1. Management of Aggressive Histologic Variants of Endometrial
Carcinoma at the Tom Baker Cancer Centre between 1984 and 1994
Peter S. Craighead, FFRad(T), FRCPC,* Khalid Sait, M.D., FRCPC,† Gavin C. Stuart, M.D., FRCPC,†
Keith Arthur, M.D., FRCPC,* Jill Nation, M.D., FRCPC,† Maire Duggan, M.D., FRCPC,‡ and
Dianlin Guo, Ph.D. (Epidem)§
*Department of Radiation Oncology, †Department of Gynecologic Oncology, and §Department of Preventive Oncology and Epidemiology, Tom Baker
Cancer Centre; and ‡Department of Pathology, Foothills Hospital and University of Calgary, 1331 29th Street NW, Calgary, Canada T3G 3A6
Received July 26, 1999
Objective. The aim of this study was to determine the patient
characteristics and outcome of patients with aggressive histologic
variants (AV) of endometrial carcinoma, including uterine papil-
lary serous carcinoma (UPSC), uterine clear cell carcinoma
(UCCC), and mixed type.
Methods and materials. All cases with AV histological type of
endometrial carcinoma from January 1984 to December 1994 at
the Tom Baker Cancer Centre were identified using the Alberta
Cancer Registry. Relevant data from the charts of these patients
were entered into a study database (Microsoft Excel) and analyzed
for presentation, demography, treatment parameters, and out-
come of treatment. All pathology was reviewed at the time of
diagnosis. Statistical analysis was performed using the S-plus
statistics computer program. Univariate and multivariate analyses
were used to assess independent prognostic factors using the Cox
proportional hazards model.
Results. A total of 103 patients with AV histological type were
identified and analyzed; there were 61, 31, and 11 cases of UPSC,
CCC, and mixed tumors, respectively. Sixty-three patients had
Stage I, 11 had Stage II, 15 had Stage III, and 14 had Stage IV
disease. The median age of patients was 67 years with a range of
36 to 86 years. Median follow-up was 60 months with a range of 36
to 156 months. The Cox proportional hazards model showed that
lymphvascular space invasion and stage are the two independent
prognostic factors affecting recurrence and survival. Forty six
percent of all cases underwent surgery alone, 39% underwent
treatment which included pelvic RT, and 17% underwent treat-
ment which included chemotherapy. Pelvic recurrence was re-
duced significantly by radiotherapy in Stages I, II, and III (19%
recurrence with no RT vs 7% recurrence with RT, P < 0.005).
Chemotherapy improved overall survival, but made little differ-
ence in distant relapse rates.
Conclusions. Stage Ia cases treated by surgery alone have a low
risk of relapse and need not be offered adjuvant systemic therapy
or pelvic radiation. Patients with Ib, Ic, II, and III have signifi-
cantly lower pelvic failure rates if treated with pelvic radiation, but
still have a high distant failure rate. Systemic therapy did not
significantly improve distant relapse-free survival, but did extend
overall survival. Stage IV patients usually died within 6 months
with a few responding to systemic chemotherapy. These results
suggest that there is a need for randomized trials for these
patients. © 2000 Academic Press
Key Words: uterine serous papillary carcinoma; endometrial
carcinoma; adjuvant radiotherapy.
INTRODUCTION
Endometrial cancer is the most common female genital tract
malignancy in North America and traditionally is regarded as
having a good prognosis. The uterine papillary serous variant
of endometrial cancer (UPSC) was first regarded as a distinct
entity in 1982 [1]. Several retrospective reviews have sug-
gested that UPSC is an aggressive variant of endometrial
cancer, with a clinical course similar to that of epithelial
ovarian cancer [2–6]. Carcangui and Chambers have shown
that uterine clear cell (UCCC) and mixed variants which con-
sist of greater than 75% UCCC or UPSC histology all exhibit
an aggressive natural history, with similar patterns of relapse to
UPSC [7]. For this reason many centers treat UPSC, UCCC,
and mixed tumors as aggressive variants (AV), with ap-
proaches that attempt to control potential pelvic and extrapel-
vic disease.
Since 1984 the Gynecologic Oncology Tumor Board at the
Tom Baker Cancer Centre has recommended that all patients
with AV undergo staging laparotomy to remove the uterus,
fallopian tubes and ovaries and assess pelvic and paraortic
nodes, omentum, and peritoneal cytology. A proportion did not
undergo full staging because of comorbid illness. However,
there has been less consensus on what adjuvant therapies to use
postsurgery. The standard recommendation for surgical Stage
Ia cases at this center has been observation. Surgically staged
Ib, Ic, and II cases usually have been recommended for whole
pelvic plus intracavitary radiotherapy (pelvic RT) after sur-
gery.
It was the recommendation that certain Stage II cases who
did not undergo full staging and all Stage III cases be offered
chemotherapy plus pelvic radiotherapy. This involved the ad-
Gynecologic Oncology 77, 248–253 (2000)
doi:10.1006/gyno.2000.5746, available online at http://www.idealibrary.com on
2480090-8258/00 $35.00
Copyright © 2000 by Academic Press
All rights of reproduction in any form reserved.

2. ministration of four cycles of cisplatin, Adriamycin, and cy-
clophosphamide (PAC), followed by pelvic RT and a further
four cycles of PAC after RT. Stage IV patients were recom-
mended for PAC chemotherapy at presentation only if they had
low bulk or symptomatic disease and were able to tolerate this
regime.
Multiple claims of treatment success have been published in
small uncontrolled series. Management in these series has
ranged from observation to adjuvant pelvic RT in Stage I cases
and from whole abdominal RT to chemotherapy in Stage II and
III cases. A randomized study currently underway by the
Gynecology Oncology Group (GOG) is comparing whole ab-
dominal RT with systemic chemotherapy in higher risk UPSC
cases. We performed this retrospective study in an attempt to
increase our understanding of the clinical outcome of this
disease, especially as it relates to a combination of UPSC,
CCC, or mixed variants. We also assessed prognostic factors
and audited compliance to the treatment guidelines.
METHOD AND MATERIALS
The demographics, details of pathology, treatment parame-
ters, and results of treatment for patients with UPSC, UCCC, or
mixed uterine carcinoma between 1984 and 1995 were ex-
tracted from registry charts and entered into a database (see
Table 1), excepting one case who also had adenosquamous
carcinoma of the cervix. Only AV cases were included in this
review. All cases had their pathology reviewed within the
Gynecology Tumor Board Forum.
Treatment
Surgery. The preferred surgery included total abdominal
hysterectomy, bilateral salpingo-oophorectomy, omentectomy,
lymph node sampling, and peritoneal cytology assessment.
Radiotherapy. Radiotherapy consisted of whole pelvic ra-
diation to a dose of 45 Gy in 25 daily fractions, utilizing high
energy photon therapy in a four-field arrangement to include
the vaginal cuff as well as the obturator and internal/external
iliac nodes bilaterally. This was followed 1 week later by a
single low-dose-rate cesium insert to a vaginal surface dose of
20 Gy.
Chemotherapy. In selected cases chemotherapy was deliv-
ered using 50 mg/m2
cisplatin, 50 mg/m2
doxorubicin, and 600
mg/m2
cyclophosphamide intravenously every 3 weeks, to-
gether with antiemetics and hydration for eight cycles. When
administered in a sandwich approach, the pelvic radiotherapy
was delivered between the fourth and fifth cycles.
Statistical Analysis
Statistical analysis was performed using the S-plus statistics
program. Analysis of differences between baseline character-
istics was performed using the Kruskall–Wallis rank sum test.
The difference between proportions was tested using the ␹2
test. Survival estimates were calculated using the Kaplan–
Meier method and the significance between survival tested
using the log-rank test. Multivariate analysis was performed
for overall survival and recurrence-free survival using the Cox
proportional hazards model.
RESULTS
Between 1984 and 1994 there were 936 patients with endo-
metrial cancer registered at the Tom Baker Cancer Centre, with
888 (95%) of them undergoing pathology review and tumor
board discussion. One hundred four cases (11%) were shown
TABLE 2
Sites of Recurrence
Stage I Stages II and III
Surgery
(36 cases)
S ϩ RT
(27 cases)
Surgery
(5 cases)
S ϩ RT
(9 cases)
S ϩ Chem
(6 cases)
S/Ch/RT
(6 cases)
No. of pts with recurrence 5 5 3 4 4 3
Pelvic recurrence
5/36
(14%)
1/27
(4%)
1/5
(20%)
1/8
(12.5%)
3/6
(50%)
1/6
(17%)
Distant disease
4/36
(11%)
4/27
(15%)
2/5
(40%)
4/8
(50%)
1/6
(17%)
3/6
(50%)
TABLE 1
Treatment Characteristics (Total ‫؍‬ 103 Cases)
Stage I Ia Ib Ic II III IV
Surgery only 36 18 14 4 2 3 6
Surgery and RT 27 5 12 10 5 3 1
Surgery ϩ ch 0 0 0 0 2 4 8
Surgery ϩ RT/ch 0 0 0 0 2 4 0
Total 63 23 26 14 11 14 15
Note. Surgery usually included total abdominal hysterectomy, bilateral
salpingo-oophorectomy, and peritoneal cytology. In a proportion it included
omentectomy and nodal assessment. RT, pelvic radiotherapy; Ch, systemic
chemotherapy using cisplatin, doxorubicin, and cyclophosphamide.
249MANAGEMENT OF ENDOMETRIAL CARCINOMA VARIANTS

3. to have aggressive variants. One of these cases was also shown
to have primary adenosquamous carcinoma of the cervix and
was thus removed from survival analysis.
Patient Characteristics
The median age of patients was 67 years, with a range of 36
to 86 years. Nulliparity was evident in 63 cases, but there was
no association between the development of this with exoge-
nous estrogens, diabetes mellitus, hypertension, or obesity.
Median follow-up was 60 months (range 36 to 156 months).
There were 61, 31, and 11 cases of UPSC, UCCC, and mixed
tumors, respectively. Division by stage showed 63 patients
with Stage I, II with Stage II, 15 with Stage III, and 14 with
Stage IV. There was no substantial difference between propor-
tions of patients with different variants per stage.
Treatment Results
Table 1 shows treatment by stage. Forty-seven (46%) of all
cases underwent surgery only, 42 cases (41%) underwent treat-
ment which included pelvic RT, and 20 cases (19%) underwent
treatment which included chemotherapy. In Stage Ia cases
managed by postsurgery surveillance there was an 11% recur-
rence rate, with all of these being either vault or pelvic in
nature. Pelvic recurrence was reduced significantly by radio-
therapy in Stages Ib/Ic, II, and III (24% of 29 pts treated
without RT vs 13% of 36 pts who received RT, P Ͻ 0.001).
Chemotherapy was associated with an improvement in overall
survival but had little effect on pelvic control or distant relapse
rates (see Table 2).
Adherence to Guidelines (See Table 2)
Sixty-five percent of the cases were treated according to the
guidelines and 35% outside of them. Surgery alone was used
appropriately in most Stage Ia and IV cases, and the majority
of Stage Ib, Ic, and II cases (88%) were treated within the
guidelines. There was adherence to the recommendation for RT
with/without PAC in 64% of Stage II cases and for RT/PAC in
58% of Stage III patients. In Stage I to III cases there was a
statistically significant reduction in local recurrence (8.5 vs
38%, P Ͻ 0.001), but a nonsignificant difference in distant
disease (27 vs 34%, P ϭ 0.4) for those treated within and
outside the guidelines. Other than in Stage Ia, where no reason
was given why 20% of cases received radiotherapy, the major
reason for failure to comply with recommendations was age/
poor performance status.
Overall and Recurrence-Free Survival, Local Control, and
Risk Factors (Tables 3 and 4 and Figs. 1–4)
Stage IV patients [14] had a median survival of 4.5 months,
with the difference between those who did not receive pallia-
tive chemotherapy (2.48 months) against those who did (7.96
months) being significant (P ϭ 0.0046, log rank). Whether this
is due to treatment or that only better performance status
patients were given chemotherapy is difficult to determine. The
remainder of the analysis was limited to Stage I to III cases.
The Cox proportional hazards model for death showed in-
creased stage, lymphvascular invasion (LVS), and the lack of
chemotherapy to be significant predictive factors for death.
Stages II, III, and IV had relative hazards of death of 3.9 (CI
1.1 to 14), 5.3 (CI 1.7 to 17), and 42 (CI 9.4 to 188) compared
TABLE 3
Multivariate Analysis for Recurrence-Free Survival
Coef Exp (coef) SE (coef) Z P
LVS 1.59 4.912 0.480 3.315 0.00092
Staging s2 1.69 5.446 0.591 2.866 0.00420
Staging s3 1.93 6.873 0.619 3.116 0.00180
Chemotherapy Ϫ0.54 0.583 0.577 Ϫ0.936 0.35000
Radiation Ϫ1.19 0.306 0.485 Ϫ2.444 0.01500
TABLE 4
Multivariate Analysis for Overall Survival
Coef Exp (coef) SE (coef) z P
LVS 1.645 5.183 0.518 3.18 0.015
Staging s2 1.365 3.917 0.653 2.09 0.037
Staging s3 1.672 5.322 0.586 2.86 0.043
Staging s4 3.739 42.059 0.767 4.88 0.0000011
Chemotherapy Ϫ1.509 0.221 0.552 Ϫ2.74 0.0062
Radiation Ϫ0.523 0.593 0.515 Ϫ1.02 0.31
250 CRAIGHEAD ET AL.

4. with Stage I (P ϭ 0.0000011). The presence of LVS had a
relative hazard of death of 5.2 (P Ͻ 0.0015, C.I. 1 to 14), and
chemotherapy significantly reduced the relative risk (RR) of
death to 0.22 (P Ͻ 0.034, CI 0.08 to 0.65).
Thirty-five percent of Stage Ib to III cases (23 of 66)
experienced relapse, with 72% of these having a form of
extrapelvic component. A Cox model for recurrence-free sur-
vival showed the presence of LVS to transfer a relative risk of
recurrence of 4.9 (P Ͻ 0.00092, 1.9 to 12.6) and Stages II and
III to have an RR of 5.4 and 6.9 compared with Stage I (P ϭ
0.0042). Radiotherapy had a relative risk of recurrence of 0.31
compared with no pelvic RT (P Ͻ 0.015, CI 0.11 to 0.79).
DISCUSSION
UPSC traditionally has been quoted as having a prevalence
of approximately 2% of endometrial cancers, but little has been
published on the prevalence of a combination of UPSC,
UCCC, or mixed variants. These results show that they con-
stituted 11% of malignant endometrial cases in our center, with
UPSC being more common than the two other diagnoses (60 vs
40%). Age and other patient characteristics are comparable to
other types of endometrial cancer. Because of small numbers
we did not compare outcome data according to UPSC, UCCC,
or mixed variant categories.
Significant risk factors for survival within this series in-
cluded increasing stage, lymphvascular invasion, and the lack
of chemotherapy, which have been reported previously [6–8].
Of interest is the finding that pelvic RT significantly improved
pelvic relapse-free rates in Stages Ib, Ic, II and III, making it an
important consideration for patients in these stages. Given that
patients in Stages I to III had significantly improved local
control when managed within the guidelines, it is necessary for
us to find ways of addressing lack of compliance. This may be
partly addressed by offering vault brachytherapy alone in poor
performance cases.
The results for intermediate risk cases (Stages Ib, Ic, and II)
support the use of radiotherapy postsurgery, but show that
distant relapses remain a problem (14% in Stage I and 35% in
Stage II). The use of pelvic RT has also been supported by the
results of the Gynecology Oncology Group 99 protocol, which
revealed a significant reduction in pelvic/vaginal recurrences
with radiotherapy in normal endometrial subtypes [9]. Despite
FIG. 2. Survival by treatment Log-rank test, P ϭ 0.000124.
FIG. 1. Survival by staging Log-rank test, P Ͻ 0.000001.
251MANAGEMENT OF ENDOMETRIAL CARCINOMA VARIANTS

5. the reduction in pelvic recurrence rate with radiotherapy in the
GOG 99 trial, there was no difference in overall survival, with
a similar finding in our study.
Several retrospective studies have recommended the use of
whole abdominal radiotherapy (WART) in early UPSC, but
have not provided sufficient data to determine where this could
be applied safely [10–12]. We have not used this approach
because of the high risk of distant relapse and the reasonable
pelvic control achieved by pelvic RT. The benefit of WART is
questionable in these patients, given the extra-abdominal re-
lapse rates of 30%.
No randomized trials comparing adjuvant treatments for
UPSC or AV have yet to be reported. Our results have affirmed
the use of observation only for Stage Ia cases treated by full
staging laporotomy, with relapse occurring in only 11% of
these cases. Although this could be reduced by vaginal brachy-
therapy, approximately 60% of these would be salvageable by
pelvic RT and/or surgery. This questions the need to apply
adjuvant radiotherapy in this subgroup of patients. Several
others have published excellent results for observation in Stage
Ia UPSC cases [13, 14].
Although some centers treat all AV cases as if they were
ovarian cancer, and therefore use adjuvant chemotherapy
alone, our results suggest that chemotherapy needs to be sup-
plemented by pelvic RT if pelvic control is to be optimized in
Stages II and III [15]. From our experience it is recommended
that Stage Ic and II cases whose tumors have lymphvascular
space involvement and all Stage III cases be considered for
systemic treatment plus pelvic RT, for they are at a reasonable
risk of both local and distant relapse. Until a systemic approach
that obviates local recurrences in a high proportion of cases can
be found, it would be pertinent to continue this combination of
modalities.
Other studies show low response rates to systemic therapy in
patients with macroscopic tumor and that the major benefit
from chemotherapy appears to be in patients who have low-
bulk disease [16–18]. Our results show that the use of chemo-
therapy alone did not reduce the risk of relapse (either local or
distant). However, it was associated with a small improvement
in overall survival. This was probably related to most patients
in our series treated with chemotherapy having advanced or
bulky disease.
FIG. 4. Recurrence-free survival by treatment (Stages I to III). Log-rank
test, P ϭ 0.00043.
FIG. 3. Recurrence-free survival by staging. Log-rank test, P ϭ
0.000174.
252 CRAIGHEAD ET AL.

6. Although the standard adjuvant and palliative systemic regime
throughout North America has usually been PAC [16], our center
now has changed to single-agent carboplatin. This was done
because of the significant morbidity experienced on PAC in this
older patient group and with evidence suggesting that carboplatin
has efficacy in other Mu¨llerian tract malignancies. The morbidity
of PAC included severe myelosuppression and emesis, which
resulted in prolonged delays during treatment. We intend to con-
duct an analysis after the first 25 patients on carboplatin to assess
the control and morbidity rates associated with this approach. We
await the results of studies assessing other systemic agents, but
until such trials have been approved will continue to use systemic
agent/s with the least toxicity.
In conclusion, Stage Ia cases should not be offered adjuvant
therapy routinely after surgery, since they have a low risk of
relapse, most of which are in the pelvis. There is a definite local
benefit to pelvic radiotherapy in intermediate-risk cases (Stages
Ib, Ic and II) who have had comprehensive surgery. Finally,
cases with lymphvascular invasion in Stages Ib, Ic, II and all
Stage III cases are at a higher risk of both pelvic and distant
relapse and should be considered for combined modality ther-
apy. This should include surgery, chemotherapy, and pelvic
radiotherapy, although the latter affects only pelvic control.
This study does not support the use of such an approach for
Stage Ib and Ic cases who don’t have LVS. We strongly
recommend that multicenter clinical trials be performed, for we
believe that it is only through this that we will be able to
advance the treatment of this group of patients in the future.
REFERENCES
1. Hendrickson M, Ross J, Eifel P, Martinez A, Kempson R: Uterine papil-
lary serous carcinoma: a highly malignant form of endometrial adenocar-
cinoma. Am J Surg Pathol 6:93–108, 1982
2. Goff BA, Kato D, Schmidt RA, Ek M, Ferry J, et al.: Uterine papillary
serous carcinoma: patterns of metastatic spread. Gynecol Oncol 54:264–
267, 1994
3. Dunton C, Gulnar G, McFarland M, Hernandez E: Uterine papillary serous
carcinoma: a review. Obstet Gynecol Surv 46(2):97–102, 1991
4. Nicklin J, Copeland L: Endometrial papillary serous carcinoma: patterns
of spread and treatment. Clin Obstet Gynecol 39(3):680–695, 1996
5. Sutton G, Brill L, Michael H, Stehman F, Ehrlich C: Malignant papillary
lesions of the endometrium. Gynecol Oncol 27:294–304, 1987
6. Carcangui M, Chambers J: Uterine papillary serous carcinoma: a study of
108 cases with emphasis on the prognostic significance of associated
endometrial carcinoma, absence of invasion, and concomitant ovarian
carcinoma. Gynecol Oncol 47:298–305, 1992
7. Carcangiu M, Chambers J: Early pathologic stage clear cell carcinoma and
uterine papillary serous carcinoma of the endometrium: comparison of
clinico pathological features and survival. Int J Gyn Pathol 14:30–38,
1995
8. Gitsch G, Friedhandler M, Wain GV, Hacker N: Uterine papillary serous
carinoma: a clinical study. Cancer 75(9):2239–2243, May 1995
9. Roberts J, Brunetto V, Keys H, Zaino R, Spiros N, et al.: A phase III
randomized study of surgery vs surgery plus adjuvant radiotherapy in
intermediate risk endometrial carcinoma (GOG 99), Abstract 35, Proceed-
ings of the Society of Gynecol Oncologists, 1998
10. Mallipeddi P, Kapp D, Teng N: Long term survival with adjuvant whole
abdominopelvic irradiation for uterine papillary serous carcinoma. Cancer
71:3076–3081, 1993
11. Parkash L, Carcangui M. Uterine papillary serous carcinoma and radio-
therapy. Cancer 69(2):494–501, 1992
12. Frank A, Tseng P, Haffty B, Papadopoulos DP, Kacinski B, et al.:
Adjuvant whole abdominal radiation therapy in uterine papillary serous
carcinoma. Cancer 68:1516–1519, 1991
13. Grice J, Ek M, Greer B, Koh WJ, Muntz H, et al.: Uterine papillary serous
carcinoma: evaluation of long term survival in surgically staged patients.
Gynecol Oncol 69:69–73, 1998
14. Aquino Parsons C, Lim P, Wong F, Milenberger M: Papillary serous and
clear cell carcinoma limited to endometrial curettings in FIGO Stage Ia
and Ib endometrial adenocarcinoma: treatment implications. Gynecol On-
col 71:83–86, 1998
15. Christman E, Kapp D, Hendrickson M, Howes A, Ballon S: Therapeutic
approaches to uterine papillary serous carcinoma: a preliminary report.
Gynecol Oncol 26:228–235, 1987
16. Levenback C, Burke T, Silva E, Morris M, Gershenson D, et al.: Uterine
papillary serous carcinoma (UPSC) treated with cisplatin, doxorubicin and
cyclophosphamide (PAC). Gynecol Oncol 46:317–321, 1992
17. Rosenberg P, Boeryd B, Simonsen E: A new aggressive approach to high
grade endometrial cancer of possible benefit to Stage I uterine papillary
cancer. Gynecol Oncol 48:32–37, 1993
18. Price F, Chambers S, Carcangui ML, Kohorn E, Schwartz P, et al.:
Intravenous cisplatin, doxorubicin, and cyclophosphamide in the treatment
of uterine papillary serous carcinoma (UPSC). Gynecol Oncol 51:383–
389, 1993
253MANAGEMENT OF ENDOMETRIAL CARCINOMA VARIANTS