GYNECOLOGIC
ONCOLOGY
36, 13-18 (I!?%)
Ovarian Cancer: Relationship between in Vitro Chemosensitivity and Clinical Response L. CARENZA, C. VILLANI, II Department
of Obstetrics
F. L. LABI, AND R. PROSPERI PORTA
G. RULLI, and Gynaecology,
University
“La
Sapienza,”
Rome,
Italy
Received May 3, 1988
Thirty-nine samplesfrom 30 patients with epithelial ovarian cancer were collected. Using the method describedby A. Hamburger (.Z.NutZ. Cancer Inst. 66, 981 (1981)) and P. Salmonet al. (AT. Engl. J. Med. 308, 129(1983)), tumor cellswereexposed to anthreoplasticagents(Adriamycm and &-platinum) and then cultured in double-layeragar. There were 23 evaluablepatients. The percentageof completeand partial responsesfor patients whosecells were sensitivein vitro, whether or not treated with the samedrugs, was68, vs 44% comparedto patientswhosecells were resistantin vitro and treated in viva with therapeutic regimenswhich may or may not have included drugs used in the test. More patients with residual tumor greater than 2 cm were resistantto chemotherapy. D IWO Academic press, I~C.
INTRODUCTION Epithelial ovarian carcinoma continues to be the gynecologic neoplasia with the worst prognosis and is the fourth most frequent cause of cancer death in women. Overall 5-year survival is less than 40% because ovarian cancers are most frequently detected in advanced stages; increased relapse-free and median survival have been obtained by applying an aggressive multidisciplinary therapeutic approach. Alkylators as single agents have produced response rates of 25-50% [I]. Doxorubicin yields results similar to those of alkylating agents in patients not previously treated with drugs, but the mean duration of response is 3 months because of rapid development of resistance, probably because of the development of new metabolic structures which permit an increased clearance of doxorubicin from the neoplastic cells [2]. The most active drug in ovarian cancer is c&platinum, which can induce complete remission even in previously treated patients. Although polychemotherapy has shown higher response percentages, it has not been clearly proven that the median survival time compared to monochemotherapy is increased. However, a higher rate of
complete response occurs, mainly in patients with minimal residual disease after surgery. For this group (less than 30% of all patients with advanced ovarian cancer), complete response to cytotoxic therapy is about 40-70% versus lo-20% in the patient group with residual tumors greater than 2 cm [l]. The development of drug resistance to antineoplastic agents is a major clinical problem that limits the overall effectiveness of chemotherapy. Frequently, acquired resistance to single agents is associated with broad crossresistance to structurally dissimilar drugs, resulting in only a 6-10% response rate to salvage chemotherapy [3]. Moreover, individual patients at the same clinical stage and with the same histologic type often do not respond uniformly to antineoplastic agents [4]. Combined treatments for ovarian cancer are needed: first, an aggressive surgical approach for maximum cytoreduction, followed by identification, and administration of therapeutic regimens to which the neoplasia shows greatest sensitivity. The search for a useful tumor chemosensitivity test dates back to at least 1954 (51. A number of techniques have been proposed: in vivo with xenotransplantation and in vitro with short-, medium-, and long-term cell cultures. The human tumor stem cell clonogenic assay (HTSCCA) has been one of the most widely studied techniques during the last 10 years. It is based on the capacity of neoplastic cells to generate colonies when cultured in semisolid agar. Stem cells have the capacity to renew and represent only a small part of the total population. The HTSCCA makes it possible to study only those cells which are responsible for tumor repopulation after debulking, chemotherapy, and metastatic growth, and therefore should be the primary targets for any cytotoxic therapy [6]. The correlation between the response in vitro to antineoplastic agents and clinical response has proved correct in 95% of drug resistance cases and in 65% of cases
13 00!%8258/90 $1.50 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
14
CARENZA ET AL.
predicting drug sensitivity [6]. Using the HTSCCA it should be possible to select “the right drug for the right patient” in order to apply personalized chemotherapy or to retrospectively study the clinical outcome in relation to parameters such as cloning efficiency and growth inhibition rates obtained with the drugs tested. In this study we report the chemosensitivity, using HTSCCA, of tumor stem cells in patients affected by ovarian cancer.
TABLE 2 Growth of Cell Coloniesin Soft Agar No. of colonies/5 x 16 cells plated Sample site
No.
Median
Range
Colony-forming efficiency (%I
Ovary Metastases Relapse Ascites
15 5 1 7
119.7 71 106.1 337.7
116-129 42-102 81-131 258-422
0.02 0.01 0.02 0.06
METHODS Sample and Patients
Thirty-nine samples from 30 patients with epithelial ovarian cancer were collected (24 primary tumors, 6 metastases, 1 pelvic relapse, 8 ascitic effusions). In one case there was insufficient tissue to assay; five samples were not plated because viable cells were absent. Of the 33 remaining cultures, no growth was observed in two cases, and three specimens were contaminated. Twentyeight cultures were evaluated: 15 for primary ovarian cancer, 5 from intestinal or lymph node metastases, 1 from relapse, and 7 from ascitic effusions (Table 1). The median cloning efficiency (number of colonies/5 x 10’ cells plated x 100) in the 28 evaluated specimens was 0.02 (range 0.01-0.06) (Table 2). Samples from 23 patients were evaluated. The patients’ characteristics are shown in Table 3. Complete response was defined as the disappearance of all manifestation of disease for more than 1 month. Partial response was defined as 50% or greater decrease in the product of the diameters of all measureable lesions. In Vitro Clonogenic Assay
The tumor tissue was prepared under sterile conditions in a laminar vertical flow system. Solid tumor samples were obtained during surgery; specimens were mechanically dissociated with sterile scissors. The samples subsequently underwent enzymatic disaggregation in a mixture of 0.5% collagenase II and 0.02% DNase (both from Sigma) at 37°C for 1 hr and then were filtered
Number
TABLE 3 Characteristicsof Patients with Ovarian Carcinoma No. of patients Age (years) Median
TABLE 1 Tumor SampleCharacteristics Tumor samples collected Tumor samples processed Samples with cell yield Sufficient for plating Contaminated samples Evaluable samples
through sterile gauze and washed twice by centrifugation at 15Og for IO min at 10°C. Cells were suspended in fresh RPM1 (Gibco) and passed through 20-, 22-, and 25-gauge needles. The viable cell count was determined by the trypan blue (0.05%) exclusion test. Ascitic effusions were collected in heparinized bottles (100 U/ml), passed through sterile gauze, centrifuged at 15Og for 20 min at 10°C and resuspended in RPMI. Viable cells were determined by trypan blue. Cells were then exposed to drugs for 1 hr at 37°C. The drugs used were Adriamycin (Adriblastina, FarmitaliaCarlo Erba) and cis-diaminodichloroplatinum (Platamine, Farmitalia-Carlo Erba) at O.Ol-, O.l-, and 1 pg/ml doses. Cells were washed twice with HBSS (Gibco) and plated. Hamburger’s [4] and Salmon’s methods [7] were used to culture tumor cells in double-layer agar. One milliliter of under-layer agar (0.5%), diluted in enriched McCoy’s 5A medium (Gibco), was solidified in a 35 x 20-mm petri dish at room temperature. The cells were plated (5 x 10’ cells/dish) and 1 ml of upper-layer agar, (0.3%) with enriched medium CMRL 1066 (Gibco), was added. Both media contained serum, electrolytes, carbohydrates, amino acids, vitamins, and insulin. All experiments were performed in triplicate. Cultures were left for 1 hr at room temperature to solidify the agar.
Percentage
39 38
91
33 3 28
87 9 85
Range Patients treated previously Residual disease Less than 2 cm Greater than 2 cm Stage (FIGO) I III IV
23 63 2% 17 6123 10123 13123 4/23 14/23 5123
CHEMOSENSITIVITY
AND OVARIAN
Plated cultures were controlled by an inverted phase microscope to ascertain that cell clumps were not present; then the cultures were incubated in a humidified atmosphere of 5% CO2 at 37°C. Plates were examined within 3 days for clusters (15-30 cells) and 7 and 21 days later to count colonies (30 or more cells aggregates). In vitro sensitivity was defined as a 50% or more decrease in the number of colonies [7].
RESULTS Patients’ responses to chemotherapy were divided into three groups: complete clinical response (CR), partial clinical response (PR), and no response (NR). Follow up was at 4-34 months. Table 4 shows data regarding patients in CR. The median age was 63 years, with a range of 25-72 years. Four were at stage I (FIGO) and 7 at stage III. Of 11 patients, 9 showed residual tumor less than 2 cm after surgery. In one case (I.R.) the sample was resistant in vitro, but the patient had been previously treated with Adriamycin (A) and cyclophosphamide (P). Another patient (D.L.P.), previously treated with HEXA-CAF, was sensitive in vitro to the highest concentration of Adriamycin. Both patients were responsive to second-choice chemotherapy (interferon; folic acid; SFU). Eight patients received at least one of the drugs to which they were sensitive in vitro. Of the four patients in PR (median age 69 years), one was at stage III and three were at stage IV. All had a residual tumor greater than 2 cm after surgery (Table 5). Two were sensitive to the tested drugs, both in vitro and
15
CANCER
in vivo, although patient B.C. had previously undergone a therapeutic regimen including c&platinum. The sample of ascitic fluid obtained from patient C.L. was sensitive to c&-platinum, whereas the metastatic biopsy sample was resistant. This patient (CL.) underwent another therapeutic regimen to which she was responsive. Finally 8 patients (median age 59 years) were not responsive to chemotherapy (Table 6). Three patients received at least one of the tested drugs in vitro. One of the two patients previously treated with A-P was sensitive in vitro to high doses of c&platinum, and neither was responsive to the second-choice regimen with folic acid and 5FU. Patients P.R. and A.M. were resistant in vitro to both A and P and had no in vivo response to other therapeutic protocols. Patient P.A. was very interesting. The primary tumor was sensitive in vitro to A-P, but because of renal toxicity, the patient underwent a regimen with Adriamycin and cyclophosphamide. The ascitic fluid sampled during treatment was sensitive only to the highest dose of c&platinum; an additional sampling performed on ascitic fluid formed during the course of disease showed resistance toward both P and A. It was possible to observe the development of in vivo and in vitro resistance to Adriamycin and cross-resistance to c&-platinum in this patient who survived 25 months. The highest percentage of CR and PR (73%) was obtained in the group of patients treated with the therapeutic regimen containing at least one of the two drugs to which they were sensitive in vitro. A clinical response of 60% was also obtained in patients resistant in vitro but treated with other drugs. Fifty percent of the subjects sensitive in vitro had a clinical response, although treated with other drugs, whereas only 25% of the patients re-
TABLE 4
Clinical Characteristics and Chemosensitivity of Patients with a Complete Response 4%
Patient
(years)
Stage
R.B. O.L. M.A.M. C.L.
62 68 25 12
lil
D.L.P. V.A. I.R. B.O. G.F.
66 62 47 71 47
Ill
S.R. M.T.
Previous chemotherapy”
-
Residual disease
Site of sample
In vitro sensitivity A-P A-P A-P A-P AU ah4 A-P A-P A A-P P AU
-
2 cm
E
G.C.’ A.M. A.R.
46 62 51
III III IV
A-P
>2 cm >2 cm >2 cm
0 M E 0
’ Abbreviations: BLEO, bleomycin. b Died. ’ Only Adriamycin was tested. ’ Only k-platinum was tested.
BLEO-MTX
See Table 4 for other abbreviations.
In vitro
In vivo
sensitivity
chemotherapy
A-P P(1 &ml) A”’ PC1 &ml) pu b4ml) s
A-C
PAC BLEO-MTX A-P
Survival (months) 25
9 5 I
Af-P
13
A-P Af-5FU P
12 5 4
CHEMOSENSITIVITY
AND OVARIAN
pointed out an unexpected correlation between in vitro resistance to P and failure of other alkylates administered singly. In our study the most evident observation is that patients sensitive in vitro to at least one of the tested drugs are more frequently responsive to chemotherapy, without regard to the drugs administered. In fact, 68% CR and PR patients were sensitive in vitro, whether or not they were treated with the same drugs, versus 44% of patients resistant in vitro but treated in vivo with therapeutic regimens which might have included drugs used in the test. The mechanisms at the root of the development of resistance to chemotherapy depend on interactions between host and epigenetic factors, for example, a progressive decrease of growth rate and an increase of the intermitotic gap of the proliferating cells, and alterations in the host’s immune response and growth of neoplastic cells in the “pharmacologic sanctuaries” where it is difficult to obtain therapeutic concentrations of the drugs [20]. For this reason cellular death due to the drug is inferior to tumor growth between the two following cycles of treatment, resulting in clinically resistant neoplasia. Furthermore, the presence of specific selection processes in the tumoral population, already genetically and phenotipically heterogeneous, can determine somatic mutations in the neoplastic cells. Naturally precocious somatic mutations in clonal expansion will produce a higher proportion of resistant cells. For a certain period, therefore, no resistant cells are present in tumors undergoing precocious growth. They will appear when the first resistant cell becomes phenotypically stable. This concept, mathematically expressed by the Goldie-Coldman model, points to the direct relationship between the statistic probability of having resistant cells and tumor dimensions [21]. In this group of patients and in the trials considered, we noticed a greater number of cases with residual tumor greater than 2 cm in patients resistant to chemotherapy. The success or failure of chemotherapy is due to the stage (tumor mass) at the time of first treatment. This element probably determines the appearance and growth of chemoresistant cells and therefore the clinical outcome [20].
CONCLUSIONS The HTSCCA still requires technique modifications which will allow the identification of growth factors leading to cloning efficiency increases. The evaluation of the cellular sensitivity could help explain the clinical failure of treatment which is due both to the typical resistance of the cells and to cells being in a phase of the mitotic
17
CANCER
cycle not sensitive to the drug, or to the use of an inadequate drug. The use of different agents with different mechanisms of action should increase our knowledge of the resistance mechanisms of the tumoral cells and of cross-resistance. The HTSCCA can be considered a good method for the isolation and study of tumoral stem cells. The analysis of their kinetics, of markers, and of their cellular and stromal interactions should lead to a greater knowledge in the field of tumoral biology. On the other hand we have pointed out a direct relationship between decreased survival of the unities forming colonies and a percentage increase in clinical response [22]. The absence of in vitro growth could be not only a technical problem, but could also have biologic significance. The in vitro development of tumoral colonies could therefore be a negative prognostic parameter [23,24]. It seems necessary to clarify the relationship between fraction of tumoral growth (the neoplastic cells kinetic evaluated by labeling index) and the data provided by the clonogenic test [25]. REFERENCES 1. Abrams, J. Present optimal therapy in ovarian cancer, Eur. J. Cancer Clin. Oncol. 22, 9-12 (1986). 2. Inaba, M., et al. Active efllux of daunomycins and Adriamycin in sensitive and resistant sublines of P 388 leukemia, Cancer Res. 39, 2200-2203 (1979). 3. Ozols, R. F. Pharmacological reversal of drug resistance in ovarian cancer, Sem. Oncol. 12, 7-11 (1985). 4. Hamburger, A. Use of in vitro tests in predictive cancer chemotherapy, J. Narl. Cancer Inst. 66, 981-988 (1981). 5. Weisenthal, L. M. Clones, dyes, nuclides, mouse kidneys, and . . . virions: A new-clonogenic assay for tumor chemosensitivity, Eur. J. Cancer Clin. Oncol. 23, 9-12 (1987). 6. Selby, P., et al. A critical appraisal of the human tumor stem-cell assay, N. Engl. .Z. Med. 308, 129-134 (1983). 7. Salmon, S. E., et al. Quantitation of differential sensitivity of human-tumor stem cells to anticancer drugs, N. Engl. J. Med. 298, 1321-1327 (1978). 8. MacKintosh, F. R., et al. Methodologic problems in clonogenic assay of spontaneous tumors, Cancer Chemother. Pharmacol. 6, 205-210 (1981). 9. Weisenthal, L. M., and Lippman, M. E. Clonogenic and nonclonogenic in vitro chemosensitivity assays, Cancer Treat. Rep. 69, 615-632 (1985). 10. Umbach, G. E., et al. Experiences with the human tumor colonyforming assay in gynecologic malignancies, J. Cancer Res. Clin. Oncol. 110, 234-237 (1985). 11. Welander, C. E., et al. In vitro chemotherapy testing of gynecologic tumors: Basis for planning therapy? Amer. J. Obstet. Gynecol. 147, 188-195 (1983). 12. Welander, C. E., and Alberts, D. S. Clonogenic assay studies of tumor cells from gynecologic malignancies, in Cancer investigation and management: Female reproductive system (C. Wilson and J. Whitehouse, Eds.), Wiley, Chichester (1985). 13. Williams, C. J. The usefulness of the human tumor stem cell assay
18
14.
15.
16.
17. 18.
19.
CARENZA in Ovarian cancer (N. M. Bleehen, Ed.), Springer Verlag, New York (1985). Alberts, D. S., et al. In vitro clonogenic assay for predicting response of ovarian cancer to chemotherapy, Lancer 2, 340-342 (1980). Alberts, D. S., et al. Chemotherapy of ovarian cancer directed by the human tumor stem cell assay, Cancer Chemother. Pharmacol. 6, 279-285 (1981). Moon, T. E., et al. Quantitative association between the in vitro human tumor stem cell assay and clinical response to cancer chemotherapy, Cancer Chemother. Pharmacol. 6, 211-218 (1981). Alonso, K. Human tumor stem cell assay: A prospective clinical trial, Cancer 54, 2475-2479 (1984). Arbuck, S. G., et al. Limitations of drug sensitivity testing in soft agar for clinical management of patients with ovarian carcinoma, Obstet. Gynecol. 66, 115-120 (1985). Simmonds, A. P., and McDonald, E. C. Ovarian carcinoma cells
ET AL. in culture: Assessment of drug sensitivity by clonogenic assay, Bit. J. Cancer 50, 317-326 (1984). 20. De Vita, V. T., Jr. The relationship between tumor mass and resistance to chemotherapy, Cancer 51, 1209-1220 (1983). 21. Goldie, J. H., and Coldman, A. J. A mathematic model for relating the drug sensitivity to tumors to their spontaneous mutation rate, Cuncer Treat. Rep. 63, 1727-1731 (1979). 22. Von Hoff, D. D., et al. Prospective clinical trial of a human tumor cloning system, Cancer Res. 43, 1926-1931 (1983). 23. Bertoncello, I., ef al. Limitations of the clonal agar assay for the assessment of primary human ovarian tumor biopsies, Bit. J. Cancer 43, 803-811 (1982). 24. Aapro, M. S. Growth of solid tumor cells in clonogenic assay: A prognostic factors? Eur. J. Cancer C/in. Oncol. 21,397-400 (1985). 25. Paradiso, A., et al. Cell kinetics of human epithelial ovarian cancers, Bas. Appl. Histochem. 30, 215-220 (1986).
ONCOLOGY
36, 13-18 (I!?%)
Ovarian Cancer: Relationship between in Vitro Chemosensitivity and Clinical Response L. CARENZA, C. VILLANI, II Department
of Obstetrics
F. L. LABI, AND R. PROSPERI PORTA
G. RULLI, and Gynaecology,
University
“La
Sapienza,”
Rome,
Italy
Received May 3, 1988
Thirty-nine samplesfrom 30 patients with epithelial ovarian cancer were collected. Using the method describedby A. Hamburger (.Z.NutZ. Cancer Inst. 66, 981 (1981)) and P. Salmonet al. (AT. Engl. J. Med. 308, 129(1983)), tumor cellswereexposed to anthreoplasticagents(Adriamycm and &-platinum) and then cultured in double-layeragar. There were 23 evaluablepatients. The percentageof completeand partial responsesfor patients whosecells were sensitivein vitro, whether or not treated with the samedrugs, was68, vs 44% comparedto patientswhosecells were resistantin vitro and treated in viva with therapeutic regimenswhich may or may not have included drugs used in the test. More patients with residual tumor greater than 2 cm were resistantto chemotherapy. D IWO Academic press, I~C.
INTRODUCTION Epithelial ovarian carcinoma continues to be the gynecologic neoplasia with the worst prognosis and is the fourth most frequent cause of cancer death in women. Overall 5-year survival is less than 40% because ovarian cancers are most frequently detected in advanced stages; increased relapse-free and median survival have been obtained by applying an aggressive multidisciplinary therapeutic approach. Alkylators as single agents have produced response rates of 25-50% [I]. Doxorubicin yields results similar to those of alkylating agents in patients not previously treated with drugs, but the mean duration of response is 3 months because of rapid development of resistance, probably because of the development of new metabolic structures which permit an increased clearance of doxorubicin from the neoplastic cells [2]. The most active drug in ovarian cancer is c&platinum, which can induce complete remission even in previously treated patients. Although polychemotherapy has shown higher response percentages, it has not been clearly proven that the median survival time compared to monochemotherapy is increased. However, a higher rate of
complete response occurs, mainly in patients with minimal residual disease after surgery. For this group (less than 30% of all patients with advanced ovarian cancer), complete response to cytotoxic therapy is about 40-70% versus lo-20% in the patient group with residual tumors greater than 2 cm [l]. The development of drug resistance to antineoplastic agents is a major clinical problem that limits the overall effectiveness of chemotherapy. Frequently, acquired resistance to single agents is associated with broad crossresistance to structurally dissimilar drugs, resulting in only a 6-10% response rate to salvage chemotherapy [3]. Moreover, individual patients at the same clinical stage and with the same histologic type often do not respond uniformly to antineoplastic agents [4]. Combined treatments for ovarian cancer are needed: first, an aggressive surgical approach for maximum cytoreduction, followed by identification, and administration of therapeutic regimens to which the neoplasia shows greatest sensitivity. The search for a useful tumor chemosensitivity test dates back to at least 1954 (51. A number of techniques have been proposed: in vivo with xenotransplantation and in vitro with short-, medium-, and long-term cell cultures. The human tumor stem cell clonogenic assay (HTSCCA) has been one of the most widely studied techniques during the last 10 years. It is based on the capacity of neoplastic cells to generate colonies when cultured in semisolid agar. Stem cells have the capacity to renew and represent only a small part of the total population. The HTSCCA makes it possible to study only those cells which are responsible for tumor repopulation after debulking, chemotherapy, and metastatic growth, and therefore should be the primary targets for any cytotoxic therapy [6]. The correlation between the response in vitro to antineoplastic agents and clinical response has proved correct in 95% of drug resistance cases and in 65% of cases
13 00!%8258/90 $1.50 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
14
CARENZA ET AL.
predicting drug sensitivity [6]. Using the HTSCCA it should be possible to select “the right drug for the right patient” in order to apply personalized chemotherapy or to retrospectively study the clinical outcome in relation to parameters such as cloning efficiency and growth inhibition rates obtained with the drugs tested. In this study we report the chemosensitivity, using HTSCCA, of tumor stem cells in patients affected by ovarian cancer.
TABLE 2 Growth of Cell Coloniesin Soft Agar No. of colonies/5 x 16 cells plated Sample site
No.
Median
Range
Colony-forming efficiency (%I
Ovary Metastases Relapse Ascites
15 5 1 7
119.7 71 106.1 337.7
116-129 42-102 81-131 258-422
0.02 0.01 0.02 0.06
METHODS Sample and Patients
Thirty-nine samples from 30 patients with epithelial ovarian cancer were collected (24 primary tumors, 6 metastases, 1 pelvic relapse, 8 ascitic effusions). In one case there was insufficient tissue to assay; five samples were not plated because viable cells were absent. Of the 33 remaining cultures, no growth was observed in two cases, and three specimens were contaminated. Twentyeight cultures were evaluated: 15 for primary ovarian cancer, 5 from intestinal or lymph node metastases, 1 from relapse, and 7 from ascitic effusions (Table 1). The median cloning efficiency (number of colonies/5 x 10’ cells plated x 100) in the 28 evaluated specimens was 0.02 (range 0.01-0.06) (Table 2). Samples from 23 patients were evaluated. The patients’ characteristics are shown in Table 3. Complete response was defined as the disappearance of all manifestation of disease for more than 1 month. Partial response was defined as 50% or greater decrease in the product of the diameters of all measureable lesions. In Vitro Clonogenic Assay
The tumor tissue was prepared under sterile conditions in a laminar vertical flow system. Solid tumor samples were obtained during surgery; specimens were mechanically dissociated with sterile scissors. The samples subsequently underwent enzymatic disaggregation in a mixture of 0.5% collagenase II and 0.02% DNase (both from Sigma) at 37°C for 1 hr and then were filtered
Number
TABLE 3 Characteristicsof Patients with Ovarian Carcinoma No. of patients Age (years) Median
TABLE 1 Tumor SampleCharacteristics Tumor samples collected Tumor samples processed Samples with cell yield Sufficient for plating Contaminated samples Evaluable samples
through sterile gauze and washed twice by centrifugation at 15Og for IO min at 10°C. Cells were suspended in fresh RPM1 (Gibco) and passed through 20-, 22-, and 25-gauge needles. The viable cell count was determined by the trypan blue (0.05%) exclusion test. Ascitic effusions were collected in heparinized bottles (100 U/ml), passed through sterile gauze, centrifuged at 15Og for 20 min at 10°C and resuspended in RPMI. Viable cells were determined by trypan blue. Cells were then exposed to drugs for 1 hr at 37°C. The drugs used were Adriamycin (Adriblastina, FarmitaliaCarlo Erba) and cis-diaminodichloroplatinum (Platamine, Farmitalia-Carlo Erba) at O.Ol-, O.l-, and 1 pg/ml doses. Cells were washed twice with HBSS (Gibco) and plated. Hamburger’s [4] and Salmon’s methods [7] were used to culture tumor cells in double-layer agar. One milliliter of under-layer agar (0.5%), diluted in enriched McCoy’s 5A medium (Gibco), was solidified in a 35 x 20-mm petri dish at room temperature. The cells were plated (5 x 10’ cells/dish) and 1 ml of upper-layer agar, (0.3%) with enriched medium CMRL 1066 (Gibco), was added. Both media contained serum, electrolytes, carbohydrates, amino acids, vitamins, and insulin. All experiments were performed in triplicate. Cultures were left for 1 hr at room temperature to solidify the agar.
Percentage
39 38
91
33 3 28
87 9 85
Range Patients treated previously Residual disease Less than 2 cm Greater than 2 cm Stage (FIGO) I III IV
23 63 2% 17 6123 10123 13123 4/23 14/23 5123
CHEMOSENSITIVITY
AND OVARIAN
Plated cultures were controlled by an inverted phase microscope to ascertain that cell clumps were not present; then the cultures were incubated in a humidified atmosphere of 5% CO2 at 37°C. Plates were examined within 3 days for clusters (15-30 cells) and 7 and 21 days later to count colonies (30 or more cells aggregates). In vitro sensitivity was defined as a 50% or more decrease in the number of colonies [7].
RESULTS Patients’ responses to chemotherapy were divided into three groups: complete clinical response (CR), partial clinical response (PR), and no response (NR). Follow up was at 4-34 months. Table 4 shows data regarding patients in CR. The median age was 63 years, with a range of 25-72 years. Four were at stage I (FIGO) and 7 at stage III. Of 11 patients, 9 showed residual tumor less than 2 cm after surgery. In one case (I.R.) the sample was resistant in vitro, but the patient had been previously treated with Adriamycin (A) and cyclophosphamide (P). Another patient (D.L.P.), previously treated with HEXA-CAF, was sensitive in vitro to the highest concentration of Adriamycin. Both patients were responsive to second-choice chemotherapy (interferon; folic acid; SFU). Eight patients received at least one of the drugs to which they were sensitive in vitro. Of the four patients in PR (median age 69 years), one was at stage III and three were at stage IV. All had a residual tumor greater than 2 cm after surgery (Table 5). Two were sensitive to the tested drugs, both in vitro and
15
CANCER
in vivo, although patient B.C. had previously undergone a therapeutic regimen including c&platinum. The sample of ascitic fluid obtained from patient C.L. was sensitive to c&-platinum, whereas the metastatic biopsy sample was resistant. This patient (CL.) underwent another therapeutic regimen to which she was responsive. Finally 8 patients (median age 59 years) were not responsive to chemotherapy (Table 6). Three patients received at least one of the tested drugs in vitro. One of the two patients previously treated with A-P was sensitive in vitro to high doses of c&platinum, and neither was responsive to the second-choice regimen with folic acid and 5FU. Patients P.R. and A.M. were resistant in vitro to both A and P and had no in vivo response to other therapeutic protocols. Patient P.A. was very interesting. The primary tumor was sensitive in vitro to A-P, but because of renal toxicity, the patient underwent a regimen with Adriamycin and cyclophosphamide. The ascitic fluid sampled during treatment was sensitive only to the highest dose of c&platinum; an additional sampling performed on ascitic fluid formed during the course of disease showed resistance toward both P and A. It was possible to observe the development of in vivo and in vitro resistance to Adriamycin and cross-resistance to c&-platinum in this patient who survived 25 months. The highest percentage of CR and PR (73%) was obtained in the group of patients treated with the therapeutic regimen containing at least one of the two drugs to which they were sensitive in vitro. A clinical response of 60% was also obtained in patients resistant in vitro but treated with other drugs. Fifty percent of the subjects sensitive in vitro had a clinical response, although treated with other drugs, whereas only 25% of the patients re-
TABLE 4
Clinical Characteristics and Chemosensitivity of Patients with a Complete Response 4%
Patient
(years)
Stage
R.B. O.L. M.A.M. C.L.
62 68 25 12
lil
D.L.P. V.A. I.R. B.O. G.F.
66 62 47 71 47
Ill
S.R. M.T.
Previous chemotherapy”
-
Residual disease
Site of sample
In vitro sensitivity A-P A-P A-P A-P AU ah4 A-P A-P A A-P P AU
-
2 cm
E
G.C.’ A.M. A.R.
46 62 51
III III IV
A-P
>2 cm >2 cm >2 cm
0 M E 0
’ Abbreviations: BLEO, bleomycin. b Died. ’ Only Adriamycin was tested. ’ Only k-platinum was tested.
BLEO-MTX
See Table 4 for other abbreviations.
In vitro
In vivo
sensitivity
chemotherapy
A-P P(1 &ml) A”’ PC1 &ml) pu b4ml) s
A-C
PAC BLEO-MTX A-P
Survival (months) 25
9 5 I
Af-P
13
A-P Af-5FU P
12 5 4
CHEMOSENSITIVITY
AND OVARIAN
pointed out an unexpected correlation between in vitro resistance to P and failure of other alkylates administered singly. In our study the most evident observation is that patients sensitive in vitro to at least one of the tested drugs are more frequently responsive to chemotherapy, without regard to the drugs administered. In fact, 68% CR and PR patients were sensitive in vitro, whether or not they were treated with the same drugs, versus 44% of patients resistant in vitro but treated in vivo with therapeutic regimens which might have included drugs used in the test. The mechanisms at the root of the development of resistance to chemotherapy depend on interactions between host and epigenetic factors, for example, a progressive decrease of growth rate and an increase of the intermitotic gap of the proliferating cells, and alterations in the host’s immune response and growth of neoplastic cells in the “pharmacologic sanctuaries” where it is difficult to obtain therapeutic concentrations of the drugs [20]. For this reason cellular death due to the drug is inferior to tumor growth between the two following cycles of treatment, resulting in clinically resistant neoplasia. Furthermore, the presence of specific selection processes in the tumoral population, already genetically and phenotipically heterogeneous, can determine somatic mutations in the neoplastic cells. Naturally precocious somatic mutations in clonal expansion will produce a higher proportion of resistant cells. For a certain period, therefore, no resistant cells are present in tumors undergoing precocious growth. They will appear when the first resistant cell becomes phenotypically stable. This concept, mathematically expressed by the Goldie-Coldman model, points to the direct relationship between the statistic probability of having resistant cells and tumor dimensions [21]. In this group of patients and in the trials considered, we noticed a greater number of cases with residual tumor greater than 2 cm in patients resistant to chemotherapy. The success or failure of chemotherapy is due to the stage (tumor mass) at the time of first treatment. This element probably determines the appearance and growth of chemoresistant cells and therefore the clinical outcome [20].
CONCLUSIONS The HTSCCA still requires technique modifications which will allow the identification of growth factors leading to cloning efficiency increases. The evaluation of the cellular sensitivity could help explain the clinical failure of treatment which is due both to the typical resistance of the cells and to cells being in a phase of the mitotic
17
CANCER
cycle not sensitive to the drug, or to the use of an inadequate drug. The use of different agents with different mechanisms of action should increase our knowledge of the resistance mechanisms of the tumoral cells and of cross-resistance. The HTSCCA can be considered a good method for the isolation and study of tumoral stem cells. The analysis of their kinetics, of markers, and of their cellular and stromal interactions should lead to a greater knowledge in the field of tumoral biology. On the other hand we have pointed out a direct relationship between decreased survival of the unities forming colonies and a percentage increase in clinical response [22]. The absence of in vitro growth could be not only a technical problem, but could also have biologic significance. The in vitro development of tumoral colonies could therefore be a negative prognostic parameter [23,24]. It seems necessary to clarify the relationship between fraction of tumoral growth (the neoplastic cells kinetic evaluated by labeling index) and the data provided by the clonogenic test [25]. REFERENCES 1. Abrams, J. Present optimal therapy in ovarian cancer, Eur. J. Cancer Clin. Oncol. 22, 9-12 (1986). 2. Inaba, M., et al. Active efllux of daunomycins and Adriamycin in sensitive and resistant sublines of P 388 leukemia, Cancer Res. 39, 2200-2203 (1979). 3. Ozols, R. F. Pharmacological reversal of drug resistance in ovarian cancer, Sem. Oncol. 12, 7-11 (1985). 4. Hamburger, A. Use of in vitro tests in predictive cancer chemotherapy, J. Narl. Cancer Inst. 66, 981-988 (1981). 5. Weisenthal, L. M. Clones, dyes, nuclides, mouse kidneys, and . . . virions: A new-clonogenic assay for tumor chemosensitivity, Eur. J. Cancer Clin. Oncol. 23, 9-12 (1987). 6. Selby, P., et al. A critical appraisal of the human tumor stem-cell assay, N. Engl. .Z. Med. 308, 129-134 (1983). 7. Salmon, S. E., et al. Quantitation of differential sensitivity of human-tumor stem cells to anticancer drugs, N. Engl. J. Med. 298, 1321-1327 (1978). 8. MacKintosh, F. R., et al. Methodologic problems in clonogenic assay of spontaneous tumors, Cancer Chemother. Pharmacol. 6, 205-210 (1981). 9. Weisenthal, L. M., and Lippman, M. E. Clonogenic and nonclonogenic in vitro chemosensitivity assays, Cancer Treat. Rep. 69, 615-632 (1985). 10. Umbach, G. E., et al. Experiences with the human tumor colonyforming assay in gynecologic malignancies, J. Cancer Res. Clin. Oncol. 110, 234-237 (1985). 11. Welander, C. E., et al. In vitro chemotherapy testing of gynecologic tumors: Basis for planning therapy? Amer. J. Obstet. Gynecol. 147, 188-195 (1983). 12. Welander, C. E., and Alberts, D. S. Clonogenic assay studies of tumor cells from gynecologic malignancies, in Cancer investigation and management: Female reproductive system (C. Wilson and J. Whitehouse, Eds.), Wiley, Chichester (1985). 13. Williams, C. J. The usefulness of the human tumor stem cell assay
18
14.
15.
16.
17. 18.
19.
CARENZA in Ovarian cancer (N. M. Bleehen, Ed.), Springer Verlag, New York (1985). Alberts, D. S., et al. In vitro clonogenic assay for predicting response of ovarian cancer to chemotherapy, Lancer 2, 340-342 (1980). Alberts, D. S., et al. Chemotherapy of ovarian cancer directed by the human tumor stem cell assay, Cancer Chemother. Pharmacol. 6, 279-285 (1981). Moon, T. E., et al. Quantitative association between the in vitro human tumor stem cell assay and clinical response to cancer chemotherapy, Cancer Chemother. Pharmacol. 6, 211-218 (1981). Alonso, K. Human tumor stem cell assay: A prospective clinical trial, Cancer 54, 2475-2479 (1984). Arbuck, S. G., et al. Limitations of drug sensitivity testing in soft agar for clinical management of patients with ovarian carcinoma, Obstet. Gynecol. 66, 115-120 (1985). Simmonds, A. P., and McDonald, E. C. Ovarian carcinoma cells
ET AL. in culture: Assessment of drug sensitivity by clonogenic assay, Bit. J. Cancer 50, 317-326 (1984). 20. De Vita, V. T., Jr. The relationship between tumor mass and resistance to chemotherapy, Cancer 51, 1209-1220 (1983). 21. Goldie, J. H., and Coldman, A. J. A mathematic model for relating the drug sensitivity to tumors to their spontaneous mutation rate, Cuncer Treat. Rep. 63, 1727-1731 (1979). 22. Von Hoff, D. D., et al. Prospective clinical trial of a human tumor cloning system, Cancer Res. 43, 1926-1931 (1983). 23. Bertoncello, I., ef al. Limitations of the clonal agar assay for the assessment of primary human ovarian tumor biopsies, Bit. J. Cancer 43, 803-811 (1982). 24. Aapro, M. S. Growth of solid tumor cells in clonogenic assay: A prognostic factors? Eur. J. Cancer C/in. Oncol. 21,397-400 (1985). 25. Paradiso, A., et al. Cell kinetics of human epithelial ovarian cancers, Bas. Appl. Histochem. 30, 215-220 (1986).
