GYNECOLOGIC ONCOLOGY 2,368-376 (1974)

Adjuvant Progestogen Therapy in the

Primary Definitive Treatment of Endometrial Cancer1

GEORGE C. LEWIS, JR., M.D., NELSON H. SLACK, PH.D.,

RODRIQUE MORTEL, M.D., AND IRWIN D. J. BROSS, PH.D.

Received March 18, 1974

On the basis of reported successful hormone therapy for recurrent cancer of the endometrium, a multi-institutional program was established in 1963 and termi- nated in 1968 to study the adjuvant use of a progestogen in the primary definitive treatment of uterine carcinoma. The research protocol restricted entry to pa- tients with disease limited to the uterine corpus as established by clinical evalua- tion, surgery, and pathology. Option of surgery alone or surgery plus irradiation was permitted. Depo-Provera and placebo given over 14 wk were randomly as- signed in such a way that 285 patients received hormone; 287 served as controls. Survival analysis at 4 yr indicated no significant contribution to results by the ad- juvant therapy. Overall excellent survival for treated patients and controls ap- peared to be the result of the high degree of patient selection incorporated in the protocol design. The influence of patient selection and of variations of disease- related parameters upon the results is discussed especially as it may affect future programs dealing with comparable patient material.

Repeated clinical experiences have underscored the value of proges- terone-like drugs in the first line of attack upon recurrent cancer of the uterine corpus. Recurrent cancer of the uterine corpus means to the therapist and patient that every reasonable modality of treatment including hormones must be considered for use though the benefit may only be partial or tran- sient. In contrast, for the majority of instances the primary management of cancer of the uterine corpus requires a vigorous all-out effort to erradicate the disease by well-established definitive therapeutic procedures, consisting of some form of radiation with surgery or surgery alone. The question that arises naturally is whether anything may supplement or even replace time honored techniques for tumor control. The relative safety of progestogens

’ This cooperative investigation was supported by the U.S. Public Health Services under Grants IRIOCA 12478 CA07353, and CA10378. There were 37 collaborating institutions and the collaborators were as follows: Drs. E. J. Alban Jr., D. P. Bernard, E. R. Carrington, E. A. Chasteney, R. E. Collier, R. J. Collins, P. S. Copit, J. D. Corbit, B. Czernobilsky, R. R. JeAlvarez, R. E. Dean, D. G. Decker, J. delRegato, J. H. Ferguson, L. V. Goye, L. Hamilton, D. M. Haynes, W. M. Heyl, E. C. Hill, M. M. Hreshchyshyn, R. R. Hughes, H. A. Kaminetsky, W. C. Keettel, W. P. Kosar, J. H. Lee, Jr., B. P. Lelich, E. B. Linton, F. R. Lock, J. E. Lynch, B. V. MacFadyen, J. T. Mallams, D. J, Marchant, H. C. McDuff, L. McGown, A. G. W. McLeod, J. J. Mikuta, J. B. Montgomery, S. Monticelli, C. W. Muckle, J. F. Nolan, D. Palermino, W. H. Pearse, J. Price, C. L. Randall, J. W. Roddick, J. Rominger, F. C. Schaeffer, P. Schreier, R. Scott, C. B. Sigel, M. L. Stone, 0. E. Talledo, E. S. Taylor, N. J. Thompson, L. VanVoorhis, L. H. Wardner, L. L. Weber, J. C. Wells, C. J. Woeppel

Copyright @ 1974 by Academic Press, Inc. All rights of reproduction in any form reserved.

ADJUVANT PROGESTOGEN THERAPY 369

and the apparent favorable responses of recurrent endometrial cancer to these agents suggest that such an approach could be worthwhile in primary treatment. It is intended in this article to consider the relative impact of a form of hormone therapy and other parameters of management upon results and to provide the findings from an investigation of adjuvant chemotherapy utilizing Depo Provera.’

MATERIALS AND METHODS

Between October 1, 1963 and November 30, 1968, 956 patients with adenocarcinoma or adenoacanthoma of the uterine corpus (Table I) were en- tered into the study of the adjuvant use of Depo Provera according to a pro- tocol described in an earlier report [I]. Physicians in 37 participating institu- tions used one of two options for definitive therapy. These were: Option A, intracavitary radium and surgery, and Option B, surgery alone.

A third option, C, involving external therapy for stages 2, 3, and 4 was dropped because of inadequate enrollment. As noted in the preliminary report, the protocol limited the surgery to total hysterectomy and bilateral salpingo-oophorectomy for patients considered eligible. Eligibility also required meeting a range of specifications for radiation therapy. Exclusion resulted if the patient had cancer, either prior or concomittant, other than adenocarcinoma or adenoacanthoma of the uterus, except for cancer of the skin. Patients were also excluded because of prior chemotherapy or radia- tion.

The adjuvant therapy varied according to option. In option A, 500 mg Depo Provera or placebo was administered the first day of radium therapy, and then once weekly thereafter for a total of 14 wk. Chemotherapy for op- tion B varied from that for option A in that the initial dose was 1 g of Depo Provera or placebo administered 3-4 days before hysterectomy. Thereafter for 14 wk the administration of drug or placebo was identical to option A. Randomization of drug and placebo and data processing were carried out at the Roswell Park Memorial Institute under the direction of two of the co- authors (NHS and IDJB).

A combination of lag in follow-up and termination of funding in 1972

TABLE I

DISPOSITION OF ADJUVANT

CHEMOTHERAPY PATIENTS

category No. patients

Entered 956

Excluded 318

Acceptable 638

Option C (excluded) 9

Insu5cient information 53

Available for analysis 574

2 The Upjohn Company, Kalamazoo, MI.

370 LEWIS ET AL.

made it impossible to provide adequate data for an analysis based on 5 yr of follow-up. Sufficient information was available to permit an analysis of ma- terial covering 4 yr of observation; for this reason the conclusions are based upon 4-yr analysis although the illustrations represent the accumulated 5-yr data for patients when it was available.

As each patient was entered into the project, a wide range of observations pertaining to history, physical evaluation, laboratory findings, and treatment were recorded. A preliminary report has presented basic information relative to 187 patients. The material for the overall series has been collected, re- viewed, and compared with the observations provided for the preliminary report. The result of this comparison is the observation that the entire series differs little in basic findings from the group of patients described in the pre- liminary report. There is in progress a detailed analysis of a variety of factors involved in the radiation therapy. It is intended to report this additional phase of the study later.

RESULTS

As the investigation progressed, a number of patients were dropped from the registry because of deviations from the protocal (Table II). Of those eliminated, about 43% were dropped because the extent of tumor was found to be beyond the uterine corpus. With the elimination of 55 patients for grossly insufficient information, the number of patients available for analysis was 574. It should be noted that the amount and type of data recorded for each patient varied. These scattered irregularities in recording did result in some variations in totals and subtotals as will be noted in the tables and illustrations.

In the preliminary report, the most striking observation was the contrast in distribution of tumor extent in the surgically removed uteri between ir- radiated and nonirradiated patients. The same observation holds for the en- tire group of 574 patients. In option A (Table III) nearly 46% of patients had no residual tumor as opposed to 7% in option B. The incidence of deep myometrial involvement was 25% in option B, 8% in option A.

TABLE II REASONS FOR EXCLUDING PATIENTS FROM ENDOMETRIUM STUDY

Reason for exclusion Number of Percentage

patients of total

Ineligible-protocol para. 3.1.2; Stage > I 136 42.8 Therapy refused by patient, or removed by invest. 62 19.5 No cervical biopsy, Option A 39 12.3 Death before completion of initial study 18 5.7 Unacceptable surgical procedures 19 6.0 Unacceptable radiotherapy-protocol dev. 28 8.8 Unacceptable drug therapy-protocol dev. 13 4.1 Postadjuvant therapy without recurrence 3 0.9

ADJUVANT PROGESTOGEN THERAPY 371

TABLE III LEVEL OF UTERINE TUMOR PENETRATION NOTED IN SURGICAL SPECIMENS

AFTER HYSTERECTOMY FOR ENDOMETRIAL CANCER IN PATIENTS IRRADIATED AND NOT IRRADIATED PREOPERATIVELY

Level of penetration

Option A Option B preoperative irradiation no irradiation

Patients Percentage Patients Percentage

No tumor 173 46 13 7 Mucosa 89 24 55 28 Myometrium < l/3 85 22 78 40 Myometrium 2 l/3 31 8 48 25

Total 378 100 194 100

There were 10 patients with a second primary tumor. Three were alive at last contact. These patients were included in analysis of survival and treated as withdrawals at the time the second primary was diagnosed. The types of second primaries are given in Table IV.

The accuracy of pathological verification of diagnosis for tumors of patients entered into the study was investigated through the assistance of Dr. Alex- ander Sedlis and members of his pathology committee in the Gynecologic Oncology Group. The slides of 149 patients were checked by four pathol- ogists who received subsections so that each subsection was seen by two of the pathologists. The reviewers agreed on only 10 patients that failed to meet eligibility requirements (no tumor on dilatation and curettage or Stage II). This constituted a 2: 1 judgment as the institution’s pathologist con- sidered the patients eligible. There were 34 cases in which one reviewer questioned eligibility making a 1: 2 judgment for the patient’s inclusion. The questionable cases did not have a different recurrence or survival experience from the remainder of the patients; hence, it would seem not justified to remove these patients from the analysis of the results or to be greatly con- cerned about such problems in the remainder of the series not subject to special review.

TABLE IV SECOND PRIMARY CANCER

Site Patients

Breast Esophagus Common bile duct A.M. leukemia Hodgkin’s disease Pancreas Bladder Transverse colon

372 LEWIS ET AL.

TABLE V RESULTS OF TREATMENT BY OPTION

Option No. pt

% Free of disease

4 yr

% Alive with recurrence or

dead of disease %

Non Ca deaths

A 350 86.5 8.3 5.2 B 192 86.5 9.4 4.1

The results of treatment by option is given in Table V. Survival free of cancer, for the two options, at 4 yr was identical. The differences set forth in Table V, though not significant, indicate that a slightly greater proportion in option B had cancer or died of it as compared to option A.

The distribution of patients according to tumor penetration and therapy (Table VI) revealed discrepancies in incidence of patients within subcat- egories receiving Depo-Provera or placebo. In all 285 patients received Depo-Provera and 287 patients placebo. Table VI indicates that in both options A and B more patients in the category “myometrium <l/3” had pla- cebo. More patients in the category “a l/3” received Depo-Provera. Addi- tional analysis of the relation of tumor extent and survival (Figs. 1 and 2) in- dicate that in the nonirradiated patients there is little difference for survival by depth of penetration except that the “no residual” group had a 100% sur- vival. For the irradiated patients the “3 l/3” penetration category had much less favorable results. No other distribution differences that might bias the results relative to chemotherapy were apparent from other analyses. Thus, it might be possible that the impact of tumor extent would be reflected in the survival data pertaining to chemotherapy.

Figure 3 depicts the probability of survival by option and by the form of adjuvant therapy. At 4 yr the survival by option and by chemotherapy (Table VII) indicates essentially no difference between options. However, in both options the placebo group has a somewhat better survival, Depo-Provera 87%, placebo 91%, option A; Depo-Provera 87%, placebo 93%, option B.

TABLE VI DISTRIBUTION OF TUMOR PENETRATION BY THERAPY

Number of patients

Depth of tumor penetration

Option A

Provera Placebo

Option B

Provera Placebo

Mucosa only 47 42 29 26 Myometrium < l/3 37 48 30 48 Myometrium > l/3 19 8 28 17 Serosa 3 1 0 3 No residual tumor 89 84 3 10

Total 195 183 90 104

------------___ --S_____

BO- l “-..............................”.....

70 - x NO TUMOR 173

60- - MUCOSA 86

50- MYOMETRIUM < 113 B2 __-

. MYOMETRIUM > l/j 40-

31

30-

20

IO 1 I

6 12 I6 24 30 36 42 46 54 60

MONTHS AFTER SURGERY

FIG. 1. Probability of survival by depth of tumor penetration at surgery for irradiated patients.

NO TUMOR I2

_ MUCOSA 54

MYOMETRIUM <l/3 73

MYOMETRIUM >I/3 46 -

0

6 I2 I6 24 30 36 42 46 54 60

MONTHS AFTER SURGERY

FIG. 2. Probability of survival by depth of tumor penetration at surgery for nonirradiated pa-

IO0

so - x

80

70

60

t OPTION

- PROVERA

---- I A

193

PLACEBO 180

X PROVER0

I

so B

o PLACEBO loo

563

./ , , ) , , , , , , ,

6 I2 I8 24 30 36 42 46 54 60

MONTHS ON STUDY

FIG. 3. Probability of survival for patients in Options A and B by therapy.

373

374 LEWIS ET AL.

TABLE VII SURVIVAL IN OPTIONS A AND B BY

THERAPY AT 4 YR

Total no. treated

Surviving

no. %

Option A Provera Placebo

Option B Provera Placebo

193 168 87 180 164 91

90 78 87 100 93 93

The overall figures are 87% for Depo-Provera; 92% for placebo. While the difference is not significant at this level of survival, the results favoring placebo really suggest that tumor extent is more reflected than is chemo- therapy effect, if there is any. Certainly, if adjuvant chemotherapy has a role to play, it is probably hidden by the effects of one or more of the parameters of the disease process.

Further analysis relative to radium dosage and results, to be reported later, do not appear to be correlated with the outcome of chemotherapy.

DISCUSSION AND CONCLUSIONS

Prior to the onset of the investigation of adjuvant chemotherapy the pub- lished results for treatment of carcinoma of the uterine corpus Stage I in the operable patient were noted on the average to be about 71% 121 for survival at 5 yr. This meant that the potential for improvement from altered treatment might have up to a 30% limit. The successful initial use of progestogens 13-51 in several varied programs led to the interest on the part of the Cancer Chemotherapy National Service Center and a collaberating group of inves- tigators in the possibility of increasing survival by hormone therapy. An es- timated lo%-15% improvement in the quality of results did not seem un- reasonable if the response rates for advanced endometrial cancer can be considered applicable [6-g].

In considering projects for a most appropriate investigation of primary chemotherapy of endometrial cancer the design committee after eliminating the medically inoperable patient separated potential patients into three cat- egories as follows:

1. Patients with disease confined to the uterus as established by clinical study, surgery, and pathology.

2. Patients with disease beyond the corpus uteri, but confined to the pelvis as established by clinical study, surgery, or pathology.

3. Patients with disease beyond the pelvis. The first category was selected for the study because it provided the most

“uniform” group of cases and it permitted a test of control of disease spread by trauma of diagnosis and therapy, at least in theory. For patients with

ADJUVANT PROGESTOGEN THERAPY 375

disease apparently limited to the uterus, the risk of distant disease es- tablished prior to diagnosis is balanced by the hazard of iatrogenic spread of tumor. For the other two categories known disease spread beyond the uterus overshadows the importance of tumor dissemination by the operative proce- dures. The use of hormone therapy for the first category was, therefore, set as an adjuvant procedure of relatively short duration. It was timed to include or come as close as possible to the manipulative procedures normally em- ployed.

In this series representing the contributions of many physicians it appears at first glance that placebo produces better results than specific adjuvant hor- mone therapy. A check of the observations suggests that variations in drug and placebo allocation to certain categories of tumor extent could have con- cealed the benefits of chemotherapy if they existed.

This project also demonstrates that a high degree of selectivity, plus the increased attention to patient care generated by a clinical research project led to a survival rate for a refined Stage I much better than anticipated. This, in turn, meant that significance of differences in study groups such as irradia- tion vs no irradiation, or chemotherapy vs placebo would be more difficult to demonstrate. This difficulty necessarily increased the number of patients needed to achieve significant results beyond the bounds of being reason- able. It means for others who employ similar patient material that it is impractical even in relatively complex cooperative programs to achieve large enough series of patients to prove anything except that many patients are “cured” by well-performed standard therapy. Variations in technique, addi- tions or deletions relative to treatment in these highly selected “Stage I” patients may have their effects, but they will probably not be apparent because of lack of numbers of patients or because minor variations in the disease characteristics could have more profound results than the therapy al- teration

The information gleaned from this collaborative study leads to the con- clusion that, under the circumstances specified, it appears impossible to conclude that there was a beneficial effect in terms of tumor control from the use of a progestogen. It is recognized that possibly a different agent given in another manner might have had benefits that would have been recognized. However, in such a select group of patients as was chosen for this study it is unlikely that a significant improvement could be obtained by any variation from our standard therapeutic armamentarium in a reasonable time with a practical number of patients.

REFERENCES

1. LEWIS, G. C., JR., NADLER, S. H., BROSS, I. D. J,, AND SLACK, N. H. Adjuvant chemotherapy for cancer of the corpus uteri, Obstet. GynecoE. 29,797-802 (1967).

2. TAYLOR, E. S. Essentials of gynecology, 2nd ed. Lea & Febiger, Philadelphia, p. 197 (1962). 3. KELLEY, R. M., AND BAKER, W. H. Progestational agents in the treatment of carcinoma of the

endometrium, N. Engl. J. Med. 264,216-222 (1961). 4. KISTNER, ft. W. Histological effects of progestins on the hyperplasia and carcinoma in situ of

the endometrium, Cancer 12, 1106-1122 (1959).

376 LEWIS ET AL.

5. VARGA, A., AND HENFUKSEN, E. Clinical and histopathological evaluation of the effect of 17- whydroxyprogesterone 17-n-caproate on endometrial carcinoma, O&et. Gynecol. 18, 658-672 (1961).

6. REIFENSTEIN, E. C., JR. Hydroxyprogesterone caproate therapy in advanced endometrial cancer, Cancer 25,485-502 (1971).

7. VARGA A., AND HENRIKSEN, E. Histologic observation of effect of 17-whydroxyprogesterone 17-n-caproate on endometrial carcinoma, Obstet. Gynecol., 26,656-664 (1965).

8. WATERMAN, E. A., AND BENSON, R. C. Medrogestone therapy in advanced endometiial adenocarcinoma, Obstet. Gynecol. 30,626-634 (1967).

9. SYKES, M. P. Management of endometrial cancer, Med. Clin. N. Amer., 50, 833-844 (1966).