Int. J. Radiation Oncology Biol. Phys., Vol. 44, No. 1, pp.

37– 45, 1999

Copyright © 1999 Elsevier Science Inc.
Printed in the USA. All rights reserved
0360-3016/99/$–see front matter

PII S0360-3016(98)00530-6

CLINICAL INVESTIGATION Female Genitalia

FACTORS AFFECTING LONG-TERM OUTCOME OF IRRADIATION IN

CARCINOMA OF THE VAGINA

CARLOS A. PEREZ, M.D.,* PERRY W. GRIGSBY, M.D.,* MELAHAT GARIPAGAOGLU, M.D.,‡

DAVID G. MUTCH, M.D.,† AND MARY ANN LOCKETT, M.B.A.*
*Radiation Oncology Center, Mallinckrodt Institute of Radiology, and †Division of Gynecologic Oncology, Department of Obstetrics
and Gynecology, Washington University Medical Center, St. Louis, MO; and ‡Radiation Oncology Department, Ankara University
Medical School, Dikimevi, Ankara, Turkey

Objective: This report evaluates prognostic and technical factors affecting outcome of patients with primary
carcinoma of the vagina treated with definitive radiation therapy.
Methods and Materials: A retrospective analysis was performed on records of 212 patients with histologically
confirmed carcinoma of the vagina treated with irradiation.
Results: Tumor stage was the most significant prognostic factor; actuarial 10-year disease-free survival was 94%
for Stage 0 (20 patients), 80% for Stage I (59 patients), 55% for Stage IIA (63 patients), 35% for Stage IIB (34
patients), 38% for Stage III (20 patients), and 0% for Stage IV (15 patients). All in situ lesions except one were
controlled with intracavitary therapy. Of the patients with Stage I disease, 86% showed no evidence of vaginal
or pelvic recurrence; most of them received interstitial or intracavitary therapy or both, and the addition of
external-beam irradiation did not significantly increase survival or tumor control. In Stage IIA (paravaginal
extension) and IIB (parametrial involvement) 66% and 56% of the tumors, respectively, were controlled with a
combination of brachytherapy and external-beam irradiation; 13 of 20 (65%) Stage III tumors were controlled
in the pelvis. Four patients with Stage IV disease (27%) had no recurrence in the pelvis. The total incidence of
distant metastases was 13% in Stage I, 30% in Stage IIA, 52% in Stage IIB, 50% in Stage III, and 47% in Stage
IV. The dose of irradiation delivered to the primary tumor or the parametrial extension was of relative
importance in achieving successful results. In patients with Stage I disease, brachytherapy alone achieved the
same local tumor control (80 –100%) as in patients receiving external pelvic irradiation (78 –100%) as well. In
Stage II and III there was a trend toward better tumor control (57– 80%) with combined external irradiation and
brachytherapy than with the latter alone (33–50%) (p 5 0.42). The incidence of grade 2–3 complications (12%)
correlated with the stage of the tumor and type of treatment given.
Conclusion: Radiation therapy is an effective treatment for patients with vaginal carcinoma, particularly Stage
I. More effective irradiation techniques, including optimization of dose distribution combining external irradi-
ation and interstitial brachytherapy in tumors beyond Stage I, are necessary to enhance locoregional tumor
control. The high incidence of distant metastases emphasizes the need for earlier diagnosis and effective systemic
cytotoxic agents to improve survival in these patients. © 1999 Elsevier Science Inc.

Vagina, Radiation therapy, Brachytherapy, Tumor control.

INTRODUCTION analyzes factors affecting tumor control and survival in a

Primary carcinoma of the vagina is rare; it accounts for
about 2% of all gynecologic malignancies at Washington
University Medical Center. Radiation therapy is the treat-
ment of choice in the majority of patients with primary
carcinoma of the vagina, with wide local excision or vagi-
nectomy sometimes being used for in situ or early invasive
tumors (1). We previously reported that higher doses of
irradiation are required to control more advanced stages (2);
judicious use of brachytherapy techniques will yield optimal
tumor control in patients with smaller lesions at the same
time that optimal functional results are achieved.
The present report updates a previous publication (2) and
larger patient population. Careful dosimetric assessment of
doses of radiation delivered to the primary lesion or the
parametria was correlated with tumor control.
METHODS AND MATERIALS
The records of 192 patients with primary invasive carci-
noma and 20 patients with in situ carcinoma of the vagina
treated at the Radiation Oncology Center, Mallinckrodt
Institute of Radiology, Washington University Medical
Center, between January 1953 and December 1991, were
reviewed. Over 60% of the patients were over 60 years of
age; 171 patients had invasive epidermoid carcinoma, 14

Reprint requests to: Carlos A. Perez, M.D., Radiation Oncology 63108. E-mail: perez@roc.wustl.edu
Center, 4511 Forest Park Boulevard, Suite 200, St. Louis, MO Accepted for publication 4 December 1998.

37
38 I. J. Radiation Oncology ● Biology
had adenocarcinoma, 2 had clear cell carcinoma not related
to diethylstilbestrol, and 5 had undifferentiated carcinoma;
20 patients had carcinoma in situ. In all cases, the bona fide
primary nature of the vaginal tumor was documented, and
patients with other primary tumors in the uterus or the
cervix or other sites that may metastasize to the vagina were
excluded. However, patients with prior history of carcinoma
of the cervix or endometrium previously treated, who had
been disease-free for more than 5 years after initial treat-
ment, were included in the analysis, because over 95% of
the failures from these primary lesions develop within 5
years from treatment; the survival of this selected group of
patients has been reported to be equivalent to that of patients
with primary vaginal carcinoma and is certainly superior to
that of recurrent lesions of the cervix or endometrium.
Patients were jointly evaluated and staged by the staff of
Radiation Oncology and Gynecologic Oncology. A routine
physical and pelvic examination, chest X-ray, and intrave-
nous pyelogram were performed. Patients with tumors be-
yond Stage IIA frequently had cystoscopy and barium en-
ema. Based on the clinical description and anatomic
drawings in the radiation oncology records, all tumors were
staged following the classification adopted by the Interna-
tional Federation of Gynecology and Obstetrics (3) with a
minor modification for Stage II (2), which consisted of
dividing these tumors into IIA when there was paravaginal
tumor extension or IIB when there was parametrial involve-
ment.
Follow-up information was updated in all patients within
the 3 months before the current review either by examina-
tion at the Medical Center or from information received
from the referring physician or, in a few instances, by direct
correspondence with the patients or relatives. The minimum
follow-up was 3 years, and the mean follow-up at last visit
for surviving patients was 12 years. Patients lost to fol-
low-up were considered dead of tumor if there was any
suspicion of recurrence. Patients dying of intercurrent dis-
ease are included in the overall analysis of survival. When
the site of recurrence was known, it was classified as local,
pelvic, distant metastases (including paraaortic nodes), or
combinations thereof.
In patients on whom isodose distributions were available,
or in a group of 65 patients treated after 1964 on whom
there was adequate description of the implants in the chart,
the dose of irradiation delivered by intracavitary or intersti-
tial techniques was calculated. The external-beam doses
were recalculated in patients treated before 1971 because of
minor dosimetric variations in calibration procedures. Ex-
ternal-beam and brachytherapy doses were added without
regard for possible differences for biologic effect because of
dose rate. Total minimum doses, at the greatest estimated
tumor depth (clinical examination) and to the lateral pelvic
wall, were calculated.
Disease-free survival and overall survival curves were
calculated by the actuarial life table as applied by Cutler and
Ederer (4). Times for survival or failure were calculated
from the onset of definitive therapy. Tests of differences
● Physics Volume 44, Number 1, 1999
between curves were made using the statistical method of
Mantel (5). All other comparisons used the Yates-corrected
Chi-square test. All analyses were performed on a VAX
computer using BMDP statistical software (6).
Methods of irradiation
The techniques of radiation therapy have been previously
described (2). Briefly, all but two patients with carcinoma in
situ were treated with intracavitary insertions to deliver
60 –70 Gy mucosal dose. In most instances, only a cylinder
was used to cover the entire vagina. In one patient, the
vaginal vault was treated with an ovoid containing a 20-mg
source, and tumor recurrence was noted distal to this area;
the patient was salvaged by an anterior exenteration.
In 32 patients with Stage I tumors, either intracavitary or
interstitial brachytherapy (single or double-plane implant)
or both, combined with external irradiation, were used. In
25 patients, only brachytherapy without external-beam irra-
diation was used. One patient who had a vaginectomy
received postoperative external-beam irradiation. Doses in
the range of 60 –70 Gy were administered in the majority of
patients (calculated at 0.5 cm beyond the plane of the
implant or vaginal mucosa). The vaginal mucosa received
doses in the range of 80 –120 Gy. In patients receiving
external irradiation, 10 or 20 Gy to the whole pelvis was
delivered, and additional parametrial dose with a midline
block was administered to give a total of 40 –50 Gy to the
lateral pelvic wall using high-energy photon beams (6 –25
MV from linear accelerators or betatron). The treatment
portals measured 15 cm in width, to cover the entire true
pelvis; the upper margin extended to the upper sacral area
(L5–S1), and distally, the entire vagina was treated, usually
including the inguinal/femoral lymph nodes in the irradia-
tion portal.
Patients with Stage II tumors were treated with similar
combinations of external irradiation and brachytherapy. The
whole-pelvis dose was 20 – 40 Gy, with additional parame-
trial irradiation given using midline block to deliver a total
of 50 – 60 Gy to the lateral pelvic wall using 18 –25 MV or
4 – 6 MV X-rays as indicated. In eight patients, only brachy-
therapy was used, and in two patients external-beam irradi-
ation alone was delivered. Patients with Stage IIB, III, and
IV disease received 40 Gy to the whole pelvis and addi-
tional parametrial dose with midline shielding to complete
minimum tumor doses of 65–75 Gy and 55– 60 Gy to the
lateral pelvic wall, in combination with interstitial or intra-
cavitary insertions, in most instances; a few of these patients
received brachytherapy or external-beam irradiation only
because of poor clinical condition. Intracavitary brachyther-
apy was administered with vaginal cylinders or Fletcher-
Suit ovoids, or both, and in some patients, with the
Mallinckrodt Institute of Radiology afterloading vaginal
applicator (MIRALVA) (7).
Dose to the tumor was calculated at the clinically per-
ceived depth of the vaginal lesion. Medial parametrial dose
was calculated 2 cm lateral to the vaginal intracavitary
 Carcinoma of the vagina ● C. A. PEREZ et al. 39

Fig. 1. Disease-free survival for all patients with primary carcinoma of the vagina (Stages 0 through IVA). From: Perez
CA, Garipagaoglu M: Vagina. In Perez CA, Brady LW, editors: Principles and Practice of Radiation Oncology, ed 3,
pp 1891–1914. Philadelphia, Lippincott-Raven, 1998.

sources (similar to point A for cervical carcinoma). The
lateral parametrial dose was calculated 6 cm from the mid-
line, at a site where the pelvic lymph nodes are generally
visualized on the lymphangiogram.
RESULTS
Disease-free and overall survival
Patients who died of intercurrent disease were excluded
as censored observations. The 10-year disease-free survival
was 94% for Stage 0, 80% for Stage I, 55% for Stage IIA,
35% for Stage IIB, and 38% for Stage III (Fig. 1). The
overall actuarial survival was significantly lower because a
number of patients died of intercurrent disease (Table 1),
and there was less impact of the clinical stage of the tumor
on probability of overall survival.
because of inadequate coverage of the entire vagina with the
intracavitary therapy. In Stage I, 8 of 59 patients (14%)
failed in the pelvis, 3 of them combined with distant me-
tastases.
In Stage IIA, 22 patients had local or parametrial failures
(34%), 11 of them with distant metastases. In Stage IIB, 15
of 34 patients (44%) had local/parametrial failures, 10 of
them in combination with distant metastases. Seven of 20
patients (35%) with Stage III and 11 of 15 (73%) with Stage
IVA had local/pelvic failures.
The total incidence of distant metastases was 13% in
Stage I, 30% in Stage IIA, 52% in Stage IIB, 50% in Stage
III, and 47% in Stage IV. This pattern of failure points
toward a slightly more aggressive clinical behavior in Stage
II than in epidermoid carcinoma of the uterine cervix (Fig. 2).

Anatomic sites of failure Tumor control and technique of irradiation

Data are shown in Table 1. Only 1 of 20 patients with There was only 1 local failure distal to the vaginal vault
carcinoma in situ failed locally, as described previously, in the 20 patients with carcinoma in situ treated with intra-

Table 1. Carcinoma of the vagina: Anatomic sites of failure

No. of Local/Parametrial Local/Parametrial Distant metastases Dead of

Stage Patients only and distant metastases only intercurrent disease

0 20 1 (5) 0 0 8 (40)
I 59 5 (8) 3 (5) 5 (8) 23 (39)
} }}}}

(14)
IIA 64 11 (17) 11 (17) 8 (13) 17 (27)
(34)
IIB 34 5 (15) 10 (29) 9 (26) 7 (21)
(44)
III 20 1 (5) 6 (30) 4 (20) 5 (25)
(35)
IVA 15 4 (27) 7 (47) 0 3 (20)
(73)

() Percent.
40 I. J. Radiation Oncology ● Biology ● Physics Volume 44, Number 1, 1999

Fig. 2. Patterns of failure (pelvic recurrence, alone or combined with distant metastasis, or distant metastasis only) in
patients reported with primary carcinoma of the vagina compared with 1499 patients with invasive carcinoma of the
cervix treated at Washington University during the same period of time. A slightly higher incidence of pelvic failures
is noted in Stage IIA and IIB vaginal carcinoma. Also, in these stages, the overall incidence of failure is somewhat
higher in vaginal cancer than in primary cervix carcinoma. From: Perez CA, Garipagaoglu M: Vagina. In Perez CA,
Brady LW, editors: Principles and Practice of Radiation Oncology, ed 3, pp 1891–1914. Philadelphia, Lippincott-Raven,
1998.

cavitary brachytherapy; the tumor was inadequately treated
initially with an ovoid containing a 20-mg radium source.
In Stage I, there was no significant correlation between
the technique of irradiation used and the incidence of re-
currence locally or in the pelvis. The addition of external
irradiation did not increase tumor control (78 –92%) in
comparison with the patients treated with brachytherapy
only (80 –100%). However, in Stage IIA the local tumor
control rate was 70% (37 of 53) in patients receiving
brachytherapy combined with external-beam irradiation in
comparison with 40% (4 of 10) in patients treated with
either brachytherapy or external-beam irradiation alone
( p 5 0.42). In Stage IIB and III there was also slightly
higher local tumor control with combinations of external-
beam irradiation and brachytherapy, in contrast to that in a
few patients who were treated with intracavitary or intersti-
tial irradiation alone, usually because of medical reasons.
Local tumor control was achieved in only 4 of 15 patients
with Stage IVA disease. Results are summarized in Table 2.
Pelvic tumor control (including local) correlated with
dose of irradiation
Analysis of irradiation doses or even techniques and
impact on local or pelvic tumor control is fraught with
possible bias, as this is a retrospective, not a prospective,
dose-escalation study; from clinical experience we know
that radiation oncologists prescribe higher doses of irra-
diation for tumors with increased risk of local failure or

Table 2. Carcinoma of the vagina: Tumor control correlated with type of therapy

Stage

Therapy 0 I IIA IIB III IVA

Intracavitary 13/14 (93) 12/15 (80) 1/2 (50) 0/1 0 0

Interstitial 6
intracavitary 1/1 (100) 10/10 (100) 2/6 (33) 1/2 (50) 1/3 (33) 0
Intracavitary 6
interstitial and
external
irradiation 1/1 (100) 14/18 (78) 17/22 (77) 11/15 (73) 4/7 (57) 1/2 (50)
Intracavitary and
interstitial and
external
irradiation 0 12/13 (92) 20/31 (65) 6/13 (46) 8/10 (80) 3/11 (27)
External
irradiation 0 1/1 (100) 1/2 (50) 1/3 (33) 0 0/2

() Percent.
 Carcinoma of the vagina ● C. A. PEREZ et al. 41

Table 3. Carcinoma of the vagina: Pelvic tumor control Table 5. Carcinoma of the vagina: Pelvic tumor control
correlated with minimum dose to vaginal tumor correlated with lateral parametrial dose

Stage Dose to
Dose Stage
lateral
(Gy) I IIA IIB III IVA parametrium
(Gy) I IIA IIB III IVA
#55 4/4 (100) 2/6 (33) 2/4 (50) 1/5 (20) 0/1
55.01–65 10/12 (83) 8/11 (73) 2/5 (40) 3/4 (75) 0/1 #40 28/32 (88) 19/30 (63) 2/7 (29) 2/4 (50) 2/3 (67)
65.01–75 10/11 (91) 12/17 (71) 2/5 (40) 6/7 (86) 0/1 40.01–50 4/5 (80) 9/12 (75) 1/5 (20) 2/7 (29) 0/1
75.01–85 7/9 (78) 7/10 (70) 4/5 (80) 1/1 (100) 3/3 (100) 50.01–60 3/4 (75) 8/11 (73) 6/9 (67) 3/3 (100) 1/1 (100)
85.01–95 6/7 (86) 5/6 (83) 4/6 (67) 1/1 (100) — .60 1/1 (100) 2/4 (50) 5/5 (100) 5/5 (100) 0/1
.95 11/12 (92) 7/10 (70) 3/4 (75) 0/1 —
() Percent.
() Percent. Table includes only patients with complete dosimetry data.
Table includes only patients with complete dosimetry data.

tumor control in comparison with 21 of 25 (84%) treated in

when the tumor does not respond to “standard doses.”
The correlation with doses delivered to the vaginal tumor
suggests that, in Stages II and III, doses in the range of
70 –75 Gy to the primary tumor or the medial parame-
trium are necessary to achieve better tumor control. Table
3 shows that patients with Stage IIA tumors receiving
doses higher than 55 Gy exhibited better local and pelvic
tumor control. As noted in Table 4, in patients with Stage
IIB and III disease, it appears that a dose to the medial
parametrium higher than 65 Gy results in significantly
higher tumor control than with lower doses, although the
differences are not statistically significant because of the
small number of patients. Likewise, in Stages IIB and III
doses to the lateral parametrium (pelvic lymph nodes)
higher than 50 Gy are associated with marked improve-
ment in pelvic tumor control compared with lower doses
(Table 5).
A more definite statement cannot be made concerning
dose and tumor control correlation because of the small
number of patients in the various stages.
Pelvic tumor control correlated with elapsed treatment
time
Prolongation of treatment time did not have a significant
impact on pelvic tumor control in these patients. In Stage I,
23 of 27 patients (85%) treated in 49 days or less had pelvic
tumor control in comparison with 6 of 8 (75%) treated in a
longer time period (up to 63 days). In Stage II, 34 of 60
patients (57%) treated in fewer than 49 days had pelvic
Table 4. Carcinoma of the vagina: Pelvic tumor control
correlated with medial parametrial dose
a longer period of time. In Stage III, the corresponding
figures are 6 of 10 (60%) and 6 of 9 (75%), respectively.
Pelvic tumor control and location, size, or morphology of
primary tumor
Other authors such as Whelton and Kottmeier (8) re-
ported decreased tumor control in patients with lesions
located in the posterior wall. Our analysis showed no sig-
nificant correlation with this parameter, although Stage II
lesions with more extensive involvement of the upper/mid-
dle/lower vagina may have lower tumor control (20 – 60%)
compared with those in other locations (50 – 80%). This is
most likely related to the extent of the disease (bulk of
tumor) rather than anatomic location.
As noted in Table 6, whereas tumor stage was an impor-
tant predictor of pelvic tumor control and 5-year disease-
free survival, the size of the tumor in Stage I patients was
not a significant prognostic indicator. However, in Stage IIA
disease, lower pelvic tumor control and survival were noted
with tumors larger than 4 cm. In Stages IIB and III, tumor
size was not a significant prognostic factor.
Analysis of tumor recurrence and survival according to
gross appearance of the lesion showed that 39 patients with
Stage I superficially ulcerated exophytic tumors had fewer
local recurrences (8%) and distant metastases (8%) than
those with infiltrating or necrotic lesions (20% local recur-
rence and 20% distant metastases) ( p 5 0.14). The differ-
ence, however, is statistically significant only for disease-
free survival (84% vs. 58%, respectively, p 5 0.05). In
patients with Stage II disease, pelvic failure rates were
similar (21% and 18%, respectively), as was 5-year disease-
free survival (53% and 50%, respectively).

Dose to Control of tumor in inguinal lymph nodes

Stage
medial Of 7 patients who received elective irradiation to clini-
parametrium
(Gy) I IIA IIB III IVA cally negative inguinal lymph nodes, only 1 with a large
lesion involving the entire vagina developed an inguinal
#65 24/28 (86) 22/33 (67) 2/10 (20) 3/9 (33) 2/3 (67) lymph node recurrence after 50 Gy. Of 100 patients with
65.01–85 11/13 (85) 15/21 (71) 7/11 (64) 5/6 (83) 1/3 (33) primary tumors in the upper and middle third of the vagina
.85.01 2/2 (100) 2/4 (50) 7/7 (100) 4/4 (100) —
who received no elective irradiation, none developed met-
() Percent. astatic inguinal lymph nodes, in contrast to 3 of 29 (10%)
Table includes only patients with complete dosimetry data. with lower-third primary tumors and one of 20 with tumor
42 I. J. Radiation Oncology ● Biology ● Physics Volume 44, Number 1, 1999

Table 6. Carcinoma of the vagina: Pelvic tumor control and 5-year disease-free survival correlated with stage and tumor size

Stage

Tumor size (cm) I IIA IIB III IVA

Tumor control
#2 16/17 (94) 8/8 (100) 0/2 — 1/1 (100)
2.1–4 12/15 (80) 20/29 (69) 6/12 (50) 2/2 (100) 3/5 (60)
.4 16/18 (89) 12/24 (50) 11/18 (61) 10/16 (62) 0/9
5-Year disease-free survival
#2 87% 74% 0% — 100%
2.1–4 73% 71% 31% 50% 80%
.4 83% 41% 39% 49% 0%

() Percent.

involving the entire length of the vagina. Of 7 patients with
initially palpable inguinal lymph nodes treated with doses in
the range of 60 Gy, only 1 developed a nodal recurrence.
As before, we recommend that elective irradiation of the
inguinal lymph nodes be carried out only in patients with
primary tumors in the lower third of the vagina or those
involving the entire organ.
Results in patients with history of previously treated
uterine tumors
There were 49 patients with carcinoma of the vagina who
had a history of previously treated primary carcinoma of the
cervix or the endometrium (irradiation, hysterectomy, or
combination of both); they had received treatment over 5
years before the diagnosis of vaginal carcinoma, and none
had evidence of tumor activity at any other site. The pos-
sibility of the vaginal lesion being a local recurrence or
metastasis was considered unlikely, because 95% of the
recurrences after treatment of primary carcinoma of the
cervix or endometrium occur within 5 years after therapy.
Table 7 shows sites of failure and the survival in these
patients, which are equivalent to those in patients with bona
fide primary carcinoma of the vagina. As before (2), we
conclude that these lesions are most likely second primary
tumors and should be treated with definitive radiation ther-
apy.
Multivariate analysis
On multivariate analysis, stage of the primary tumor and
medial parametrial dose were significant prognostic factors
for local failure. Clinical stage of the disease and histologic
grade were significant factors for predicting the develop-
ment of distant metastases. Both the stage of the primary
tumor and the dose to the medial parametrium were signif-
icant prognostic factors for disease-free survival ( p values
all ,0.05). Analysis is summarized in Table 8.
Analysis of our patients failed to demonstrate any definite
relationship between the location of the primary tumor in
the various thirds or walls of the vagina and the incidence of
complications (data not shown).
Complications of therapy
The incidence of complications has been relatively low
(Table 9) and are comparable to those observed in carci-
noma of the uterine cervix at our institution. In 73 patients
with Stage 0 and I disease, 5 Grade 2–3 complications were
noted that could be ascribed to radiation therapy (7%). In
the Stage II patients, 4 rectovaginal and 2 vesicovaginal
fistulae were noted (7%), as well as 2 cases of proctitis, 1
rectal stricture, and 1 rectal ulcer. An enterocutaneous fis-
tula was noted in a patient who underwent pelvic exenter-
ation for local/pelvic recurrence. In the patients with Stage

Table 7. Carcinoma of the vagina: Patients with previous cervix or endometrial primary carcinoma (over 5 years): Site of
post-treatment failure

5-Year
Local/Parametrial disease-free
No. of Local/Parametrial and distant Distant survival
Stage patients only metastasis metastasis only (%)

0 8 1 (13) — — 88
I 15 2 (13) 2 (13) — 76
IIA 13 2 (15) 4 (31) 2 (15) 49
IIB 7 2 (29) 2 (29) 2 (29) 14
III 3 — 1 (33) 1 (33) 33
IVA 3 1 (33) 1 (33) — 33

() Percent.
 Carcinoma of the vagina ● C. A. PEREZ et al. 43

Table 8. Carcinoma of the vagina: Multivariate analysis ported for patients with carcinoma of the cervix, which
( p values) range from 20% to 80% at 5 years, depending on stage of
Local Distant Disease-free disease (9).
Variable failure metastasis survival Brown et al. (10) and Perez et al. (2) reported excellent
tumor control and survival rates in patients with carcinoma
Stage 0.031 0.053 0.045 in situ and Stage I invasive carcinoma. These authors cau-
Medial parametrial dose 0.009 0.620 0.050
Histologic grade 0.282 0.044 0.071 tioned against overly aggressive therapy in these early le-
Treatment technique* 0.592 0.517 0.533 sions because of the possibility of producing mucosal injury
Age 0.625 0.669 0.665 and interference with sexual function. In patients with le-
Lateral parametrial dose 0.265 0.541 0.751 sions thicker than 0.5 cm in diameter, it is advisable to
Elapsed days of irradiation 0.219 0.898 0.821
combine intracavitary therapy with interstitial irradiation
Primary tumor irradiation dose 0.998 0.652 0.822
Primary tumor site† 0.715 0.626 0.922 (single-plane implant) because this enhances the probability
of tumor control without exposing the entire mucosa of the
* Brachytherapy (intracavitary, interstitial, or both) alone or vagina to high doses of irradiation. Hintz et al. (11), after
combined with external beam irradiation.
† analysis of 16 patients with carcinoma of the vagina who
Upper or lower vagina.
had local tumor control for a minimum of 18 months,
indicated that doses of irradiation greater than 140 Gy to the
III and IV tumors, the most significant morbidity was 2 upper vaginal mucosa and 98 Gy to the lower vaginal
rectovaginal and 1 vesicovaginal fistulae (about 10%). mucosa (both external-beam and brachytherapy contribu-
tions) result in a higher incidence of complications with
dose rates less than 0.8 Gy/h. In Stage I disease, external-
DISCUSSION beam irradiation, in addition to brachytherapy, is advocated
Radiation therapy is the treatment of choice for carci- only for infiltrating or poorly differentiated tumors that may
noma of the vagina, since it provides good tumor control, have a higher probability of lymph node metastases.
particularly in the small superficial tumors, and satisfactory The high incidence of pelvic failures in Stages II and III
functional results. Tumor control in Stage 0 and I is ade- is of major concern and suggests that a more radical ap-
quate with brachytherapy alone, and the addition of external proach is necessary in these advanced tumors. Either the
irradiation does not improve probability of local tumor external-beam dose should be increased or there should be
control. With adequate therapy, the survival rates of patients wider use of parametrial interstitial implants. Templates
with carcinoma of the vagina are comparable to those re- such as the Martinez multiple-site perineal template

Table 9. Carcinoma of the vagina: Morbidity of therapy

Stages 0 and I Stage II Stages III and IV

BT EB 6 BT BT EB 6 BT BT EB 6 BT
6 (n 5 40) (n 5 33) (n 5 11) (n 5 86) (n 5 3) (n 5 32)

Rectovaginal fistula 1 3 2
Vesicovaginal fistula 2 1
Rectal stricture 1* 1
Proctitis 2†
Rectal ulcer 1
Diverticulitis 1
Small bowel obstruction 1‡
Enterocutaneous fistula 1§
Bladder neck contracture 1¶
Urethral stricture 2 1
Cardiac arrest 1
Neuritis 1
Cystitis 1
Vaginal necrosis 1
Leg edema 1
Vaginal stenosis 1

BT 5 brachytherapy; EB6 5 external-beam irradiation with or without brachytherapy.

* Patient required partial colectomy.
†
One patient required colostomy.
‡
Colostomy required.
§
After pelvic exenteration (complication of surgery).
¶
Patient required transurethral resection.
44 I. J. Radiation Oncology ● Biology
(MUPIT) (12) are frequently used in an effort to improve
the dose distribution; they help to achieve better placement
of the radioactive sources in the pelvis. Integrated preoper-
ative irradiation and surgical resection when feasible should
be judiciously used (13). However, an aggressive approach
must be weighed against an expected greater incidence of
complications, and caution should be exercised because
many of these patients are of advanced age and may have
had previous pelvic surgery or may have medical conditions
that preclude overly aggressive therapy that will jeopardize
the patient’s welfare. Also, a greater frequency of compli-
cations has been observed with combined high doses of
irradiation and radical surgery in the treatment of pelvic
tumors (14). The high incidence of distant metastases
strongly emphasizes the need for effective chemotherapeu-
tic agents to improve survival in these patients. Efforts to
enhance the biologic effects of irradiation with chemical
sensitizers, chemotherapeutic agents such as cisplatin or
5-fluorouracil, which are being used in patients with ad-
vanced cervical and endometrial carcinoma, or hyperther-
mia, which is used in the management of patients with
advanced carcinoma of the uterine cervix, may be helpful.
We noted that, in carcinoma in situ and Stage I vaginal
carcinoma, tumor control in the pelvis was approximately
the same with brachytherapy alone or when combined with
external-beam irradiation. However, although not statisti-
cally significant, better tumor control (65%) was observed
in Stage II and III disease with the addition of external
irradiation in comparison with brachytherapy only (40%)
( p 5 0.11). These observations are similar to those of
MacNaught et al. (15), who in an analysis of 78 cases of
primary vaginal carcinoma (61 squamous cell carcinoma),
reported better tumor control and survival with combined
external-beam and interstitial treatment compared with
brachytherapy alone, despite the staging distribution being
in favor of the patients treated with brachytherapy only.
Similar observations were described by Leung and Sexton
(16) in a review of 103 patients.
Prempree and Amornmarn (9) and Puthawala et al. (17)
described tumor control rates in the pelvis ranging from
65% to 80% with a combination of external irradiation, and
when appropriate, paravaginal or parametrial interstitial im-
plant, in addition to intracavitary brachytherapy.
Chyle et al. (18) of the M.D. Anderson Cancer Center
updated the results in 301 patients with primary vaginal
carcinoma (271 squamous cell and 30 adenocarcinoma)
treated with radiation therapy. Brachytherapy alone was
used in 25 Stage I patients, external beam and brachyther-
apy in 38, and transvaginal orthovoltage X-rays in 2 pa-
tients. In Stage II patients, brachytherapy alone was used in
20, external beam and brachytherapy in 66, and external
irradiation alone in 36 patients. In patients with Stage III
disease, external-beam irradiation and brachytherapy were
used in 15, and external-beam irradiation alone was used in
45 patients. Mean dose was 74.7 Gy, and median dose was
70 Gy. The 10-year local recurrence rate was 17% in 37
Stage 0 patients, 15% in 59 Stage I patients, 18% in 104
● Physics Volume 44, Number 1, 1999
Stage II patients, 35% in 55 Stage III patients, and 60% in
16 patients with Stage IVA disease. As in our patients, the
10-year local recurrence rate was 22% in 80 patients with a
history of previous gynecologic malignancy and 22% in 191
without such a history. Ten-year survival rates were 78% in
Stage 0, 55% in Stage I, 51% in Stage II, 37% in Stage III,
and 40% in Stage IVA disease. As in our patients, the
10-year survival rate was 59% in patients with a history of
previous gynecologic malignancy and 49% in those without
such a history ( p 5 0.15).
Likewise, Urbanski et al. (19) reported on 125 patients
with vaginal carcinoma treated with a combination of ex-
ternal-beam and intracavitary irradiation. Five-year survival
was the same in 16 patients who had previous hysterectomy
(43.8%) and 109 without such a previous history (42.2%).
Kucera and Vavra (20) reported on 434 patients treated
with irradiation for invasive vaginal carcinoma, with more
details for 110 treated in more recent years. Intracavitary
radium was the standard primary treatment, to deliver
60 –90 Gy to the tumor and 30 Gy to the surrounding
vagina. For more advanced lesions, external irradiation (56
Gy to the lateral pelvic wall) was delivered. Contrary to our
experience, 5-year survival was higher in patients with
tumors in the upper third (21 of 35, 60%) compared with
those in the middle or lower third (19 of 51, 37.5%) or those
involving the whole vagina (5 of 24, 20.8%).
Kirkbride et al. (21) described results in 153 patients with
vaginal carcinoma, 15 of whom had carcinoma in situ; 128
were treated with radiation therapy (10 received irradiation
postoperatively and 26 concomitant chemotherapy). The
5-year cause-specific survival rates were 100% for Stage 0
(CIS), 77% for Stages I and II, and 56% for Stages III and
IV, results similar to ours. Late complications from treat-
ment were infrequent and were classified as severe in only
12 patients (10%). Univariate analysis indicated that size
and stage of tumor, histologic grade, patient age, and irra-
diation dose greater than 70 Gy were significant factors in
predicting survival, although in a multivariate analysis only
size and stage retained significance.
Fine et al. (22), in 55 patients with various stages of
vaginal carcinoma treated with radiation therapy, noted
fewer recurrences in patients who received doses higher
than 75 Gy. Eight patients developed complications.
Lee et al. (23) reported on the importance of overall
treatment time in 65 patients with primary carcinoma of the
vagina treated with radiation therapy. Five-year cause-spe-
cific survival was 100% in 6 patients with Stage 0, 94% in
17 with Stage I, 80% in 6 with Stage IIA, 39% in 10 with
Stage IIB, 29% in 10 with Stage III, and 62% in 6 patients
with Stage IVA disease. If the entire course of irradiation
was completed within 9 weeks, pelvic tumor control was
97%, in contrast to only 54% when treatment time extended
beyond 9 weeks ( p 5 0.0003). The incidence of major
treatment morbidity was 12%; in patients receiving less than
80 Gy, severe complications were noted in 9% (5 of 53),
whereas in those receiving doses higher than 80 Gy, mor-
bidity incidence was 25% (3 of 12).
 Carcinoma of the vagina
Boronow et al. (24) advocated a combination of surgery
and irradiation for treatment of patients with advanced vulvo-
vaginal cancer. Most of the 37 patients treated with this tech-
nique had advanced primary tumors of the vulva. The 5-year
survival rate for the primary cases was 75.6%. For patients
with recurrent disease, the 5-year survival rate was
62.6%. In most patients (94.8%), the bladder and rectum
were preserved.
Stock et al. (25) reported on 100 cases of vaginal carci-
noma; 50% of patients had hysterectomy before the diag-
nosis of vaginal cancer. Treatment consisted of surgery in
40 patients, radiation therapy in 47, and surgery plus irra-
diation in 13. With a median follow-up of 11.2 years, 5-year
disease-free survival was 67% for Stage I (23 patients), 53%
for Stage II (58 patients), 0% for Stage III (9 patients), and
15% for Stage IV disease (10 patients). The same authors (26),
in an analysis of 49 patients, emphasized the role of brachy-
therapy in the treatment of these patients.
REFERENCES
1. Hoffman MS, DeCesare SL, Roberts WS, et al. Upper vagi-
nectomy for in situ and occult, superficially invasive carci-
noma of the vagina. Am J Obstet Gynecol 1992;166:30 –33.
2. Perez CA, Camel HM, Galakatos AE, et al. Definitive irradi-
ation in carcinoma of the vagina: Long term evaluation of
results. Int J Radiat Oncol Biol Phys 1988;15:1283–1290.
3. Kottmeier HL. The classification and clinical staging of car-
cinoma of the uterus and vagina. J Int Fed Gynecol Obstet
1963;1:83–93.
4. Cutler SJ, Ederer F. Maximum utilization of the life table
method in analyzing survival. J Chron Dis 1958;8:699 –713.
5. Mantel N. Evaluation of survival data and two new rank order
statistics arising in its consideration. Cancer Chemother Rep
1966;50:163–170.
6. Dixon WJ, editor. NEWS for 1987 release of BMDP statistical
software. Los Angeles (CA): University of California Press;
1986.
7. Perez CA, Slessinger E, Grigsby PW. Design of an afterload-
ing vaginal applicator (MIRALVA). Int J Radiat Oncol Biol
Phys 1990;18:1503–1508.
8. Whelton J, Kottmeier HL. Primary carcinoma of the vagina.
Acta Obstet Gynecol Scand 1962;41:22– 40.
9. Prempree T, Amornmarn R. Radiation treatment of primary
carcinoma of the vagina: Patterns of failure after definitive
therapy. Acta Radiol Oncol 1985;24:51–56.
10. Brown GR, Fletcher GH, Rutledge FN. Irradiation of “in situ”
and invasive squamous cell carcinomas of the vagina. Cancer
1971;28:1278 –1283.
11. Hintz BL, Kagan AR, Chan P, et al. Radiation tolerance of the
vaginal mucosa. Int J Radiat Oncol Biol Phys 1980;6:711–
716.
12. Martinez A, Edmundson GK, Cox RS, et al. Combination of
external beam irradiation and multiple-site perineal applicator
(MUPIT) for treatment of locally advanced or recurrent pros-
tatic, anorectal, and gynecologic malignancies. Int J Radiat
Oncol Biol Phys 1985;11:391–398.
13. Perez CA, Korba A, Sharma S. Dosimetric considerations in
irradiation of carcinoma of the vagina. Int J Radiat Oncol Biol
Phys 1977;2:639 – 649.
14. Chau PM, Fletcher GH, Rutledge FN, et al. Complications in
high dose whole pelvis irradiation in female pelvic cancer.
Am J Roentgenol Radium Ther Nucl Med 1962;87:22– 40.
15. MacNaught R, Symmonds RP, Hole D, et al. Improved con-
● C. A. PEREZ et al. 45
Patients treated with radiation therapy for carcinoma of
the vagina should be carefully followed, because early de-
tection of recurrences allows a surgical procedure for sal-
vage that may result in subsequent cure of a significant
proportion of patients with recurring lesions.
As noted by Gore et al. (27), high-dose-rate vaginal
cylinders may provide dose distributions that have less
variation than low-dose-rate applicators; experience with
this modality is minimal.
Although radiation therapy is an effective method of treat-
ment for patients with early-stage vaginal carcinoma, there is
still a significant recurrence rate in patients with Stage IIB and
III tumors. More effective irradiation techniques, including
optimized dose distribution with external irradiation, and in-
terstitial as well as intercavitary brachytherapy or concurrent
administration of cytotoxic agents with irradiation, may be
necessary to improve the survival of these patients. Additional
clinical trials are strongly encouraged.
trol of primary vaginal tumors by combined external beam and
interstitial radiotherapy. Clin Radiol 1986;37:29 –32.
16. Leung S, Sexton M. Radical radiation therapy for carcinoma
of the vagina: Impact of treatment modalities on outcome:
Peter MacCallum Cancer Institute experience 1970 –1990. Int
J Radiat Oncol Biol Phys 1993;25:413– 418.
17. Puthawala A, Syed AMN, Nalick R, et al. Integrated external
and interstitial radiation therapy for primary carcinoma of the
vagina. Obstet Gynecol 1983;62:367–372.
18. Chyle V, Zagars GK, Wheeler JA, et al. Definitive radiother-
apy for carcinoma of the vagina: Outcome and prognostic
factors. Int J Radiat Oncol Biol Phys 1996;35:891–905.
19. Urbanski K, Kojs Z, Reinfuss M, et al. Primary invasive
vaginal carcinoma treated with radiotherapy: Analysis of
prognostic factors. Gynecol Oncol 1996;60:16 –21.
20. Kucera H, Vavra N. Radiation management of primary carci-
noma of the vagina: Clinical and histopathological variables
associated with survival. Gynecol Oncol 1991;40:12–16.
21. Kirkbride P, Fyles A, Rawlings GA, et al. Carcinoma of the
vagina: Experience at the Princess Margaret Hospital (1974 –
1989) (review). Gynecol Oncol 1995;56:435– 443.
22. Fine BA, Piver MS, McAuley M, Driscoll D. The curative
potential of radiation therapy in the treatment of primary
vaginal carcinoma. Am J Clin Oncol (CCT) 1996;19:39 – 44.
23. Lee WR, Marcus RB, Sombeck MD, et al. Radiotherapy alone
for carcinoma of the vagina: The importance of overall treat-
ment time. Int J Radiat Oncol Biol Phys 1994;29:983–988.
24. Boronow RC, Hickman BT, Reagan MT, et al. Combined
therapy as an alternative to exenteration for locally advanced
vulvovaginal cancer. II. Results, complications, and dosimet-
ric and surgical considerations. Am J Clin Oncol 1987;10:
171–181.
25. Stock RG, Chen AS, Seski J. A 30-year experience in the
management of primary carcinoma of the vagina: Analysis of
prognostic factors and treatment modalities. Gynecol Oncol
1995;56:45–52.
26. Stock RG, Mychalczak B, Armstrong JG, et al. The impor-
tance of brachytherapy technique in the management of pri-
mary carcinoma of the vagina. Int J Radiat Oncol Biol Phys
1992;24:747–753.
27. Gore E, Gillin MT, Albano K, et al. Comparison of high
dose-rate and low dose-rate dose distributions for vaginal
cylinders. Int J Radiat Oncol Biol Phys 1995;31:165–170.