity criteria. Latetoxicity was recorded retrospectively by a thorough review ofthe patient’s hospital charts.

RESULTSThe patient characteristics of 200 patients who

underwent image-guided IC BRT are summarized in Table 1.Of the 200 patients, 17 (8.5%) had uterine perforation. A totalof 626 applications had CT images for BRT planning aftertandem placement. Of the 626 applications with CT images,30 (4.8%) resulted in uterine perforation.

The median age of the entire group was 60 years (range,21Y89 years). Themedian age of patients with perforation washigher (68 years; range, 44Y89 years) than those patientswithout perforation (59 years; range, 21Y87 years). The mean(SD) tumor size at diagnosis was larger in patients withperforation (7.0 [1.5] cm) than in patients without perforation(5.0 [1.5] cm; Fig. 1).

The most frequent perforation site was the posterioruterine wall (8 patients), followed by the fundus (5 patients)and anterior wall (4 patients). In all these patients, anatomicdistortion resulted in uncertain tandem positioning during theprocedure with resultant suspected perforation. Eight patientswere perforated only once, 8 patients were perforated twice,and 2 patients were perforated 3 times. Of the 7 patients with aretroverted uterus, 4 (57%) had uterine perforation during IC,whereas only 10 (5%) of the 193 patients with an anteverteduterus had uterine perforation.

In 67 patients with MRI delivered before BRT, only 3(4%) had uterine perforation; 2 of these 3 patients with uterineperforation had a retroverted uterus (Fig. 2). However, of the133 patients with no MRI evaluation before BRT, 14 (11%)had uterine perforation.

All perforations were detected by CT after applicatorplacement. After each perforation, the applicator was re-moved, and prophylactic antibiotics were administered inpatients at high risk or with necrotic lesions. No significant

FIGURE 1. Tumor size according to uterine perforation.

FIGURE 2. A, T1-weighted sagittal images of a representative case taken after external radiotherapy. No grossresidual tumor was observed in the retroverted uterus. B, The uterus was perforated at the fundus after tandeminsertion. C, The applicator was removed and reinserted appropriately with no perforation.

Onal et al International Journal of Gynecological Cancer & Volume 24, Number 2, February 2014

348 * 2014 IGCS and ESGO

Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.

clinical sequelae resulted. In all patients, the second attempt toplace the IC BRT device was successfully confirmed on CTafter applicator placement, and no additional perforationswere noted (Fig. 3). No operative consultations with otherservices were required secondary to difficulty with deviceplacement. In 2 patients with a retroverted uterus, uterineperforation occurred 3 times, and the application wascorrected thereafter. These 2 patients had stage IIIB diseaseand residual disease completely obstructing the cervical os.

According to the Radiation Therapy Oncology Group/European Organization for Research and Treatment of Cancerlate toxicity criteria, 2 patients with uterine perforation de-veloped grade 2 chronic rectal toxicity, including diarrhea,tenesmus, nausea, and vomiting, whereas only 1 patient de-veloped grade 2 chronic bladder and urethra toxicity (dysuria,polyuria, and incontinence). All of these patients responded toconservative management. Four patients developed grade 3bladder toxicity with severe hematuria and frequency, and 2patients developed grade 3 and 4 gastrointestinal systemtoxicity (1 patient with bowel obstruction and 1 patient withfistula formation). Among patients who developed fistulaformation, only 1 patient had retroverted uterus. The inci-dence of vaginal lacerations was not systematically recorded.At the time of the last follow-up, there were no signs of in-traperitoneal tumor cell seeding in any of the patients in whomuterine perforation occurred.

DISCUSSIONIn the current study, we reviewed the data concerning

BRT applications for cervical cancer and analyzed uterineperforation during IC tandem applications. We found uterineperforation occurred more frequently in patients with ad-vanced age, larger tumors, a distorted or stenotic cervical os,and a retroverted uterus; results that were similar to literaturefindings. In addition, MRI delivered before BRT for as-sessing the ERT response, resulted in a lower rate of uterineperforation.

A modern approach in treatment planning for cervicalcarcinoma is based on CT sections and 3D dose distribution.Advantages of 3D imaging in gynecologic BRT that may leadto improved patient outcome, irrespective of the dose rate,include avoidance or early detection of a uterine perforation,ensuring target coverage, and avoidance of an excessive doseto the surrounding organs. Thus, better assessment of dosedistributions in the target volumes and surrounding normaltissues is achieved. In 2004, guidelines were published forproposing image-based BRT for cervical cancer, and CT-guided BRT planning is frequently used.15

Large retrospective series without either routine IUS orCT indicate that the incidence of perforation ranges from 1%to 3%.3,8,16,17 However, in the absence of routine postoper-ative CT imaging, these series likely underestimated the in-cidence of perforation. Importantly, the 2 series reported thatthe incidence of unrecognized perforations ranged from 10%to 13.7% of procedures as assessed by routine postoperativeCT in the absence of IUS.4,10 Davidson et al10 reported a 10%incidence of uterine perforation in 21 women with 35 in-sertions. Barnes et al4 detected uterine perforation in 13.7% ofthe cases with 124 sequential tandem insertions in 114 pa-tients. Our perforation rates were less than those in previouslyreported series. The reason for the lower rates of uterineperforation in our series is that we treated the patient using 4fractions of BRT, making the applications after the first in-sertion easier. Second, nearly one third of our patients hadpreoperative MRI.

There is evidence in the literature to indicate that the useof IUS decreases the uterine perforation rate.5,7,11 Granaiet al7 reported about routine IUS for 72 patients and notedno clinically evident perforations. Schaner et al11 reporteda 1.4% incidence of uterine perforation in 356 IUS-guidedapplicator placements. Segedin et al5 found uterine perfora-tions in 13 (3.0%) of the 428 applications in 10 (4.6%) ofthe 219 patients. All these studies demonstrated the feasi-bility of IUS-guided BRT, and the authors suggested that theuse of IUS diminishes the risk of perforation 5- to 10-fold.

FIGURE 3. A, T1-weighted sagittal images of a representative case with a large cervical tumor completely obstructingthe cervical os and causing vaginal infiltration (arrows). B, A small residual tumor (arrow) at the cervical os aftercompletion of external radiotherapy before BRT application. C, Computed tomographic images taken after insertionof a tandem, demonstrating appropriate application without uterine perforation.

International Journal of Gynecological Cancer & Volume 24, Number 2, February 2014 Uterine Perforation in Cervical Cancer

* 2014 IGCS and ESGO 349

Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.

However, the use of ultrasonography (USG) requires someexperience, and not all departments have the capability toperform USG-guided BRT, particularly in departments with ahigh patient overload. In our clinic, after 2009, we began touse MRI for evaluating ERT response, which is a significantprognostic factor, and for delineating the residual tumors toachieve better dose coverage. In addition, MRI clearly dem-onstrated uterine position, whether retroverted or highlyanteverted. In a retroverted uterus, we tried to insert the ap-plicator according to the position of the uterus. Unfortunately,in 2 patients with a retroverted uterus, although they hadpreoperative MRI, uterine perforation was observed in 3applications. Thus, more intense care should be given in suchcases. In patients with a highly anteverted uterus, we preferredto inflate the bladder with saline for better positioning of theuterus. Besides these 2 patients with a retroverted uterus, only1 (1.5%) patient with preoperative MRI had uterine perfo-ration; a finding that is similar to IUS-guided BRTapplicationseries. However, in patients with uterine perforation, therewere not more adverse effects because these patients weretreated with corrected tandem insertion resulting in appro-priate dose distribution.

In the treatment of cervical cancer, the importance ofoptimal applicator placement in local control and toxicity iswell established.18,19 However, perforation remains a frequentcomplication, even in cases that appear straightforward.Classic standards for assessing the adequacy of applicatorplacement are insufficiently sensitive to consistently detectperforation. New modalities, including laparoscopy, USG,and CT, are effective in detecting uterine perforation; how-ever, only intraoperative USG provides cost-effective, real-time guidance and confirmation of tandem placement.However, the limitation of USG is its low-resolution capacityin detecting solid tumors compared with MRI and CT. Inaddition, USG has been widely used in complicated cases, butthe role of routine USG is less well established.

The present study possesses limitations. The retro-spective nature of our study is the largest limitation. Second,particularly in cases with difficult application, such as aretroverted uterus, IUS could be much more feasible, andappropriate tandem insertion could be performed withoutcausing uterine perforation.20 Finally, we only reported ourexperience with cervical BRT; however, the clinical outcomesof uterine perforation regarding disease control and survivalshould also be performed in a large prospective series.

CONCLUSIONSIn the present study, we demonstrated that uterine

perforation is a complication that could be detected duringuterine tandem insertion. Older age, larger tumors, aretroverted uterus, and a stenotic cervical os were allpredisposing factors for uterine perforation during BRT. Ourstudy is also important in demonstrating that preoperativeMRI, which evaluates the treatment response of ERT anddemonstrates the uterus position, decreases the risk of uterineperforation during tandem insertion. Preoperative MRI is afeasible and safe method and could be used preoperatively atcenters where IUS is not used in routine practice.

REFERENCES1. Atahan IL, Onal C, Ozyar E, et al. Long-term outcome and

prognostic factors in patients with cervical carcinoma: aretrospective study. Int J Gynecol Cancer. 2007;17:833Y842.

2. Eifel PJ, Moughan J, Erickson B, et al. Patterns of radiotherapypractice for patients with carcinoma of the uterine cervix: apatterns of care study. Int J Radiat Oncol Biol Phys. 2004;60:1144Y1153.

3. Jhingran A, Eifel PJ. Perioperative and postoperativecomplications of intracavitary radiation for FIGO stage I-IIIcarcinoma of the cervix. Int J Radiat Oncol Biol Phys. 2000;46:1177Y1183.

4. Barnes EA, Thomas G, Ackerman I, et al. Prospectivecomparison of clinical and computed tomography assessment indetecting uterine perforation with intracavitary brachytherapyfor carcinoma of the cervix. Int J Gynecol Cancer. 2007;17:821Y826.

5. Segedin B, Gugic J, Petric P. Uterine perforationV5-yearexperience in 3-D image guided gynaecological brachytherapyat Institute of Oncology Ljubljana. Radiol Oncol. 2013;47:154Y160.

6. Irvin W, Rice L, Taylor P, et al. Uterine perforation at the time ofbrachytherapy for carcinoma of the cervix. Gynecol Oncol.2003;90:113Y122.

7. Granai CO, Doherty F, Allee P, et al. Ultrasound for diagnosingand preventing malplacement of intrauterine tandems. ObstetGynecol. 1990;75:110Y113.

8. Corn BW, Hanlon AL, Pajak TF, et al. Technically accurateintracavitary insertions improve pelvic control and survivalamong patients with locally advanced carcinoma of the uterinecervix. Gynecol Oncol. 1994;53:294Y300.

9. Small W Jr, Strauss JB, Hwang CS, et al. Should uterine tandemapplicators ever be placed without ultrasound guidance? No:a brief report and review of the literature. Int J Gynecol Cancer.2011;21:941Y944.

10. Davidson MT, Yuen J, D’Souza DP, et al. Optimization ofhigh-dose-rate cervix brachytherapy applicator placement: thebenefits of intraoperative ultrasound guidance. Brachytherapy.2008;7:248Y253.

11. Schaner PE, Caudell JJ, De Los Santos JF, et al. Intraoperativeultrasound guidance during intracavitary brachytherapyapplicator placement in cervical cancer: the University ofAlabama at Birmingham experience. Int J Gynecol Cancer.2013;23:559Y566.

12. Viswanathan AN, Erickson BA. Three-dimensional imagingin gynecologic brachytherapy: a survey of the AmericanBrachytherapy Society. Int J Radiat Oncol Biol Phys. 2010;76:104Y109.

13. Onal C, Arslan G, Topkan E, et al. Comparison of conventionaland CT-based planning for intracavitary brachytherapy forcervical cancer: target volume coverage and organs at risk doses.J Exp Clin Cancer Res. 2009;28:95.

14. Oymak E, Onal C, Arslan G. Gynecological tumorbrachytherapy: Baskent University Experience. Turk OnkolojiDergisi. 2012;27:12Y17.

15. Nag S, Cardenes H, Chang S, et al. Proposed guidelines forimage-based intracavitary brachytherapy for cervicalcarcinoma: report from Image-Guided Brachytherapy WorkingGroup. Int J Radiat Oncol Biol Phys. 2004;60:1160Y1172.

16. Lanciano R, Corn B, Martin E, et al. Perioperative morbidityof intracavitary gynecologic brachytherapy. Int J Radiat OncolBiol Phys. 1994;29:969Y974.

Onal et al International Journal of Gynecological Cancer & Volume 24, Number 2, February 2014

350 * 2014 IGCS and ESGO

Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.

17. Petereit DG, Sarkaria JN, Chappell RJ. Perioperative morbidityand mortality of high-dose-rate gynecologic brachytherapy.Int J Radiat Oncol Biol Phys. 1998;42:1025Y1031.

18. Cunningham DE, Stryker JA, Velkley DE, et al. Routine clinicalestimation of rectal, rectosigmoidal, and bladder doses fromintracavitary brachytherapy in the treatment of carcinomaof the cervix. Int J Radiat Oncol Biol Phys. 1981;7:653Y660.

19. Katz A, Eifel PJ. Quantification of intracavitary brachytherapyparameters and correlation with outcome in patients withcarcinoma of the cervix. Int J Radiat Oncol Biol Phys. 2000;48:1417Y1425.

20. Mayr NA, Montebello JF, Sorosky JI, et al. Brachytherapymanagement of the retroverted uterus using ultrasound-guidedimplant applicator placement. Brachytherapy.2005;4:24Y29.

International Journal of Gynecological Cancer & Volume 24, Number 2, February 2014 Uterine Perforation in Cervical Cancer

* 2014 IGCS and ESGO 351

Copyright © 2014 by IGCS and ESGO. Unauthorized reproduction of this article is prohibited.