Ca icer mesearch Clinical 9

J Cancer Res Clin Oncol (1984)107:106-110

9 Springer-Verlag 1984

Pathologic Data of Prognostic Significance for Remission Induction in Advanced Ovarian Carcinoma* Ben W. Davis 1, Aron Goldhirsch 1, Gottfried W. Locher 2, Eckhard Dreher 2, Richard Greiner 3, Kurt Burki 4, and Kurt W. Brunner 5 1 LudwigInstitute for Cancer Research- Berne Branch, Inselspital,CH-3010 Berne 2 Universit/its-Frauenspital,CH-3012 Berne 3 Departmentof Radiation Therapy, Inselspital,CH-3010 Berne 4 PathologyInstitute, Inselspital, CH-3010 Berne s Institute for MedicalOncology,Inselspital,CH-3010 Berne, Switzerland Summary. Sixty-eight patients with "advanced ovarian

carcinoma" were entered into an ongoing phase-II trial for remission induction with cis-platinum (DDP) 80 mg/m 2 i.v. on day 1 followed by forced saline diuresis, melphalan (L-PAM) 12 mg/m 2 i.v. on day 2 and hexamethylamine (HMM) 130 mg/m 2 p.o. • 14 days from days 8-21 in six monthly cycles following operative resection and/or staging. Fifty-one patients were evaluable for response, ten had not completed six courses and could not be assessed, two patients died early (one probably of toxicity), and five patients refused treatment and follow-up. Thirty-Two patients had serous, endometrioid or undifferentiated carcinomas of the ovary. Of these, 11 (35%) achieved a pathologically proven complete remission (CR), five (16%) were NED after second-look (residual disease in ovary or removed omentum with all other biopsies and cytology washings negative), eight (32%) achieved a partial remission (PR), and three (12%) had progressive disease. None of the seven patients with clear-cell carcinoma and none of the three patients with mixed Mullerian tumor -of the ovary responded. Six of nine patients with tumors of uncertain origin or proven metastasis to ovary did not respond to treatment, These preliminary results indicate that advanced ovarian carcinomas form a heterogeneous group of recognizable neoplastic diseases with striking variation in response to treatment. Key words: Advanced ovarian carcinoma - Phase-II trial Introduction

During the last decade, chemotherapy has been used effectively to induce remission in patients with "advanced ovarian carcinoma." Clinical studies have demOffprint requests shouldbe addressed to BenW. Davis,M. D. (address

see above) * Paper presented at the Annual Meetingof the Swiss Societyfor Oncology, Basel, March 1983

onstrated that only those patients who have an absence of detected cancer (i.e., complete response) after therapy are potentially cured [1]. Recently, the addition of cis-platinum to combination chemotherapy has resuited in an improved complete response rate [2]. Since the dose-response curve for almost all known treatment modalities is steep for both the toxicity and therapeutic effects, any remission induction schedule that maximizes the reduction in detectable tumor mass will be toxic. It therefore becomes important to find better ways to select those patients who are likely to benefit from a chemical debulking regimen and exclude those patients who will have no benefit, but may suffer from toxicity. As a result of the therapeutic trials, certain prognostic factors for survival have been retrospectively identified for ovarian cancer. Patient age, stage of disease, completeness of the primary surgical operation, residual tumor volume after therapy, and histologic grade of the tumor have all been found to have an impact on the prognosis of patients with ovarian cancer. Recently, Dembo et al. [3] have illustrated the added strength of combining histologic grade with tumor type to select patients for radiation therapy. Currently, there is little information about the usefulness of pathologic data for prediction of response to combination chemotherapy. In 1979, an aggressive combined modality protocol for treatment of advanced ovarian cancer was initiated [4]. Those patients who obtained a significant debulking of detectable tumor volume by cytoreductive surgery and combination chemotherapy involving cis-platinum, were subsequently treated with whole-abdomen radiation. Here we report the analysis of pathologic data obtained from the initial staging surgical operation which may be useful in predicting remission by chemotherapy. Patients and Methods

Sixty-eight patients with "advanced ovarian carcinoma" were entered into an ongoingphase-II trial for remissioninductionwith cis-

107

B.W. Davis et al.: Remission in Advanced Ovarian Cancer T a b l e L The effect of,combination chemotherapy on the induction of second look laparotomy assessed remission (51 evaluable patients)

Response

No. patients

Percentage of evaluable patients

Complete response Partial response with no evidence of disease after second look Partial response No change Progressive disease

12 5

25 10

13 2 19

25 4 37

stage of disease was determined according to the FIGO criteria prior to the initiation of chemotherapy. The volume of residual disease was estimated as the maximum diameter of the largest residual tumor mass. The statistical significance of the variables was estimated using the chi-square test with Yates correction.

Results

The results of the remission induction regimen of combination chemotherapy in the 51 evaluable patients are presented in Table 1. Partial responders whose residual disease at second-look operation was totally resectable and was limited to one site, including ovary, omentum, or pelvic peritoneum only, were designated as PRned. A partial response in which there was a 50% reduction in measurable disease was seen in an additional 13 (25%) patients. The response of twelve of these patients was assessed by second-look operation and one patient refused additional operation. Thus, the overall response rate in these evaluable patients was 58% (30/51) and 34% (17/51) of the patients had no detectable disease (pathologically confirmed) after chemotherapy and second-look operation. The responses to chemotherapy by disease characteristics of the primary tumor prior to initiation of the remission induction protocol are listed in Table 2. Since three distinct prognostic categories were noted, the previously determined histologic tumor types were grouped into the three categories designated A (serous, endometrioid, undifferentiated cancers), B (clear-cell cancer and malignant mixed Mullerian tumors), and C (proven metastatic lesions and cancers of uncertain origin).

platinum (DDP) 80 mg/m 2 i.v. on day 1 followed by forced saline diuresis, melphalan (L-PAM) 12 mg/m z i.v. on day 2 and hexamethylmelamin (HMM) 130 mg/m2 p.o. for 14 days from days 8-21 in six monthly cycles following cytoreductive surgery and/or staging. Seventeen patients were not evaluable for this report: three refused treatment, two refused evaluation, one died probably of toxicity, one had an early death, and ten were too early for evaluation. The median age of the evaluable 51 patients was 53 years (range 28-73 years). Thirty patients had a complete hysterectomy and bilateral adnexeetomy. Patients who showed an objective clinical response after six cycles of the chemotherapy regimen, underwent a second-look operation to confirm remission and with the intent of maximal cytoreduction of residual disease. Complete response to the regimen was defined as total absence of detectable disease at second-look after pathologic examination of multiple biopsies of peritoneum including the diaphramatic surfaces and usually retroperitoneal and pelvic lymph nodes. Pathological examination included classification of histologic tumor type [5] and histologic grade based on a 1 4 scale of the modified Broder's system. This was performed by two pathologists without knowledge of the treatment results. Immunohistochemistry was performed by established techniques using antisera for human chorionic gonadotrophin (HCG), alpha fetoprotein (AFP), and carcinoembryonic antigen (CEA) obtained from Immulok. The

T a b l e 2. Response to the remission induction protocol and disease characteristics of the primary ovarian cancer

No.

CR a No. (%)

PRnedb No. (%)

PRc No. (%)

NC d No. (%)

PD e No. (%)

Tumor grade 1 2+3 4

2 24 25

2 (100) 7 (29) 3 (12)

2 (9) 3 (12)

7 (29) 6 (24)

1 (4) 1 (4)

7 (29) 12 (48)

Residual size _<2 cm __>2 cm

18 33

7 (40) 5 (15)

1 (15) 4 (12)

4 (22) 9 (27)

2

(6)

6 (33) 13 (40)

FIGO stage IIb+c III IV

5 35 11

4 (80) 7 (20)

4

(11)

10 (29)

2

(6)

1 (20) 2 (34)

1

1

(9)

Tumor type A B C

32 10 9

11 (35)

(9)

1 (11)

3

(27)

-

5 (16)

11 (34)

2

-

-

-

-

a Complete response b Partial response with no evidence of disease after second look Partial response

2

6

(6)

(22) d No change e Progressive disease

(55)

3 (9) 10 (100) 6 (67)

108

B.W. Davis et al.: Remission in Advanced Ovarian Cancer

Fig. 1 a, b. a Clear-cellcarcinoma of the ovary in a 47-year-oldfemalepresenting with stage-IIc diseasewhich progressed during chemotherapy. The clear cellswith luminal orientation of nuclei are characteristic of this tumor type. x 1025. b Malignant mixed Mullerian tumor of the ovary in a 54-year-oldwoman presenting with stage-III disease which progressed during chemotherapy. Note the presence of malignant cartilage and epithelial tubules, x 125

The effects o f each disease characteristic o n the degree o f response to c o m b i n a t i o n c h e m o t h e r a p y are illustrated in Table 3. The distributions of responses were n o t different for t u m o r grade, F I G O stage, and residual t u m o r size w h e n patients with either a C R or

a P R n e d were c o m p a r e d with those patients with only a PR, N C or PD. I n contrast, patients with t u m o r s of type B ( P = 0 . 0 0 1 ) or C ( P = 0 . 0 5 ) were less likely to achieve either a C R or P R n e d t h a n those who h a d type A tumors. Type-B t u m o r s consisted of seven clear-cell

109

B.W. Davis et al.: Remission in Advanced Ovarian Cancer Table 3. The effectofdiseasecharacteristicson the degreeof response to combinationchemotherapy Characteristics

Distribution tested

Tumor grade FIGO stage Residual size Tumor type

P value

2+ 3 vs 4 II+III vs IV <2 cm vs > 2 cm A vs B A vs C

NS NS NS 0.001 0.05

For the purpose of this analysis, CR and PRned were combinedto comparewith responsesdesignatedas PR, NC and PD by chi-square test with Yates correction Table 4. Distribution of tumor associated antigens in tumor type B which failed to show response to therapy

Clear-cell carcinoma [7] Mixed Mullerian tumor [3]

Glycogen Mucin AFP

HCG

CEA

7/7

0/7

0/6

0/6

1/3

2/3

2/3

0/3

0/3

1/2

carcinomas and three malignant mixed MuUerian tumors (MMT) (Fig. 1). To investigate the possibility that malignant germ-cell tumors may have a morphological appearance indistinguishable from these tumor types, immunohistochemical localization of oncofetal antigens associated with germ cell neoplasia of the ovary was performed (Table 4). None of the type-B tumors tested contained staining for human chorionic gonadotropin (HCG) or alpha fetoprotein (AFP). Intracellular glycogen, identified by periodic acid Schiff's reaction with a diastase control, was present in all seven clear-cell cancers and 2/3 MMT. None of the clear-cell cancers contained mucin. Type-C (Table 2) tumors consisted of four cases of proven metastatic tumor (two colon, two breast) to the ovary and five cancers of uncertain origin. One woman with a previous mastectomy for lobular carcinoma of the breast attained a pathologically confirmed complete remission after resection of bilateral ovarian and omental metastases followed by the protocol chemotherapy. Two patients had colectomies for invasive polypoid mucosal mucinous carcinomas with regional mesenteric lymph-node and liver metastases. One of these patients had a partial remission after chemotherapy. The five patients with tumors of uncertain origin presented with diffuse abdominal disease and had limited staging operations. Discussion

The findings in this study demonstrate that different specific histologic types of ovarian cancer may have

strikingly different responses to chemotherapy. Previous authors have demonstrated that histologic type may influence overall survival when considering all stages [6, 7]. However, this correlation was not found by investigators when the histologic grade and stage of disease were considered [8, 9]. Recently, Dembo et al. [3] have been able to identify a high-risk group of patients with rather limited disease by combining histologic type with grade. These patients did not benefit from whole-abdomen radiotherapy alone. Similar information for chemotherapy is not available. We therefore analyzed the pathologic data in 51 evaluable patients who received combination chemotherapy with platinum for "advanced ovarian carcinoma" and found that 7/7 clear-cell carcinomas and 3/3 malignant mixed Mullerian tumors did not respond to the remission-induction protocol. Not surprisingly, patients who had ovarian masses which were subsequently proven to be metastatic lesions did not usually benefit from the chemotherapy. Omura [10] has raised the question of whether all patients presenting as diffuse abdominal carcinomatosis and entered into a treatment trial of "advanced ovarian cancer" really have stage-III (or IV) ovarian cancers. Our pathology review identified nine evaluable cases which could not be considered as definite ovarian primary tumors (type C, Table 2). Cancers of uncertain origin are known to carry a poor prognosis [11] and are usually diagnosed from minimal amounts of tissue removed from unresectable tumors. The problem of massive bulk residual tumor in this context may certainly contribute to the poor response to chemotherapy of these tumors. Histologic grade was not associated with the degree of response to chemotherapy. Overall survival has been found by other investigators to be influenced by the grade of tumor [12, 13] and this influence was seen most profoundly in low stages [12]. In addition, Ozols et al. [13] concluded that the natural history of high-grade tumors was not altered by chemotherapy. While the rate of significant response ( C R + PRned) was not influenced by tumor grade (2 + 3 vs 4, Table 3), the follow-up to date is too short to exclude the possibility that, as Ozols et al. [13] suggest, the duration of response may be short-term in patients with grade-4 tumors. Similar reasons may account for the lack of a correlation between the degree of response and residual tumor size. While a trend is apparent in which better responses are seen in patients with small residual disease (=< 2 cm), the number of events is as yet too small to demonstrate statistical significance. However, this aggressive chemotherapy regimen has demonstrated effective chemical debulking of residual tumor volumes of more than 2 cm. When the degree of response for type-A tumors only was analyzed (i.e., CR + PRned vs PR + NC + PD) for tumor grade, stage, and size of re-

110

sidual disease, the results were not statistically significant (data not shown). Immunohistochemical detection of tumor markers was used in this study to exclude germ-cell neoplasia. The prognostic information gleaned from classical morphology may be enhanced by addition of functional investigations. The heterogenous appearance of ovarian cancer certainly results from a heterogenous expression of biochemical and genetic properties which may have an impact on the clinical behavior of tumors. In the future, identification of specific cellular antigens or nucleic acids may provide useful diagnostic and prognostic information that will aid in better selection of patients for specific therapeutic protocols. Acknowledgements. The authors would like to thank Ursula Gygli for aid in preparation of the manuscript.

References 1. Goldhirsch A, Greiner R, Davis B Chemotherapie followed by whole-abdominal radiation for ovarian cancer. Alberts DS, Surwit EA (eds) Gyn Oncology II (to be published) 2. Neijt JP, Ten Bokkel Huiniuk WM, van der Burg MEL, van Oosterom AT, Vriesendorp R, Boven E, Kooyman CD, van Honurelingen JC, Pinedo HM (1983) Combination chemotherapy with hexa-CAF and CHAP-5 in advanced ovarian carcinoma:

B.W. Davis et al.: Remission in Advanced Ovarian Cancer A randomized study of the Netherlands joint study group for ovarian cancer. Proc ASCO 2:148 3. Dembo AJ, Brown TC, Bush RS, Sturgeon JFG (1982) Prognostic significance of pathology subtype and differentiation in epithelial carcinoma of the ovary. Proc ASCO Abstract C406 4. Greiner R, Goldhirsch A, Dreher E, Davis B, Locher G, Payer T, Neuenschwander H, Joss R, Brunner K, Veraguth P Wholeabdominal radiation in patients with advanced ovarian carcinoma after surgery, chemotherapy and second-look laparotomy. Eur J Cancer Clin Oncol (in press) 5. Serov SF, Scully RE, Sobin LH (1973) International histological classification of tumors, No. 9. Histological Typing of Ovarian Tumors. Geneva, World Health Organization 6. Aure JC, Hoeg K, Kolstad P (1971) Clinical and histological studies of ovarian carcinoma: long term follow-up of 900 cases. Obstet Gynecol 37:1 7. Kottmeier HL (1979) Annual report on the results of treatment in gynecological cancer, vol. 17. Statement of results obtained in 1969 to 1972. FIGO. Stockholm 8. Young RC, Hubbard SP, DeVita VT (1974) The chemotherapy of ovarian carcinoma. Cancer Treat Rev 1:99 9. Smith JP, Rutledge F, Wharton JT (1972) Chemotherapy of ovarian cancer: new approaches to treatment. Cancer 30:1565 10. Omura G (1981) Are all stage III cancers of the ovary really cancers of the ovary? Cancer Clin Trials 4:219 11. Nissenblatt MJ (1981) The CUP syndrome (carcinoma of unknown primary). Cancer Treat Rev 8:211 12. Julian C, Woodruff JD (1969) The role ofchemotherapy in treatment of primary ovarian malignancy. Obstet Gynecol Surg 24:1307 13. Ozols RF, Garvin AJ, Cost J (1975) Histological grade in advanced ovarian cancer. Cancer Treat Rep 63:255