Characterization of in vitro chemosensitivity of perioperative human ovarian malignancies by adenosine triphosphate chemosensitivity assay Robert T. Gerhardt, MPH, James P. Perras, PhD, Bernd-Uwe Sevin, MD, PhD, Edgar Petru, MD, Reinaldo Ramos, BS, Lucy Guerra, and Hervy E. Averette, MD Miami, Florida We report the in vitro chemosensitivity of primary and recurrent human ovarian tumor samples analyzed by adenosine triphosphate chemosensitivity assay. We defined sensitivity as a ~70% decrease in intracellular adenosine triphosphate versus control at 20% of the reported peak plasma concentration per agent tested. Twenty of 21 assays (95.24%) were completed successfully. Single-agent and combined dose-response patterns consisting of decreasing viability with increasing drug concentration were observed consistently. Thirteen primary tumors were assayed, with 15.4% sensitive to cisplatin, 7.7% sensitive to 4-hydroxycyclophosphamide and 53.8% sensitive to their combination. Seven recurrent tumors were assayed, with 14.3% sensitive to cisplatin, 28.6% sensitive to 5-fluorouracil, and 42.9% sensitive to their combination. Dose-response characteristics and in vitro sensitivity rates reported in this article are consistent with reports of patient response in the literature. We conclude that adenosine triphosphate chemosensitivity assay is an efficient and reliable instrument for the in vitro chemosensitivity assessment of human tumors and warrants further clinical investigation. (AM J OSSTET GYNECOL 1991 ;165:245-55.)

Key words: Adenosine triphosphate chemosensitivity assay, bioluminescence, chemosensitivity, ovarian cancer

Researchers and clinical oncologists have long recognized the potential value of pretreatment assessment of chemosensitivity for patients with cancer. 1 Efforts in this direction to date have produced mixed results! The issues of in vitro evaluability, heterogeneity of tumor subpopulations," emerging drug resistance: and the questionable role of clonogenic or stem cells in malignancy have provided formidable obstacles to the development of an in vitro assay possessing in vivo predictive value. Assays such as the human clonogenic assay, succinate dehydrogenase inhibition, and radioactive thymidine uptake index are often inconclusive, and they possess a high rate of nonevaluability.5.7 Consequently, in vitro chemosensitivity testing to date has remained primarily within the purview of the basic scientist, leaving the clinician to select chemotherapeutic regimens on an empiric basis. From the Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, University of Miami School of Medicine. Supported in part by the John R. McCain Medical Student Research Fellowship, South Atlantic Association of Obstetricians and Gynecologists, and the Doris Kleiber Memorial Fund, University of Miami. John R. McCain Fellowship Lecture, presented at the Fifty-third Annual Meeting of the South Atlantic Association of Obstetricians and Gynecologists, Hot Springs, Virginia, January 27-30,

1991.

Reprint requests: Bernd-Uwe Sevin, MD, PhD, Division of Gynecologic Oncology, Department of Obstetrics and Gynecology (D-52), University of Miami School of Medicine, P.O. Box 016190, Miami, FL 33101. 6/6/30756

Ovarian malignancy is a formidable disease for the patient and a challenge for the clinician. Its insidious onset with relatively late detection, high malignant potential, tendency toward early metastasis, and rapid development of resistance to chemotherapy contribute to a generally poor prognosis for the afflicted patient. 8 The standard treatment for both primary and recurrent malignancy includes cytoreductive surgery followed by adjuvant chemotherapy with one or a combination of several agents, commonly including cisplatin and cyclosphosphamide. 5-Fluorouracil (5-FU) is sometimes used in the salvage therapy of recurrent malignancies. Prognosis is related to clinical stage, histologic grade, tumor type, the amount of unresectable residual tumor at the time of cytoreductive surgery, and response to postoperative adjuvant chemotherapy. The adenosine triphosphate (ATP) chemosensitivity assay has been reported as an efficient, highly evaluable, and cost-effective means of characterizing the in vitro chemosensitivity of cell lines and fresh human tumors, with promising preliminary clinical correlations. g· ll With ATP extraction by acid cytolysis and measurement by luciferin-Iuciferase bioluminometry, ATP chemosensitivity assay assesses chemotherapeutic-induced cell kill of the entire tumor sample (including proliferative and non proliferative subpopulations) by comparing total ATP concentration in treated samples versus untreated controls. Total ATP concentration as

245

246

Gerhardt et al.

assessed by ATP chemosensitivity assay demonstrates a strong correlation with in vitro viable cell count (R 2 > 0.95).9-11 This study was designed to assess critically the use of the ATP chemosensitivity assay as an in vitro model of human tumor chemosensitivity. Both primary and recurrent human ovarian carcinomas obtained intraoperatively were analyzed over a wider range of chemotherapeutic dosages than in prior studies to more accurately characterize chemosensitivity and the related phenomena of drug resistance, heterogeneity, and single- versus combined-agent effects. In addition, in vitro chemosensitivity rates obtained by ATP chemosensitivity assay were compared with clinical reports of patient response to identical single agents and combinations to determine whether the in vitro results were consistent with clinical experience. Material and methods

Tumor tissue specimens were obtained from 21 patients who underwent exploratory laparotomy for surgical staging and cytoreduction of ovarian carcinoma at the University of Miami/Jackson Memorial Medical Center between Feb. 12, 1988, and Nov. 31, 1989. Primary ovarian carcinoma was diagnosed in 14 of these patients whereas recurrent ovarian cancer was diagnosed in seven. The biopsy specimens were obtained under sterile conditions from fresh tumors immediately after resection. Samples were subjected to ATP chemosensitivity assay, which is described in detail in a prior publication. lo Briefly, ATP chemosensitivity assay begins with mechanical disaggregation followed by gentle enzymatic disaggregation, which yields a suspension of single cells and small clusters of up to 30 cells each. The cell suspension is then plated over a soft agar underlayer in standard 24-well culture plates. The agar underlayer allows growth and proliferation of neoplastic cells, but impedes the survival of nonneoplastic tissues by preventing anchorage. Varying concentrations of individual chemotherapeutic agents and combinations are then applied to the wells in triplicate. After a 6-day incubation period, intracellular ATP is extracted by trichloroacetic acid cytolysis, followed by buffering of the resultant solution and luminometry using a luciferin-luciferase enzyme complex. This complex generates quantifiable light energy when bound to ATP molecules. The average luminosity of drugtreated wells versus controls is then compared to characterize the cytotoxic effect of the drug or combination in question. Primary tumor samples were tested for sensitivity to cisplatin, 4-hydroxycyclophosphamide (an active metabolite of cyclophosphamide) and to a combination of these two agents; recurrent tumor samples were tested for sensitivity to cisplatin, 5-FU, and to their combi-

August 1991 Am J Obstet Gynecol

nation. All chemotherapeutic agents and combinations were tested at final peak plasma concentration equivalents l2 of 10%, 20%, 50%, 100%, and 500% after taking into consideration a final volume composed of cell suspension, agar underlayer, and the aliquot of chemotherapeutic agent. The drug concentration calculations were based on the following 100% peak plasma concentration values reported in micrograms of agent per milliliter of plasma: 2.5 /-lg/ml of cisplatin, 6 /-lg/ml of 4-hydroxycyclophosphamide, and 50 /-lg/ml of 5-FU. The value for 4-hydroxycyclophosphamide represents 20% of the reported peak plasma concentration of cyclophosphamide (30 /-lg / ml); this reflects the resultant plasma level of this metabolite after its intrahepatic conversion. After obtaining the bioluminescence counts and calculating the ATP concentration in each control and treated culture well, average ATP concentration and coefficient of variance were calculated for each control and treatment type-dose group. The difference in average ATP concentration between treatment and control groups was then calculated. Dose-response curves were constructed from these data. In this study, sensitivity to a chemotherapeutic agent or combination was defined as a ~70% average reduction in ATP concentration versus control at a dose of 20% of the peak plasma concentration for the agent or combination. If a chemotherapeutically treated tumor sample failed to meet these criteria it was designated resistant to that respective agent or combination. The concentrations of chemotherapeutic agents necessary to produce a 50% decrease of cellular ATP (IC 5o ) were calculated by median effect analysis. This method uses linear regression of the log of drug concentration versus the log of the quotient of the fraction of affected cells divided by the fraction of unaffected cells [Log (C) versus Log(f,/fu)] to produce a least-squares equation from which IC 50 may be calculated. I' These calculations were also used to determine the type of combined drug response pattern. Results

During this study, 21 subject tumor tissue specimens eligible at time of surgery underwent ATP chemosensitivity assay. Twenty assays were completed successfully, yielding an evaluability rate of95.24%. One assay displayed evidence of bacterial contamination and was excluded from the study. Among the triplicate control wells studied, coefficients of variation ranged from 2.3% to 15.9%. Sixty-five percent of all triplicate controls analyzed possessed coefficients of variation of ~1O%, whereas 95% possessed coefficients of variation of ~15%. Greater variation was noted among chemotherapeutically treated triplicates, with coefficients of variation at 20% peak plasma concentration drug con-

In vitro chemosensitivity of human ovarian cancers

Volume 165 Number 2

247

Table I. Fifty percent inhibitory chemotherapeutic concentration (lC;o) for primary human ovarian carcinomas treated with cisplatin, 4-hydroxycyclophosphamide, and their combination Tumor

IC,o DDP

IC m 4HC

IC,o Combined

Ag-l Cah-l Cer-l Fr- l Prov-l Ric-l Rol- l Wei-l Wes-l HG-l HC-l IS-l CL-l

12.73 109.82 461.88 59.94 75.15 273.53 21.30 46.29 180.22 21.04 42.46 17.30 26.60

16.98 9.87 22 .85 29.08 70.51 56.23 30.70 51.66 160.33 21.12 12.95 10.62 19.77

8.51 5. 15 NA* 0.74 NA* NA* NA* NA* NA* 13.60 6.24 4.80 17.0

Average IC", SD R2 (DDP vs 4-He)

103.7 131.7 0.05

39.43 40.93

8.01* 5.58*

DDP, Cisplatin; 4HC, 4-hydroxycydophosphamide. IC ;o is reported as percentage of peak plasma concentration. *Because of a change in laboratory procedure while this study was still in progress, combined drug concentrations for these tumors were prepared only for 10%/ 10%, 20%/ 20 %,50%/ 20%, and 100%110% of cisplatin and 4-hydroxycydophosphamide, respectively. The lack of three consecutive equivalent drug concentrations rendered it impossible to calculate IC;o values, thus the combination statistics were calculated without them.

Table II. Fifty percent inhibitory chemotherapeutic concentration (IC;o) for recurrent human ovarian carcinomas treated with cisplatin, 5-FU, and their combination Tumor

IC,oDDP

IC,o 5-FU

IC,o Combined

MC-l MS-l PO-I LP-I MR-I La- I YK-l

30.91 0.0004 39.92 15.38 9.93 45.88 62.90

42.48 8.88 0.013 1l.l9 12.41 38.27 25.97

21 .59 0.0004 8.49 3. 13 8.64 20.07 21.89

Average IC,o SD R2 (DDP vs 5-FU)

27.5 23.72 0.45

19.88 15.98

11.97 9. 13

DDP, Cisplatin. IC,o is reported as percentage of peak plasma concentration.

centration ranging from 6.01% to 67. 19%. Among these triplicates, 40% possessed coefficients of variation of

Characterization of in vitro chemosensitivity of perioperative human ovarian malignancies by adenosine triphosphate chemosensitivity assay.

We report the in vitro chemosensitivity of primary and recurrent human ovarian tumor samples analyzed by adenosine triphosphate chemosensitivity assay...
1MB Sizes 0 Downloads 0 Views