- Email: [email protected]
Pelvic fractures after definitive and postoperative radiotherapy for cervical cancer: A retrospective analysis of risk factors
YGYNO-976904; No. of pages: 4; 4C: Gynecologic Oncology xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno
Pelvic fractures after definitive and postoperative radiotherapy for cervical cancer: A retrospective analysis of risk factors Kasumi Yamamoto, Shoji Nagao ⁎, Kazuhiro Suzuki, Ai Kogiku, Tokihiro Senda, Hiroko Yano, Miho Kitai, Takaya Shiozaki, Kazuko Matsuoka, Satoshi Yamaguchi Department of Gynecologic Oncology, Hyogo Cancer Center, 13-70 Kitaoji-cho, Akashi-city, Hyogo 673-8558, Japan
H I G H L I G H T S • • • •
The incidence of pelvic fracture was 15.8% in women with cervical cancer received radiotherapy. Pelvic fracture was diagnosed after a median of 14 months from the completion of radiotherapy. The predisposing factors were postmenopausal status, rheumatoid arthritis, and brachytherapy. Active interventions are necessary for women with high-risk factors.
a r t i c l e
i n f o
Article history: Received 26 July 2017 Received in revised form 24 September 2017 Accepted 30 September 2017 Available online xxxx Keywords: Cervical cancer Radiation therapy Pelvic insufficiency fractures
a b s t r a c t Objectives. This study clarified the incidence of and identified the risk factors for post-radiation pelvic insufficiency fractures (PIFs) in women who received postoperative definitive or adjuvant radiotherapy (RT) for cervical cancer. Patients and methods. The medical records and data of imaging studies, including computed tomography scan and magnetic resonance imaging, of women with cervical cancer who received external-beam RT for the entire pelvic area between January 2003 and December 2012 at our institution were reviewed. Results. A total of 533 patients with histologically diagnosed cervical cancer who received RT (298: definitive RT, 235: adjuvant RT) were included in this study. Eighty-four patients (15.8%) developed PIF in the irradiated field. Median age at onset of PIF was 72.5 years (range: 54–95 years), and 82 of them (98%) were postmenopausal women. Sixty-nine patients (80%) developed PIF within 3 years from the completion of RT. The median time for the development of PIF was 14 months (range: 1–81 months). The most commonly involved fracture site was the sacral bone. Postmenopausal state, coexistence of rheumatoid arthritis, and high-dose-rate intracavitary brachytherapy (HDR-ICBT) use were significant predisposing factors for the development of PIF, according to multivariate analysis. Conclusions. The incidence rate of PIF among patients who received RT for locally advanced cervical cancer was 15.8%. The principal predisposing factors for post-radiation PIF were postmenopausal state, rheumatoid arthritis, and HDR-ICBT use. Active interventions, including bone density screening followed by medication, should be considered during the early stage of RT for women with high-risk factors of PIF. © 2017 Published by Elsevier Inc.
1. Introduction The incidence of radiotherapy (RT)-induced pelvic insufficiency fractures (PIFs) in patients with gynecologic cancer is 10%–29% [1–5]. PIF is a late-onset complication of RT in patients with cervical cancer. It may cause refractory pain and frequently impair the quality of life of
⁎ Corresponding author. E-mail address: [email protected] (S. Nagao).
patients for a prolonged period. Hence, RT-induced PIF in patients with cervical cancer is one of the urgent issues to be resolved. The first step to preventing PIF is to appropriately identify high-risk women for PIF. However, most previous studies on PIF among patients with cervical cancer have confined their research to definitive RTinduced PIF. Information on PIF, including both adjuvant and definitive RT-induced PIF, is lacking. The objective of this study was to clarify the incidence of and identify the risk factors for RT-induced PIF in patients who received definitive or postoperative adjuvant RT for locally advanced cervical cancer.
https://doi.org/10.1016/j.ygyno.2017.09.035 0090-8258/© 2017 Published by Elsevier Inc.
Please cite this article as: K. Yamamoto, et al., Pelvic fractures after definitive and postoperative radiotherapy for cervical cancer: A retrospective analysis of risk factors, Gynecol Oncol (2017), https://doi.org/10.1016/j.ygyno.2017.09.035
2
K. Yamamoto et al. / Gynecologic Oncology xxx (2017) xxx–xxx
2. Patients and methods
3. Results
2.1. Study population
Of 561 patients with cervical cancer who underwent whole pelvic radiation therapy at Hyogo Cancer Center between January 2003 and December 2012, 533 patients were included in this study. Twentyeight patients who received CT or MRI only once during the follow-up period were excluded from the study. Among the study patients, 298 received definitive RT and 235 received adjuvant RT after surgery. Patient characteristics are shown in Table 1. Eighty-four (15.8%) of 533 patients developed PIF in the irradiated field during a median follow-up period of 49 months (range: 6–143 months). The median age of the patients who developed PIF was 72.5 years (range: 54–95 years), and 82 of them (98%) were postmenopausal women. By contrast, the median age of patients without PIF was 55 years. No fracture occurred in patients who received hormone replacement therapy. Only 39% of non-fracture cases received definitive RT, whereas 70% of the fracture cases received definitive RT. In total, 30 patients underwent irradiation for the enlargement of para-aortic lymph nodes. PIF was diagnosed after a median of 14 months (range, 1–81 months) from the completion of RT (Fig. 1). Sixty-nine patients (82.1%) developed PIF within 3 years. The distribution of PIFs (includes overlap cases) was as follows: sacrum in 39 (46.4%), pubis in 27 (32.1%), lumbar spinal vertebra in 25 (29.8%), sacroiliac joint in 22 (26.2%), ilium in 11 (13.1%), and femoral head in 4 (4.8%) women. Multiple fractures developed at initial diagnosis in 51(60.7%) patients. In 84 women with PIF, 25 (29.7%) required NSAIDs and 20 (23.8%) required an opioid combined with NSAIDs. These 45 women (53.5%) may have suffered from pain. Furthermore, 6 (9.4%) of 84 women required hospitalization for pain control. No woman underwent surgical treatment. In contrast, 16 women with PIF, identified base on imaging findings, had no symptoms relevant to PIF-induced pain. In the univariate analysis using the log-rank test, age ≥ 70 years (p b 0.0001), postmenopausal state (p b 0.0001), HDR-ICBT use (p b 0.0001), history of previous delivery (p = 0.0260), body mass index ≥20 kg/m2
After institutional review board approval was obtained, we reviewed the medical records of women with cervical cancer who had received external-beam RT (EBRT) for the entire pelvis between January 2003 and December 2012 at our institution. These women included both patients who received postoperative adjuvant RT and definitive RT. Patients with PIF at the beginning of RT were excluded. We reviewed the medical records and extracted the patient clinical information according to the STROBE statement [6].
2.2. Follow-up and diagnosis of PIF After the completion of RT, all patients received radiological imaging tests, including computed tomography (CT) and magnetic resonance imaging (MRI), twice a year for the first 2 years and yearly for the next 3 years. We identified PIFs using the following diagnostic criteria: a bony lesion with an apparent fracture line with or without sclerotic changes within the irradiated field. Fractures secondary to bone metastasis were excluded.
2.3. Radiation therapy The definitive RT protocol followed at our hospital comprises a combination of EBRT and high-dose-rate intracavitary brachytherapy (HDR-ICBT) using a remote afterloading system, concurrent with or without chemotherapy. A total dose of 30–41.4 Gy of EBRT administered at 1.8 Gy per fraction, followed by HDR-ICBT and EBRT with central shielding (CS) up to a total external dose of 50–50.4 Gy, was delivered over 5 to 6 weeks. Usually, EBRT was delivered using the four-field box technique, except for elderly or obese patients, in whom the anteroposterior/posteroanterior (AP/ PA) field technique was used. Two-to-four fractions of HDR-ICBT at 5–6 Gy per fraction were administered to point A. Patients with normal organ function received 5–6 cycles of concurrent chemotherapy comprising weekly cisplatin (40 mg/m2) or weekly nedaplatin (30 mg/m2). Intravenous and oral steroids were administered before chemotherapy. For postoperative adjuvant therapy, patients with a high risk of pelvic recurrence received 50.4 Gy at 1.8 Gy per fraction of EBRT for the entire pelvic field, using the four-field box technique. Patients with a high risk of vaginal stump recurrence received 30.6 Gy at 1.8 Gy per fraction of EBRT using the four-field box technique, followed by 19.8 Gy at 1.8 Gy per fraction of EBRT using the AP/PA field technique for the same field, with CS and 24 Gy at 6 Gy per fraction of HDR-ICBT to point A. In patients with positive para-aortic nodes, the RT fields were extended up to the levels of the involved lymph nodes. Chemotherapy was administered concurrently in patients with positive pelvic nodes, pathological parametrial invasion, and non-squamous cell carcinoma histology.
2.4. Data analysis We calculated the temporal rates of PIF using the Kaplan–Meier method. Risk factors for PIF, including age, menopausal status, diabetes mellitus, rheumatoid arthritis, RT technique, chemotherapy use, and HDR-ICBT use, were assessed using the log-rank test (univariate analysis) and Cox regression analysis with hazard models (multivariate analysis). Factors with a difference in p-values of b 0.005 were considered significant. Statistical analysis was performed using IBM SPSS statistics Version 22.0 (IBM Corp.; Armonk, New York, USA).
Table 1 Patient characteristics.
Median age (range) Median body mass index (range) Postmenopausal state Prior HRT Smoking Current/past Never Unknown Medical history Diabetes mellitus Rheumatoid arthritis FIGO stage I II III IV Histopathology Squamous cell carcinoma Adenocarcinoma Adenosquamous carcinoma Neuroendocrine carcinoma Unknown Radiation therapy Adjuvant RT Adjuvant CCRT Definitive RT Definitive CCRT HDR-ICBT
Non-PIF (N = 449)
PIF (N = 84)
p-Value
55.0 (25–89) 21 (14–36) 250 (57%) 33 (7%)
72.5 (54–95) 22 (15–31) 82 (98%) 0 (0%)
b0.0001 0.290 0.737 0.014
91 (20%) 250 (56%) 108 (24%)
9 (11%) 61 (72%) 14 (17%)
0.014
24 (5%) 3 (1%)
12 (14%) 5 (6%)
0.897 0.768
151 (34%) 152 (34%) 97 (22%) 49 (10%)
9 (11%) 34 (40%) 28 (33%) 13 (16%)
328 (73%) 71 (17%) 11 (2%) 10 (2%) 29 (6%)
65 (78%) 6 (7%) 3 (4%) 1 (1%) 9 (10%)
0.408
96 (21%) 177 (40%) 44 (10%) 132 (29%) 178 (40%)
9 (11%) 16 (19%) 21 (25%) 38 (45%) 60 (71%)
b0.0001
0.004
b0.0001
PIF: pelvic insufficiency fracture, HRT: hormone replacement therapy, RT: radio therapy, CCRT: concurrent chemoradiotherapy, HDR-ICBT: high-dose-rate intracavitary brachytherapy.
Please cite this article as: K. Yamamoto, et al., Pelvic fractures after definitive and postoperative radiotherapy for cervical cancer: A retrospective analysis of risk factors, Gynecol Oncol (2017), https://doi.org/10.1016/j.ygyno.2017.09.035
K. Yamamoto et al. / Gynecologic Oncology xxx (2017) xxx–xxx
3
Table 3 Multivariate analysis of risk factors associated with PIF (N = 533). Factors
HR
95%CI
p-Value
Postmenopausal state Rheumatoid arthritis HDR-ICBT HR: hazard ratio, CI: confidence interval, HDR-ICBT: High-dose-rate intracavitary brachytherapy.
Fig. 1. Incidence of insufficiency fractures after pelvic radiation therapy among patients who suffered fractures.
(p = 0.0222), smoking history (p = 0.0402), absence of hormonal replacement therapy (p = 0.0066), rheumatoid arthritis (p = 0.0020), and diabetes mellitus (p = 0.0027) were significantly related to PIF (Table 2). Concurrent chemotherapy (p = 0.6916) was not significant. Three factors, namely age ≥ 70 years, postmenopausal state, and no hormonal replacement therapy, were strongly correlated, and they are summarized into a factor of the postmenopausal state. Logistic regression analysis of seven factors revealed that rheumatoid arthritis and HDR-ICBT were significant predisposing factors for the development of PIF (Table 3). 4. Discussion In our study, 84 of 533 patients (15.5%) developed PIF after definitive and postoperative pelvic RT for cervical cancer, which was in accordance with the findings of other studies [1–5]. A retrospective cohort study by Baxter et al. evaluated women aged ≥65 years with pelvic malignancies using the Surveillance, Epidemiology, and End Results cancer registry data linked to Medicare claims data [7]. They reported that among 1139 cervical cancer patients, the incidences of PIF in women treated with or without RT were 8.2% and 5.9%, respectively. A subsequent study by Ikushima et al. reported a PIF rate of 11% among 158 patients with gynecologic malignancies treated with RT [2]. A recent study by Kwon et al. retrospectively evaluated post-treatment MRI studies on 510 women treated with RT for cervical cancer, and PIF was noted in 20% of patients [4]. As described above, the frequency of PIFs induced by RT is increasing in recent years. Blomlie et al. prospectively studied 18 Norwegian women who were treated with RT for advanced cervical cancer [8]. They performed pre- and post-treatment MRI studies and detected PIF in 89% of patients. This higher incidence may be because Table 2 Univariate analysis of risk factors associated with PIF (N = 533). Factors
HR
95% CI
p-Value
Age ≥ 70 (years) Postmenopausal state Birth history BMI ≥ 20 kg/m2 Smoking history No prior HRT Diabetes mellitus Rheumatoid arthritis HDR-ICBT Concurrent chemotherapy
16.28 5.21 2.01 1.66 1.79 2.99 3.87 37.60 4.80 1.10
9.60–27.62 3.21–8.46 1.09–3.70 1.08–2.57 1.03–3.13 1.36–6.60 1.60–9.37 5.68–249.10 0.13–0.33 0.70–1.73
b0.0001 b0.0001 0.0260 0.0222 0.0402 0.0066 0.0027 0.0020 b0.0001 0.6916
HR: hazard ratio, CI: confidence interval, BMI: body mass index, HRT: hormone replacement therapy, HDR-ICBT: High-dose-rate intracavitary brachytherapy.
of the wider use of high-performance imaging modalities, including CT, MRI, and bone scintigraphy. These modalities are more sensitive to the detection of PIF than is conventional radiography. We identified three principal predisposing factors of post-radiation PIF: postmenopausal state, rheumatoid arthritis, and HDR-ICBT use. According to many studies, the most important risk factor for PIF is osteoporosis [1,9]. We could not evaluate the importance of osteoporosis, because only few women had undergone bone mass assessment before and during RT. Blomlie et al. reported that 95% of women with PIF in their study were in the postmenopausal state, and 98% of women with PIF were in the postmenopausal state in our study [8]. It is presumed that elderly women who already have osteoporosis prior to treatment initiation are likely to suffer from PIF due to RT. We were concerned that the addition of chemotherapy would be a high-risk factor for PIF, because steroid use during chemotherapy may cause a decrease in bone mass. However, the addition of chemotherapy was not identified as a high-risk factor in our study. Patient older than 75 years who are at an extremely high risk of osteoporosis consistently received RT without chemotherapy. This might have counteracted the negative effect of chemotherapy. Rheumatoid arthritis is a risk factor for PIF [1,3,5]. Chronic inflammation of the joints, long-term use of adrenocortical hormones, and low momentum compared with healthy individuals may cause bone mineral density loss. The four-field box EBRT irradiation method, followed by AP/PA EBRT with CS and HDR-ICBT, has been known to increase the risk of PIF compared with the four-field box EBRT method alone. HDR-ICBT was used in combination with AP/PA EBRT with CS, and it was performed in 87% of patients who received definitive RT. The dose–volume analysis in the current study suggests that 5%–20% of prescribed brachytherapy doses scatter to the pubic bone [10,11]. Thus, the usage of HDR-ICBT may increase the irradiance level of the pelvic bones. In addition, we were concerned regarding the increase in PIF because of early artificial menopause due to treatment. However, in this study, only 2 of 196 (1%) women who were premenopausal at the initiation of treatment experienced PIF. Among 196 women, only 33 (16.8%) received hormone replacement therapy during or after treatment. Bone mass at the initiation of RT may be a more important factor to determine the risk of PIF than is the subsequent decrease in bone mass. As just described, it is important to evaluate bone mineral density before RT in postmenopausal women and provide pharmacological interventions. However, drugs recommended for the prevention and treatment of PIF induced by RT have not yet been determined. Bisphosphonate, which induces apoptosis in osteoclasts and inhibits bone resorption, has been widely used for osteoporosis [12]. Several reports have described that bisphosphonate is useful for bone loss associated with bone metastasis and is also effective for treating and preventing PIF after RT [3,13]. However, bisphosphonate has an antiangiogenic effect similar to that of RT and suppresses bone turnover. Therefore, it may confound osteonecrosis and induce pathologic fractures [13–15]. Selective estrogen receptor modulator (SERM) and vitamin preparations are also used for postmenopausal osteoporosis. Previous studies have demonstrated that SERM normalizes the crosslinked structure of collagen by improving estrogenic action and homocysteine metabolism [16,17]. Consequently, it improves bone quality and increases bone strength [18–20]. In addition, through an antioxidant effect, SERM is
Please cite this article as: K. Yamamoto, et al., Pelvic fractures after definitive and postoperative radiotherapy for cervical cancer: A retrospective analysis of risk factors, Gynecol Oncol (2017), https://doi.org/10.1016/j.ygyno.2017.09.035
4
K. Yamamoto et al. / Gynecologic Oncology xxx (2017) xxx–xxx
expected to suppress oxidative stress caused by irradiation of the pelvic bone. It may be suitable for the prevention of PIF caused by RT. We identified postmenopausal state and rheumatoid arthritis as high-risk factors of PIF that are related to low bone mass. Low bone mass at the beginning of RT may be an important predictive factor of PIF. Unfortunately, bone mass data are insufficient in this study. However, the influence of RT on the bone is thought to change bone quality, including the bridging structure, and not bone mass. Therefore, it is difficult to predict PIF based on changes in the bone mass. We identified three high-risk factors of PIF, namely postmenopausal status, coexistent rheumatoid arthritis, and usage of HDR-ICBT. Active interventions, such as drug administration from an early stage of treatment, are necessary for women with high-risk factors. Conflict of interest statement None of the authors report a conflict of interest.
References [1] I. Ogino, N. Okamoto, Y. Ono, T. Kitamura, H. Nakayama, Pelvic insufficiency fractures in postmenopausal woman with advanced cervical cancer treated by radiotherapy, Radiother. Oncol. 68 (1) (2003) 61–67. [2] H. Ikushima, K. Osaki, S. Furutani, K. Yamashita, Y. Kishida, T. Kudoh, H. Nishitani, Pelvic bone complications following radiation therapy of gynecologic malignancies: clinical evaluation of radiation-induced pelvic insufficiency fractures, Gynecol. Oncol. 103 (3) (2006) 1100–1104. [3] D. Oh, S.J. Huh, H. Nam, W. Park, Y. Han, H. Lim do, Y.C. Ahn, J.W. Lee, B.G. Kim, D.S. Bae, J.H. Lee, Pelvic insufficiency fracture after pelvic radiotherapy for cervical cancer: analysis of risk factors, Int. J. Radiat. Oncol. Biol. Phys. 70 (4) (2008) 1183–1188. [4] J.W. Kwon, S.J. Huh, Y.C. Yoon, S.H. Choi, J.Y. Jung, D. Oh, B.K. Choe, Pelvic bone complications after radiation therapy of uterine cervical cancer: evaluation with MRI, AJR Am. J. Roentgenol. 191 (4) (2008 Oct) 987–994, https://doi.org/10.2214/AJR. 07.3634. [5] K.M. Schmeler, A. Jhingran, R.B. Iyer, C.C. Sun, P.J. Eifel, P.T. Soliman, P.T. Ramirez, M. Frumovitz, D.C. Bodurka, A.K. Sood, Pelvic fractures after radiotherapy for cervical cancer: implications for survivors, Cancer 116 (3) (2010) 625–630, https://doi. org/10.1002/cncr.24811. [6] J.P. Vandenbroucke, E. von Elm, D.G. Altman, P.C. Gøtzsche, C.D. Mulrow, S.J. Pocock, C. Poole, J.J. Schlesselman, M. Egger, Strengthening the reporting of observational studies in epidemiology (STROBE): explanation and elaboration, Epidemiology 18 (6) (2007) 805–835.
[7] N.N. Baxter, E.B. Habermann, J.E. Tepper, S.B. Durham, B.A. Virnig, Risk of pelvic fractures in older women following pelvic irradiation, JAMA 294 (20) (2005) 2587–2593. [8] V. Blomlie, E.K. Rofstad, K. Talle, K. Sundfør, M. Winderen, H.H. Lien, Incidence of radiation-induced insufficiency fractures of the female pelvis: evaluation with MR imaging, Am. J. Roentgenol. 167 (5) (1996) 1205–1210. [9] J.W. Hopewell, Radiation-therapy effects on bone density, Med. Pediatr. Oncol. 41 (3) (2003) 208–211. [10] A.L. Fu, K.M. Greven, Y. Maruyama, Radiation osteitis and insufficiency fractures after pelvic irradiation for gynecologic malignancies, Am. J. Clin. Oncol. 17 (3) (1994) 248–254. [11] P. Bliss, C.A. Parsons, P.R. Blake, Incidence and possible aetiological factors in the development of pelvic insufficiency fractures following radical radiotherapy, Br. J. Radiol. 69 (822) (1996) 548–554. [12] H. Fleisch, Bisphosphonates in Bone Disease, Fourth Edition Academic Press, San Diego, 2000 From the Laboratory to the Patient. [13] W. Small Jr, L. Kachnic, Postradiotherapy pelvic fractures: cause for concern or opportunity for future research? JAMA 294 (20) (2005) 2635–2637. [14] J. Wood, K. Bonjean, S. Ruetz, A. Bellahcène, L. Devy, J.M. Foidart, V. Castronovo, J.R. Green, Novel antiangiogenic effects of the bisphosphonate compound zoledronic acid, J. Pharmacol. Exp. Ther. 302 (3) (2002) 1055–1061. [15] H. Uezono, K. Tsujino, K. Moriki, F. Nagano, Y. Ota, R. Sasaki, T. Soejima, Pelvic insufficiency fracture after definitive radiotherapy for uterine cervical cancer: retrospective analysis of risk factors, J. Radiat. Res. 54 (6) (2013) 1102–1109. [16] M. Saito, K. Marumo, S. Soshi, Y. Kida, C. Ushiku, A. Shinohara, Raloxifene ameliorates detrimental enzymatic and nonenzymatic collagen cross-links and bone strength in rabbits with hyperhomocysteinemia, Osteoporos. Int. 21 (2010) 655–666. [17] M. Saito, Y. Kida, T. Nishizawa, S. Arakawa, H. Okabe, A. Seki, K. Marumo, Effects of 18-month treatment with bazedoxifene on enzymatic immature and mature cross-rink factors and non-enzymatic advanced glycation end products, mineralization, and trabecular microarchitecture of vertebra in ovariectomized monkeys, Bone 81 (2015) 573–580. [18] N. Iikuni, E. Hamaya, S. Nihojima, S. Yokoyama, W. Goto, M. Taketsuna, A. Miyauchi, H. Sowa, Safety and effectiveness profile of raloxifene in long-term, prospective, postmarketing surveillance, J. Bone Miner. Metab. 30 (6) (2012) 674–682, https:// doi.org/10.1007/s00774-012-0365-1. [19] E. Seeman, G.G. Crans, A. Diez-Perez, K.V. Pinette, P.D. Delmas, Anti-vertebral fracture efficacy of raloxifene: a meta-analysis, Osteoporos. Int. 17 (2) (2006) 313–316. [20] S.L. Silverman, C. Christiansen, H.K. Genant, S. Vukicevic, J.R. Zanchetta, T.J. de Villiers, G.D. Constantine, A.A. Chines, Efficacy of bazedoxifene in reducing new vertebral fracture risk in postmenopausal women with osteoporosis: results from a 3year, randomized, placebo-, and active-controlled clinical trial, J. Bone Miner. Res. 23 (12) (2008) 1923–1934.
Please cite this article as: K. Yamamoto, et al., Pelvic fractures after definitive and postoperative radiotherapy for cervical cancer: A retrospective analysis of risk factors, Gynecol Oncol (2017), https://doi.org/10.1016/j.ygyno.2017.09.035
Contents lists available at ScienceDirect
Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno
Pelvic fractures after definitive and postoperative radiotherapy for cervical cancer: A retrospective analysis of risk factors Kasumi Yamamoto, Shoji Nagao ⁎, Kazuhiro Suzuki, Ai Kogiku, Tokihiro Senda, Hiroko Yano, Miho Kitai, Takaya Shiozaki, Kazuko Matsuoka, Satoshi Yamaguchi Department of Gynecologic Oncology, Hyogo Cancer Center, 13-70 Kitaoji-cho, Akashi-city, Hyogo 673-8558, Japan
H I G H L I G H T S • • • •
The incidence of pelvic fracture was 15.8% in women with cervical cancer received radiotherapy. Pelvic fracture was diagnosed after a median of 14 months from the completion of radiotherapy. The predisposing factors were postmenopausal status, rheumatoid arthritis, and brachytherapy. Active interventions are necessary for women with high-risk factors.
a r t i c l e
i n f o
Article history: Received 26 July 2017 Received in revised form 24 September 2017 Accepted 30 September 2017 Available online xxxx Keywords: Cervical cancer Radiation therapy Pelvic insufficiency fractures
a b s t r a c t Objectives. This study clarified the incidence of and identified the risk factors for post-radiation pelvic insufficiency fractures (PIFs) in women who received postoperative definitive or adjuvant radiotherapy (RT) for cervical cancer. Patients and methods. The medical records and data of imaging studies, including computed tomography scan and magnetic resonance imaging, of women with cervical cancer who received external-beam RT for the entire pelvic area between January 2003 and December 2012 at our institution were reviewed. Results. A total of 533 patients with histologically diagnosed cervical cancer who received RT (298: definitive RT, 235: adjuvant RT) were included in this study. Eighty-four patients (15.8%) developed PIF in the irradiated field. Median age at onset of PIF was 72.5 years (range: 54–95 years), and 82 of them (98%) were postmenopausal women. Sixty-nine patients (80%) developed PIF within 3 years from the completion of RT. The median time for the development of PIF was 14 months (range: 1–81 months). The most commonly involved fracture site was the sacral bone. Postmenopausal state, coexistence of rheumatoid arthritis, and high-dose-rate intracavitary brachytherapy (HDR-ICBT) use were significant predisposing factors for the development of PIF, according to multivariate analysis. Conclusions. The incidence rate of PIF among patients who received RT for locally advanced cervical cancer was 15.8%. The principal predisposing factors for post-radiation PIF were postmenopausal state, rheumatoid arthritis, and HDR-ICBT use. Active interventions, including bone density screening followed by medication, should be considered during the early stage of RT for women with high-risk factors of PIF. © 2017 Published by Elsevier Inc.
1. Introduction The incidence of radiotherapy (RT)-induced pelvic insufficiency fractures (PIFs) in patients with gynecologic cancer is 10%–29% [1–5]. PIF is a late-onset complication of RT in patients with cervical cancer. It may cause refractory pain and frequently impair the quality of life of
⁎ Corresponding author. E-mail address: [email protected] (S. Nagao).
patients for a prolonged period. Hence, RT-induced PIF in patients with cervical cancer is one of the urgent issues to be resolved. The first step to preventing PIF is to appropriately identify high-risk women for PIF. However, most previous studies on PIF among patients with cervical cancer have confined their research to definitive RTinduced PIF. Information on PIF, including both adjuvant and definitive RT-induced PIF, is lacking. The objective of this study was to clarify the incidence of and identify the risk factors for RT-induced PIF in patients who received definitive or postoperative adjuvant RT for locally advanced cervical cancer.
https://doi.org/10.1016/j.ygyno.2017.09.035 0090-8258/© 2017 Published by Elsevier Inc.
Please cite this article as: K. Yamamoto, et al., Pelvic fractures after definitive and postoperative radiotherapy for cervical cancer: A retrospective analysis of risk factors, Gynecol Oncol (2017), https://doi.org/10.1016/j.ygyno.2017.09.035
2
K. Yamamoto et al. / Gynecologic Oncology xxx (2017) xxx–xxx
2. Patients and methods
3. Results
2.1. Study population
Of 561 patients with cervical cancer who underwent whole pelvic radiation therapy at Hyogo Cancer Center between January 2003 and December 2012, 533 patients were included in this study. Twentyeight patients who received CT or MRI only once during the follow-up period were excluded from the study. Among the study patients, 298 received definitive RT and 235 received adjuvant RT after surgery. Patient characteristics are shown in Table 1. Eighty-four (15.8%) of 533 patients developed PIF in the irradiated field during a median follow-up period of 49 months (range: 6–143 months). The median age of the patients who developed PIF was 72.5 years (range: 54–95 years), and 82 of them (98%) were postmenopausal women. By contrast, the median age of patients without PIF was 55 years. No fracture occurred in patients who received hormone replacement therapy. Only 39% of non-fracture cases received definitive RT, whereas 70% of the fracture cases received definitive RT. In total, 30 patients underwent irradiation for the enlargement of para-aortic lymph nodes. PIF was diagnosed after a median of 14 months (range, 1–81 months) from the completion of RT (Fig. 1). Sixty-nine patients (82.1%) developed PIF within 3 years. The distribution of PIFs (includes overlap cases) was as follows: sacrum in 39 (46.4%), pubis in 27 (32.1%), lumbar spinal vertebra in 25 (29.8%), sacroiliac joint in 22 (26.2%), ilium in 11 (13.1%), and femoral head in 4 (4.8%) women. Multiple fractures developed at initial diagnosis in 51(60.7%) patients. In 84 women with PIF, 25 (29.7%) required NSAIDs and 20 (23.8%) required an opioid combined with NSAIDs. These 45 women (53.5%) may have suffered from pain. Furthermore, 6 (9.4%) of 84 women required hospitalization for pain control. No woman underwent surgical treatment. In contrast, 16 women with PIF, identified base on imaging findings, had no symptoms relevant to PIF-induced pain. In the univariate analysis using the log-rank test, age ≥ 70 years (p b 0.0001), postmenopausal state (p b 0.0001), HDR-ICBT use (p b 0.0001), history of previous delivery (p = 0.0260), body mass index ≥20 kg/m2
After institutional review board approval was obtained, we reviewed the medical records of women with cervical cancer who had received external-beam RT (EBRT) for the entire pelvis between January 2003 and December 2012 at our institution. These women included both patients who received postoperative adjuvant RT and definitive RT. Patients with PIF at the beginning of RT were excluded. We reviewed the medical records and extracted the patient clinical information according to the STROBE statement [6].
2.2. Follow-up and diagnosis of PIF After the completion of RT, all patients received radiological imaging tests, including computed tomography (CT) and magnetic resonance imaging (MRI), twice a year for the first 2 years and yearly for the next 3 years. We identified PIFs using the following diagnostic criteria: a bony lesion with an apparent fracture line with or without sclerotic changes within the irradiated field. Fractures secondary to bone metastasis were excluded.
2.3. Radiation therapy The definitive RT protocol followed at our hospital comprises a combination of EBRT and high-dose-rate intracavitary brachytherapy (HDR-ICBT) using a remote afterloading system, concurrent with or without chemotherapy. A total dose of 30–41.4 Gy of EBRT administered at 1.8 Gy per fraction, followed by HDR-ICBT and EBRT with central shielding (CS) up to a total external dose of 50–50.4 Gy, was delivered over 5 to 6 weeks. Usually, EBRT was delivered using the four-field box technique, except for elderly or obese patients, in whom the anteroposterior/posteroanterior (AP/ PA) field technique was used. Two-to-four fractions of HDR-ICBT at 5–6 Gy per fraction were administered to point A. Patients with normal organ function received 5–6 cycles of concurrent chemotherapy comprising weekly cisplatin (40 mg/m2) or weekly nedaplatin (30 mg/m2). Intravenous and oral steroids were administered before chemotherapy. For postoperative adjuvant therapy, patients with a high risk of pelvic recurrence received 50.4 Gy at 1.8 Gy per fraction of EBRT for the entire pelvic field, using the four-field box technique. Patients with a high risk of vaginal stump recurrence received 30.6 Gy at 1.8 Gy per fraction of EBRT using the four-field box technique, followed by 19.8 Gy at 1.8 Gy per fraction of EBRT using the AP/PA field technique for the same field, with CS and 24 Gy at 6 Gy per fraction of HDR-ICBT to point A. In patients with positive para-aortic nodes, the RT fields were extended up to the levels of the involved lymph nodes. Chemotherapy was administered concurrently in patients with positive pelvic nodes, pathological parametrial invasion, and non-squamous cell carcinoma histology.
2.4. Data analysis We calculated the temporal rates of PIF using the Kaplan–Meier method. Risk factors for PIF, including age, menopausal status, diabetes mellitus, rheumatoid arthritis, RT technique, chemotherapy use, and HDR-ICBT use, were assessed using the log-rank test (univariate analysis) and Cox regression analysis with hazard models (multivariate analysis). Factors with a difference in p-values of b 0.005 were considered significant. Statistical analysis was performed using IBM SPSS statistics Version 22.0 (IBM Corp.; Armonk, New York, USA).
Table 1 Patient characteristics.
Median age (range) Median body mass index (range) Postmenopausal state Prior HRT Smoking Current/past Never Unknown Medical history Diabetes mellitus Rheumatoid arthritis FIGO stage I II III IV Histopathology Squamous cell carcinoma Adenocarcinoma Adenosquamous carcinoma Neuroendocrine carcinoma Unknown Radiation therapy Adjuvant RT Adjuvant CCRT Definitive RT Definitive CCRT HDR-ICBT
Non-PIF (N = 449)
PIF (N = 84)
p-Value
55.0 (25–89) 21 (14–36) 250 (57%) 33 (7%)
72.5 (54–95) 22 (15–31) 82 (98%) 0 (0%)
b0.0001 0.290 0.737 0.014
91 (20%) 250 (56%) 108 (24%)
9 (11%) 61 (72%) 14 (17%)
0.014
24 (5%) 3 (1%)
12 (14%) 5 (6%)
0.897 0.768
151 (34%) 152 (34%) 97 (22%) 49 (10%)
9 (11%) 34 (40%) 28 (33%) 13 (16%)
328 (73%) 71 (17%) 11 (2%) 10 (2%) 29 (6%)
65 (78%) 6 (7%) 3 (4%) 1 (1%) 9 (10%)
0.408
96 (21%) 177 (40%) 44 (10%) 132 (29%) 178 (40%)
9 (11%) 16 (19%) 21 (25%) 38 (45%) 60 (71%)
b0.0001
0.004
b0.0001
PIF: pelvic insufficiency fracture, HRT: hormone replacement therapy, RT: radio therapy, CCRT: concurrent chemoradiotherapy, HDR-ICBT: high-dose-rate intracavitary brachytherapy.
Please cite this article as: K. Yamamoto, et al., Pelvic fractures after definitive and postoperative radiotherapy for cervical cancer: A retrospective analysis of risk factors, Gynecol Oncol (2017), https://doi.org/10.1016/j.ygyno.2017.09.035
K. Yamamoto et al. / Gynecologic Oncology xxx (2017) xxx–xxx
3
Table 3 Multivariate analysis of risk factors associated with PIF (N = 533). Factors
HR
95%CI
p-Value
Postmenopausal state Rheumatoid arthritis HDR-ICBT HR: hazard ratio, CI: confidence interval, HDR-ICBT: High-dose-rate intracavitary brachytherapy.
Fig. 1. Incidence of insufficiency fractures after pelvic radiation therapy among patients who suffered fractures.
(p = 0.0222), smoking history (p = 0.0402), absence of hormonal replacement therapy (p = 0.0066), rheumatoid arthritis (p = 0.0020), and diabetes mellitus (p = 0.0027) were significantly related to PIF (Table 2). Concurrent chemotherapy (p = 0.6916) was not significant. Three factors, namely age ≥ 70 years, postmenopausal state, and no hormonal replacement therapy, were strongly correlated, and they are summarized into a factor of the postmenopausal state. Logistic regression analysis of seven factors revealed that rheumatoid arthritis and HDR-ICBT were significant predisposing factors for the development of PIF (Table 3). 4. Discussion In our study, 84 of 533 patients (15.5%) developed PIF after definitive and postoperative pelvic RT for cervical cancer, which was in accordance with the findings of other studies [1–5]. A retrospective cohort study by Baxter et al. evaluated women aged ≥65 years with pelvic malignancies using the Surveillance, Epidemiology, and End Results cancer registry data linked to Medicare claims data [7]. They reported that among 1139 cervical cancer patients, the incidences of PIF in women treated with or without RT were 8.2% and 5.9%, respectively. A subsequent study by Ikushima et al. reported a PIF rate of 11% among 158 patients with gynecologic malignancies treated with RT [2]. A recent study by Kwon et al. retrospectively evaluated post-treatment MRI studies on 510 women treated with RT for cervical cancer, and PIF was noted in 20% of patients [4]. As described above, the frequency of PIFs induced by RT is increasing in recent years. Blomlie et al. prospectively studied 18 Norwegian women who were treated with RT for advanced cervical cancer [8]. They performed pre- and post-treatment MRI studies and detected PIF in 89% of patients. This higher incidence may be because Table 2 Univariate analysis of risk factors associated with PIF (N = 533). Factors
HR
95% CI
p-Value
Age ≥ 70 (years) Postmenopausal state Birth history BMI ≥ 20 kg/m2 Smoking history No prior HRT Diabetes mellitus Rheumatoid arthritis HDR-ICBT Concurrent chemotherapy
16.28 5.21 2.01 1.66 1.79 2.99 3.87 37.60 4.80 1.10
9.60–27.62 3.21–8.46 1.09–3.70 1.08–2.57 1.03–3.13 1.36–6.60 1.60–9.37 5.68–249.10 0.13–0.33 0.70–1.73
b0.0001 b0.0001 0.0260 0.0222 0.0402 0.0066 0.0027 0.0020 b0.0001 0.6916
HR: hazard ratio, CI: confidence interval, BMI: body mass index, HRT: hormone replacement therapy, HDR-ICBT: High-dose-rate intracavitary brachytherapy.
of the wider use of high-performance imaging modalities, including CT, MRI, and bone scintigraphy. These modalities are more sensitive to the detection of PIF than is conventional radiography. We identified three principal predisposing factors of post-radiation PIF: postmenopausal state, rheumatoid arthritis, and HDR-ICBT use. According to many studies, the most important risk factor for PIF is osteoporosis [1,9]. We could not evaluate the importance of osteoporosis, because only few women had undergone bone mass assessment before and during RT. Blomlie et al. reported that 95% of women with PIF in their study were in the postmenopausal state, and 98% of women with PIF were in the postmenopausal state in our study [8]. It is presumed that elderly women who already have osteoporosis prior to treatment initiation are likely to suffer from PIF due to RT. We were concerned that the addition of chemotherapy would be a high-risk factor for PIF, because steroid use during chemotherapy may cause a decrease in bone mass. However, the addition of chemotherapy was not identified as a high-risk factor in our study. Patient older than 75 years who are at an extremely high risk of osteoporosis consistently received RT without chemotherapy. This might have counteracted the negative effect of chemotherapy. Rheumatoid arthritis is a risk factor for PIF [1,3,5]. Chronic inflammation of the joints, long-term use of adrenocortical hormones, and low momentum compared with healthy individuals may cause bone mineral density loss. The four-field box EBRT irradiation method, followed by AP/PA EBRT with CS and HDR-ICBT, has been known to increase the risk of PIF compared with the four-field box EBRT method alone. HDR-ICBT was used in combination with AP/PA EBRT with CS, and it was performed in 87% of patients who received definitive RT. The dose–volume analysis in the current study suggests that 5%–20% of prescribed brachytherapy doses scatter to the pubic bone [10,11]. Thus, the usage of HDR-ICBT may increase the irradiance level of the pelvic bones. In addition, we were concerned regarding the increase in PIF because of early artificial menopause due to treatment. However, in this study, only 2 of 196 (1%) women who were premenopausal at the initiation of treatment experienced PIF. Among 196 women, only 33 (16.8%) received hormone replacement therapy during or after treatment. Bone mass at the initiation of RT may be a more important factor to determine the risk of PIF than is the subsequent decrease in bone mass. As just described, it is important to evaluate bone mineral density before RT in postmenopausal women and provide pharmacological interventions. However, drugs recommended for the prevention and treatment of PIF induced by RT have not yet been determined. Bisphosphonate, which induces apoptosis in osteoclasts and inhibits bone resorption, has been widely used for osteoporosis [12]. Several reports have described that bisphosphonate is useful for bone loss associated with bone metastasis and is also effective for treating and preventing PIF after RT [3,13]. However, bisphosphonate has an antiangiogenic effect similar to that of RT and suppresses bone turnover. Therefore, it may confound osteonecrosis and induce pathologic fractures [13–15]. Selective estrogen receptor modulator (SERM) and vitamin preparations are also used for postmenopausal osteoporosis. Previous studies have demonstrated that SERM normalizes the crosslinked structure of collagen by improving estrogenic action and homocysteine metabolism [16,17]. Consequently, it improves bone quality and increases bone strength [18–20]. In addition, through an antioxidant effect, SERM is
Please cite this article as: K. Yamamoto, et al., Pelvic fractures after definitive and postoperative radiotherapy for cervical cancer: A retrospective analysis of risk factors, Gynecol Oncol (2017), https://doi.org/10.1016/j.ygyno.2017.09.035
4
K. Yamamoto et al. / Gynecologic Oncology xxx (2017) xxx–xxx
expected to suppress oxidative stress caused by irradiation of the pelvic bone. It may be suitable for the prevention of PIF caused by RT. We identified postmenopausal state and rheumatoid arthritis as high-risk factors of PIF that are related to low bone mass. Low bone mass at the beginning of RT may be an important predictive factor of PIF. Unfortunately, bone mass data are insufficient in this study. However, the influence of RT on the bone is thought to change bone quality, including the bridging structure, and not bone mass. Therefore, it is difficult to predict PIF based on changes in the bone mass. We identified three high-risk factors of PIF, namely postmenopausal status, coexistent rheumatoid arthritis, and usage of HDR-ICBT. Active interventions, such as drug administration from an early stage of treatment, are necessary for women with high-risk factors. Conflict of interest statement None of the authors report a conflict of interest.
References [1] I. Ogino, N. Okamoto, Y. Ono, T. Kitamura, H. Nakayama, Pelvic insufficiency fractures in postmenopausal woman with advanced cervical cancer treated by radiotherapy, Radiother. Oncol. 68 (1) (2003) 61–67. [2] H. Ikushima, K. Osaki, S. Furutani, K. Yamashita, Y. Kishida, T. Kudoh, H. Nishitani, Pelvic bone complications following radiation therapy of gynecologic malignancies: clinical evaluation of radiation-induced pelvic insufficiency fractures, Gynecol. Oncol. 103 (3) (2006) 1100–1104. [3] D. Oh, S.J. Huh, H. Nam, W. Park, Y. Han, H. Lim do, Y.C. Ahn, J.W. Lee, B.G. Kim, D.S. Bae, J.H. Lee, Pelvic insufficiency fracture after pelvic radiotherapy for cervical cancer: analysis of risk factors, Int. J. Radiat. Oncol. Biol. Phys. 70 (4) (2008) 1183–1188. [4] J.W. Kwon, S.J. Huh, Y.C. Yoon, S.H. Choi, J.Y. Jung, D. Oh, B.K. Choe, Pelvic bone complications after radiation therapy of uterine cervical cancer: evaluation with MRI, AJR Am. J. Roentgenol. 191 (4) (2008 Oct) 987–994, https://doi.org/10.2214/AJR. 07.3634. [5] K.M. Schmeler, A. Jhingran, R.B. Iyer, C.C. Sun, P.J. Eifel, P.T. Soliman, P.T. Ramirez, M. Frumovitz, D.C. Bodurka, A.K. Sood, Pelvic fractures after radiotherapy for cervical cancer: implications for survivors, Cancer 116 (3) (2010) 625–630, https://doi. org/10.1002/cncr.24811. [6] J.P. Vandenbroucke, E. von Elm, D.G. Altman, P.C. Gøtzsche, C.D. Mulrow, S.J. Pocock, C. Poole, J.J. Schlesselman, M. Egger, Strengthening the reporting of observational studies in epidemiology (STROBE): explanation and elaboration, Epidemiology 18 (6) (2007) 805–835.
[7] N.N. Baxter, E.B. Habermann, J.E. Tepper, S.B. Durham, B.A. Virnig, Risk of pelvic fractures in older women following pelvic irradiation, JAMA 294 (20) (2005) 2587–2593. [8] V. Blomlie, E.K. Rofstad, K. Talle, K. Sundfør, M. Winderen, H.H. Lien, Incidence of radiation-induced insufficiency fractures of the female pelvis: evaluation with MR imaging, Am. J. Roentgenol. 167 (5) (1996) 1205–1210. [9] J.W. Hopewell, Radiation-therapy effects on bone density, Med. Pediatr. Oncol. 41 (3) (2003) 208–211. [10] A.L. Fu, K.M. Greven, Y. Maruyama, Radiation osteitis and insufficiency fractures after pelvic irradiation for gynecologic malignancies, Am. J. Clin. Oncol. 17 (3) (1994) 248–254. [11] P. Bliss, C.A. Parsons, P.R. Blake, Incidence and possible aetiological factors in the development of pelvic insufficiency fractures following radical radiotherapy, Br. J. Radiol. 69 (822) (1996) 548–554. [12] H. Fleisch, Bisphosphonates in Bone Disease, Fourth Edition Academic Press, San Diego, 2000 From the Laboratory to the Patient. [13] W. Small Jr, L. Kachnic, Postradiotherapy pelvic fractures: cause for concern or opportunity for future research? JAMA 294 (20) (2005) 2635–2637. [14] J. Wood, K. Bonjean, S. Ruetz, A. Bellahcène, L. Devy, J.M. Foidart, V. Castronovo, J.R. Green, Novel antiangiogenic effects of the bisphosphonate compound zoledronic acid, J. Pharmacol. Exp. Ther. 302 (3) (2002) 1055–1061. [15] H. Uezono, K. Tsujino, K. Moriki, F. Nagano, Y. Ota, R. Sasaki, T. Soejima, Pelvic insufficiency fracture after definitive radiotherapy for uterine cervical cancer: retrospective analysis of risk factors, J. Radiat. Res. 54 (6) (2013) 1102–1109. [16] M. Saito, K. Marumo, S. Soshi, Y. Kida, C. Ushiku, A. Shinohara, Raloxifene ameliorates detrimental enzymatic and nonenzymatic collagen cross-links and bone strength in rabbits with hyperhomocysteinemia, Osteoporos. Int. 21 (2010) 655–666. [17] M. Saito, Y. Kida, T. Nishizawa, S. Arakawa, H. Okabe, A. Seki, K. Marumo, Effects of 18-month treatment with bazedoxifene on enzymatic immature and mature cross-rink factors and non-enzymatic advanced glycation end products, mineralization, and trabecular microarchitecture of vertebra in ovariectomized monkeys, Bone 81 (2015) 573–580. [18] N. Iikuni, E. Hamaya, S. Nihojima, S. Yokoyama, W. Goto, M. Taketsuna, A. Miyauchi, H. Sowa, Safety and effectiveness profile of raloxifene in long-term, prospective, postmarketing surveillance, J. Bone Miner. Metab. 30 (6) (2012) 674–682, https:// doi.org/10.1007/s00774-012-0365-1. [19] E. Seeman, G.G. Crans, A. Diez-Perez, K.V. Pinette, P.D. Delmas, Anti-vertebral fracture efficacy of raloxifene: a meta-analysis, Osteoporos. Int. 17 (2) (2006) 313–316. [20] S.L. Silverman, C. Christiansen, H.K. Genant, S. Vukicevic, J.R. Zanchetta, T.J. de Villiers, G.D. Constantine, A.A. Chines, Efficacy of bazedoxifene in reducing new vertebral fracture risk in postmenopausal women with osteoporosis: results from a 3year, randomized, placebo-, and active-controlled clinical trial, J. Bone Miner. Res. 23 (12) (2008) 1923–1934.
Please cite this article as: K. Yamamoto, et al., Pelvic fractures after definitive and postoperative radiotherapy for cervical cancer: A retrospective analysis of risk factors, Gynecol Oncol (2017), https://doi.org/10.1016/j.ygyno.2017.09.035