Breast Cancer Research and Treatment 8: 19%204, 1986 © Martinus Nijhoff Publishers, Boston - Printed in the Netherlands Report Metastatic pattern and response to endocrine therapy in human breast cancer Claus Kamby 1 and Carsten Rose Department of Oncology ONA, The Finsen Institute, 49, Strandboulevarden, DK-2100 Copenhagen 0, Denmark (1address for offprints) Keywords: endocrine responsiveness, extent of metastases, site of metastases, survival after recurrence, tamoxifen Abstract The effect of endocrine therapy in 465 postmenopausal patients with advanced breast cancer who entered four consecutive, randomized trials has been related to the site of the metastases. Patients received either tamoxifen (T) alone or T in combination with medroxyprogesterone acetate, diethylstilbestrol, halotestin, or aminoglutethimide. The overall response rate was 40%. Responses were most frequently seen in patients with metastases in soft tissue, and the duration of response to endocrine therapy in these patients was longer than for those with metastases in bone or viscera (p<0.00001). In addition, the response rate was inversely correlated with the number of main metastatic sites in patients with soft tissue metastases, whereas the response rate was not associated with the number of metastatic sites in patients with metastases in bone and viscera. Survival after first recurrence was significantly longer in responding patients with soft tissue lesions compared to those with recurrence in bone or viscera. In contrast, survival after first recurrence was identical in patients with nonresponding disease, irrespective of dominant site of metastases. The outcome of endocrine therapy depends partially upon the dominant site of metastases. This may reflect a difference in biological characteristics of human breast cancer tumor cells that metastasize to different sites. Introduction Differences in survival among patients with breast cancer can be explained by diversities in the biological aggressiveness [1-3] of the tumors and by differences in response to antineoplastic treatment [4-6]. Furthermore, the survival after recurrence (SAR) in patients with metastatic disease is influenced both by the location and the extent of metastases [7]. Factors responsible for the pattern of metastases are still unknown [8], but it has been suggested that the estrogen receptor (ER) content of the primary breast tumors is associated with the subsequent location of metastases at various sites [9-12]. Accordingly, the E R content of the meta- stases seems to vary in different anatomical sites [13-15]. Previous studies have shown that response rates to endocrine therapy (ET) are positively correlated with the E R contents in the tumors [16-18]. Since a response to E T at various metastatic locations can be considered as a clinical test for hormone dependency of the tumor, we have analyzed the response to E T in relation to the metastatic pattern in patients with advanced breast cancer. Furthermore, the impact of the metastatic pattern upon survival in responders and non-responders to endocrine therapy has been studied. 198 Claus Kamby et al. Patients and methods From September 1976 until February 1983, postmenopausal patients with advanced breast cancer entered one of four consecutive randomized studies of ET at the Department of Oncology ONA at the Finsen Institute [19-23]. These patients received either tamoxifen alone (30 or 90 mg daily) [19] or tamoxifen (30 rag) in combination with either diethylstilbestrol [20], medroxyprogesterone acetate [21], aminoglutethimide plus hydrocortisone [22], or halotestin [23]. The criteria of eligibility were identical for all four protocols and were the following: Patients must have measurable and/or evaluable progressive disease according to UICC [24]. Patients must have a performance status ~<3. Patients must be functionally postmenopausal (>12 months of menostasia). Patients >65 years of age received ET as firstline therapy for their advanced disease, whereas patients ~<65 years all had received prior cytotoxic therapy. In most patients (92%) this treatment comprised the combination of 3 or 4 cytotoxics (cyclophosphamide, methotrexate, 5-fluorouracil, adriamycin, vincristin and/or CCNU). The study program prior to onset of therapy consisted of patient history, physical examination, blood tests (hemoglobin, leukocytes, thrombocytes, liver enzymes, and calcium), chest X-rays, and radiographic bone survey. The diagnosis of liver metastases required ultrasound scanning, and patients were included irrespective of the liver function. Additional diagnostic procedures were carried out when indicated. Patients were assessed I month after the initiation of therapy and at intervals of i to 2 months thereafter. The time to treatment failure was defined as the period from initiation until progressive disease (PD). The recurrence-free interval (RFI) was defined as the time from the initial diagnosis until the time of recurrence; and the survival after recurrence (SAR) was defined as the time from start of ET until death. Patients who died or for whom treatment had to be discontinued due to rapid progression within the first 3 months of treatment are included in the analysis as having progressive dis- ease. Patients who achieved either a partial or a complete response [24] are designated as responders, while patients with progression or no change are designated as non-responders. The anatomical sites of metastases were grouped as follows: SOFT TISSUE (skin, subcutis, lymph nodes, contralateral breast); BONE (verified by X-rays); and VISCERA (all other sites). Dominant site of metastasis was defined as recurrence in viscera, bone, and soft tissue in decreasing order of priority. Qualitative data were compared using the Chi square test with Yates correction or the MannWhitney rank-sum test for unpaired data. Times to treatment failure and SAR were calculated according to the Kaplan-Meier product limit procedure and were compared using the log-rank test [25]. A two-tailed p value of <0.05 was considered to be significant. Results At the time of evaluation (May 1984) 485 patients were allocated to the trials. Of these, 20 patients (4%) were excluded due to either violation of protocol (16 patients) or other malignant disease (4 patients). Of the 465 evaluable patients, 361 patients (78%) had completed ET, and 260 patients (56%) had died. The median age (range) at the time of entry to the protocol was 69 years (34-91), and 21 patients (5%) were under 50 years of age. Chemotherapy for advanced disease was given to 137 patients (29%) prior to ET. Response (PR + CR) to ET was obtained in 184 patients (40%, 95% confidence limits (C.L.): 35-45%). No change (NC) was observed in 129 patients (27%, 95% C.L.: 22-32%), while 152 patients (33%, 95% C.L.: 28-38%) had PD. Three hundred sixty-three (78%) patients had soft tissue metastases, 158 (34%) patients had bone metastases, and 174 (37%) patients had visceral metastases. On e hundred forty-four patients (31%) had metastases in two of the three sites, while 43 (9%) had metastases in all three sites. The distribution of responding and non-responding patients for each of the sites of metastases is presented in Table Metastatic pattern and endocrine therapy 1. As can be seen, 45% of the 363 patients with soft tissue metastases responded, and 55% did not. In contrast, only 46 patients (29%) with bone metastases and 47 patients (27%) with visceral metastases achieved a response (Table 1). The response rate to endocrine therapy in previously untreated patients was 48%, while the rate of response in patients previously treated with chemotherapy was 21% (p<0.05). Since a greater proportion of nonresponders than responders had been treated with chemotherapy, we analyzed the data after exclusion of patients previously treated with cytotoxic therapy and found the same distribution of responders and non-responders at the 3 sites (p<0.05). Table 2 shows the number of metastatic sites in responders and non-responders to ET. Patients with one site of metastasis had a higher response rate than patients with two or three sites (p -- 0.01). Thus, the differences in response rate to E T according to metastatic site presented in Table 1 could be due to differences in the number of metastatic sites in responders and non-responders (Table 2). The data have therefore been further analyzed with respect to site of metastasis and number of sites. The response rate to E T in patients 199 with soft tissue metastases (Table 3a) was influenced by the occurrence of concomitant involvement in the other main metastatic sites (p = 0.001). In contrast, the response rate in patients with bone or visceral metastases was not associated with the concomitant metastatic sites (Tables 3b and 3c). The time to treatment failure according to dominant site of metastasis in responding patients is shown in Fig. t. The median time to treatment failure in patients with soft tissue metastases was 33 months (range: 4-48). The corresponding values for patients with dominant site of metastasis in bone and viscera were 11.5 (range: 4-45) and 9 months (range: 4-46), respectively. The median time to treatment failure in non-responding patients was 4 months (range: 0-34) for patients with soft tissue metastases only, and 3 months (range: 0-21) for patients with dominant site of disease in bone and viscera. The survival after recurrence (SAR) was analyzed in relation to dominant site of metastasis in non-responding (Fig. 2a) and responding patients (Fig. 2b). Although the non-responding patients with only soft tissue metastases had a longer SAR than patients with visceral metastases (p<0.05), the poor prognosis for non-responding patients Table 1. Distribution of patients accordingto site of metastasis and response to endocrine therapy. Site of metastasis Non-responders (%) Responders(%) Total (%) Soft tissue Bone Viscera 199 (55) 112 (71) 127 (73) 164 (45) 46 (29) 47 (27) 363 (100) 158 (100) 174 (100) )~2 (d.f.: 2) = 22.1; p<0.00002. Table 2. Distribution of patients accordingto number of sites (soft tissue, bone and/or viscera) and response to endocrine therapy. No. of sites Non-responders (%) Responders (%) Total (%) 1 2 3 156 (55) 93 (64) 32 (74) 122 (45) 51 (36) 11 (26) 278 Total no. of pts. 281 (60) 184 (40) 465 (100) p = 0.01 (Mann-Whitney). (100) 144 (100) 43 (100) 200 C l a u s K a m b y et al. seems to apply to all patients irrespective of dominant site of metastasis. M o r e o v e r , if S A R of patients with i n v o l v e m e n t of o n e site only (soft tissue, b o n e or viscera) is considered (data not shown), the survival of n o n - r e s p o n d e r s with soft tissue disease is of the same length as that of patients with b o n e ( p > 0 . 5 0 ) and visceral metastases (p>0.20). In responding patients (Fig. 2b) the p r o l o n g e d S A R seems to be due to a longer survival in patients with soft tissue metastases ( p < 0 . 0 5 ) , whereas the prognosis for patients with b o n e and visceral metastases is c o m p a r a b l e to that obtained for n o n - r e s p o n d i n g patients. These differences of S A R are also seen, if patients with only one site of r e c u r r e n c e are considered (soft tissue vs b o n e or viscera: p < 0 . 0 2 5 - data not shown). Discussion This study shows that both the anatomical localization and the n u m b e r of metastatic sites involved are factors which influence the response rate to E T in patients with a d v a n c e d breast cancer. Patients with metastases in soft tissue were m o r e likely to res p o n d to E T than patients with b o n e and visceral involvement. M o r e o v e r , the response rate seems to be inversely correlated with the n u m b e r of metastatic sites. This relation, however, applies only to patients with soft tissue metastases, since the number of sites involved was almost equal in responders and n o n - r e s p o n d e r s with metastases in b o n e and viscera. T h e higher efficacy of E T in patients with soft tissue metastases has been described in other Table 3a. Concomitant metastatic pattern in patients with soft tissue metastases and response to endocrine therapy. Non-responders (%) Responders (%) Total (%) Soft tissue (ST) only ST + bone (BO) or viscera (VI) ST + BO + V! 101 (48) 66 (59) 32 (74) 108 (52) 45 (41) 11 (26) 209 (100) 111 (100) 43 (100) Total 199 (55) 164 (45) 363 (100) p = 0.001 (Mann-Whitney). Table 3b. Concomitant metastatic pattern in patients with bone metastases and response to endocrine therapy. Non-responders (%) Bone (BO) only BO + soft tissue (ST) or viscera (VI) BO + ST + VI Total Responders (%) Total (%) 22 (81) 58 (66) 32 (74) 5 (19) 30 (34) 11 (26) 27 (100) 88 (100) 43 (100) 112 (71) 46 (29) 158 (100) p = 0.78 (Mann-Whitney). Table 3c. Concomitant metastatic pattern in patients with visceral metastases and response to endocrine therapy. Non-responders (%) Viscera (VI) only VI + soft tissue (ST) or bone (BO) VI + ST + BO Total p = 0.63 (Mann-Whitney). Responders (%) Total (%) 34 (77) 61 (69) 32 (74) 9 (23) 27 (31) 11 (26) 43 (100) 88 (100) 43 (100) 127 (73) 47 (27) 174 (100) Metastatic pattern and endocrine therapy 201 TIME TO TREATMENT FAILURE (r-l) S o f t tissue vs (X) Bone vs (~7) V i s c e r o 1 O0 80 ~t "~" 6 0 U3 Z 0 4O 20 0 10 20 30 40 50 MONTHS [] 108 )< 29 47 40 7 12 4 2 2 Fig. 1. Time to treatment failure (TTF) by dominant site of metastasis for responding patients (E3 soft tissue (ST), × bone (BO), V viscera (VI)). Log rank: ST vs BO, p<0.0005; BOvs VI, p>0.30; ST vs VI, p<0.0005. The number of patients at risk is indicated under the abcissa. studies concerning response to ET in advanced breast cancer [26]. However, our results indicate that the response rate also depends upon the number of sites prior to treatment. Traditionally, response rates of clinical trials are presented according to the dominant site of metastasis. The present study shows that different response rates obtained in this hierarchic grouping of major sites may be due both to the anatomical site of recurrence and to the number of main sites involved. The effect of ET is positively correlated with the E R content of the tumor cells [16-18], and both the response rate and the duration of response increase with increasing E R content [27]. The demonstration of a higher response rate and a prolonged time to treatment failure in patients with soft tissue metastases may, therefore, indicate that the ER content of these metastases is higher than that of metastases in bone and viscera. Previous studies have shown that the E R content in metastases varies in different anatomical sites [13-15], and that E R positivity in the primary tumors is associated with occurrence of metastases in bone, while ER negativity is associated with occurrence of metastases in viscera [9-11, 28, 29]. Other studies have, however, failed to demonstrate such a correlation [30, 31], and the present study does not indicate a difference in hormone-responsiveness of metastases located in bone and viscera. The prolonged SAR of responding patients is mainly due to a prolongation of SAR in patients with soft tissue metastases only (Fig. 2b). In contrast, the shorter SAR in non-responding patients seems to apply to all patients, irrespective of dominant site of disease (Fig. 2a). Using linear logistic regression analysis, Paridaens et al. [16] showed that variations of ER content, age, and menopausal status in 49 patients were significantly related to the outcome of ET, whereas the localization and the total number of metastases were unassociated with the response to ET. However, the data presented here, derived from a larger number of functionally postmenopausal patients, suggest that the hormone dependency of meta- 202 Claus Kamby et al. A: N O N - R E S P O N D E R S (n) Soft ~_ I00 tissue v s (;~) B o n e vs (v) Viscero %. o 0 20 60 40 rn 101 MONTHS 21 X V 53 127 8 13 80 100 2 I 2 B: R E S P O N D E R S (!1) S o f t lOOJ tissue v s (;~) B o n e v s (V) V i s c e r o laJ 8O z L=J tY tY ::~ 60 IJJ n." IZ I.iJ 20 0'3 0 20 40 60 80 100 MONTHS U 108 x 19 8 3 29 47 5 0 0 Fig. 2. Survival after recurrence (SAR) in non-responding (A) and in responding patients (B) related to the dominant site of metastasis ([3 soft tissue (ST), x bone (BO), V viscera (VI)). Log rank: (A): ST vs BO, p>0.90; BOvs VI, p>0.15; ST vs VI, p<0.025. (B): ST vs BO, p<0.000001; B O v s VI, p>0.80; ST vs VI, p<0.0000001. The number of patients at risk is indicated under the abcissa. Metastatic pattern and endocrine therapy s t a s i z i n g b r e a s t c a n c e r is r e l a t e d t o t h e a n a t o m i c a l location of metastases, indicating a difference in biological characteristics between breast cancer tum o r c e l l s m e t a s t a s i z i n g t o d i f f e r e n t sites. 12. Acknowledgement 13. Supported by the Danish Medical Research Council, n o . 12-5342 a n d t h e A r v i d N i e l s s o n F o u n d a - 14. tion. References 1. Sainsbury JRC, Farndon JR, Sherbet GV, Harris AL: Epidermal-growth-factor receptors and oestrogen receptors in human breast cancer. Lancet i: 364-366, 1985 2. Paterson AHG, Szafran O, Lees AW, Hanson J: Prognostic factors at presentation and their relevance to the natural history of breast cancer. Rev Endocrine-Related Cancer, suppl 14: 35-39, 1984 3. Silvestrini R: Cell kinetic and ER assays in breast cancer. Rev Endocrine-Related Cancer 18: 15-19, 1984 4. Howell A, Bush H, George WD, Howat JMT, Crowther D, Sellwood RA, Rubens RD, Hayward JL, Bulbrook RD, Fentiman IS, Chaudary M: Controlled trial of adjuvant chemotherapy with cyclophosphamide, methotrexate, and fluorouracil for breast cancer. Lancet ii: 307-311, 1984 5. Rose C, Thorpe SM, Andersen KW, Pedersen BV, Mouridsen HT, Blichert-Toft M, Rasmussen BB: Beneficial effect of adjuvant tamoxifen therapy in primary breast cancer patients with high oestrogen receptor values. Lancet i: 16-19, 1985 6. Howat JMT, Harris M, Swindell R, Barnes DM: The effect of oestrogen and progesterone receptors on recurrence and survival in patients with carcinoma of the breast. Br J Cancer 51: 263-270, 1985 7. Swenerton KD, Legha SS, Smith T, Hortobagyi GN, Gehan EA, Yap H-Y, Gutterman JU, Blumenschein GR: Prognostic factors in metastatic breast cancer treated with combination chemotherapy. Cancer Res 39: 1552-1562, 1979 8. Roos E, Dingemans KP: Mechanisms of metastasis. Biochim Biophys Acta 560: 135-166, 1979 9. Singhakowinta A, Potter HG, Samal B, Brooks SC, Vaitkevicius VK: Estrogen receptors and natural course of breast cancer. Ann Surg 183: 84-88, 1976 10. Stewart JF, King RJB, Sexton SA, Millis RR, Rubens RD, Hayward JL: Oestrogen receptors, sites of metastatic disease and survival in recurrent breast cancer. Eur J Cancer 17: 449-453, 1981 11. Campbell FC, Blarney RW, Elston CW, Nicholson RI, 15. 16. 17. 18. 19. 20. 21. 22. 23. 203 Griffiths K, Haybittle JL: Oestrogen-receptor status and sites of metastasis in breast cancer. Br J Cancer 44: 456459, 1981 Lippman ME, Allegra JC, Thompson EB, Simon R, Barlock A, Green L, Huff KK, DO HMT, Aitken SC, Warren R: The relation between estrogen receptors and response rate to cytotoxic chemotherapy in metastatic breast cancer. N Engl J Med 298: 1233-1238, 1978 de la Monte SM, Hutchins GM, Moore GW: Estrogen and progesterone receptors in prediction of metastatic behavior of breast carcinoma. Amer J Med 76: 11-17, 1984 Rosen PP, Menendez-Botet CJ, Urban JA, Fracchia A, Schwartz MK: Estrogen receptor protein (ERP) in multiple tumor specimens from individual patients with breast cancer. Cancer 39: 2194-2200, 1977 Holdaway IM, Bowditch JW: Variation in receptor status between primary and metastatic breast cancer. Cancer 52: 479-485, 1983 Paridaens R, Sylvester RJ, Ferrazzi E, Legros N, Leclercq G, Heuson JC: Clinical significance of the quantitative assessment of estrogen receptors in advanced breast cancer. Cancer 46: 2889-2895, 1980 Fernandez MD, Alaghband-Zadeh J, Burn JI: Quantitative oestrogen and progesterone receptor values in primary breast cancer and predictability of response to endocrine therapy. Clin Oncol 9: 245-250, 1983' Rose C, Thorpe SM, LOber J, Daenfeldt JL, Palshof T, Mouridsen HT: Therapeutic effect of tamoxifen related to estrogen receptor level. In: Henningsen B, Linder F, Steichele C (eds): Endocrine Treatment of Breast Cancer. Springer-Verlag, Berlin- Heidelberg - New York, 1980, pp 134-141. (Recent Results in Cancer Research). Rose C, Theilade K, Boesen E, Salimtschik M, Dombernowsky P, Brunner N, Kjaer M, Mouridsen HT: Treatment of advanced breast cancer with tamoxifen: Evaluation of the dose-response relationship at two dose levels. Breast Cancer Res Treat 2: 395-400, 1982 Mouridsen HT, Salimtschik M, Dombernowsky P, Gelshoj K, Palshof T, Rorth M, Daehnfeldt JL, Rose C: Therapeutic effect of tamoxifen versus combined tamoxifen and diethylstilboestrol in advanced breast cancer in postmenopausal women. In: Mouridsen HT, Palshof T (eds): Breast Cancer - Experimental and Clinical Aspects. Pergamon Press, Oxford and New York, 1980, pp 107-110 Mouridsen HT, Ellemann K, Mattsson W, Palshof T, Daehnfeldt JL, Rose C: Therapeutic effect of tamoxifen versus tamoxifen combined with medroxyprogesterone acetate in advanced breast cancer in postmenopansal women. Cancer Treat Rep 63: 171-175, 1979 Rose C, Kamby C, Mouridsen HT, Bastholt L, Brincker H, Skovgaard-Poulsen H, Andersen AP, Loft H, Dombernowsky P, Andersen KW: Combined endocrine treatment of postmenopausal patients with advanced breast cancer. Breast Cancer Res Treat 7 (suppl): 45-50, 1986 Rose C, Kamby C, Mouridsen HT, Bastholt L, Brincker H, Skovgaard-Poulsen H, Andersen AP, Loft H, Domber- 204 Claus Kamby et al. nowsky P on behalf of DBCG: Combined endocrine treatment of postmenopausal patients with advanced breast cancer. A randomized trial of tamoxifen vs. tamoxifen plus aminoglutethimide and hydrocortisone vs. tamoxifen plus fluoxymesterone. (Abstract, submitted to Am Assoc Clin Oncol 1986) 24. Hayward JL, Carbone PP, Heuson JC, Kumaoka S, Segaloff A, Rubens RD: Assessment of response to therapy in advanced breast cancer. Eur J Cancer 13: 89-94, 1977 25. Peto R, Pike MC, Armitage P, Breslow E, Cox DR, Howard SV, Mantel N, McPherson K, Peto J, Smith PG: Design and analysis of randomized clinical trials requiring prolonged observation of each patient. II. Analysis and examples. Br J Cancer 35: 1-39, 1977 26. Maass H, Jonat W: Site factor affecting response of metastases. In: Stoll BA (ed): Systemic control of breast cancer, ch. 9, pp 188-203, vol. 4 in the series: New Aspects of Breast Cancer, William Heinemann Medical Books, London 1980 27. Campbell FC, Blarney RW, Elston CW, Morris AH, Nicholson RI, Griffiths K, Haybittle JL: Quantitative oestradiol receptor values in primary breast cancer and response of metastases to endocrine therapy. Lancet ii: 131%1319, 1981 28. Qazi R, Chuang J-L, Drobyski W: Estrogen receptors and the pattern of relapse in breast cancer. Arch Intern Med 144: 2365-2367, 1984 29. Budd GT: Estrogen receptor profile of patients with breast cancer metastatic to bone marrow. J Surg Oncol 24: 167169, 1983 30. Kamby C, Rose C, Iversen H, Holm NV, Andersen KW, Thorpe SM: Pattern of metastases in human breast carcinoma in relation to estrogen receptor status. Anticancer Res 6: 10%112, 1986 31. Hahnel R, Woodings T, Vivian AB: Prognostic value of estrogen receptors in primary breast cancer. Cancer 44: 671-675, 1979