Radiotherapy alone for invasive vaginal cancer: outcome with intracavitary high dose rate brachytherapy versus conventional low dose rate brachytherapy

Acta Obstet Gynecol Scand 2001; 80: 355–360 Copyright C Acta Obstet Gynecol Scand 2001

Printed in Denmark ¡ All rights reservedActa Obstetricia et

Gynecologica ScandinavicaISSN 0001-6349

ORIGINAL ARTICLE

Radiotherapy alone for invasive vaginalcancer: outcome with intracavitary high doserate brachytherapy versus conventional lowdose rate brachytherapyHERWIG KUCERA1, ULRIKE MOCK2, THOMAS H. KNOCKE2, ELISABETH KUCERA1 AND RICHARD PÖTTER2

From the 1Departments of Gynecology and 2Radiotherapy, University of Vienna, Vienna, Austria

Acta Obstet Gynecol Scand 2001; 80: 355–360. C Acta Obstet Gynecol Scand 2001

Purpose. Our aim was to compare the role of remote afterloaded high-dose-rate brachyther-apy (HDRB) with traditional low-dose-rate brachytherapy (LDRB) for patients with invasiveprimary vaginal carcinoma.Methods. The study group comprised 190 patients with invasive carcinoma of the vagina. Thepatients were staged according to the International Federation of Gynecology and Obstetrics(FIGO) staging system. Eighty patients were treated with intracavitary high-dose rate iridium192 brachytherapy with or without external beam therapy. These patients are compared witha historical group of 110 patients treated with intracavitary low-dose-rate radium 226 orcesium 137 brachytherapy with or without external beam therapy.Results. No significant differences were found for stages, tumor grade or location betweenthe two groups. Crude 5-year survival for all patients was 41% in the former LDRB group,81% in stage I and 43% in stage II. Overall actuarial 3-year survival and disease-specificsurvival rates for all patients in the HDRB series were 51% and 66%, respectively. Disease-specific 3-year survival attained 83% in stage I and 66% in stage II. There were no significantdifferences in local and distant recurrences between the treatment modalities. The comparisonof treatments with or without external beam radiation and of complications showed no sig-nificant differences between the HDRB and LDRB series.Conclusion. With HDRB and its advantages of decreased radiation exposure and patientimmobilization and precise positioning, treatment results to be obtained are at least similarto traditional LDRB for primary vaginal cancer.

Key words: high-dose-rate brachytherapy; low-dose-rate brachytherapy; treatment outcome;vaginal cancer

Submitted 16 August, 2000Accepted 10 November, 2000

Primary vaginal cancer is one of the rarest of allmalignancies and accounts for approximately 2%of all female genital neoplasms (1). Radiation ther-apy is the treatment of choice in the majority ofpatients with carcinoma of the vagina, with wide

Abbreviations:HDRB: high-dose-rate brachytherapy; LDRB: low-dose-ratebrachytherapy; EORTC: European Organisation for Researchand Treatment of Cancer.

C Acta Obstet Gynecol Scand 80 (2001)

local excision or vaginectomy sometimes beingused for intraepithelial or early invasive tumors(2). Both external beam radiation therapy and bra-chytherapy have been used alone and in combi-nation in the treatment of this disease. We pre-viously reported results of 110 patients treated bymeans of low-dose-rate brachytherapy with con-ventional radium or cesium sources (3). The tran-sition from low-dose-rate brachytherapy (LDRB)to high-dose-rate remote afterloading treatment is

356 H. Kucera et al.

in progress in many centers of radiation therapy.High-dose-rate brachytherapy (HDRB) is deliver-ed in a few minutes while LDRB requires severalhours. HDRB is defined as exceeding 1200 cGy/h,while LDRB is in the range of 40 to 200 cGy/h(4). Advantages of HDRB include a decreasedradiation exposure of the medical staff, a reduc-tion of patient immobilization and discomfort, aswell as precise positioning and better dose distri-bution (5).

Another advantage of HDRB is reduced costsecondary to elimination of in-patient stays andhospitalization. Although there has been an in-creased use of HDRB in the treatment of gyne-cologic malignancies, experience with this mo-dality in the treatment of vaginal carcinoma is stillminimal (6). The purpose of the present report isto update our previous publication and to presentour experience with the use of remote afterloadingintracavitary HDRB for treatment of the first 80patients with primary invasive vaginal cancer.

Materials and methods

This is a retrospective review of the charts of 80patients with primary carcinoma of the vaginatreated with intracavitary HDRB alone or in com-bination with external beam therapy at the De-partment of Radiation Oncology of the ViennaMedical School in Austria during the period 1986–1998. Most patients had been treated since 1993when radium sources were withdrawn. The pre-viously reported 110 cases from 1975 to 1984 weretreated by means of conventional intracavitaryLDRB with radium or cesium insertions and rep-resent the historical group for comparison (3).

All patients were staged according to the Inter-national Federation of Gynecology and Obstetrics(FIGO) classification: stage I is limited to the va-ginal wall, stage II means involvement of the sub-vaginal tissues but without extension of carcinomaonto the pelvic wall. In stage III, the carcinomahas extended onto the pelvic wall, while in stageIV extension goes beyond the true pelvis or hasinvolved the mucosa of the bladder or rectum. Pa-tients with in situ carcinoma of the vagina wereexcluded from the study. The tumor stage was de-termined clinically, and the tumor location withinthe vagina, the inguinal lymph node status and themanifestation of distant metastases were defined.Patients with tumors beyond stage II frequentlyhad cystoscopy and barium enema. In the past fewyears, computed tomography and ultrasound wereobtained whenever deemed necessary.

In order to histologically confirm the diagnosis,our initial surgical procedure consisted in a biopsyand all patients were treated with radiotherapy.


Radiation management was individualized basedon tumor size, location and stage as well as thepatients’ age and general health.

The technique of conventional LDRB with ra-dium or cesium have been previously described (3).Briefly, all patients were treated with intracavitaryinsertions to deliver 60–90 Gy mucosal dose. Intra-cavitary insertion in a modification of the Stock-holm method, with three applications in the vaginaand in some cases in the uterus, was the standardtreatment in the former LDRB group. In the re-cent HDRB group, patients were treated with vari-ous types of intravaginal cylinders using either thesingle line channel of the Buchler or the Nucletronsystem remotely afterloaded with a 10 Ci Iridium192 source. The number of insertions ranged from2 to 6 with a median of five applications. In ourattempt to quantify the delivered doses despite thevariable reference systems, a specific point at theouter contour of the applicator (equivalent to thevaginal surface) was defined. Correspondingvalues of the different applications ranged from 9.1Gy–17.2 Gy per fraction and the applied totalphysical dose of median 54 Gy was assessed (range15 Gy–71.4 Gy).

Treatment techniques of external beam ir-radiation also varied over the time period. In allpatients in the LDRB group and in 28 patientsin the HDRB group external beam therapy wasemployed using Cobalt 60 unit and a biaxial rota-tional treatment technique. With regard to the ro-tational treatment technique, fraction size per axisranged from 1.0–3.0 Gy and a mean dose of 39 Gy(range 6–45 Gy) per axis was applied. As from1993, patients were treated by a four-field box orusing opposed fields with a beam energy of 25MEV from a linear accelerator. The mean dose atthe reference point was 48 Gy (range 25–50 Gy).Depending on the number of HDRB applications,external beam irradiation was performed with orwithout midline shielding. Where six HDRB inser-tions were performed, blocks were used in order tolimit the central dose to 25 Gy, while a parametrialdose of 50 Gy was applied. In case of externalbeam irradiation without midline shielding,HDRB was restricted to three applications. Theinguinal lymph nodes were irradiated in case of tu-mor manifestation in the lower vaginal third orwith tumor involvement. The administered dosesof both treatment techniques were simply added topresent an applied mean total dose at the referencepoint (vaginal surface) of 91.9 Gy (range 18 Gy–151.9 Gy).

The minimum follow-up in the LDRB groupwas five years while the mean follow-up in theHDRB series was three years. Since most vaginalcancers recur within the first two years a shorter

Radiation alone for vaginal cancer 357

Table I. Vaginal cancer: Patient characteristics

LDRB-group (1975–85) HDRB-group (1986–98)nΩ110 nΩ80

Age (in years) 67.4 (32–88) 70.5 (40–92)Histologic type

squamous 92% 87%adenocarcinoma 7% 5%other 2% 8%

StageI 16 (14%) 16 (20%)II 23 (21%) 39 (48%)III 46 (42%) 20 (25%)IV 25 (23%) 5 (7%)

Grade differentiationG1 15 17G2 37 31G3 39 29

No grading 19 3Location

upper vagina 32 36middle 22 21lower 24 24whole 22 19

medium follow-up may be adequate. Disease-freesurvival and overall survival curves for the HDRBseries were calculated using the Kaplan-Meiermethod, and tests in respect of the significance ofdifferences were based on the long-rank statisticsmodel. For the historical LDRB group treatedwith conventional radium or cesium insertions,crude 5-year survival was documented.

Treatment-related acute and late side effects ofthe HDRB group were assessed according to thecriteria of the EORTC grading system (7).

Results

Patients’ age ranged from 32 to 92 years in thewhole patient group although the median age inthe HDRB group was higher than in the LDRBgroup (70.5 versus 67.4). The most common histo-logic type of vaginal tumors was squamous cellcarcinoma in both treatment groups (Table I).While 35% of the patients in the LDRB groupwere suffering from stage I and II tumors, the cor-responding value was 68% in the HDRB group.Except the higher rate of stage I and II in theHDRB group, there were no significant differencesregarding tumor grade or location of the tumorbetween the two groups (Table I).

The crude 5-year survival rate for the formerLDRB group reached 41%. The actuarial overall3-year survival rate for the HDRB group was 51%while the disease-specific survival rate was 66%(Table II). As expected, survival was related to thestage of disease. In the HDRB group, we found astatistically significant difference in survival


(p,0.03) only when early stage disease (stage I)was compared to advanced disease (stage IV). Thecomparison between traditional LDRB and mod-ern HDRB showed no significant differences ofcure rates with both treatment forms, but there isa clear trend in all stages for a better outcome withHDRB treatment.

Furthermore, there were no significant differ-ences in local and distant failures between theLDRB and the HDRB group, although the ratefor local recurrence in LDRB group was lowerthan in the HDRB group (14.5% versus 26.3%)(Table III). About two-thirds of all relapses werelocated within the vagina. In the series withHDRB, tumor recurrences occurred in 27 of 80 pa-tients (33.7%), four of them are alive after ad-ditional radiotherapy. The mean interval from pri-mary treatment to recurrence was 10.7 months forall sites, whereas 89% of the relapses occurredwithin 12 months. Analyzing the number of recur-rences by stage in both treatment groups showedan increasing incidence of relapses with higherstages. No apparent differences in incidence andsite of relapses between the LDRB and HDRBgroups were identified.

Comparing brachytherapy with or without ad-ditional external beam irradiation in both treat-ment groups, the impact of combined treatment

Table II. Vaginal cancer: Treatment results (%)

HDRB-group

LDRB-group overall disease-specificcrude 5-yr survival 3-yr survival 3-yr survival

Total 41 51 66

StageI 81 54 83II 43 53 66III 35 53 63IV 24 20 27

Table III. Vaginal cancer: Recurrences

LDRB-group HDRB-group

Total 27/110 (24.8%) 27/80 (33.7%)

Local 16 21Pelvis 5 2Inguinal nodes 2 2Distant 4 2

StageI 1 2II 5 13III 12 9IV 9 3


Table IV. Vaginal cancer: Results (%) with brachytherapy alone or in combi-nation with external radiotherapy in early stages

HDRB-groupLDRB group disease-specific

crude 5-yr survival 3-yr survival

Stage IBT alone 85 90BTπext. RT 67 83

Stage IIBT alone 50 63BTπext. RT 42 53

Table V. Vaginal cancer: Results (%) related to the location of the primarytumor

HDRB-groupLDRB-group disease-specific

crude 5-yr survival 3-yr survival

Tumor locationupper vagina 60 62middle 37.5 63lower 37.5 20whole 20.8 61

Table VI. Vaginal cancer: Treatment related side effects of the HDRB group(nΩ80)

Vagina Rectum Bladder

Acute side effectsG1 11 6 5G2 3 7 2G3 – 1 1G4 – – –

Late side effectsG1 5 2 2G2 4 2 1G3 3 2 1G4 1 2 –

modalities on outcome for early stages of diseasewas tested (Table IV). In stage I and II, additionalpelvic irradiation could not improve outcome, anddifferences were not significant.

The site of the tumor in the vagina was assessedwith respect to the position in the upper, middleor lower third of the vagina (Table V). Althoughtumor involvement of the upper third obviouslyhad the best prognosis in the LDRB group, thedifferences concerning tumor location were notsignificant in the HDRB series.

In the former LDRB group six patients de-veloped severe complications (5.4%) while threepatients suffered from rectovaginal and three fromvesicovaginal fistulae (3). It was difficult to com-pare the rate of adverse side effects. An assessmentof vaginal function was not carried out in the


LDRB group. Toxicity of HDRB is demonstratedin Table VI. Treatment-related side effects of theHDRB group were assessed according to the cri-teria of the EORTC grading system (7). In theHDRB series the majority of complications oc-curred in the vagina, but only three out of 80 pa-tients (3.7%) developed severe late complications:one patient with vesicovaginal and two patientswith rectovaginal fistulae.

Discussion

Radiation therapy is the treatment of choice forinvasive vaginal cancer. It provides good tumorcontrol and satisfactory functional results. Few re-ports in the literature discuss surgical treatmentfor the disease. The proximity of the bladder andrectum make vaginal resections with preservationof these organs often problematic (8–10).

For many decades, the only form of brachyther-apy was LDRB with radium 226 and cesium 137sources. Remote-control afterloading devices usingHDRB-cobalt 60 sources were later developed toprovide radiation protection. However, it was thedevelopment of miniaturized iridium 192 sourcesthat revolutionized remote afterloading. HDRB in-volves a single high-intensity source that movesstepwise through the applicator to deliver treat-ment in a period of several minutes. When the ef-fective strength of the source is determined by howlong it spends at each of the dwelling positions, afar better control of the radiation dose distri-butions is achieved (5). In addition, there is agreatly improved radiation safety and an avoid-ance of the problems associated with prolongedLDRB-applicator placements within hours ordays. The use of HDRB techniques for the treat-ment of gynecological malignancies is being usedwith increased frequency. To the authors’ knowl-edge, this is the first report comparing LDRB-ra-dium treatment with HDRB-iridium treatment ina larger series of patients with primary invasive va-ginal carcinoma.

Although there is likely to be some heterogeneitywithin the population of patients in this survey ob-served for over two decades, it is still possible tocompare the treatment group receiving traditionalradium LDRB with the modern iridium HDRBgroup. The most common histological type of tu-mor in both LDRB and HDRB groups wassquamous cell carcinoma. Tumor differentiationand location within the vagina were similar, too.There is general agreement in the literature thattumor bulk specified by FIGO stage is the mostinfluential variable of prognostic factors. Whencomparing our series to others, we found no ap-parent difference in tumor characteristics.

Radiation alone for vaginal cancer 359

Treatment of all patients in this series withLDRB produced a crude 5-year survival rate of41%, while the actuarial 3 year overall survivalwith HDRB was 51% and 3-year disease-specificsurvival was 66%. Our results from treating va-ginal cancer with LDRB are consistent with thefindings by other authors who published theirexperience over the past 10 years (6, 8, 11–17).The Annual Report (18), which includes morethan 6000 cases of carcinoma of the vagina, re-ported a crude 5-year survival of 38.6% for theperiod from 1979–81 and an actuarial 5-year sur-vival rate of 46.8% for the period of 1987–89.There are few data concerning HDRB and carci-noma of the vagina. Nanavati et al. reported 13patients in stage I and II, all of them had acomplete response while local control wasachieved in 92% at a median follow-up of 2.6years (19). In our HDRB series of 55 patients instage I and II, disease-specific survival at 3 yearswas 83% for stage I and 66% for stage II. In theHDRB group being analyzed a relatively highmedian age of 70.5 years was determinedwhereas the median age in our LDRB series wasminus 3 years corresponding to other authorswho reported a median age of 55–67 years (8,12, 15, 17). Regarding our results, there is atrend towards a better outcome with HDRB butit is difficult to say whether there has been anobvious improvement between the LDRB andHDRB eras. Survival results for HDRB are ex-pressed in actuarial 3-year overall and diseasespecific survival, while the LDRB survival is ex-pressed in crude 5-year overall survival. Hence,comparison between outcome of the two groupsis not totally satisfactory and could be mislead-ing in terms of treatment success and longtermtoxicity. Perhaps the better outcome with HDRBcould have been influenced by higher externalbeam radiotherapy doses for the 52 HDRB pa-tients treated with 25 MEV. However, resultswith the modern iridium-HDRB technique are atleast equal to the more antiquated radium-LDRB technique.

In our HDRB series, metastases in the inguinalnodes or at distant sites seemed to be a rare siteof recurrence. Therefore inguinal nodes were notroutinely included in the treatment volume exceptin case of tumor involvement or with tumor mani-festation in the lower vaginal third. Perhaps, bycombining radiation with chemotherapy results forpatients with stage III and IV, tumors might beimproved (1).

With both treatment modalities, brachytherapyalone appears to be adequate for stage I and IIpatients. HDRB was applied with or without ex-ternal beam radiation, resulting in similar survival


and with local failure rates without significant im-provements attained by combined treatment op-tions. Recently, Perez et al. noted that vaginal car-cinoma tumor control in the pelvis in stage I wasapproximately the same with brachytherapy aloneas in combination with external beam irradiation.In stages II and III, there was a trend toward abetter tumor control with combined external ir-radiation and brachytherapy than with the latteralone although the difference was not significant(6). This is in keeping with the results of otherauthors (12, 20–22) but it should be stressed thatonly small-volume, superficial tumors should betreated with brachytherapy alone.

It is noteworthy that good results for vaginalcarcinoma are achieved with radiotherapy in tu-mors of the upper third of the vagina since patientswith lesions in this location are often candidatesfor an upper partial vaginectomy. In the formerLDRB series, the best treatment results wereachieved in the proximal vagina, which is consist-ent with the experience of Ali et al. (23). However,no significant variations in response to the locationof the primary tumor were found in the HDRBseries.

HDRB toxicity compares favorably to thestudies using a conventional LDRB technique. Inour former LDRB group, serious complicationswere reported in 5.4% of patients (3). This is con-sistent with other reports in the literature (11, 12,16, 20–22, 24). In the HDRB series, grade-4 sideeffects were found in three out of 80 patients(3.7%). In a series of 13 stage I and II patientstreated with HDRB, Nanavati et al. reported noacute or chronic intestinal or bladder grade 3 or 4toxicity (19). However, the incidence rates of re-ported complications are difficult to compare be-cause of varying definitions of major compli-cations.

In conclusion, our study shows that radiationtherapy is an effective treatment for patients withvaginal carcinoma, particularly at the early stages.With remote afterloaded HDRB treatment resultsare at least similar if not better than with tra-ditional LDRB and toxicity is acceptable. The bestway to compare HDRB to LDRB in the treatmentof vaginal carcinoma would be by means of a pros-pective, randomized study. The rarity of vaginalcancer and the disadvantages of long-lastingLDRB, however, make such a comparative study adifficult undertaking.

References

1. Creasman WT, Phillips JL, Menk HR. The national cancerdata base report on cancer of vagina. Cancer 1998; 83:1033–40.

2. Hoffman MS, DeCesare SL, Roberts WS, Fiorica JV, Fin-


an MA, Cavanagh D. Upper vaginectomy for in situ andoccult, superficially invasive carcinoma of the vagina. AmJ Obstet Gynecol 1992; 166: 30–3.

3. Kucera H, Vavra N. Radiation management of primarycarcinoma of the vagina: Clinical and histopathologicalvariables associated with survival. Gynecol Oncol 1991; 40:12–14.

4. ICRU (International Commission on Radiation Units andMeasures). Dose and volume specification for reporting in-tracavitary therapy in gynaecology. ICRU Report 1985; Nr.38, ICRU, Bethesda, MD.

5. Gore E, Gillin MT, Albano K, Erickson B. Comparison ofhigh dose rate and low dose rate dose distribution for va-ginal cylinders. Int J Radiat Oncol Biol Phys 1995; 31: 165–70.

6. Perez CA, Grigsby PW, Garipagaolu M, Mutch DG,Lockett MA. Factors affecting long term outcome of ir-radiation in carcinoma of the vagina. Int J Radiat OncolBiol Phys 1999; 44: 37–45.

7. Chassagne D, Sismondi P, Horiot JC, Sinistero G, Bey P,Pernot H et al. A glossary for reporting complications oftreatment in gynecologic cancers. Radiother Oncol 1993;26: 165–202.

8. Stock GR, Chen ASJ, Seski J. A 30 year experience in themanagement of primary carcinoma of the vagina: Analysisof prognostic factors and treatment modalities. GynecolOncol 1995; 56: 45–52.

9. Davis KP, Stanhope CR, Garton GR, Atkinson EJ, O’Bri-en PC. Invasive vaginal carcinoma: analysis of early stagedisease. Gynecol Oncol 1991; 42: 131–6.

10. Rubin SC, Young J, Mikuta JJ. Squamous carcinoma ofthe vagina: Treatment complications and long term follow-up. Gynecol Oncol 1985; 20: 346–53.

11. Manetta A, Gutrecht E, Berman ML, DiSaia PJ. Primaryinvasive carcinoma of the vagina. Obstet Gynecol 1990;76(4): 639–42.

12. Leung S, Sexton M. Radical radiation therapy for carci-noma of the vagina. Impact of treatment modalities on out-come. Peter MacCallum Cancer Institute experience 1970–1990. Int J Radiat Oncol Biol Phys 1993; 25: 413–18.

13. Lee WR, Marcus RB, Sombeck MD, Mendenhall WM,Morgan LS, Freeman DE et al. Radiotherapy alone forcarcinoma of the vagina: The importance of overall treat-ment time. Int J Radiat Oncol Biol Phys 1994; 29: 983–8.

14. Kirkbride P, Flyes A, Rawlings GA, Manchul L, Levin W,Murphy KJ et al. Carcinoma of the vagina. Experience atthe Princess Margaret Hospital (1974–1989). Gynecol On-col 1995; 56: 435–43.


15. Urbanski K, Kojs Z, Reinfuss M, Fabisiak W. Primary in-vasive vaginal carcinoma treated with radiotherapy: analy-sis of prognostic factors. Gynecol Oncol 1996; 60: 16–20.

16. Fine BA, Piver MS, McAuley M, Discroll D. The curativepotential of radiation therapy in the treatment of primaryvaginal carcinoma. Am J Clin Oncol (CTT) 1996; 19: 39–44.

17. Chyle V, Zagars G, Wheeler JA, Wharton JT, Delclos L.Definitive radiotherapy for carcinoma of the vagina: Out-come and prognostic factors. Int J Radiat Oncol Biol Phys1996; 35: 891–905.

18. Petterson F, ed. Annual report on the results of treatmentin gynecologic cancer. Radiumhemmet Stockholm 1995:22: 111–17.

19. Nanavati PJ, Fanning J, Hilgers R, Hallstrom J, CrawfordD. High dose rate brachytherapy in primary stage I and IIvaginal cancer. Gynecol Oncol 1993; 51: 67–71.

20. Kucera H, Langer M, Smekal G, Weghaupt K. Radio-therapy of primary carcinoma of the vagina. Managementand results of different therapy schemes. Gynecol Oncol1985; 21: 87–93.

21. Perez CA, Camel HM, Galakatos AE, Grigsby PW, KuskeRR, Buchsbaum G et al. Definitive irradiation in carci-noma of the vagina: Long term evaluation of results. Int JRadiat Oncol Biol Phys 1988; 15: 1283–90.

22. Prempree T, Amornmarn R. Radiation treatment of pri-mary carcinoma of the vagina: Patterns of failure after de-finitive therapy. Acta Radiol Oncol 1985; 24: 51–6.

23. Ali MM, Huang DT, Goplerud DR, Howells R, Lu J. Radi-ation alone for carcinoma of the vagina. Cancer 1996; 77:1934–9.

24. Dancuart F, Delclos L, Wharton JT, Silva EG. Primarysquamous cell carcinoma of the vagina treated by radio-therapy: a failure analysis. The M.D. Anderson Hospitalexperiences 1955–1982. Int J Radiat Oncol Biol Phys 1988;14: 745–9.

Address for correspondence:

Herwig Kucera, M.D.General Hospital ViennaDepartment of GynecologyWähringer Gürtel 18–201090 ViennaAustria

Documents

Radiotherapy alone for invasive vaginal cancer: outcome with intracavitary high dose rate brachytherapy versus conventional low dose rate brachytherapy