ORIGINAL ARTICLE
Outcomes of Definitive Radiation Therapy for Primary Vaginal Carcinoma Julie C. Greenwalt, MD,* Robert J. Amdur, MD,* Christopher G. Morris, MS,* Linda S. Morgan, MD,w Jacqueline Castagno, MD,w Merry Jennifer Markham, MD,z Shayna Rich, MD, PhD,* and Anamaria R. Yeung, MD*
Objective(s): The aim of this study was to review treatment and outcomes of patients with primary vaginal cancer treated with definitive radiotherapy. Materials and Methods: We retrospectively reviewed medical records of 71 patients with primary vaginal adenocarcinoma or squamous cell carcinoma treated with definitive radiotherapy with at least 2 years of follow-up (median follow-up, 6.24 y). Results: Ninety-three percent of patients were treated with externalbeam radiotherapy plus brachytherapy (median dose, 7540 cGy); 4 patients with stage I disease and 1 patient with stage II disease were treated with brachytherapy alone (median dose, 6000 cGy). The causespecific 5- and 10-year survival rates, respectively, were 96% and 96% for stage I patients, 75% and 68% for stage II patients, 69% and 64% for stage III patients, and 53% and 53% for stage IVA patients. The 5and 10-year local-regional control rates for all patients were 79% and 75%, respectively. The 5- and 10-year distant metastasis-free survival rates for all patients were 87% and 85%, respectively. Sixteen patients had tumors involving the distal one third of the vagina. Of the 7 who received elective inguinal node irradiation, 0 failed in the inguinal nodes. Of the 9 who did not receive elective inguinal node irradiation, 2 failed in the inguinal nodes. Severe complications (grades 3 to 4) occurred in 16 patients (23%). Conclusions: Radiotherapy provides excellent results as definitive treatment for primary vaginal cancer, although the risk of severe complications is high. Generally, treatment should consist of both external-beam radiation therapy and brachytherapy. Inguinal nodes should be irradiated electively when the primary tumor involves the distal one third of the vagina. Key Words: definitive radiotherapy, brachytherapy, vagina, outcomes
(Am J Clin Oncol 2015;38:583–587)
P
adenocarcinoma being the second-most prevalent histologic subtype (14% to 18%).4 Approximately 40% of cases of vaginal SCC are associated with human papilloma virus. A link to diethylstilbestrol exposure in utero has been associated with vaginal adenocarcinoma for many years.4–8 Although there is certainly a role for surgery in the treatment of primary vaginal cancers, especially in early-stage disease, definitive radiotherapy has been the mainstay of treatment.9,10 Optimizing treatment to increase tumor control and survival while minimizing toxicity has been difficult because of the rarity of this disease. No randomized clinical trials have been performed; therefore, institutional retrospective data have been essential in better understanding patterns of failure and determining optimal treatment. With this in mind, the purpose of this study was to review our experience treating primary vaginal cancer with definitive radiotherapy.
MATERIALS AND METHODS The records of 115 patients with vaginal carcinoma treated with radiotherapy between August 1965 and March 2008 at the University of Florida were reviewed under institutional review board approval. The inclusion criteria were as follows: primary SCC or adenocarcinoma of the vagina, treatment with curativeintent radiotherapy, and a minimum of 2 years of follow-up. Patients with prior history of carcinoma of the cervix or endometrium treated within 15 years of the diagnosis of vaginal cancer and patients who underwent surgery before radiation to debulk their disease were excluded. Fifteen patients were excluded because of a history of gynecologic malignancy, 2 patients were excluded because of treatment with external-beam radiotherapy (EBRT) without brachytherapy, 8 patients were excluded with vaginal carcinoma in situ disease, 6 patients were excluded because they underwent surgery as their primary treatment, 5 patients were excluded because of histologies other than SCC or adenocarcinoma, and 8 patients were excluded because they were lost to follow-up. The 2 EBRT-alone patients were excluded because they were unable to complete the planned course of curative treatment because of patient-specific factors. This left 71 patients whose data were used in this analysis. Patient characteristics are summarized in Table 1.
rimary vaginal carcinoma is an extremely rare diagnosis, representing only 1% to 2% of all gynecologic malignancies.1 The International Federation of Gynecology and Obstetrics declared vaginal carcinoma as a diagnosis of exclusion once cervix, urethra, and vulva have been ruled out as primary sites.2 Approximately 2890 new cases of primary vaginal cancer are diagnosed every year in the United States and 840 deaths occur as a result.3 The most common histology is squamous cell carcinoma (SCC; 75% to 85%), with
Radiotherapy Details
From the Departments of *Radiation Oncology; wObstetrics and Gynecology; and zDivision of Hematology & Oncology, College of Medicine, University of Florida, Gainesville, FL. The authors declare no conflicts of interest. Reprints: Anamaria R. Yeung, MD, 2000 SW Archer Rd., P.O. Box 100385, Gainesville, FL 32610-0385. E-mail: [email protected]. Copyright r 2013 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0277-3732/15/3806-0583 DOI: 10.1097/COC.0000000000000002
All patients were treated with definitive radiotherapy. The majority of patients (N = 66; 93%) were treated with a combination of EBRT and brachytherapy. The remaining 5 patients (7%) were treated with brachytherapy alone. No patient received chemotherapy as part of their definitive treatment. In patients receiving EBRT plus brachytherapy, the median dose to the pelvis was 5000 cGy with the external beam and 2540 cGy delivered with brachytherapy, for a total median dose of 7540 cGy. The median dose per fraction for
American Journal of Clinical Oncology
Volume 38, Number 6, December 2015
www.amjclinicaloncology.com |
Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.
583
Greenwalt et al
American Journal of Clinical Oncology
TABLE 1. Patient Characteristics
Characteristics Patient characteristics Age at presentation, years (range) Race White Black Smoking status Nonsmoker Smoker Unknown Disease characteristics at presentation RGO stage I II III IVA Positive nodes None Inguinal Regional Unknown Tumor characteristics at presentation Histology Squamous cell carcinoma Adenocarcinoma Adenosquamous Grade Poorly differentiated Moderately differentiated Well differentiated Unknown Tumor size Proximal one third Middle one third Distal one third Entire vagina Tumor size (cm) Tumor thickness r5 mm > 5 mm Unknown
Values, n (%) 64 (32-87) 60 (82) 11 (18) 21 (30) 22 (32) 27 (38) 24 17 25 5
(34) (24) (35) (7)
52 13 5 1
(73) (18) (7) (2)
61 (86) 9 (13) 1 (1) 21 17 13 20
(30) (24) (18) (14)
37 4 19 11 3
(52) (6) (27) (15) (0.1-7)
6 (8) 54 (76) 11 (16)
EBRT was 167 cGy (120 to 189 cGy). The majority of patients were treated with the 4-field box technique; however, 8 patients (12%) were treated with anterior-posterior/posterioranterior fields. The beam energy was cobalt-60 in about half the patients and Z6 MV photons in the other half. The majority (79%) of implants were interstitial with Syed templates or were performed by freehand needle placement. Only 21% of the implants were performed using an endocavitary applicator, either tandem and ovoid or tandem and vaginal cylinder. Of the endocavitary implants, 10 used a tandem and vaginal cylinder applicator, whereas 6 used a tandem and ovoid applicator. Sixty-three patients had a single implant, all but one of which was low dose rate (LDR). Eight patients had 2 implants, all of which were LDR. Five patients were treated with brachytherapy alone, with a median dose of 6000 cGy. One of these patients had stage II disease, but the other 4 patients had stage I disease. Three patients had 1 implant, and 2 patients had 2 implants. One patient received endocavitary implants; all others received interstitial implants, usually using a template. All implants were LDR. Patients were followed up every 3 to 6 months with serial pelvic examinations or as often as possible. Computed tomography/positron emission tomography imaging was not
Volume 38, Number 6, December 2015
performed routinely at this time. The median follow-up time for all patients was 6.3 years (range, 0.4 to 35 y). The median follow-up period for 21 patients alive at last contact was 9.6 years.
Statistical Methods SAS and JMP software were used for all statistical calculations (SAS Institute, Cary, NC). The Kaplan-Meier product-limit method provided estimates of overall survival, cause-specific survival (CSS), local control (LC), localregional control (LRC), and freedom from distant metastasis.11 Stratification of these endpoints by levels of selected prognostic factors was evaluated with the log-rank test statistic.
RESULTS Survival Figure 1 shows the Kaplan-Meir curves for LC, LRC, distant metastasis-free survival, CSS, and overall survival (OS). The 5- and 10-year survival rates are listed in Figure 1. The distant metastasis-free survival rate was 87% at 5 years and 85% at 10 years, showing that fairly few patients developed metastases after treatment, and most of them did so within the first 5 years. The LC rate was 83% at 5 years and 79% at 10 years. The LRC rate, which includes pelvic and inguinal nodal failures, was 79% at 5 years and 75% at 10 years. The CSS rate was 78% at 5 years and 73% at 10 years, suggesting that when LRC is achieved, long-term cure is attained. Figure 2 shows the CSS stratified by stage. Our data reveal that stage I patients do extremely well, with a 5-year CSS rate of 96% compared with 70% to 75% for stage II and III patients. Moreover, about 85% of relapses occurred in the first 2 years after treatment.
Patterns of Failure Table 2 lists the site of first failure stratified by stage and treatment. Of the patients treated with brachytherapy alone, only 1 of the 5 patients developed a recurrence. The recurrence occurred in a stage I patient treated with an interstitial implant to 6000 cGy, who 6 months later failed in an inguinal node. The other 4 patients treated with brachytherapy alone had no documented failures. The majority of patients (93%) were treated with EBRT plus brachytherapy. There were 20 stage 1 patients treated with these combined modalities; 18 (90%) were successfully treated and had no recurrent disease. One patient had a local failure and another patient failed in the inguinal nodes. Of the 16 stage II patients, only 63% were cured with no recurrent disease. More than a quarter (31%) of these stage II patients had recurrent local disease, and 1 patient had a distant relapse. Of the 25 stage III patients, 60% were cured with no recurrent disease. Nearly a quarter (24%) of the stage III patients failed locally, whereas 4 patients (16%) failed distantly. Of the 5 stage IV patients, 3 (60%) were cured with no recurrent disease, whereas one had a local failure and another failed with distant metastasis.
Elective Nodal Treatment Patients were divided in this study into those who received elective nodal treatment and those who did not, as seen in Table 3. Fourteen patients in this study presented with inguinal node involvement and were excluded from this subset evaluation. Elective nodal treatment in this study was defined as treatment of the inguinal nodes during the primary RT course when the nodes were clinically negative at presentation.
584 | www.amjclinicaloncology.com Copyright r 2013 Wolters Kluwer Health, Inc. All rights reserved. Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.
American Journal of Clinical Oncology
Volume 38, Number 6, December 2015
Definitive RT for Vaginal Carcinoma
TABLE 2. The Sites of First Failure
Stage I II
FIGURE 1. The Kaplan-Meier survival curves for local control, local-regional control, distant metastasis-free survival, causespecific survival, and overall survival.
Clinically negative was defined as no palpable lymphadenopathy that was also negative on imaging, when available. In total, 19 patients in our study received elective nodal irradiation. Seven of these patients presented with vaginal tumors in the distal one third of the vagina. Of these 19 patients who received elective inguinal node irradiation, none of them failed in the inguinal nodes. There were 38 patients who did not receive elective nodal treatment, 9 of whom had tumors in the distal one third of the vagina. Of these 38 patients who did not receive elective inguinal nodal irradiation, 2 patients (representing 5%) failed in the inguinal nodes. The 2 patients who failed in the inguinal nodes did in fact have tumors in the distal one third of the vagina. One of these patients was treated with brachytherapy alone (an interstitial implant to 6000 cGy) and the other was treated with EBRT at an outside facility to 4500 cGy and then treated at our institution only for brachytherapy. Both of their inguinal node relapses occurred within 6 months of finishing treatment.
Univariate Analysis Univariate analysis was performed on factors thought to have clinical relevance, including stage, histology, tumor differentiation, race, tumor site, smoker, and dose. The only statistically significant factors with a P value < 0.05 were histology for CSS and stage for age (r64 vs. > 64 y) for OS. In terms of histology, adenocarcinoma had a 38% CSS rate at 5 years, whereas SCC had a 78% CSS rate at 5 years. Age was a significant factor in terms of OS with patients aged 64 years or
External-beam Radiotherapy + Brachytherapyw (N = 66)
Brachytherapy Alone* (N = 5) 1 (25%) inguinal node 3 (75%) none 1 (100%) none
III
—
IV
—
1 (5%) local 1 (5%) inguinal node 18 (90%) none 5 (31%) local 1 (6%) metastasis 10 (63%) none 6 (24%) local 4 (16%) metastasis 15 (60%) none 1 (20%) local 1 (20%) metastasis 3 (60%) none
*Median dose for brachytherapy alone group was 6000 cGy. wMedian dose for external-beam radiotherapy + brachytherapy group was 7540 cGy.
younger having an OS rate of 74% at 5 years; patients older than 64 years of age had an OS rate of only 56%. In terms of stage, the more advanced the stage at diagnosis, the lower the OS rate, with stage I at 77%, stage II at 69%, stage III at 55%, and stage IV at 40% at 5 years.
Complications In this study, complications were graded per the National Cancer Institute’s Common Terminology Criteria for Adverse Events, version 4.0. The rate of grade 3 + toxicity was 23%. There were no grade 5 complications. In total, 12 patients (17%) experienced grade 3 complications requiring treatment: 4 patients with soft tissue necrosis of the vagina, 3 patients with bowel incontinence, 3 patients with radiation proctitis, 1 patient with radiation osteitis, and 1 patient who developed a urethral stricture. Four patients (5%) developed grade 4 complications: 1 patient had a urethral stricture that had to be surgically repaired, 1 patient developed a deep venous thrombosis while undergoing treatment and underwent a below-the-knee amputation, 1 patient developed osteoradionecrosis of the hip that led to internal fixation surgery, and 1 patient had soft tissue necrosis that required cryosurgery. There was no difference in complication rates between patients treated in earlier years with cobalt-60 as compared with patients in the non-cobalt group (patients treated with energies Z6 MV photons). There were two grade 4 complications in both the cobalt-60 group and the non-cobalt group. Seven out of 41 patients had grade 3 complications in the noncobalt group whereas 5 out of 30 patients had grade 3 complications in the cobalt group. TABLE 3. Elective Inguinal Node Irradiation and Inguinal Failures
Primary tumor in distal 1/3 vagina Inguinal node failure
Patients Receiving Elective Inguinal Node RT, N = 19
Patients Not Receiving Elective Inguinal Node RT, N = 38
7
9
0
2*
*Both patients had primary tumors involving the distal 1/3 of vagina. RT indicates radiotherapy.
FIGURE 2. Cause-specific survival rates by stage.
Copyright
r
2013 Wolters Kluwer Health, Inc. All rights reserved.
www.amjclinicaloncology.com |
Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.
585
American Journal of Clinical Oncology
Greenwalt et al
DISCUSSION This report describes the University of Florida’s experience treating primary vaginal cancer with definitive radiotherapy over the last 30 years. Our standard of care has been to treat vaginal cancer patients with EBRT plus brachytherapy, using brachytherapy alone for patients with limitations to a prolonged treatment course and reserving surgery for recurrences or complications. Our results show good control with radiation therapy alone. Few patients developed metastases after treatment, and LC, LRC, and CSS were all nearly identical at 5 and 10 years, suggesting that long-term cure is attained if LRC is achieved. Recurrence generally occurred within 5 years of treatment but usually occurred within 2 years. No treatment complications in this study led to death. Our results demonstrate that primary radiotherapy for vaginal cancer is an excellent modality of definitive treatment, especially in early-stage disease. As in similar studies (Table 4), the overwhelming majority of patients with stage I disease were cured with radiotherapy. This finding is consistent with a study of 212 patients with primary vaginal cancer treated at Washington University (St. Louis, MO), which also concluded that radiotherapy is a definitive treatment for stage I disease.10 Furthermore, however, it appears that brachytherapy alone is insufficient to provide definitive treatment. In our study, 1 of 4 patients treated with brachytherapy alone developed recurrence in the inguinal nodes after a single interstitial implant, whereas 3 of 15 of the Washington University patients treated with intracavitary implants alone developed recurrence in the pelvis. In the Washington University series, the percentage of patients cured with radiotherapy decreased by stage at diagnosis, with less than half of patients with stage III/IV disease being cured.10 A substantial number of patients who presented with stage II-IV disease failed locally despite treatment with EBRT plus brachytherapy, demonstrating the need to increase the effectiveness of radiotherapy for local disease. That study also concluded that local and pelvic failures were frequent for stage II and III patients treated with EBRT. Thus, improved radiation treatment should be investigated for patients presenting with more advanced disease, and may include dose escalation, wider fields, or improved brachytherapy techniques.9 Another consideration is the addition of concurrent chemotherapy.13,14 Although there is no phase 3 data to support chemotherapy for the treatment of vaginal cancer, we currently add weekly cisplatin during EBRT based on the evidence of its effectiveness in cervix cancer. The present study also supports electively irradiating the inguinal nodes in patients with tumors in the distal one third of the vagina. None of the patients who received elective nodal treatment failed in the inguinal nodes, whereas 2 who did not
Volume 38, Number 6, December 2015
receive elective nodal treatment subsequently developed recurrence in the groin. Both patients had primary tumors in the distal one third of the vagina. Our results are consistent with prior studies showing that elective inguinal node treatment is beneficial in pelvic carcinomas15 and reinforces the need to irradiate regional lymphatic channels at risk. In particular, the proximal two thirds of the vagina usually drains to the pelvic lymph nodes covered in a standard treatment field, whereas the distal one third of the vagina usually drains to inguinal lymph nodes and requires elective inguinal node radiation to prevent local-regional recurrences. In our study, stage was associated with OS but not CSS, likely because higher stages (II to IV) present with more extensive disease but appear to recur at similar rates locally, regionally, and distantly. Histology was not associated with OS, but adenocarcinoma had an increased risk of recurrence compared with SCCA. Younger patients had a better prognosis by OS, consistent with other studies of primary vaginal cancer.16,17 Patient race, smoking status, tumor site, tumor grade, and total treatment dose were not associated with OS or CSS, although others have proposed that race can affect stage at diagnosis and that smoking is associated with an increased risk of gynecologic cancers such as cervical cancer.18 There are several limitations to this study. One major limitation is the significant loss to follow-up, likely because of the fact that our institution is a tertiary referral center and most of our patients travel long distances for treatment. Thus, yearly follow-up is logistically difficult for many patients, and local gynecologists or radiation oncologists often perform follow-up. We attempted to obtain follow-up information from local physicians as much as was possible. This study is also limited because of lack of imaging studies at diagnosis, as computed tomography, positron emission tomography, and magnetic resonance imaging were not routinely available when many of the patients were treated. This imaging significantly improves the ability to identify and treat the full extent of disease, and it would likely provide more precise clinical staging compared with physical examination, which can vary by physician. It is challenging to improve treatment of vaginal cancer, given a lack of randomized clinical trials comparing treatment modalities directly. Improvements have relied instead on institutional retrospective studies highlighting problems with treatment and advances at individual cancer centers. Given the results of this study, the standard treatment used at our institution for all stages of vaginal cancer is EBRT plus brachytherapy with a total dose goal of 75 to 80 Gy. EBRT is given with doses ranging from 45 to 50.4 Gy in 1.8 Gy fractions to the pelvis. The inguinal nodes are treated if the distal one third of the vagina is involved with tumor or if the nodes are positive
TABLE 4. Five-year Disease-free Survival (DFS) Rates in the Literature by Stage
Study MD Anderson Cancer Center (Frank et al12w) Washington University (Perez et al10) Indiana University (Sinha et al13z) University of Florida (Greenwalt and colleagues), This Study
N
5-y DFS, Stage I (%)
193 212*
85 B78
45* 71
92 87
5-y DFS, Stage II (%) 78 B55 (IIA) B38 (IIB) 88 71
5-y DFS, Stage III/IVA (%) 58 B40 (III) < 20 (IVA) 30 67 (III) 60 (IVA)
*Carcinoma in situ patients included. wDisease-specific survival. zProgression-free survival.
586 | www.amjclinicaloncology.com Copyright r 2013 Wolters Kluwer Health, Inc. All rights reserved. Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.
American Journal of Clinical Oncology
Volume 38, Number 6, December 2015
on presentation. Brachytherapy is determined based on the depth of tumor invasion in the vaginal cavity at the time of the implant: if the depth is
Outcomes of Definitive Radiation Therapy for Primary Vaginal Carcinoma Julie C. Greenwalt, MD,* Robert J. Amdur, MD,* Christopher G. Morris, MS,* Linda S. Morgan, MD,w Jacqueline Castagno, MD,w Merry Jennifer Markham, MD,z Shayna Rich, MD, PhD,* and Anamaria R. Yeung, MD*
Objective(s): The aim of this study was to review treatment and outcomes of patients with primary vaginal cancer treated with definitive radiotherapy. Materials and Methods: We retrospectively reviewed medical records of 71 patients with primary vaginal adenocarcinoma or squamous cell carcinoma treated with definitive radiotherapy with at least 2 years of follow-up (median follow-up, 6.24 y). Results: Ninety-three percent of patients were treated with externalbeam radiotherapy plus brachytherapy (median dose, 7540 cGy); 4 patients with stage I disease and 1 patient with stage II disease were treated with brachytherapy alone (median dose, 6000 cGy). The causespecific 5- and 10-year survival rates, respectively, were 96% and 96% for stage I patients, 75% and 68% for stage II patients, 69% and 64% for stage III patients, and 53% and 53% for stage IVA patients. The 5and 10-year local-regional control rates for all patients were 79% and 75%, respectively. The 5- and 10-year distant metastasis-free survival rates for all patients were 87% and 85%, respectively. Sixteen patients had tumors involving the distal one third of the vagina. Of the 7 who received elective inguinal node irradiation, 0 failed in the inguinal nodes. Of the 9 who did not receive elective inguinal node irradiation, 2 failed in the inguinal nodes. Severe complications (grades 3 to 4) occurred in 16 patients (23%). Conclusions: Radiotherapy provides excellent results as definitive treatment for primary vaginal cancer, although the risk of severe complications is high. Generally, treatment should consist of both external-beam radiation therapy and brachytherapy. Inguinal nodes should be irradiated electively when the primary tumor involves the distal one third of the vagina. Key Words: definitive radiotherapy, brachytherapy, vagina, outcomes
(Am J Clin Oncol 2015;38:583–587)
P
adenocarcinoma being the second-most prevalent histologic subtype (14% to 18%).4 Approximately 40% of cases of vaginal SCC are associated with human papilloma virus. A link to diethylstilbestrol exposure in utero has been associated with vaginal adenocarcinoma for many years.4–8 Although there is certainly a role for surgery in the treatment of primary vaginal cancers, especially in early-stage disease, definitive radiotherapy has been the mainstay of treatment.9,10 Optimizing treatment to increase tumor control and survival while minimizing toxicity has been difficult because of the rarity of this disease. No randomized clinical trials have been performed; therefore, institutional retrospective data have been essential in better understanding patterns of failure and determining optimal treatment. With this in mind, the purpose of this study was to review our experience treating primary vaginal cancer with definitive radiotherapy.
MATERIALS AND METHODS The records of 115 patients with vaginal carcinoma treated with radiotherapy between August 1965 and March 2008 at the University of Florida were reviewed under institutional review board approval. The inclusion criteria were as follows: primary SCC or adenocarcinoma of the vagina, treatment with curativeintent radiotherapy, and a minimum of 2 years of follow-up. Patients with prior history of carcinoma of the cervix or endometrium treated within 15 years of the diagnosis of vaginal cancer and patients who underwent surgery before radiation to debulk their disease were excluded. Fifteen patients were excluded because of a history of gynecologic malignancy, 2 patients were excluded because of treatment with external-beam radiotherapy (EBRT) without brachytherapy, 8 patients were excluded with vaginal carcinoma in situ disease, 6 patients were excluded because they underwent surgery as their primary treatment, 5 patients were excluded because of histologies other than SCC or adenocarcinoma, and 8 patients were excluded because they were lost to follow-up. The 2 EBRT-alone patients were excluded because they were unable to complete the planned course of curative treatment because of patient-specific factors. This left 71 patients whose data were used in this analysis. Patient characteristics are summarized in Table 1.
rimary vaginal carcinoma is an extremely rare diagnosis, representing only 1% to 2% of all gynecologic malignancies.1 The International Federation of Gynecology and Obstetrics declared vaginal carcinoma as a diagnosis of exclusion once cervix, urethra, and vulva have been ruled out as primary sites.2 Approximately 2890 new cases of primary vaginal cancer are diagnosed every year in the United States and 840 deaths occur as a result.3 The most common histology is squamous cell carcinoma (SCC; 75% to 85%), with
Radiotherapy Details
From the Departments of *Radiation Oncology; wObstetrics and Gynecology; and zDivision of Hematology & Oncology, College of Medicine, University of Florida, Gainesville, FL. The authors declare no conflicts of interest. Reprints: Anamaria R. Yeung, MD, 2000 SW Archer Rd., P.O. Box 100385, Gainesville, FL 32610-0385. E-mail: [email protected]. Copyright r 2013 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0277-3732/15/3806-0583 DOI: 10.1097/COC.0000000000000002
All patients were treated with definitive radiotherapy. The majority of patients (N = 66; 93%) were treated with a combination of EBRT and brachytherapy. The remaining 5 patients (7%) were treated with brachytherapy alone. No patient received chemotherapy as part of their definitive treatment. In patients receiving EBRT plus brachytherapy, the median dose to the pelvis was 5000 cGy with the external beam and 2540 cGy delivered with brachytherapy, for a total median dose of 7540 cGy. The median dose per fraction for
American Journal of Clinical Oncology
Volume 38, Number 6, December 2015
www.amjclinicaloncology.com |
Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.
583
Greenwalt et al
American Journal of Clinical Oncology
TABLE 1. Patient Characteristics
Characteristics Patient characteristics Age at presentation, years (range) Race White Black Smoking status Nonsmoker Smoker Unknown Disease characteristics at presentation RGO stage I II III IVA Positive nodes None Inguinal Regional Unknown Tumor characteristics at presentation Histology Squamous cell carcinoma Adenocarcinoma Adenosquamous Grade Poorly differentiated Moderately differentiated Well differentiated Unknown Tumor size Proximal one third Middle one third Distal one third Entire vagina Tumor size (cm) Tumor thickness r5 mm > 5 mm Unknown
Values, n (%) 64 (32-87) 60 (82) 11 (18) 21 (30) 22 (32) 27 (38) 24 17 25 5
(34) (24) (35) (7)
52 13 5 1
(73) (18) (7) (2)
61 (86) 9 (13) 1 (1) 21 17 13 20
(30) (24) (18) (14)
37 4 19 11 3
(52) (6) (27) (15) (0.1-7)
6 (8) 54 (76) 11 (16)
EBRT was 167 cGy (120 to 189 cGy). The majority of patients were treated with the 4-field box technique; however, 8 patients (12%) were treated with anterior-posterior/posterioranterior fields. The beam energy was cobalt-60 in about half the patients and Z6 MV photons in the other half. The majority (79%) of implants were interstitial with Syed templates or were performed by freehand needle placement. Only 21% of the implants were performed using an endocavitary applicator, either tandem and ovoid or tandem and vaginal cylinder. Of the endocavitary implants, 10 used a tandem and vaginal cylinder applicator, whereas 6 used a tandem and ovoid applicator. Sixty-three patients had a single implant, all but one of which was low dose rate (LDR). Eight patients had 2 implants, all of which were LDR. Five patients were treated with brachytherapy alone, with a median dose of 6000 cGy. One of these patients had stage II disease, but the other 4 patients had stage I disease. Three patients had 1 implant, and 2 patients had 2 implants. One patient received endocavitary implants; all others received interstitial implants, usually using a template. All implants were LDR. Patients were followed up every 3 to 6 months with serial pelvic examinations or as often as possible. Computed tomography/positron emission tomography imaging was not
Volume 38, Number 6, December 2015
performed routinely at this time. The median follow-up time for all patients was 6.3 years (range, 0.4 to 35 y). The median follow-up period for 21 patients alive at last contact was 9.6 years.
Statistical Methods SAS and JMP software were used for all statistical calculations (SAS Institute, Cary, NC). The Kaplan-Meier product-limit method provided estimates of overall survival, cause-specific survival (CSS), local control (LC), localregional control (LRC), and freedom from distant metastasis.11 Stratification of these endpoints by levels of selected prognostic factors was evaluated with the log-rank test statistic.
RESULTS Survival Figure 1 shows the Kaplan-Meir curves for LC, LRC, distant metastasis-free survival, CSS, and overall survival (OS). The 5- and 10-year survival rates are listed in Figure 1. The distant metastasis-free survival rate was 87% at 5 years and 85% at 10 years, showing that fairly few patients developed metastases after treatment, and most of them did so within the first 5 years. The LC rate was 83% at 5 years and 79% at 10 years. The LRC rate, which includes pelvic and inguinal nodal failures, was 79% at 5 years and 75% at 10 years. The CSS rate was 78% at 5 years and 73% at 10 years, suggesting that when LRC is achieved, long-term cure is attained. Figure 2 shows the CSS stratified by stage. Our data reveal that stage I patients do extremely well, with a 5-year CSS rate of 96% compared with 70% to 75% for stage II and III patients. Moreover, about 85% of relapses occurred in the first 2 years after treatment.
Patterns of Failure Table 2 lists the site of first failure stratified by stage and treatment. Of the patients treated with brachytherapy alone, only 1 of the 5 patients developed a recurrence. The recurrence occurred in a stage I patient treated with an interstitial implant to 6000 cGy, who 6 months later failed in an inguinal node. The other 4 patients treated with brachytherapy alone had no documented failures. The majority of patients (93%) were treated with EBRT plus brachytherapy. There were 20 stage 1 patients treated with these combined modalities; 18 (90%) were successfully treated and had no recurrent disease. One patient had a local failure and another patient failed in the inguinal nodes. Of the 16 stage II patients, only 63% were cured with no recurrent disease. More than a quarter (31%) of these stage II patients had recurrent local disease, and 1 patient had a distant relapse. Of the 25 stage III patients, 60% were cured with no recurrent disease. Nearly a quarter (24%) of the stage III patients failed locally, whereas 4 patients (16%) failed distantly. Of the 5 stage IV patients, 3 (60%) were cured with no recurrent disease, whereas one had a local failure and another failed with distant metastasis.
Elective Nodal Treatment Patients were divided in this study into those who received elective nodal treatment and those who did not, as seen in Table 3. Fourteen patients in this study presented with inguinal node involvement and were excluded from this subset evaluation. Elective nodal treatment in this study was defined as treatment of the inguinal nodes during the primary RT course when the nodes were clinically negative at presentation.
584 | www.amjclinicaloncology.com Copyright r 2013 Wolters Kluwer Health, Inc. All rights reserved. Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.
American Journal of Clinical Oncology
Volume 38, Number 6, December 2015
Definitive RT for Vaginal Carcinoma
TABLE 2. The Sites of First Failure
Stage I II
FIGURE 1. The Kaplan-Meier survival curves for local control, local-regional control, distant metastasis-free survival, causespecific survival, and overall survival.
Clinically negative was defined as no palpable lymphadenopathy that was also negative on imaging, when available. In total, 19 patients in our study received elective nodal irradiation. Seven of these patients presented with vaginal tumors in the distal one third of the vagina. Of these 19 patients who received elective inguinal node irradiation, none of them failed in the inguinal nodes. There were 38 patients who did not receive elective nodal treatment, 9 of whom had tumors in the distal one third of the vagina. Of these 38 patients who did not receive elective inguinal nodal irradiation, 2 patients (representing 5%) failed in the inguinal nodes. The 2 patients who failed in the inguinal nodes did in fact have tumors in the distal one third of the vagina. One of these patients was treated with brachytherapy alone (an interstitial implant to 6000 cGy) and the other was treated with EBRT at an outside facility to 4500 cGy and then treated at our institution only for brachytherapy. Both of their inguinal node relapses occurred within 6 months of finishing treatment.
Univariate Analysis Univariate analysis was performed on factors thought to have clinical relevance, including stage, histology, tumor differentiation, race, tumor site, smoker, and dose. The only statistically significant factors with a P value < 0.05 were histology for CSS and stage for age (r64 vs. > 64 y) for OS. In terms of histology, adenocarcinoma had a 38% CSS rate at 5 years, whereas SCC had a 78% CSS rate at 5 years. Age was a significant factor in terms of OS with patients aged 64 years or
External-beam Radiotherapy + Brachytherapyw (N = 66)
Brachytherapy Alone* (N = 5) 1 (25%) inguinal node 3 (75%) none 1 (100%) none
III
—
IV
—
1 (5%) local 1 (5%) inguinal node 18 (90%) none 5 (31%) local 1 (6%) metastasis 10 (63%) none 6 (24%) local 4 (16%) metastasis 15 (60%) none 1 (20%) local 1 (20%) metastasis 3 (60%) none
*Median dose for brachytherapy alone group was 6000 cGy. wMedian dose for external-beam radiotherapy + brachytherapy group was 7540 cGy.
younger having an OS rate of 74% at 5 years; patients older than 64 years of age had an OS rate of only 56%. In terms of stage, the more advanced the stage at diagnosis, the lower the OS rate, with stage I at 77%, stage II at 69%, stage III at 55%, and stage IV at 40% at 5 years.
Complications In this study, complications were graded per the National Cancer Institute’s Common Terminology Criteria for Adverse Events, version 4.0. The rate of grade 3 + toxicity was 23%. There were no grade 5 complications. In total, 12 patients (17%) experienced grade 3 complications requiring treatment: 4 patients with soft tissue necrosis of the vagina, 3 patients with bowel incontinence, 3 patients with radiation proctitis, 1 patient with radiation osteitis, and 1 patient who developed a urethral stricture. Four patients (5%) developed grade 4 complications: 1 patient had a urethral stricture that had to be surgically repaired, 1 patient developed a deep venous thrombosis while undergoing treatment and underwent a below-the-knee amputation, 1 patient developed osteoradionecrosis of the hip that led to internal fixation surgery, and 1 patient had soft tissue necrosis that required cryosurgery. There was no difference in complication rates between patients treated in earlier years with cobalt-60 as compared with patients in the non-cobalt group (patients treated with energies Z6 MV photons). There were two grade 4 complications in both the cobalt-60 group and the non-cobalt group. Seven out of 41 patients had grade 3 complications in the noncobalt group whereas 5 out of 30 patients had grade 3 complications in the cobalt group. TABLE 3. Elective Inguinal Node Irradiation and Inguinal Failures
Primary tumor in distal 1/3 vagina Inguinal node failure
Patients Receiving Elective Inguinal Node RT, N = 19
Patients Not Receiving Elective Inguinal Node RT, N = 38
7
9
0
2*
*Both patients had primary tumors involving the distal 1/3 of vagina. RT indicates radiotherapy.
FIGURE 2. Cause-specific survival rates by stage.
Copyright
r
2013 Wolters Kluwer Health, Inc. All rights reserved.
www.amjclinicaloncology.com |
Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.
585
American Journal of Clinical Oncology
Greenwalt et al
DISCUSSION This report describes the University of Florida’s experience treating primary vaginal cancer with definitive radiotherapy over the last 30 years. Our standard of care has been to treat vaginal cancer patients with EBRT plus brachytherapy, using brachytherapy alone for patients with limitations to a prolonged treatment course and reserving surgery for recurrences or complications. Our results show good control with radiation therapy alone. Few patients developed metastases after treatment, and LC, LRC, and CSS were all nearly identical at 5 and 10 years, suggesting that long-term cure is attained if LRC is achieved. Recurrence generally occurred within 5 years of treatment but usually occurred within 2 years. No treatment complications in this study led to death. Our results demonstrate that primary radiotherapy for vaginal cancer is an excellent modality of definitive treatment, especially in early-stage disease. As in similar studies (Table 4), the overwhelming majority of patients with stage I disease were cured with radiotherapy. This finding is consistent with a study of 212 patients with primary vaginal cancer treated at Washington University (St. Louis, MO), which also concluded that radiotherapy is a definitive treatment for stage I disease.10 Furthermore, however, it appears that brachytherapy alone is insufficient to provide definitive treatment. In our study, 1 of 4 patients treated with brachytherapy alone developed recurrence in the inguinal nodes after a single interstitial implant, whereas 3 of 15 of the Washington University patients treated with intracavitary implants alone developed recurrence in the pelvis. In the Washington University series, the percentage of patients cured with radiotherapy decreased by stage at diagnosis, with less than half of patients with stage III/IV disease being cured.10 A substantial number of patients who presented with stage II-IV disease failed locally despite treatment with EBRT plus brachytherapy, demonstrating the need to increase the effectiveness of radiotherapy for local disease. That study also concluded that local and pelvic failures were frequent for stage II and III patients treated with EBRT. Thus, improved radiation treatment should be investigated for patients presenting with more advanced disease, and may include dose escalation, wider fields, or improved brachytherapy techniques.9 Another consideration is the addition of concurrent chemotherapy.13,14 Although there is no phase 3 data to support chemotherapy for the treatment of vaginal cancer, we currently add weekly cisplatin during EBRT based on the evidence of its effectiveness in cervix cancer. The present study also supports electively irradiating the inguinal nodes in patients with tumors in the distal one third of the vagina. None of the patients who received elective nodal treatment failed in the inguinal nodes, whereas 2 who did not
Volume 38, Number 6, December 2015
receive elective nodal treatment subsequently developed recurrence in the groin. Both patients had primary tumors in the distal one third of the vagina. Our results are consistent with prior studies showing that elective inguinal node treatment is beneficial in pelvic carcinomas15 and reinforces the need to irradiate regional lymphatic channels at risk. In particular, the proximal two thirds of the vagina usually drains to the pelvic lymph nodes covered in a standard treatment field, whereas the distal one third of the vagina usually drains to inguinal lymph nodes and requires elective inguinal node radiation to prevent local-regional recurrences. In our study, stage was associated with OS but not CSS, likely because higher stages (II to IV) present with more extensive disease but appear to recur at similar rates locally, regionally, and distantly. Histology was not associated with OS, but adenocarcinoma had an increased risk of recurrence compared with SCCA. Younger patients had a better prognosis by OS, consistent with other studies of primary vaginal cancer.16,17 Patient race, smoking status, tumor site, tumor grade, and total treatment dose were not associated with OS or CSS, although others have proposed that race can affect stage at diagnosis and that smoking is associated with an increased risk of gynecologic cancers such as cervical cancer.18 There are several limitations to this study. One major limitation is the significant loss to follow-up, likely because of the fact that our institution is a tertiary referral center and most of our patients travel long distances for treatment. Thus, yearly follow-up is logistically difficult for many patients, and local gynecologists or radiation oncologists often perform follow-up. We attempted to obtain follow-up information from local physicians as much as was possible. This study is also limited because of lack of imaging studies at diagnosis, as computed tomography, positron emission tomography, and magnetic resonance imaging were not routinely available when many of the patients were treated. This imaging significantly improves the ability to identify and treat the full extent of disease, and it would likely provide more precise clinical staging compared with physical examination, which can vary by physician. It is challenging to improve treatment of vaginal cancer, given a lack of randomized clinical trials comparing treatment modalities directly. Improvements have relied instead on institutional retrospective studies highlighting problems with treatment and advances at individual cancer centers. Given the results of this study, the standard treatment used at our institution for all stages of vaginal cancer is EBRT plus brachytherapy with a total dose goal of 75 to 80 Gy. EBRT is given with doses ranging from 45 to 50.4 Gy in 1.8 Gy fractions to the pelvis. The inguinal nodes are treated if the distal one third of the vagina is involved with tumor or if the nodes are positive
TABLE 4. Five-year Disease-free Survival (DFS) Rates in the Literature by Stage
Study MD Anderson Cancer Center (Frank et al12w) Washington University (Perez et al10) Indiana University (Sinha et al13z) University of Florida (Greenwalt and colleagues), This Study
N
5-y DFS, Stage I (%)
193 212*
85 B78
45* 71
92 87
5-y DFS, Stage II (%) 78 B55 (IIA) B38 (IIB) 88 71
5-y DFS, Stage III/IVA (%) 58 B40 (III) < 20 (IVA) 30 67 (III) 60 (IVA)
*Carcinoma in situ patients included. wDisease-specific survival. zProgression-free survival.
586 | www.amjclinicaloncology.com Copyright r 2013 Wolters Kluwer Health, Inc. All rights reserved. Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.
American Journal of Clinical Oncology
Volume 38, Number 6, December 2015
on presentation. Brachytherapy is determined based on the depth of tumor invasion in the vaginal cavity at the time of the implant: if the depth is
