Brachytherapy

-

(2015)

-

Treatment results of image-guided high-dose-rate interstitial brachytherapy for pelvic recurrence of uterine cancer Ken Yoshida1,*, Hideya Yamazaki2, Tadayuki Kotsuma3, Tadashi Takenaka4, Koji Masui2, Yasuo Yoshioka5, Yasuo Uesugi1, Taiju Shimbo1, Nobuhiko Yoshikawa1, Hiroto Yoshioka1, Yoshifumi Narumi6, Keiji Tatsumi7, Eiichi Tanaka3 1

Department of Radiation Oncology, Osaka Medical College, Takatsuki, Osaka, Japan 2 Department of Radiology, Kyoto Prefectural University of Medicine, Kyoto, Japan 3 Department of Radiation Oncology, National Hospital Organization Osaka National Hospital, Osaka, Japan 4 Department of Radiology, National Hospital Organization Himeji Medical Center, Himeji, Japan 5 Department of Radiation Oncology, Osaka University Graduate School of Medicine, Osaka, Japan 6 Department of Radiology, Osaka Medical College, Takatsuki, Osaka, Japan 7 Department of Obstetrics and Gynecology, National Hospital Organization Osaka National Hospital, Osaka, Japan

ABSTRACT

PURPOSE: We analyzed clinical data to evaluate the effectiveness of image-guided high-doserate interstitial brachytherapy (HDR-ISBT) for pelvic recurrence of uterine cancer. METHODS AND MATERIALS: Between 2003 and 2011, 56 patients were treated with HDR-ISBT with or without external beam radiotherapy (EBRT). The median followup time was 33 months (range, 1e109 months). Pre-ISBT treatments included radical hysterectomy for 35 patients (Group A), radical hysterectomy with postoperative radiotherapy for 8 patients (Group B), and radical radiotherapy for 13 patients (Group C). We initiated MRI-assisted CT-based planning for the last 49 patients. The median ISBT single fraction dose was 6 Gy. The median total doses were 30 and 54 Gy with and without EBRT (range, 30e50 Gy) for Group A, respectively and 48 Gy without EBRT for Groups B and C. RESULTS: The 3-year local control (LC) rates were 85%, 75%, and 46% for Groups A, B, and C, respectively ( p 5 0.017). The 3-year LC rates were 84%, 73%, and 33% for clinical target volume at the time of HDR-ISBT of !10, 10e29, and $30 cc, respectively ( p 5 0.005). The 3-year LC results tended to be higher for patients whose D100 (clinical target volume) was equal or higher than 67.1 Gy ( p 5 0.098). A total of 13 late complications of Grades 3e5 occurred in 11 patients (20%). CONCLUSIONS: Our image-guided HDR-ISBT for pelvic recurrence of uterine cancer provided good treatment outcomes. The treatment results for patients who underwent radical surgery with or without postoperative radiotherapy are better than those for patients who underwent radical radiotherapy. Ó 2015 American Brachytherapy Society. Published by Elsevier Inc. All rights reserved.

Keywords:

Pelvic recurrence; Uterine cancer; Image guided; High-dose-rate interstitial brachytherapy; Magnetic resonance imaging

Introduction Radical treatment for pelvic recurrence of uterine cancer is difficult because many organs at risk (OARs), such as the rectum, sigmoid colon, small intestine, bladder, and Received 24 October 2014; received in revised form 12 February 2015; accepted 12 February 2015. Conflict of interest: None. * Corresponding author. Department of Radiation Oncology, Osaka Medical College, 2-7 Daigaku-machi, Takatsuki, Osaka 569-8686, Japan. Tel.: þ81-72-683-1221; fax: þ81-72-684-7219. E-mail address: [email protected] (K. Yoshida).

urethra, are near the tumor site. The indications for curative organ-sparing salvage surgery or pelvic exenteration are limited (1). External beam radiotherapy (EBRT) with or without chemotherapy is well tolerated; however, its treatment outcomes are not satisfactory (2). Interstitial brachytherapy (ISBT) features the potential to salvage such tumors while preserving the function of OARs. The American Brachytherapy Society consensus guidelines show that patients with recurrent cervical, endometrial, or vulvar carcinoma with residual vaginal lesions greater than 0.5-cm thickness are potential candidates for ISBT (3). Several clinical studies have reported good local

1538-4721/$ - see front matter Ó 2015 American Brachytherapy Society. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.brachy.2015.02.195

2

K. Yoshida et al. / Brachytherapy

control (LC) with low-dose-rate (LDR) or high-dose-rate (HDR) ISBT (4e19). Recurrent tumors are often large and with complex shapes that makes it difficult to achieve satisfactory implantation without image guidance. Therefore, image guidance and three-dimensional (3D) dose calculation have the potential to improve ISBT treatment results. The applicability of ultrasonography (4,5), CT (5e14), and MRI (7,13e15) has been investigated for applicator implantation or treatment planning. In the past, we have installed implants via transrectal ultrasonography (TRUS) guidance and CT treatment planning with MRI assistance (19) using flexible plastic applicators to allow the contours of the gross tumor volume (GTV), clinical target volume (CTV), and OARs to be easily drawn from the CT images without metal artifacts. As a result, no metallic treatment items were used in our patients, thus allowing them to undergo MRI examinations. We have already reported our preliminary experience with this method for previously untreated uterine cervical carcinoma (20). We report here the 3-year treatment results obtained with our imaging-assisted HDR-ISBT technique for pelvic recurrence of uterine cancer.

Material and Methods Between May 2003 and January 2011, 63 patients with pelvic recurrences of uterine cancer underwent HDRISBT at the Department of Radiation Oncology, National Hospital Organization, Osaka National Hospital, Osaka, Japan. About 7 patients were excluded from this study because of distant metastases or because they were lost to followup in less than 12 months after HDR-ISBT. The median followup times were 33 months (range, 1e109 months) for the remaining 56 patients (median age, 59 years; range, 27e82 years) and 58 months for the survivors (Table 1). The primary tumor site was the uterine cervix for 45 patients and the uterine corpus for 11. Histological findings revealed 30 squamous cell carcinomas, 2 adenosquamous carcinomas, 23 adenocarcinomas, and 1 endocrine tumor. The GTV was assessed on CT and MRI at the time of ISBT planning. The median maximum tumor diameter before treatment was 25 mm (range, 5e79 mm). The tumor morphological types were superficial and indurative for 9 and 47 patients, respectively. Among the latter, the maximum tumor diameter was less than 50 mm for 39 patients and 50e79 mm for 8 patients. In total, there were 55 N0 and 1 N1 patients. We divided the patients into three groups according to their previous treatment modalities: patients who underwent radical surgery without postoperative radiotherapy (Group A; n 5 35), those who underwent radical surgery with postoperative radiotherapy (Group B; n 5 8), and

-

(2015)

-

Table 1 Patient characteristics Characteristics

Values

Patient number Followup Age Histology Primary site N-stage Maximum tumor diameter CTV at the time of ISBT Previous treatment Recurrence interval Chemotherapy

56 (2003.5e2011.1) 1e109 mo (median, 33 mo) 27e82 y (median, 59 y) Sq:Ad:AdSq:Endocrine 5 30:23:2:1 Uterine cervix:Corpus 5 45:11 N0:N1 5 55:1 Superficial: !50 mm:50e79 mm 5 9:39:8 !10 cc:10e29 cc:$30 cc 5 31:19:6 Op:Op þ PostOp RT:Radical RT 5 35:8:13 #6 mo:7e24 mo:O24 mo 5 7:30:19 (þ):() 5 8:48

Sq 5 squamous cell carcinoma; Ad 5 adenocarcinoma; AdSq 5 adenosquamous cell carcinoma; CTV 5 clinical target volume; Op 5 operation; RT 5 radiotherapy.

those who underwent radical radiotherapy (Group C; n 5 13). The median time interval between the previous radical treatment and our treatment was 15 months; the intervals were 6 months or less for 7 patients, 7e24 months for 30 patients, and more than 24 months for 19 patients. About 20 patients in Group A underwent EBRT to the whole pelvis with a median prescribed dose of 30 Gy (range, 30e50 Gy). In addition, 17 patients underwent center-shielded EBRT (median, 20 Gy; range, 10e20 Gy). One patient underwent an additional boost of irradiation for pelvic lymph node metastases (10 Gy). The median overall treatment time (duration of whole EBRT and ISBT) was 46 days (range, 35e56 days). We performed ISBT after whole pelvic EBRT was essentially completed. About 8 patients (14%) received concurrent chemotherapy before or during EBRT. About 7 of them received intravenous cisplatin, and 1 patient received intravenous carboplatin. Applicator implantation Implantation was performed under lumbar anesthesia. We also initiated epidural anesthesia and continued this until the applicator was extracted. We performed a singleapplicator implantation with multifractionated HDR-ISBT for all patients. Implantation was monitored with the aid of TRUS (SSD-1000 and Prosound a7; Hitachi Aloka Medical, Ltd., Tokyo, Japan), and flexible needles (ProGuide Sharp Needle; Nucletron an Elekta Company, Veenendaal, The Netherlands) were used for all patients. Before implantation, we implanted three to four titanium markers at the distal, lateral edge of the GTV. We used a templateguided and nonambulatory implant technique for the first 5 patients and a freehand and ambulatory implant technique for the remaining 51 patients. The ambulatory applicator implantation technique has been described elsewhere (21). First, a single flexible needle applicator was inserted through the center of the vaginal stump or uterine cervix. A button stopper was then affixed to the needle and placed

K. Yoshida et al. / Brachytherapy

in contact with the stump or external os. After needle implantation, a silicon cylinder was inserted into the vagina. This cylinder was custom-made from silicon rubber, and the size was dependent on the vaginal size. The first needle and cylinder complex was sutured to the vaginal stump or uterine cervix with silk thread. After inserting the complex, we attached a custom-made vinyl template containing holes for flexible needle applicators to the perineum. The positions of the implantation holes were determined using preimplantation TRUS, CT, and MRI images. We implanted the other applicators from the vinyl template under TRUS guidance. The objective of the implantation was to adequately cover CTV. We implanted 7e15 (median, 12) applicators. After completing implantation, we fixed all needles to the vinyl template and the perineum using silk threads. The protruded connector end of the applicator was cut short enough to enable the patient to walk. Treatment planning and treatment All patients underwent CT and CT-based planning. The OARs (rectum, bladder, and urethra) were delineated using the CT images.

-

(2015)

-

3

About 49 of the 56 patients also underwent MRI examinations to provide a reference for the GTV contours. GTV was delineated using CT images with assistance from axial T2-weighted MRI (Figs. 1a and 1b). Treatment planning was performed using PLATO (software version 14.2; Nucletron an Elekta Company) and conducted via geometrical optimization with manual modification [Fig. 1c; (22)]. For Group A, we selected ISBT combined with EBRT for the patients who showed local recurrence within 2 years after previous surgery or showed nodal recurrence. We selected ISBT as a monotherapy for patients exhibiting local recurrence for more than 2 years after the previous treatment. The median single-fraction planning aim dose (PAD) of HDR-ISBT was 6 Gy (range, 3e7 Gy), and the median total PADs were 54 Gy in nine fractions (range, 48e54 Gy) for ISBT as a monotherapy and 30 Gy in five fractions (range, 27e36 Gy) for ISBT combined with EBRT. In Groups B and C, we omitted EBRT because of the previous irradiation histories and also reduced the total PADs. The median total PADs were 48 Gy in eight fractions (range, 42e51 Gy). We used the microSelectron-HDR (Nucletron an Elekta Company) for treatment and 192Ir as the treatment source.

Fig. 1. (a) A magnetic resonance image of the postoperative local recurrent lesion of a uterine cervical cancer patient was obtained after implantation. The patient has a large local recurrent lesion in the right parametrium (white arrow). (b) A computed tomography image of the same patient was obtained after implantation. This image shows the contours of clinical target volume (black line), which was delineated with the assistance of magnetic resonance imaging. (c) The dose distribution curve was calculated via computer optimization with manual modification. The clinical target volume (black line) was well covered by the 100% prescribed isodose line (white dotted line).

4

K. Yoshida et al. / Brachytherapy

It was necessary to prepare an additional tube when the transfer tube was attached to the applicator. Accordingly, we prepared a catheter with an outer diameter of 4 or 4.2F to serve as a connector between the applicator and transfer tube. One end of this catheter was inserted into the transfer tube and the other into the applicator. We initiated treatment after completing these procedures. Doseevolume histogram analysis For Groups A and B, doseevolume histogram (DVH) was calculated for the CTV (CTV is equal to GTV). The planning target volume (PTV) margin was 10e15 mm in the cranial direction because we considered applicator displacement to represent the caudal direction. For Group C, we used another CTV definition because uterine was preserved. We used the high-risk CTV (HR-CTV) as the CTV for DVH analysis because it was often difficult to judge the difference between tumor lesion inside the uterine cervix and normal cervical tissue. The HR-CTV was the whole cervix and the presumed extracervical tumor extension at the time of ISBT (23,24). We calculated the percentage of CTV covered by the prescribed dose (V100) and the doses that covered 90% and 100% of the CTV (D90 and D100). For the OARs, we calculated the minimum doses received by the maximally irradiated 0.1- and 2-cc volumes (D0.1 cc and D2 cc). We used D2 cc for the bladder and the rectum and D0.1 cc for the urethra. We used biologically equivalent doses to calculate the EBRT and ISBT doses. The biologically equivalent dose was calculated as equivalent 2-Gy fractions (EQD2) using a linear quadratic model, where a/b 5 10 for tumors and 3 for OARs. Because we did not have registration software for each plan, center-shielded doses were not added for the DVH. Toxicity was assessed according to the Common Terminology Criteria for Adverse Events version 3.0. Statistical analysis Statistical analyses were performed using the StatView v5.0 software program (SAS Institute, Cary, NC). The LC and overall survival (OS) rates were analyzed. Survival curves were estimated according to the KaplaneMeier method and examined for significance using the log-rank test. A p-value lower than 0.05 was considered statistically significant.

Results DVH for CTV and OARs The median CTV at the time of HDR-ISBT was 9.3 cc (range, 0.1e197 cc). The mean D90 (CTV) was 120.8  17.9% PADs. The mean D100 (CTV) was 95.6  20.3% PAD.

-

(2015)

-

The EQD2 of the median D90 (CTV) was 85.7 Gy (range, 52.7e198 Gy), and the median D100 (CTV) was 67.1 Gy (range, 15.1e154.5 Gy). The median D2 cc (bladder) per fraction was 80% PAD (range, 18.3e148.3% PAD). The EQD2 of the median D2 cc (bladder) for all treatments (ISBT with or without EBRT) was 65.6 Gy (range, 14e190.6 Gy). The median D2 cc (rectum) per fraction was 68.3% PAD (range, 26.7e91.7% PAD). The EQD2 of the median D2cc (rectum) for all treatments was 55.8 Gy (range, 11.8e90.4 Gy). The median D0.1 cc (urethra) per fraction was 51.7% PAD (range, 6.7e131.7% PAD). The EQD2 of the median D0.1 cc (urethra) for all treatments was 40 Gy (range, 2.2e154.2 Gy). LC and OS results About 21 patients died of carcinoma, and 9 cancerbearing patients remained alive at the time of writing this manuscript. The 3-year LC and OS rates were 75% and 68%, respectively, for all patients. Of the 30 patients showing disease progression, 16 patients presented with local failure, 5 with nodal failure, and 23 with distant metastases. Around 11 of the 16 local failure patients had central recurrences, and the other 5 had marginal recurrences. Not all local failure patients could be salvaged. Among the five marginal recurrence cases, two recurrent tumors were located very near the treated areas. One tumor was located relatively far from the treated area, and therefore, we did not consider the possibility of occult tumor invasion. One tumor was located in an area that was not treated with HDR-ISBT but instead treated with the initial radical radiotherapy. The fifth tumor was located in the perineal skin, and we suspected that tumor dissemination occurred at the time of applicator implantation. The treatment results for each treatment group were also analyzed. The 3-year LC rates were 85%, 75%, and 46% for Groups A, B, and C, respectively ( p 5 0.01; Fig. 2a). Local central recurrence was observed in 3, 2, and 6 patients belonging to Groups A, B, and C, respectively. Local marginal recurrence was observed in 2, 2, and 1 patients belonging to Groups A, B, and C, respectively. The 3year OS rates were 72%, 88%, and 42% for Groups A, B, and C, respectively; however, these differences were not significant ( p 5 0.15; Fig. 2b). The 3-year LC rates for superficial-type, indurative-type with !50 mm, and indurative-type with $50 mm were 100%, 74%, and 50%, respectively ( p 5 0.078), and the 3-year OS were 100%, 66%, and 38%, respectively ( p 5 0.023). The 3-year LC rates were 84%, 73%, and 33% for CTV at the time of HDR-ISBT of !10, 10e29, and $30 cc, respectively ( p 5 0.005). The 3-year OS rates were 79%, 59%, and 33% for CTV at the time of HDR-ISBT of !10, 10e29, and $30 cc, respectively ( p 5 0.039).

K. Yoshida et al. / Brachytherapy

-

(2015)

-

5

Fig. 2. (a) The local control rate for each treatment group is shown. Group A: Patients who underwent radical surgery without postoperative radiotherapy; Group B: Patients who underwent radical surgery with postoperative radiotherapy; Group C: Patients who underwent radical radiotherapy. (b) The overall survival rate for each treatment group is shown.

The interval between recurrence and initial treatment was not a significant factor. The 3-year LC rates were 57%, 72%, and 84% for intervals of #6, 7e24, and $25, respectively ( p 5 0.51). The 3-year OS rates were 43%, 66%, and 79% for intervals of #6, 7e24, and $25, respectively ( p 5 0.29). Similarly, age and histology were not significant factors. The 3-year LC results were 70% and 79% for a D90 (CTV) of #85.7 and O85.7 Gy, respectively ( p 5 0.35). The 3-year OS results were 63% and 71% for a D90 (HRCTV) of #85.7 and O85.7 Gy, respectively ( p 5 0.98). The D90 (CTV) was 81.3  8.1 Gy for patients with local central recurrence and 90.3  23.6 Gy for all other patients ( p 5 0.14). The 3-year LC results were 65% and 85% for a D100 (CTV) of #67.1 and O67.1 Gy, respectively ( p 5 0.098). The 3-year OS results were 67% and 69% for a D100 (HR CTV) of #67.1 and O67.1 Gy, respectively ( p 5 0.52). The D100 (CTV) was 58.1  15 Gy for patients with central recurrence and 70.3  20.4 Gy for all other patients ( p 5 0.07). Complications A total of 13 Grades 3e5 adverse events were observed in 11 patients (20%; Table 2). Among 11 patients, 4 (20%) underwent ISBT with EBRT, and 7 (19%) underwent ISBT alone. Three of the 13 adverse events were small bowel obstructions caused by EBRT. The other 10 Grades 3e5 adverse events in 8 patients were thought to have been caused by ISBT with or without EBRT. Grade 5 complications were observed in 2 patients (4%). One patient had an ileus 8 months after HDR-ISBT and she repeated admission because of feeding disorder. Finally, she died 70 months after HDR-ISBT. Another patient had a recto-vesico-vaginal fistula 3 months after HDR-ISBT and received ileotransversostomy two times. However, she suffered from pain, infection, and feeding disorder after that. She also had repeated admission and finally died 32 months after HDR-ISBT.

Grade 4 complications were observed in 2 patients (4%). One patient had a recto-vesico-vaginal fistula 6 months after HDR-ISBT and she received colostomy and nephrostomy. She was free from disease 40 months after HDRISBT at present. Another patient had a vesicovaginal fistula 11 months after HDR-ISBT and received ureterostomy. Local recurrence was observed 55 months after HDR-ISBT and she was alive with the disease 68 months after HDRISBT at present. Grade 3 complications were observed in 7 patients (13%). Three patients had a rectovaginal fistula and were judged as an indication of colostomy. Two of them received colostomy and the other 1 patient refused it and was followed without any treatment. Ileus, vaginal ulcer without fistula, and rectal bleeding were observed in 2, 1, and 1 patients, respectively. The D2 cc (rectum) for 5 patients with rectovaginal fistulae and the other 51 patients was 63.7  17.2 and 52.8  17 Gy, respectively ( p 5 0.13). The D2 cc (bladder) for 3 patients with vesicovaginal fistulae and the other 53 patients was 88.9  70.1 Gy and 69  33.7 Gy, respectively ( p 5 0.94). Among 8 patients, one (11%) had a superficial-type tumor and 7 (15%) had indurative-type tumors. Of these 8 patients, 5 (14%) belonged to Group A, 2 (25%) belonged to Group B, and 1 (8%) belonged to Group C.

Table 2 Severe late complication results Grade

No. of occurrences (%), n 5 56

5

2 (4) Small bowel obstruction Rectovaginal fistula and vesicovaginal fistula 2 (4) Rectovaginal fistula and vesicovaginal fistula Vesicovaginal fistula 7 (13) Rectovaginal fistula Small bowel obstruction Rectal bleeding Vaginal ulcer

4

3

1 1 1 1 3 2 1 1

6

Table 3 Clinical results of interstitial brachytherapy for pelvic recurrence of uterine cancer Recurrent

Previously non-irradiated

ISBT þ EBRT

Causespecific survival rate (%)

Study

Dose Imaging modality ISBT rate Primary (n) Reirradiation (n) during implantation ISBTalone (n) dose (Gy)

Overall EBRT Local control survival dose (Gy) rate (%) rate (%)

2D dose calculation Corn et al. (7)

LDR

25e33.6 40e50

45e50.4

60

40

40

0

34 44 18 12 11 2 0 22

15e55 25e75 Mean 25.5

20e60 d Mean 48

70

56

62

10

77

NA

65

17

18e35 40e50 12e45

45e50.4 d d

100

NA

77

15

NA

About 35 (2 y) NA (We calculate this number from figure)

18

32 2 20 0 6 0 9

30 NA Total 69e87 70e80 10e35

46 d 40e45 d 20e60

74 (CR) 55

63

NA

29

40

40

18

83

83

83

17

NA

NA

NA

10

47

38

43

9

70

NA

7

86 (Primary) NA 17 (Recurrent) About 50 (3 y) NA (We calculate this number from figure) 68 NA

28

Charra et al. (17)

5 0

NA

32 2 20 0 5 1 9

Nag et al. (12) Badakh et al. (16)

3D dose calculation Jensen et al. (10)

Beriwal et al. (6) Thibault et al. (13) Mabuchi et al. (11)

Present study

HDR 0 PDR 23 HDR 0 HDR 17 13 HDR 9 34 HDR 52 0

1 12 11 25 5 NA

HDR 56 0

35 21

0 52

CT Laparoscope CT MRI CT MRI US CT CT CT MRI CT

US

1 8 15 28 2 43 0 0 52 20 36

HDR: Median 24 30.6e 50.4 LDR: Median 35 14e21 43.2e50 28e42 d 18.75e21.3 24e50.4 18e35

40e50.4

42

d

27e36 42e54

30e50 d

-

Weitmann et al. (5)

XP Laparoscope XP CT NA

(2015)

PDR 22 12 Eisbruch et al. (8) LDR 14 6 Popowski et al. (15) PDR 6 0 Viswanathan et al. (14) LDR 10

Tewari et al. (18)

-

49 29 17 13 13 0 22 0

3 2

K. Yoshida et al. / Brachytherapy

LDR 78 0 LDR 30 0 LDR 13 0 HDR 22 0

Laparoscope CT MRI XP

Severe complication (%)

76 (Primary) 85 (Recurrent) 87 (Primary) 45 (Recurrent) 77 (CR) 75

25

20

ISBT 5 interstitial brachytherapy; EBRT 5 external beam radiotherapy; NA 5 not available; LDR 5 low dose rate; HDR 5 high dose rate; PDR 5 pulse dose rate; CT 5 computed tomography; MRI 5 magnetic resonance imaging; XP 5 X-ray film; US 5 ultrasonography; CR 5 complete response.

K. Yoshida et al. / Brachytherapy

Ten Grade 2 adverse events were observed in 7 patients (13%). About 2 of the 7 patients also had Grade 3 adverse events as described previously. Genitourinary complications were observed in 4 patients (urinary pain, 1; hematuria, 1; urethral stenosis, 2). Gastrointestinal complications (rectal bleeding) were observed in 4 patients. Vaginal ulcer and sciatic neuralgia were observed in 1 patient each. Discussion The LC and OS rates after ISBT ranged from 45% to 100% and 17% to 83%, respectively (Table 3). Although the LC after 3D dose calculation with an image-guided system (60%e100%) was theoretically superior to 2D dose calculation (45e85%), a wide variety of outcomes were observed, and image-guided ISBT did not provide a better LC result mainly because of the heterogeneous patient population, which featured selection bias. For example, Nag et al. (12) reported an LC rate of 100% in 13 favorableoutcome patients without 3D dose calculation. The 13 patients did not have previous irradiation, the tumor sizes were #2 and 2e4 cm for 7 and 3 patients, respectively, and the interval from hysterectomy to recurrence was more than 12 months for 9 patients. Because the tumor size (5,11,17) and recurrence interval (5,10,11) were important prognostic factors, the patient selection was biased toward a better outcome. In contrast, Weitmann et al. (5) reported an LC rate of 47% for 23 patients, although 3D dose calculation had been performed. The authors of that study treated many higher risk patients, 11 of whom had previous irradiation. Additionally, 8 had larger tumor sizes (35%) and 12 (52%) had shorter recurrence intervals. One important point related to image-guided radiotherapy is that this technique provides an opportunity to investigate the precise relationship between the irradiated dose and tumor response. In clinical situations, we could not unite the prescribed dose and PAD under a single value. Weitmann et al. (5) reported that the D100 (CTV) ranged from 54% to 100% of the PAD (5). This finding suggests that we should investigate the relationship between the dose and tumor response not according to the PAD but to the actual prescribed dose. The authors of that study reported that a prescribed total dose of higher than 64 Gy with a coverage index higher than 0.8 yielded a better treatment outcome. In the present study, the D100 (CTV) showed a slight tendency to affect the LC and central recurrence rates. To investigate DVH analyses, precise contour delineation of the GTV via imaging modalities is necessary. In this study, we used both CT and MRI. MRI could judge the contour of the tumor lesion from the neighboring normal tissues such as the colon and ureter more precisely than CT. Metal markers are also useful because superficial tumors are difficult to identify with CT and MRI. However, we experienced five local marginal recurrences despite using these imaging modalities. In such cases, we should make a greater effort to precisely delineate the

-

(2015)

-

7

GTV. In the present study period, we could not perform MRI fusion to CT because we did not have the registration software. We have recently begun to use such software and thus have solved this problem. Other imaging modalities such as positron emission tomography may be useful (25). We should also consider the CTV/PTV margin. Although it is difficult to prepare a larger margin because the neighboring normal tissue tolerance is decreased by previous radical treatments, we should investigate the adequate distance for a CTV margin. The PTV margin is also difficult because many reports have shown caudal needle applicator displacement (26e32). We also investigated and consequently reported a median caudal needle applicator displacement of 5 mm (range, 2 to 26 mm) at 93 h after implantation for recurrent gynecological tumors (26). A O10-mm displacement was observed for 26 of 122 needles (21%) at 21 h and 19 of 60 needles (32%) at 93 h. From these reports, we fixed a cranial margin of 10 mm and are presently implementing daily CT to initiate corrective action. In the present study, we experienced 11 local central recurrences. As mentioned previously, recurrent gynecological tumors include a wide variety of prognostic factors. Therefore, it is necessary to determine an adequate prescribed dose depending on these prognostic factors, especially for recurrent gynecological cancers. In this study, the previous treatment modality, tumor size, and tumor volume were significant prognostic factors. In particular, the 3-year LC rate was 46% for Group C, and 6 of 7 local recurrent patients belonging to Group C showed central recurrence. The 3-year LC rate was 75% for Group B, and 2 of 4 local recurrent patients belonging to Group B showed central recurrence. Because previous irradiation doses for Group B were lower than those for Group C, we concluded that a better LC rate was observed for Group B. The 3-year LC rate for indurative-type of $50 mm was 50%, and the 3-year LC rate was 33% for a CTV at the time of HDR-ISBT of $30 cc. We must consider CTV dose escalation in such patients with poor prognosis. However, ISBT dose escalation must be considered carefully, as we have already experienced a Grade $3 complication rate of 20%. Our dose-fractionation schedule was almost at the upper limit because EQD2 (54 Gy in nine fractions) was 97 Gy (a/b 5 3 for OARs). We have used this schedule as per the protocol of Osaka Medical Center for Cancer and Cardiovascular Diseases since 2000 (33). Our preliminary results (since 2003) showed a lower late complication rate than those obtained in this study [7% Grade 3 late complication rate; (19)]. However, the present study (longer followup and more patient numbers) showed a higher late complication rate. Therefore, we might investigate other additional treatment modalities such as chemotherapy, radiosensitization, and hypoxia sensitization (34e36). The American Brachytherapy Society consensus guidelines have introduced a Phase II clinical trial conducted by the Gynecological Oncology Group. This trial compares concurrent weekly cisplatin with EBRT and brachytherapy

8

K. Yoshida et al. / Brachytherapy

for patients with vaginal cuff recurrences (34). Yamazaki et al. (35) reported a feasibility study of a hypoxia sensitizer (Sanazole) for advanced cancers, including gynecological cancer. Ogawa et al. (36) reported the efficacy of hypoxia sensitization using hydrogen peroxide for locally advanced tumors, although gynecological tumors were not included. Future investigations are required to show a more precise relationship between the dose and tumor response as well as the individualized treatment options. This study has limitations. The patient numbers were relatively small, and there was a heterogeneous patient population with varied treatment modalities. However, DVH can unite quantitative values from different treatment schedules. Although we could only find a trend that D100 (CTV) may correlate with LC and central recurrence rates in the present study, we will analyze the prognostic factors and the best treatment schedules for patients who had different tumor factors and treatment histories using DVH values in the future. In conclusion, our image-guided HDR-ISBT for pelvic recurrence of uterine cancer provided good treatment results, especially for patients with a favorable prognoses (i.e., previously non-irradiated and small tumor volume). The treatment results for patients who previously received postoperative radiotherapy are also good compared with those for patients who received radical radiotherapy.

Acknowledgments This work was supported by JSPS Grant-in-Aid for Scientific Research (KAKENHI) (C) Grant Number 25461931, Ministry of Health, Labor and Welfare Grant-in-Aid for Scientific Research (KAKENHI) Grant Number 26270801, and Japan mHDR Research Fund. The authors would like to thank Hisakazu Okada, DDS; Mineo Yoshida, MD; Hironori Akiyama, DDS; Tadashi Takenaka, RTT; Shunsuke Miyake, RTT; Mari Mikami Ueda, RTT; Toshiro Kajihara, RTT; Kazumasa Aramoto, RTT; Masaji Fukumoto, RTT; Yoshiyuki Konishi, RTT; Tadayoshi Nagano, MD; Chiaki Ban, MD; Shigetoshi Yamada, MD; Shoji Kamiura, MD; Kenichiro Morishige, MD; Toshiki Okada, MD; Takayuki Nose, MD; Sachiko Yamada; Kanako Hamasaki; and the staff of the Departments of Radiation Oncology, Radiology, Obstetrics and Gynecology, Anesthesiology; and Operating Room for helping us in many ways during the completion of this article. References [1] Miller B, Morris M, Rutledge F, et al. Aborted exenterative procedures in recurrent cervical cancer. Gynecol Oncol 1993;50:94e99. [2] Maneo A, Landoni F, Cormio G, et al. Concurrent carboplatin/5fluorouracil and radiotherapy for recurrent cervical carcinoma. Ann Oncol 1999;10:803e807.

-

(2015)

-

[3] Beriwal S, Demanes DJ, Erickson B, et al. American Brachytherapy Society consensus guidelines for interstitial brachytherapy for vaginal cancer. Brachytherapy 2012;11:68e75. [4] Stock RG, Chan K, Terk M, et al. A new technique for performing Syed-Neblett template interstitial implants for gynecologic malignancies using transrectal-ultrasound guidance. Int J Radiat Oncol Biol Phys 1997;37:819e825. [5] Weitmann HD, Knocke TH, Waldh€ausl C, et al. Ultrasound-guided interstitial brachytherapy in the treatment of advanced vaginal recurrences from cervical and endometrial carcinoma. Strahlenther Onkol 2006;18:86e95. [6] Beriwal S, Rwigema JC, Higgins E, et al. Three-dimensional imagebased high-dose-rate interstitial brachytherapy for vaginal cancer. Brachytherapy 2012;11:176e180. [7] Corn BW, Lanciano RM, Rosenblum N, et al. Improved treatment planning for the Syed-Neblett template using endorectal-coil magnetic resonance and intraoperative (laparotomy/laparoscopy) guidance: A new integrated technique for hysterectomized women with vaginal tumors. Gynecol Oncol 1995;56:255e261. [8] Eisbruch A, Johnston CM, Martel MK, et al. Customized gynecologic interstitial implants: CT-based planning, dose evaluation, and optimization aided by laparotomy. Int J Radiat Oncol Biol Phys 1998;40:1087e1093. [9] Erickson B, Albano K, Gillin M. CT-guided interstitial implantation of gynecologic malignancies. Int J Radiat Oncol Biol Phys 1996;36:699e709. [10] Jensen PT, Roed H, Engelholm SA, Rosendal F. Pulsed dose rate (PDR) brachytherapy as salvage treatment of locally advanced or recurrent gynecologic cancer. Int J Radiat Oncol Biol Phys 1998; 42:1041e1047. [11] Mabuchi S, Takahashi R, Isohashi F, et al. Reirradiation using highdose-rate interstitial brachytherapy for locally recurrent cervical cancer. Int J Gynecol Cancer 2014;24:141e148. [12] Nag S, Yacoub S, Copeland LJ, Fowler JM. Interstitial brachytherapy for salvage treatment of vaginal recurrences in previously unirradiated endometrial cancer patients. Int J Radiat Oncol Biol Phys 2002;54:1153e1159. [13] Thibault I, Lavallee MC, Aubin S, et al. Inverse-planned gynecologic high-dose-rate interstitial brachytherapy: Clinical outcomes and dose-volume histogram analysis. Brachytherapy 2012;11:181e191. [14] Viswanathan AN, Cormack R, Holloway CL, et al. Magnetic resonance-guided interstitial therapy for vaginal recurrence of endometrial cancer. Int J Radiat Oncol Biol Phys 2006;66:91e99. [15] Popowski Y, Hiltbrand E, Joliat D, et al. Open magnetic resonance imaging using titanium-zirconium needles: Improved accuracy for interstitial brachytherapy implants? Int J Radiat Oncol Biol Phys 2000;47:759e765. [16] Badakh DK, Grover AH. Reirradiation with high-dose-rate remote afterloading brachytherapy implant in patients with locally recurrent or residual cervical carcinoma. J Cancer Res Ther 2009;5:24e30. [17] Charra C, Roy P, Coquard R, et al. Outcome of treatment of upper third vaginal recurrences of cervical and endometrial carcinomas with interstitial brachytherapy. Int J Radiat Oncol Biol Phys 1998; 40:421e426. [18] Tewari K, Cappuccini F, Brewster WR, et al. Interstitial brachytherapy for vaginal recurrences of endometrial carcinoma. Gynecol Oncol 1999;74:416e422. [19] Kotsuma T, Yoshida K, Yamazaki H, et al. Preliminary results of magnetic resonance imaging-aided high-dose-rate interstitial brachytherapy for recurrent uterine carcinoma after curative surgery. J Radiat Res 2011;52:329e334. [20] Yoshida K, Yamazaki H, Takenaka T, et al. A dose-volume analysis of magnetic resonance imaging-aided high-dose-rate image-based interstitial brachytherapy for uterine cervical cancer. Int J Radiat Oncol Biol Phys 2010;77:765e772. [21] Yoshida K, Nose T, Shiomi H, et al. New ambulatory implant technique of high-dose-rate interstitial brachytherapy for prostate cancer. Radiat Med 2006;24:595e599.

K. Yoshida et al. / Brachytherapy [22] Yoshida K, Nose T, Koizumi M, et al. The usefulness of metal markers for CTV-based dose prescription in high-dose-rate interstitial brachytherapy. J Jpn Soc Ther Radiol Oncol 2002;13:253e260. [23] Haie-Meder C, P€ otter R, Van Limbergen E, et al. Recommendations from Gynaecological (GYN) GEC-ESTRO Working Group (I): Concepts and terms in 3D image based 3D treatment planning in cervix cancer brachytherapy with emphasis on MRI assessment of GTV and CTV. Radiother Oncol 2005;74:235e245. [24] P€ otter R, Haie-Meder C, Van Limbergen E, et al. Recommendations from gynaecological (GYN) GEC ESTRO working group (II): Concepts and terms in 3D image-based treatment planning in cervix cancer brachytherapy-3D dose volume parameters and aspects of 3D image-based anatomy, radiation physics, radiobiology. Radiother Oncol 2006;78:67e77. [25] Vees H, Casanova N, Zilli T, et al. Impact of 18F-FDG PET/CT on target volume delineation in recurrent or residual gynaecologic carcinoma. Radiat Oncol 2012;7:176. [26] Yoshida K, Ueda M, Takenaka T, et al. Daily CT measurement of needle applicator displacement during multifractionated high-doserate interstitial brachytherapy for postoperative recurrent uterine cancer. J Radiat Res 2012;53:295e300. [27] Demato AL, Cormack RA, Viswanathan AN. Characterization of implant displacement and deformation in gynecologic interstitial brachytherapy. Brachytherapy 2014;13:100e109. [28] Mikami M, Yoshida K, Takenaka T, et al. Daily computed tomography measurement of needle applicator displacement during highdose-rate interstitial brachytherapy for previously untreated uterine cervical cancer. Brachytherapy 2011;10:318e324. [29] Onoe T, Nose T, Yamashita H, et al. High-dose-rate interstitial brachytherapy for gynecologic malignanciesdDosimetric changes during treatment period. J Radiat Res 2013;54:663e670.

-

(2015)

-

9

[30] Pinnaduwage DS, Cunha JA, Weinberg V, et al. A dosimetric evaluation of using a single treatment plan for multiple treatment fractions within a given applicator insertion in gynecologic brachytherapy. Brachytherapy 2013;12:487e494. [31] Rey F, Chang C, Mesina C, et al. Dosimetric impact of interfraction catheter movement and organ motion on MRI/CT guided HDR interstitial brachytherapy for gynecologic cancer. Radiother Oncol 2013; 107:112e116. [32] Shukla P, Chopra S, Engineer R, et al. Quality assurance of multifractionated pelvic interstitial brachytherapy for postoperative recurrences of cervical cancers: A prospective study. Int J Radiat Oncol Biol Phys 2012;82:617e622. [33] Nose T, Koizumi M, Yoshida K, et al. In vivo dosimetry of high-doserate interstitial brachytherapy in the pelvic region: Use of a radiophotoluminescence glass dosimeter for measurement of 1004 points in 66 patients with pelvic malignancy. Int J Radiat Oncol Biol Phys 2008; 70:626e633. [34] Gynecologic Oncology Group (GOG). GOG 238 a randomized trial of pelvic irradiation with or without concurrent weekly cisplatin in patients with pelvic-only recurrence of carcinoma of the uterine corpus. Available at: http://www.gog.org. Accessed July 2, 2011. [35] Yamazaki H, Nakamura S, Kobayashi K, et al. Feasibility trial for daily oral administrationof the hypoxic sensitizer AK-2123 (Sanazole) in radiotherapy. Anticancer Res 2013;33:643e646. [36] Ogawa Y, Kubota K, Ue H, et al. Phase I study of a new radiosensitizer containing hydrogen peroxide and sodium hyaluronate for topical tumor injection: A new enzyme-targeting radiosensitization treatment, Kochi Oxydol-Radiation Therapy for Unresectable Carcinomas, Type II (KORTUC II). Int J Oncol 2011;39: 553e560.