1934 Radiation Alone for Carcinoma of the Vagina Variation in Response Related to the Location of the Primary Tumor Moinuddin M. Ali, M.D.' David T. Huang, M.D., Ph.D.' Dean R. Goplerud, M.D? Robert Howells, M.s.c.' Jian dong Lu, Ph.0.' ' Department of Radiation Oncology, Medical College of Virginia Hospitals, Virginia Commonwealth University, Richmond, Virginia. Department of Obstetrics and Gynecology, Division of Gynecology/Oncology, Medical College of Virginia Hospitals, Virginia Commonwealth University, Richmond, Virginia. BACKGROUND. A retrospective study of 40 patients with histologically confirmed carcinoma of the vagina is reported. The patients were treated by radiation alone (a combination of external beam therapy and implants) between October 1969 and September 1991 at the Medical College of Virginia Hospital in Richmond. METHODS. Thirty-three patients (82%) had squamous cell carcinoma, 2 patients (7%) had adenocarcinoma, and 2 patients (5%) had poorly differentiated cancers (1 melanoma and 1 leiomyosarcoma). The patients were staged according to the International Federation of Gynecology and Obstetrics (FIGO) staging system; there were 13 patients (33%) in Stage 1, 21 (52%) in Stage I I , 4 (10%) in Stage 111, and 2 (5%) in Stage IV. Thirty-six patients (90%)were treated with external beam therapy and some combination of implant: cylinder, ovoid, or interstitial implants with iodine-125 or iridium-192 (afterloading). Only 4 patients (10%) received treatment by implant only. RESULTS. Based on their response, two groups of patients were identified. Group I had 23 patients with tumors predominantly located in the proximal half of the vagina; there were 8 patients in Stage I, 11 in Stage 11, 3 in Stage 111, and 1 in Stage IV. Of these, three patients failed: one each in Stages I11 and IV and one Stage I1 patient was salvaged by surgery. Three patients died due to unrelated causes but with local control. The 5-year actuarial survival in this group was 81%. Group I1 had 17 patients with tumors located in the mid to distal half of the vagina; there were 5 patients in Stage I, 10 in Stage 11, and 2 in Stage IV. Ten patients failed. Eight patients in Stage I1 had persistent disease, were lost to follow-up, and are presumed dead. Two patients with Stage IV disease also had inadequate local control. The overall actuarid survival in the distal group was 41%, which was significantly worse than the proximal group (81%),at a Pvalue of 0.05. CONCLUSIONS. This study discusses the curability of carcinoma of the vagina based on its anatomic location when predominantly similar treatment techniques and radiation doses were applied to either the proximal or the distal part of the vagina, those with cancer in the proximal half had better survival (81%) than those whose cancer was in the distal half (41%). Cancer 1996; 721934-9. 0 1996American Cancer Society. KEYWORDS: vaginal carcinoma, proximal and distal radiation doses, treatment outcome, external beam therapy, implants. Presented at the 77th Annual Meeting of the American Radium Society, Paris, April 29-May 3, 1995. Address for reprints: Moinuddin M. Ali, M.D., Radiation Oncology, Medical College of Virginia Hospitals, P. 0. Box 980058, Richmond, VA 23298-0058. Received August 3, 1995; revision received November 22,1995; accepted December 20,1995. 0 1996 American Cancer Society P rimary carcinoma of the vagina accounts for I-2% of gynecologic rnalignancie~.'-~ Although extension to the vagina from tumors of the cervix or vulva is observed frequently, tumors that involve both the vagina and the cervix are classified as cervical cancers and those that involve the vagina and vulva are considered neoplasms of the vulva?v6 Local surgical excision and partial or complete vaginectomy, which used to be the mainstay of treatment, have given way to a more individualized approach that takes into consideration the patient's age and Radiation for Carcinoma of the VaginaIAli et al. 1935 TABLE 1A Patient Characteristics Variable Statistics Proximal group Distal group Group difference Race No. (%) White No. (%) Black Mean (SD) Median Median 6 (50) 6 (50) 63.2 (10.7) 64 31 17 (60.7) 11 (39.3) 65.4 (15) 70 NSa 37 pisb Age Follow-up time (rno.) NSb SD, standard deviation; NS: not significant. 'Chi-square test at level 0.05. bKmskal-Wallis test at level 0.05. TABLE 1B Patient Characteristics (cont'd) Stagea Proximal group Distal group Histology" I I1 I11 N Total I I1 111 Iv Total SCC Adeno Others 8 0 0 8 2 1 19 2 2 2 0 0 0 14 1 0 5 0 0 1 1 2 1 0 0 8 I 0 2 Total 8 11 3 1 23 5 10 0 2 17 0 SCC smallcell carcinoma; adeno: adenocarcinoma. 'No significant difference between groups at level 0.05 by chi-square test. the extent of the lesion (whether it is localized or multicentric). There is not enough data in the literature regarding the curability of primary carcinoma of the vagina based on location (proximal versus distal). In young patients with early stage disease, treatment also depends on the desire to preserve a functional vagina. In most patients, the major treatment modality is Intracavitary and interstitial radiation is used in small superficial Stage I disease, a combination of external beam therapy and interstitial implants is used in extensive Stage I and Stage I1 disease, and a combination of internal and external radiation is used in Stages I11 and IV disease. Survival rates with this malignancy are approximately 65-90% for Stage I, 40-66% for Stage 11, and 20-50% for Stages 111 and IV.' This study discusses the curability or failure pattern of carcinoma of the vagina. Based on our study, there appears to be a difference in the local control of the tumor at different locations (proximal versus distal vagina) when similar treatment modalities and radiation doses are applied. MATERIALS AND METHODS This is a retrospective review of the charts of 40 patients with primary carcinoma of the vagina treated at the Department of Radiation Oncology at the Medical College of Virginia Hospital in Richmond during the period October 1969 through September 1991. The mean follow-up time was 45.6 months (median, 33.2 months; range, 13-155.2 months). Although these patients had confirmed invasive carcinoma of the vagina, four patients who were treated for adenocarcinoma of the endometrium more than 5 years earlier and had a different histology were included in this series. The patients were staged according to the International Federation of Gynecology and Obstetrics (FIGO) classification." Management was individualized based on the tumor size, location, and stage as well as the patient's age and general health. All patients were treated with radiotherapy. Table 2 summarizes the various radiation treatment modalities utilized. Different treatment techniques were used in the management of these patients because of the various time periods involved in this study. Histologically, 93-95% of these ma- 1936 CANCER May 1,1996 / Volume 77 / Number 9 ~ _ _ _ _ _ _ _ _23_ _ . _ _ _ _ _ _ . _ 04 0.5 02 01 LI t 1 II p=.o19 01 0 35 72 108 144 180 Time (months) FIGURE 2. Overall survival of patients by group. FIGURE 1. Age distribution of patients by group lignancies were classified as squamous cell carcinoma. Other types included adenocarcinoma, papillary and clear cell variety of adenocarcinoma, melanoma, and undifferentiated carcinoma. There were 12 patients in the proximal group with an equal racial division. There were 17 whites and 11 African-Americans in the distal tumor group, which was not statistically significant (Table 1). The age distribution is approximately the same for all the stages and also for Groups I & 11. Greater than 70% of the patients were older than 60 years of age. (Fig. 1) There was no statistical significance between the age and the survival of patients in the proximal versus the distal tumor groups (Table 1). The exact etiology of carcinoma of the vagina remains unknown. However, it is observed frequently during the sixth and the seventh decades of life and women who have had hysterectomies are believed to be at a higher risk." Twenty-two of 23 patients (96%) in the proximal group (Group I) and 14 of 17 patients (82%) in the distal tumor group (Group 11) were treated with external beam therapy with some combination of implants. Only a small percentage of patients (1 of 23 [Group I] and 3 of 17 [Group 111) were treated with implants alone without the use of external beam therapy. (Table 2) In patients with proximal lesions and surgically absent uteri, ovoid or cylinder implants were used unless the lesions were extensive or thick, in which case an interstitial implant with iridium- 192 (afterloading technique) was utilized. In the majority of patients, external pelvic irradiation was delivered using megavoltage radiation (cobalt-60,6 or 10 megavolt linear accelerator). The isotopes used were iridium-192 for temporary implants and iodine-125 for permanent implantation. Computerized dosimetry was obtained for all interstitial and intracavitary applications using the standard commercial treatment planning systems. Pelvic fields were treated either with anterior and posterior portals or with a four-field technique encompassing the entire vagina and the pelvic lymph nodes." In tumors involving the distal half or distal third of the vagina with clinically negative groins, the portal treatment included the groins. The dose of external beam therapy ranged between 40 to 50 Gray (Gy) for 4.5 weeks to 5.5 weeks, depending on the stage and extent of the disease. In the case of advanced lesions, additional parametrial boost treatments were added. In patients undergoing interstitial implants with centrally located lesions, the rectum and bladder were shielded at a dose level of 40 to 45 Gy. For temporary implants, a Syed-Neblett template was used' and the doses were prescribed at 0.5-cm depth from the plane of the implant. The implant doses ranged from about 25 Gy to 30 Gy for 3 to 4 days. Based on local control, two groups of patients were identified. To analyze the treatment results, certain statistical methods were used. Statistical Methods The outcome chosen was overall survival. Survival time was measured from the end of radiotherapy to the date of death or last follow-up, which was treated as censored observation. The differences in patient characteristics and treatment parameters between the two groups were tested by chi-square tests, if categoric data was involved, or the Kruskal-Wallis test if continuous data was given. Estimation of survival curves were obtained by the Kaplan-Meier method and comparisons of survival curves were made by log rank tests. In addition, we stratified the patients based on stage information and then tested the significance of the risk factors including tumor site (proximal tumor group vs. distal tumor group), histology, race, age, and total dose. RESULTS Group I had 23 patients with lesions predominantly located in the proximal half of the vagina. Eight patients were Stage I, 11 patients were Stage 11, and 4 patients had Stage I11 and J Y disease. Among the five patients who Radiation for Carcinoma of the Vagina/Ali et al. 1937 TABLE 2 Treatment Information Variable Statistics Proximal group Distal group Group difference External Beam EBT Dose (Gy) No. (%) Mean (SD) Median No. (%) None No. (%) Brachy No. (W)Cylin. No. (W)Ovoids Mean (SD) Median Mean (SD) Median 22 (95.7) 48.3 (7.3) 50 1 (4.4) 12 (52.1) 9 (39.1) 1 (4.4) 83.0 (18.3) 78 2.8 (1.8) 2.9 14 (82.4) 45.2 (10.8) 47 2 (11.8) 3 (17.6) 10 (58.8) 2 (11.8) 79.0 (35.4) 70 1.9 (1) 1.9 NS" NSb Implant methodology Total dose (Cy) Total treatment time (mo) NS" NSb NSbP = 0.079 EBT external beam therapy; S D standard deviation; NS: not significant; Brachy: brachytherapy; Cylin: cylinder. 'Chi-square test at level 0.05. bKruska-Wallis test at level 0.05. TABLE 3 Outcome Summary Stage Proximal Distal Local control status I I1 IIIlN Total I I1 IIllN Total NED Failure Total 8 0 8 10 2 2 4 20 3 23 5 2 8 10 0 2 2 7 10 1 11 0 5 17 N E D no evidence of disease. died, three died of intercurrent disease and two died of recurrent disease. There were only three local failures and one Stage I1 patient was salvaged by surgery. The two patients who died of disease progression were Stage 111 and Stage IV,respectively. The overall 5-year survival in this group was 81%. Group I1 had 17 patients who had tumors located in the distal half of the vagina. Five patients had Stage I disease, 10 Stage I1 disease, and 2 Stage IV disease. Ten patients had local failure; among these ten patients, four were lost to follow-up and were believed to have died because of the disease progression in this analysis. The remaining six patients died of recurrent disease. The 5year overall survival rate in this group was 41%. N o apparent differences in patient characteristics between the groups were identified. The 5-year survival by stage for the 40 patients was 100%for Stage I, 53% for Stage 11, and 22% for Stages 111 and IV (Fig. 3). Table 4 shows the 5-year overall survival for each group of patients and their corresponding stages. The differences in overall survival was statistically significant between Group I and Group I1 ( P = 0.004), after adjusting for stage. Local Control All patients who died of disease had local failure. In Group I, one patient failure (Stage 11) of the three was salvaged by surgery. Ten patients in Group I1 had local failure. Four of these patients were lost to follow-up but had persistent disease at the time of their last follow-up and so were believed to have died. There is not enough follow-up data to reach a statistical conclusion with respect to local control. Dose Response The median dose in Group I was 78 Gy. For Group 11, the median dose was 70 Gy (Table 2). There is no significant difference in doses received between patients in Group I and Group 11. 1938 CANCER May 1,1996 I Volume 77 I Number 9 11 Stage I Stage ll -Stage IllllV 01 -. 10 00 36 72 FIGURE 4. Overall survival of FIGURE 3. Overall survival of patients by stage. The ratio of brachytherapy dose to the external beam dose was also compared without significant difference between the two groups. The median of the ratio was 0.72 for the proximal tumor group (Group I) and 0.65 for the distal tumor group (Group 11).There was no statistical significant difference between the groups ( P = 0.8). 108 144 180 Time (months) Time (months) 09 i . ....... . _ _. 08 0.7 ~0 6 ~0 5 0 4 -- 03 02 Stage II patients by group. ~ A ~ : : 4 ~~ 01 0, COMPLICATIONS The Radiation Therapy Oncology Group system was utilized to grade the toxicity or complications of treatment. There was one recurrence and four severe complications. In Group I, one Stage I1 patient had parametrial recurrence and required total pelvic exenteration. Four patients in Group 11 developed Grade 3 and Grade 4 complications. Two patients developed a rectovaginal fistula (Stage I11 and Tv), and two additional patients developed small bowel obstruction, (Grade 4) without any tumor recurrence. DISCUSSION Radiation therapy is the treatment of choice for carcinoma of the vagina regardless of the anatomic location of the tumor either in the proximal or the distal part of the vag i ~ ~ a . Radiation ~-’ therapy provides excellent local tumor control in early and superficial lesions with satisfactory functional results.13 It is also important to note overly aggressive treatment may result in damage to the normal tiss u e ~ . It’ ~is clear from a review of the literature that the outcome for carcinoma of the vagina has improved over the past 50-60 years partly because of our ability to understand the natural history of the disease and partly due to the radio responsiveness of this diseasels along with improvements in treatment techniques. External pelvic irradiation, especially in patients with bulky vaginal lesions, helps to shrink the tumor and optimize the use of brachytherapy, thereby reducing the volume of irradiation,and contributes to the success of treatment. Patients with Stage I1 and IiI tumors often require a combination of external beam and interstitial therapy. In our experience (Table 3), and that of others, local tumor control has been better when both modalities were uti- FIGURE 5. Overall survival of Stage III/IV patients by group. lized.” In our series, there has been a significantlyhigher number of local failures in the distal tumor group (Table 3): 8 of 10 patients (80%)as compared with the proximal tumor group (Table 3). Although optimum radiation doses (Table 2) were delivered utilizing both external beam and implants in 82-96% of patients, the local failure rate was extremely high in the Group I1 patients. As point out by Perez et al., the higher incidence of pelvic failures in Stages Ii and 111 is of major concern and they suggested that a more radical approach be adopted in these advanced tumors?,’**In our series, the treatment results of the proximal and the distal vagina in Stage I1 patients were drastically different. (Figs. 2 and 4) The overall survival was in excess of 85% in the proximal tumor group but was very low (10%)in the distal group (Fig. 4). In our series, although the numbers are small in these two groups, it appears that the proximal tumor group has much better results than the distal tumor group (Fig. 5). There was 1 failure in a Stage I1 patient in the proximal tumor group (32 months after treatment); this patient underwent radical surgery and remains alive with no evidence of disease thus far. The poor results in the distal tumor group in Stage I1 patients may be due to factors such as poor response to radiation, lack of vascularity, inadequacy in the staging system, or the longer elapsed time to deliver the definitive treatment.I6 (Table 4 and 5) Alternatively, either a higher external beam dose or wider use of interstitial implants may help to improve the target volume dose distribution and might help to Radiation for Carcinoma of the Vagina/Ali et al. 1939 TABLE 4 2- and 5-Year Survival Rate Stage Proxlmal Distal Survival rate (W) 1 II Ill/Iv Total I I1 lllllv Total 2-year 5-year 100 100 90. I 90.1 7s 37.5 91.1 81 100 100 so 50 0 64.7 14.4 Stage and group (pmrimal 6 .distak 2 = O.OG4) are significan: factors for the ouicome at level 3.05 by the 25 log iank ies:. TABLE 5 Patients Who Failed (by Group) Stage Histology Implant methodology Total treatment time Total dose (Gy) Proximal (3) Distal (10) 1, !I; I , 111; I, !v 2, SCC; I, Sarcoma 3, Brachy 2, 3 mo; I, 6 mo 67.6- 105 8, 11; 2, !v 7, SCC; I, Adeno: 1, Melanoma; I, Pap Adeno 2, None; 2, Brachy; 6, Cylinder Median; 2 months; Range; 0.6-3.4 nio 60-180 SCC: small cell carcinoma; Adeno: adenocarcinoma; Pap Adeno: papillary adenocarcinoma; Brachy: brachytherapy. improve local control in tumors of the distal New therapeutic strategies such as high LET radiation, as well as the adjuvant use of radiosensitizers and radioprotectors should be considered for more advanced lesions as suggested by Nori et al.,15 particularly in the middle and lower third vaginal lesions. Chemotherapeutic regimens should be investigated as radiosensitizers in the management of advanced tumors. 8. 9. 10. 1. Cancer Facts and Figures - 1995. American Cancer Society, 1995. 2. Brady LW. Radiation therapy for carcinoma of the vagina. In: McGowan L, editor. Gynecologic oncology. New York: Appleton-Century-Crofts, 1978:345-59. 3. Perez CA, Camel HM. Long term follow-up in radiation therapy of carcinoma of the vagina. Cancer 1982:49:1308-15. 4. Dancuart F, Delclos I., Warton JT, Silva EG. Primary squamous cell carcinoma of the vagina treated by radiotherapy: a failures analysis-- the M. D. Anderson Hospital experience: 1955-1982. I n t ] Radiat Oncol Biol Phys 1988; 14:745-9. 5. Hubin P, McDonald S, Qazi R. Clinical oncology. A multidisciplinary approach for physicians and students. 7th edition. Philadelphia: W.B. Saunders Co., 1993:400-3. 6 . Haskel CM. Cancer treatment. 2nd edition. Philadelphia: W.B. Saunders Co., 1985. 7. Perez CA, Korba A, Sharma S. Dosimetric considerations in irradiation of carcinoma of the vagina. Int J Radial Oncol Biol Phys 1977;2:639-49. 1. 2. 13. 14. 15. 16. 17. Perez CA, Di Saia PI, Knapp RD. Gynecologic tumors. In: DeVita VT, IIellman S, Rosenberg SA, editors. Cancer: principles and practice of oncology. Philadelphia: J.B. Lippincott, 1985:1013-81. Puthawala A, Syed AM, Nalick H , iMcNamara C, Di Saia PJ. Integrated external and interstitial radiation therapy for primary carcinoma of the vagina. Obstet Gynecol1983;62:367-72. MacNaught R, Symmonds RP, Hole D, Watson ER. Improved control of primary vaginal tumors by combined external beam and interstitial radiotherapy. Clin Radio/ 1987;37:39. Deppe Q, Lawrence WD. Cancer of the vagina including DES-related lesions. In: Ausberg SB,Shingleton HM, Deppe Q, editors. Female genital cancer. New York: Churchill Livingstone, 1988253-74. Plentl AA, Friedman FA. Lymphatic system of the female genitalia: the morphologic basis of oncologic diagnosis and therapy. Philadelphia: W.B. Saunders, 1971. Kottmeir HII.. Annual report on the result of treatment in gynecological cancer. Int Fed Gynecol Obstet Sweden 1979;13,18. Brown GR, Fletcher GII, Rutledge FN. Irradiation of in situ and invasive squamous cell carcinoma of the vagina. Cancer 1971;28:1278-83. Nori I>, Hilaris BS, Stanimir G, I.ewis JL. Radiation therapy of primary vaginal carcinoma. Int / Radial Oncol Biol Phys 1983;9( 10):1471-5. Lee WR. Radiotherapy alone for carcinomas of the vagina: the importance of overall treatment time. lnt / Radiar Oizcol Biol Phys 1994;29:983-8. Prempree T, Amornmarn R. Kadiation treatment of primary carcinoma of the vagina. Patterns of failures after definitive therapy. Actu Radio1 Oncol 1985;24:51-6.