A Phase II Trial of Combination Chemotherapy and Surgical Resection for the Treatment of Metastatic Adrenocortical Carcinoma
Continuous Infusion Doxorubicin, Vincristine, and Etoposide with Daily Mitotane as a P-glycoprotein Antagonist
Jame Abraham, M.D.1 Susan Bakke, R.N.1 Ann Rutt, R.N.1 Beverly Meadows, R.N.1 Maria Merino, M.D.2 Richard Alexander, M.D.3 David Schrump, M.D.3 David Bartlett, M.D.3 Peter Choyke, M.D.4 Rob Robey, M.S. Elizabeth Hung, M.S., M.P.H.1 Seth M. Steinberg, Ph.D.5 Susan Bates, M.D.1 Tito Fojo, M.D., Ph.D.1
1 Medicine Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, Maryland.
2 Surgical Pathology Branch, Center for Cancer Research, National Cancer Institute, National Insti- tutes of Health, Bethesda, Maryland.
3 Surgery Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, Maryland.
4 Department of Radiology, Clinical Center, Na- tional Institutes of Health, Bethesda, Maryland.
5 Biostatistics and Data Management Section, Na- tional Cancer Institute, National Institutes of Health, Bethesda, Maryland.
Presented in part as a poster at the American Society of Clinical Oncology Annual Meeting, At- lanta, Georgia, May 15-18, 1999.
Address for reprints: Tito Fojo, M.D., Ph.D., Medicine Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Building 10, Room 12N226, 9000 Rockville Pike, Bethesda, MD 20892; Fax: (301) 402-1608; E-mail: tfojo@ helix.nih.gov
Received March 16, 2001; revision received Au- gust 27, 2001; accepted January 29, 2002.
BACKGROUND. Adrenocortical carcinoma (ACC) is rare, nearly always fatal, and to the authors’ knowledge has few nonsurgical treatment options. Based on in vitro studies demonstrating the efficacy of mitotane as a P-glycoprotein (Pgp) antago- nist, and expression of high levels of Pgp in ACC, the authors conducted a study of infusional doxorubicin, vincristine, and etoposide with oral mitotane ± surgical resection in patients with metastatic ACC.
METHODS. Thirty-six patients with metastatic ACC received daily oral mitotane (mean, 4.6 g/day) and 96-hour infusional doxorubicin (10 mg/m2/day), etoposide (75 mg/m2/day), and vincristine (0.4 mg/m2/day). Four responding patients (11%) underwent surgery.
RESULTS. Thirty-five patients were evaluable; all had metastatic disease. Eleven patients had not undergone resection of the primary tumor. Approximately 53% of patients had functional tumors. A total of 190 cycles were administered to 36 patients. Responses were observed in 8 patients (22%): 1 complete, 4 partial, and 3 minor responses. The mean duration of response was 12.4 months. Using a landmark method, the median survival of patients who did not respond to che- motherapy was 11.6 months from a point 4 months after the initiation of therapy, whereas that of 8 patients who demonstrated a response to chemotherapy was 34.3 months from that same landmark. High levels of Pgp expression were documented in nine of nine tumors. Mitotane levels > 10 µg/mL, previously shown to antag- onize Pgp in vitro, were achieved in 25 of 36 patients (69%). However, rhodamine efflux from CD56-positive cells was not impaired, suggesting poor in vivo Pgp inhibi- tion. The predominant Grade 3/4 toxicity (according to the Common Toxicity Criteria of the National Cancer Institute) was neutropenia in 66% of cycles; however, fever occurred in only 3% of cycles. Daily mitotane was associated with Grade 1/2 nausea, diarrhea, fatigue, and neuropsychiatric changes in 31 of 36 patients (86%).
CONCLUSIONS. Using a combination regimen of daily mitotane with infusional doxorubicin, vincristine, and etoposide in patients with metastatic ACC, responses were observed in 22% of patients. The superiority of this combination over single- agent mitotane is uncertain. The side effects of mitotane made treatment difficult. More effective Pgp antagonists are needed. Cancer 2002;94:2333-43.
@ 2002 American Cancer Society.
DOI 10.1002/cncr.10487
KEYWORDS: combination chemotherapy, drug resistance reversal, adrenocortical carcinoma (ACC), P-glycoprotein (Pgp), mitotane, doxorubicin, etoposide, vincristine.
A drenocortical carcinoma (ACC) is a rare tumor with an estimated annual incidence of 1-2 cases per million in the U.S.1 At the time of diagnosis approximately 40-70% of the tumors already have un-
dergone metastatic spread.2 ACC has a very poor prog- nosis with an overall 5-year mortality rate of 75-90%.2,3 To our knowledge, the treatment options for metastatic ACC are limited. Although ACC is po- tentially curable in early stages with surgical resection, only 30% of tumors are reported to be confined to the adrenal gland at the time of diagnosis.4,5 Mitotane (o,p’DDD), an analogue of the insecticide DDT, orig- inally was developed in 1960 and has been used alone in the treatment of ACC or in combination with other agents.6 Although early studies in the era before mod- ern imaging modalities reported high response rates, it now is clear that mitotane has at best a 19-22% response rate.7-9 In addition to the evaluation of mi- totane, numerous studies examining the efficacy of conventional chemotherapy have demonstrated either no or very low activity, indicating a limited role for these agents.10-12 Doxorubicin as a single agent or in combination chemotherapy has shown some activity against ACC.13,14 To our knowledge there are no data regarding the activity of vincristine or etoposide as single agents in the treatment of ACC.
Multidrug resistance (MDR) is a major cause of chemotherapy failure in cancer treatment. Extensive in vitro evidence has established P-glycoprotein (Pgp), the product of the MDR1 gene, as a mediator of MDR.15 Surveys of human tumors have found very high levels of Pgp expression in a majority of ACC cases.16,17 The high levels of Pgp expression in ACC and its demonstrated function as a drug efflux pump provided the basis for proposing that the treatment of ACC may benefit from the addition of a Pgp antago- nist. Work in our laboratories has shown that in vitro, mitotane can increase drug accumulation by decreas- ing drug efflux, resulting in enhanced cytotoxicity. This effect has been observed in cell lines expressing Pgp, suggesting that mitotane interferes with Pgp function.18,19
The current report describes the results of a Phase II study evaluating a multimodality approach in pa- tients with metastatic ACC. The treatment strategy was comprised of daily mitotane (o.p.‘DDD) with con- tinuous infusion doxorubicin, vincristine, and etopo- side that was administered preoperatively and post- operatively to patients with potentially resectable disease. Surgical resection was used as an integral modality in cases in which it was deemed that poten- tial benefit could be derived. Doxorubicin, vincristine, and etoposide, three natural products that are re- ported to be effective against a wide spectrum of ma- lignancies, can be transported out of cells by Pgp. Although mitotane is known to be cytotoxic to ACC, we used it as an inhibitor of Pgp. The purpose of the current report is to summarize our findings.
MATERIALS AND METHODS Patients and Methods
Thirty-six patients were enrolled on the trial; informed consent was obtained from all patients. Patients were accrued onto the study between August 1993 and July 1999. Patients with biopsy proven recurrent or meta- static ACC with measurable disease were eligible. Pathologic material from all 36 patients was reviewed and the diagnosis of ACC was confirmed by the De- partment of Pathology of the National Cancer Institute (M.M.). Eligibility criteria included: a life expectancy of at least 3 months, a Karnofsky performance status of ≥ 70%, adequate bone marrow function (leukocyte count > 3000/mm3, and platelet count > 100,000/ mm3), adequate renal function (creatinine clearance > 50 mL/minute), and adequate liver function (serum bilirubin < 1.5 mg/dL, and serum transaminase < twice the normal limit). Patients who had received prior doxorubicin were enrolled, provided they met all other entry criteria and had an ejection fraction > 40% as determined by multigated angiogram (MUGA) scan. Prior mitotane therapy was allowed provided a dose of 3 g/day had been tolerated for at least 1 week. Exclusion criteria included second malignancy other than squamous cell carcinoma of the skin, active sys- temic infection, positive serology for the human im- munodeficiency virus, or a positive pregnancy test. Patients were not allowed to receive other potential Pgp antagonists such as diltiazem, nicardipine, phe- nothiazines, phenytoin, or verapamil. The Institu- tional Review Board of the National Cancer Institute approved the protocol. All patients provided written informed consent.
The goal of the study was to determine the effi- cacy of the therapy by treating patients who were considered to be candidates for surgical resection ei- ther at the time of presentation or after a response to therapy. Although it was considered desirable that patients had disease that was surgically resectable at the time of presentation or that could be resected after an initial response to chemotherapy, it also was stated that, at the discretion of the principal investigator, a patient could be enrolled even if surgical resectability was believed to be unlikely. The latter was deemed necessary because ACC is a rare disease, and stricter entry criteria would have made accrual very difficult if not impossible. With these liberal acceptance criteria, all but three of the patients who were referred and screened as potential protocol candidates were en- rolled on the study. The three patients who were not enrolled included 1 patient whose life expectancy was believed to be < 3 months (and who died 2 months later), and 2 patients in whom surgical resection was
TABLE 1 Treatment Regimen
Doxorubicin: 10 mg/m2/day × 4 days by continuous infusionª Etoposide: 75 mg/m2/day × 4 days by continuous infusionª Vincristine: 0.4 mg/m2/day × 4 days by continuous infusion Mitotane: 2-3 g/day increasing as tolerated to achieve 10-15 mg/mL
ª Dose escalation allowed if the absolute granulocyte count did not fall below 1000 in a given cycle. Dose levels were:
1) Doxorubicin, 12.5 mg/m2.
2) Doxorubicin, 15 mg/m2.
3) Etoposide, 100 mg/m2.
4) Etoposide, 125 mg/m2.
5) Etoposide, 150 mg/m2.
deemed to be best medical practice and who conse- quently underwent surgery and were not enrolled on the study because they underwent complete resec- tions. These latter two patients had not been evalu- ated by a surgical oncologist before referral.
Treatment Plan Chemotherapy
Chemotherapy was administered on an outpatient ba- sis as a continuous intravenous infusion through an indwelling central venous catheter over 96 hours. Starting doses were: doxorubicin, 10 mg/m2/day; vin- cristine, 0.4 mg/m2/day; and etoposide, 75 mg/m2/ day. Cycles were repeated every 3 weeks if the neutro- phil count was > 1500/µL and the platelet count was > 100,000/p.L. Mitotane was initiated at a dose of 2-3 g daily, 2-3 weeks prior to the administration of the first cycle of infusional chemotherapy and was con- tinued on a daily schedule for the duration of the protocol. The dose of mitotane was increased 1-2 g per day, every cycle, as tolerated to achieve a thera- peutic level of 10-15 µg/L. Antiemetics were admin- istered as needed, but did not include phenothiazines. Diphenhydramine hydrochloride, dexamethasone, droperidol, metoclopramide, ondansetron, and loraz- epam were allowed. Dose escalation was allowed (1 increase per cycle), if the absolute granulocyte count did not fall below 1000 µL in a given cycle. Dose levels were as follows: 1) doxorubicin, 12.5 mg/m2/day; 2) doxorubicin, 15 mg/m2/day; 3) etoposide, 100 mg/m2/ day; 4) etoposide, 125 mg/m2/day; and 5) etoposide, 150 mg/m2/day (Table 1).
Surgery and postsurgical treatment
Surgical resections were performed after three or more cycles of chemotherapy in patients with stable disease or who demonstrated any response to chemotherapy, if both the principal investigator and the surgical con- sultant agreed that this was the best strategy. Guided
by findings on computed tomography (CT) scan and/or magnetic resonance imaging (MRI), resections were attempted only if the surgical consultant be- lieved that a complete resection was possible. The existence of disease in another area (e.g., pulmonary metastases in a patient under consideration for lapa- rotomy) was not considered a contraindication, pro- vided a second surgical consultant believed the dis- ease at this other site also would likely be completely resectable after an interval of 6 weeks to 3 months. In no case did a patient who was responding to chemo- therapy undergo surgery if it was believed that a com- plete resection would not be possible. In patients with evidence of progressive disease, surgical resection was performed after two or more cycles of chemotherapy. Decision-making in these patients was guided by the same criteria applied to patients responding to che- motherapy. The surgeon obtained a separate in- formed consent for the procedure. Patients with resid- ual disease who responded to chemotherapy prior to surgery received additional chemotherapy, beginning as soon as possible after surgery. A maximum of four additional cycles were administered after surgery. However, mitotane was continued indefinitely.
Replacement hormonal therapy
Previous studies have documented the occurrence of adrenal insufficiency in patients receiving mitotane therapy.11,12,20 In the current study, this complication was treated by the administration of daily hydrocorti- sone, and when indicated, supplemental fluorohydro- cortisone. In the first few patients, replacement ther- apy was not initiated until there was clinical evidence of adrenal insufficiency or a high adrenocorticotropic hormone level was documented. However, in subse- quent patients, replacement therapy usually was ini- tiated within 4-6 weeks of the start of mitotane treat- ment, before signs of adrenal insufficiency developed.
Criteria for Response and Toxicity
Radiologic follow-up of the disease in the abdomen and pelvis was performed after three cycles of chemo- therapy initially, then repeated every two cycles. A complete response was defined as no evidence of measurable disease for a minimum of 4 weeks. A par- tial response was defined as a ≥ 50% decrease in the sum of the products of the dimensions of all measur- able lesions for at least 1 month. Minor response was defined as 25-50% decrease in the sum of the products of the dimensions of all measurable lesions for at least 1 month. Any patient who experienced a complete response, a partial response, or a minor response that warranted surgical resection was considered to have achieved a response for this study. Stable disease was
a decrease of < 25% or an increase of < 25% in the sum of the products of the dimensions of all measur- able lesions over 2 cycles of chemotherapy. Progres- sive disease was defined as a ≥ 25% increase in the sum of the products of the dimension of any measur- able lesion over 1 cycle of chemotherapy or the ap- pearance of any new lesions consistent with meta- static disease. Reduction in hormone production was not considered as a criteria for response. Toxicity was graded according to the Common Toxicity Criteria of the National Cancer Institute. MUGA scans were re- peated if the initial ejection fraction was < 45% or if the total doxorubicin dose exceeded 350 mg/m2.
It should be stressed that although the response criteria summarized earlier were adhered to, the study was not designed to establish the best response but rather the optimal time to proceed with surgical inter- vention, if surgery was possible. Thus, four patients were referred for surgical resection even though they still were responding to chemotherapy and had likely not achieved a maximal response. This was followed by the administration of postoperative chemotherapy. Specifically, the protocol stated that surgical debulk- ing would be performed after three or more cycles of chemotherapy in those patients with stable disease or who demonstrated any response to chemotherapy if both the principal investigator and the surgical con- sultant agreed that this was the best strategy. Surgical debulking also was allowed for patients with progres- sive disease if this was deemed the best strategy. Five patients with stable or progressive disease underwent surgical resections.
Rhodamine Accumulation in CD56-Positive T-cells
Rhodamine accumulation in CD56-positive (CD56+) cells as a means to monitor Pgp antagonism in a surrogate assay was evaluated.21 Whole blood was ob- tained from each patient in a heparinized syringe. Rhodamine 123 (Sigma Chemical Company, St. Louis, MO) with or without PSC 833 was added to aliquots of whole blood to achieve a final rhodamine concentra- tion of 0.5 µg/mL and a PSC 833 concentration of 3 ug/mL. The blood was incubated for 30 minutes at 37 ℃ in 5% carbon dioxide (CO2). After the accumulation period, the blood was layered onto Lymphocyte Sep- aration Medium and centrifuged at 2000 revolutions per minute for 5 minutes. The mononuclear cell layer from each tube was transferred to a separate tube, washed with cold Dulbecco phosphate buffered saline (DPBS), resuspended in 200 µL cold DPBS with 2% fetal calf serum (DPBS/FCS), and held at 4 ℃ for later staining. Cells that were to be subjected to an efflux period then were resuspended in rhodamine-free complete media (phenol red-free improved minimal
essential medium [IMEM] with 10% FCS) with or with- out 3 µg/mL PSC 833 and incubated another 60 min- utes at 37 ℃ in 5% CO2. After the efflux period, cells were sedimented and washed twice with cold DPBS/ FCS. After the washings, the cells were resuspended in 200 µL cold DPBS/FCS. The cells then were stained with phycoerythrin (PE)-labeled CD56 antibody (Bec- ton Dickinson, San José, CA) or PE-labeled mouse immunoglobulin (Ig) G1 (Becton Dickinson) as a neg- ative control. After staining, the cells were washed twice and then resuspended in DPBS and kept on ice in the dark until analyzed. A FACSort flow cytometer (Becton Dickinson) with a 488-nanometer (nm) argon laser was used to analyze the samples. Rhodamine fluorescence was collected after a 520-nm bandpass filter and PE fluorescence was collected after a 585-nm bandpass filter. A minimum of 5000 events were col- lected per sample and the samples were gated on forward scatter versus side scatter to exclude clumps and debris. Dead cells were excluded based on pro- pidium iodide staining.
Statistical Methods
The protocol was designed using a three-stage accrual plan to help ensure that additional patients were en- tered only if there was potential benefit. In the first stage five patients were entered. It was planned that if no patient responded, then a 50% response was ruled out with 90% confidence and accrual would stop. However, 3 patients exhibited a clinical response and accrual proceeded to 11 patients. Once the first 11 patients were accrued and evaluated, a decision would be made regarding further accrual. If only 1 or 2 of the 11 patients responded, accrual would stop because a 50% response was ruled out successfully. However, because 3 patients responded to therapy, accrual was continued until 30 patients were entered. Although there were sufficient clinical responses observed to accrue 30 patients, the number ended up being much less than 50%. However, during the course of the study, a surrogate assay for Pgp inhibition was devel- oped and validated in the laboratory, and because of its potential value, a decision was made to evaluate a cohort of patients using this assay. Consequently, an amendment was enacted to allow up to 12 additional patients to be enrolled to obtain correlations between mitotane levels and inhibition of Pgp.
Progression-free survival and overall survival were determined from the date of diagnosis as well as the date the patient was enrolled on study, until the date of disease progression, death, or last follow-up as ap- propriate. The probability of survival or progression- free survival was determined using the Kaplan-Meier method, and the statistical significance of the differ-
| Total No. of Patients: | |
|---|---|
| 36 | |
| Gender | |
| Male | 11 (31%) |
| Female | 25 (69%) |
| Age (yrs) | |
| Median | 44.4 |
| Range | 23-70 |
| Performance status | |
| ECOG 1 | 36 (100%) |
| Stage of disease | |
| Metastatic | 36 (100%) |
| Sites of metastases | |
| Lung | 27 (75%) |
| Liver | 18 (50%) |
| Adrenal bed | 17 (47%) |
| Mediastinum | 10 (28%) |
| Retroperitoneum | 9 (25%) |
| Soft tissue | 8 (22%) |
| Bone | 5 (14%) |
ECOG: Eastern Cooperative Oncology Group.
ence between pairs of Kaplan-Meier curves was de- termined by the Mantel-Haenszel method.22 In addi- tion, a landmark method was used to determine the survival from a point 4 months after the initiation of therapy to provide an unbiased conditional estimate of the difference in survival between responders and nonresponders. Four months was selected because this was the timepoint by which 7 of 8 responses were noted, whereas the longest time to declare a response was nearly 6 months after the initiation of therapy. All P values are two-sided. The median potential follow- up was 52.3 months since diagnosis and 34.8 months since the on-study date.
RESULTS
The study began on August 20, 1993 and the last patient was accrued on July 15, 1999. A total of 36 patients were enrolled; 1 patient was not evaluable. This patient had a long history of extensive disease and died just before the start of the second cycle, after hematologic recovery from the first cycle of treatment.
A diagnosis of ACC was confirmed in all patients, after review. The clinical characteristics of the 36 pa- tients enrolled on study are summarized in Table 2. Of the 36 patients, 25 were females and 11 were males. The median age was 44.4 years (range, 23-70 years). All patients had an Eastern Cooperative Oncology Group (ECOG) performance status of 1. All 36 patients had metastatic disease. Sites of disease included the lung (75%), liver (50%), adrenal bed (47%), mediasti-
| Type of treatment | No. of (%) |
|---|---|
| Total no. of patients | 36 (100%) |
| Adrenalectomy | 25 (69%) |
| Metastatectomy | 8 (22%) |
| Mitotane | 12 (33%) |
| Radiotherapy | 4 (11%) |
| Chemotherapy | 13 (36%) |
| Paclitaxel | 6 (17%) |
| Cisplatin | 6 (17%) |
| Etoposide | 5 (14%) |
| Carboplatin | 3 (8%) |
| Doxorubicin | 2 (6%) |
| Others | 4 (11%) |
num (27%), retroperitoneum (25%), soft tissues (22%), and bone (13%). Approximately 53% of the patients had either clinical features consistent with functioning tumors or biochemical evidence of excess hormone production. The clinical features in the majority of patients (13 of 19 patients) were subtle, comprised of signs such as increased facial hair. Only six patients had the more “classic” Cushing syndrome presenta- tion. Prior treatments included surgery in 91% (adre- nalectomy in 69% and metastatectomy in 22%), che- motherapy in 36%, and radiotherapy in 11% of patients (Table 3). Prior chemotherapy included cis- platin (six patients), paclitaxel (six patients), etoposide (five patients), carboplatin (three patients), or doxoru- bicin (two patients). Other treatments included topo- tecan (one patient), PSC 833 and vinblastine (one pa- tient), and aminoglutethimide (one patient). The most commonly used combination chemotherapy regimen was cisplatin and etoposide (four patients). With the exception of one patient, objective responses were not achieved with any of these chemotherapy regimens. One-third of the patients (12 patients) had received prior mitotane therapy without success.
An objective response was documented in 8 of 36 evaluable patients (22%) with 1 complete response (3%), 4 partial responses (11%), and 3 minor responses (8%). Among responders, the median time to response was 2.7 months, with objective evidence of response first noted as early as after the first cycle, or as late as after the fifth cycle. Sites of objective responses in- cluded the lung (seven patients), adrenal bed (one patient), and retroperitoneum (three patients). Re- sponses were observed in patients with extensive dis- ease, including as many as 69 lung metastases, and abdominal masses as large as 10 cm × 5.5 cm. The mean duration of response exceeded 12.4 months. Of the responding patients, four went on to have surgical resections performed. Surgical resection also was at-
ADRENOCORTICAL CANCER SURVIVAL
100
90
80
PERCENT SURVIVAL
70
60
50
40
30
20
10
0
12
24
36
48
60
MONTHS FROM LANDMARK DATE
Legend: * non-responder
☐ responder
16/25 failed
4/8 failed
tempted in five patients who did not respond to che- motherapy. The median survival from the on-study date for the group as a whole was 13.5 months. Using a landmark method, the median survival for those patients whose tumor responded to chemotherapy was 34.3 months from a point 4 months after the initiation of therapy, whereas that of patients whose tumor did not respond to chemotherapy was 11.6 months. Among the variables examined, only the functional status of the tumor could potentially be related to overall survival. Patients with nonfunctional tumors had a longer survival from the time of diagno- sis (P = 0.02) (Fig. 1) and from the on-study date (P = 0.056). No other variable could be associated with overall survival including gender, a mitotane level < 14 µg/mL or > 14 µg/mL, and a mitotane level < 10 ug/mL or > 10 µg/mL.
A total of 190 cycles of chemotherapy were admin- istered. The median number of chemotherapy cycles per patient was 5.6 (range, 1-16 cycles). None of the patients discontinued the treatment because of toxic- ity (Table 4). The starting doses stipulated by the pro- tocol were administered to all patients in the first cycle. In subsequent cycles, a higher dose of doxoru- bicin was administered in 75 cycles and a higher eto- poside dose was administered in 38 cycles. Dose re- ductions were not necessary. The predominant Grade 3/4 toxicity was neutropenia in 66% of patients (125 episodes). Other Grade 3/4 toxicities included fever in
| Toxicity | No. of cycles (%)b |
|---|---|
| Neutropenia | 125 (66%) |
| Fever/neutropenia | 6 (3%) |
| Nausea/emesis | 5 (3%) |
| Anemia | 12 (6%) |
| Thrombocytopenia | 4 (2%) |
| Prolonged PTT | 4 (2%) |
PTT: partial prothrombin time. a Grading was performed using the Common Toxicity Criteria of the National Cancer Institute. b The total number of cycles was 190 (100%).
Other Grade 3/4 toxicities (< 1%) included diarrhea, fatigue, ataxia, neuropsychiatric changes, and paresthesia.
the setting of neutropenia (3%), nausea/emesis (3%), anemia (6%), and thrombocytopenia (2%). Very rarely, Grade 3/4 neurologic toxicities such as paresthesia, ataxia, and neuropsychiatric symptoms were ob- served. Weight loss was common, especially in the first few cycles, with patients loosing a mean of 10% of their body weight.
Mitotane was administered aggressively, seeking to achieve a serum level of ≥ 10-15 µg/mL. This dose range was targeted based on previous reports suggest- ing enhanced efficacy at levels > 10 µg/mL or 14 µg/mL and in vitro data indicating maximal inhibition of Pgp at a mitotane concentration of 15 µg/mL.10,1 10,11
| Toxicity | No. of cycles (%)b |
|---|---|
| Nausea/emesis | 90 (47%) |
| Diarrhea | 40 (21%) |
| Fatigue | 39 (21%) |
| Neurologic events | 63 (33%) |
| Neuropsychiatric changes | 31 (16%) |
| Paresthesia | 26 (14%) |
| Ataxia | 6 (3%) |
a Grading was performed using the Common Toxicity Criteria of the National Cancer Institute. b The total number of cycles was 190 (100%).
The median dose of mitotane was 6 g/day, with a mean of 4.6 g/day. The mean level was 12.9 µg/mL. A level of 10 µg/mL was reached in 25 patients, whereas a level of 14 µg/mL was attained in 20 patients. The mean daily dose of mitotane in patients reaching a level of 10 µg/mL was 5.7 g compared with a mean daily dose of 5.65 g in patients reaching a level of 14 µg/mL. The similarity between these 2 doses reflects the fact that once a level of 10 µg/mL was reached, the dose usually was not adjusted further because it was expected that with continued administration, higher levels would be attained. A mean of 2.5 cycles were administered before a level of 10 µg/mL was reached. As mentioned earlier, no association between re- sponse or overall survival and maximum mitotane level could be demonstrated when cutoffs of 10 µg/mL or 14 µg/mL were employed. Daily mitotane therapy was associated with significant chronic toxicities. Grade 1 or 2 toxicities ascribed to mitotane included nausea, diarrhea, fatigue, and neuropsychiatric changes (Table 5). These were observed in 31 of 36 patients (86%). They were ascribed to mitotane either because 1) they began during the first 2-3 weeks of treatment when only mitotane had been adminis- tered, 2) were present continuously throughout the cycle and were not determined to be related to the administration of chemotherapy, 3) persisted even af- ter chemotherapy had been discontinued and the pa- tient was receiving only mitotane, and/or 4) resolved on discontinuation of mitotane or after a reduction in the dose administered. It is important to emphasize that although the severity of these toxicities was less as judged by the Common Toxicity Criteria, they often were present continuously, without relief. The pa- tients’ overall quality of life was impaired greatly with daily mitotane therapy compared with the intermit- tent chemotherapy.
High levels of Pgp were demonstrated in the nine patients in whom expression was measured using a
quantitative polymerase chain reaction assay. The lev- els were comparable to those measured in many mul- tidrug resistant cell lines maintained in vitro, which are 10 to several hundred fold more resistant to che- motherapy than the parental cells from which they were derived.
Despite aggressive mitotane administration, and the attainment of levels > 10 µg/mL in 25 patients, we were unable to demonstrate evidence of in vivo Pgp inhibition in a subset of study patients using a surro- gate assay that employs circulating CD56+ cells.21 Figure 2 illustrates a representative result. This assay is performed by incubating mononuclear cells in a solu- tion containing the Pgp substrate, rhodamine, and determining rhodamine accumulation by measuring fluorescence. A sample obtained from a patient is divided in two. One sample is directly incubated with rhodamine and the extent of accumulation reflects the extent of Pgp inhibition by the mitotane bathing cir- culating cells. The other sample is incubated in rho- damine with exogenous PSC 833 added to a final con- centration of 3 µg/mL. This final concentration of PSC 833 results in complete inhibition of Pgp-mediated efflux and maximal rhodamine accumulation. The greater the inhibition by the circulating mitotane, the smaller the difference between the mean fluorescence obtained with and that obtained without the addition of exogenous PSC 833. Conversely, the less the inhibi- tion by the circulating mitotane, the greater the dif- ference between the result without added PSC 833 and the result after the addition of PSC 833. As shown in Figure 2, mitotane failed to inhibit Pgp in the CD56+ cells. For each patient, samples were obtained before the initiation of mitotane therapy and at 3-4-week intervals thereafter. Thus, for each patient, rhodamine accumulations over a range of mitotane concentra- tions were available. This assay was performed in a total of 14 samples, obtained from 8 patients, after extensive experience allowed for a reliable compari- son between samples obtained weeks to months apart. The mitotane levels at the time the CD56+ cells were harvested ranged from 2.2-24 µg/mL. In all but two patients the mitotane levels were > 10 µg/mL, at least during 1 harvest. However, the lack of any sub- stantial inhibition indicates that in vivo inhibition of this Pgp surrogate by mitotane was not possible.
DISCUSSION
The current report describes the results of what to our knowledge is the largest single institution Phase II trial reported to date in patients with ACC. A combined modality regimen comprised of oral mitotane and in- fusional doxorubicin, etoposide, and vincristine with surgical resection was evaluated in 36 patients with
70
PSC
BLANK
EFFLUX
0
0
200
400
600
800
1000
100
PSC
BLANK
EFFLUX
Counts
0
0
200
400
600
800
1000
120
PSC
BLANK
EFFLUX
0
0
200
400
600
800
1000
120
2
PSC
BLANK
EFFLUX
0
0
200
400
600
800
1000
Rhodamine Fluorescence
metastatic ACC. All patients received oral mitotane and the infusional chemotherapy, with a subset of the patients proceeding to undergo surgical resection. A total of 190 cycles of infusional chemotherapy were administered with only minimal Grade 3/4 toxicities in addition to neutropenia. However, chronic Grade 1 and 2 toxicities attributed to mitotane made the ther- apy difficult to administer, although no patient with- drew from the protocol because of toxicity. Responses to the combination of oral mitotane and infusional chemotherapy were observed in 22% of patients. From a landmark of 4 months after the initiation of therapy, the median survival of patients who responded to chemotherapy was 34.3 months, whereas the median survival of patients who did not respond to treatment was 11.6 months from the same landmark. Although activity was observed with this combination of mito- tane with infusional chemotherapy, the superiority of this regimen over single-agent mitotane remains un- certain. Furthermore, although mitotane was included in the regimen as Pgp antagonist, evidence obtained from a surrogate assay using CD56+ cells suggests that it likely was not an effective in vivo inhibitor of drug efflux.
ACC is a rare disease that affects only approxi- mately 250 patients in the U.S. annually. This low incidence has made it difficult to perform appropriate studies to evaluate treatment regimens. There is uni- versal agreement that surgical resection of primary disease offers the best chance for cure.3-5,20 Further- more, although the randomization of some surgical studies has not been optimal, several single institution and cooperative group studies have shown that surgi- cal resection can prolong survival in patients with metastatic or recurrent disease.24,25 Although defini- tive answers are not available, one can use these data to argue in support of the thesis that surgical resection is the treatment of choice in patients with primary, metastatic, and recurrent ACC.4,26
To our knowledge, among medical options, only mitotane has been studied extensively to date. Origi- nally described in 1960, mitotane is an adrenolytic agent that also can cause adrenal inhibition without causing cellular destruction.3,6-8 To our knowledge, its mechanism of action is not clear. Anecdotal reports, case series, and retrospective studies have reported regression of metastases, long-term disease remission, and increased survival.2,27 However, to our knowledge randomized controlled studies have not been per- formed to date. Some early studies reported response rates as high as 60-80%, but these all antedate the era of modern imaging techniques, including CT and MRI,6-8 and often relied principally on clinical obser- vations, including improvement in symptoms related
to hormone excess. Although to our knowledge early studies did not report drug levels, two retrospective studies concluded that mitotane levels > 10 µg/mL or 14 µg/mL were associated with a higher response rate.10,27,28 In the current study, no association was observed between mitotane levels > 10 µg/mL or 14 µg/mL and response, suggesting that mitotane levels did not have a significant effect on the efficacy of this treatment regimen.
With few exceptions, the majority of reports of cytotoxic therapy for ACC are anecdotal or include a small number of patients. The Southwestern Oncology Group (SWOG) published a Phase II study in 1993, using cisplatin and mitotane.29 Forty-two patients, di- vided into low-risk and high-risk groups, were treated with either 75 mg/m2 or 100 mg/m2 of cisplatin, re- spectively, in conjunction with 4 g of mitotane daily. The overall response rate was 30%, with a median duration of response of 7.9 months and a median survival of 11.8 months. This survival was similar to that reported previously, and mitotane levels were not reported.12,13,24 A more recent study from SWOG using 50 mg/m2 of cisplatin on Days 1 and 2 together with 100 mg/m2 of etoposide on Days 1, 2, and 3 achieved a response rate of 11%.30 However, as the authors cautioned, there was overlapping between the confi- dence intervals of the two trials. In an ECOG study in which moderate doses of mitotane were used to treat patients with well differentiated, functional tumors, the response rate was reported to be 22%.13 Those patients with poorly differentiated histologies and those who failed mitotane were treated with single- agent doxorubicin, with a 19% response rate. Recog- nizing the potential value of surgical resection in this disease, but cognizant of its limitations, we designed the current study trial as a combined modality regi- men. We treated our patients with a combination of three natural products whose transport is mediated by Pgp (doxorubicin, vincristine, and etoposide) and oral mitotane before and after surgical resection. Doxoru- bicin, vincristine, and etoposide were administered as a 4-day continuous infusion because in vitro evidence has shown that this mode of administration may be more effective in overcoming Pgp-mediated resis- tance.31 Although mitotane is believed to be cytotoxic to ACC, the principal reason for its inclusion in this regimen was as an inhibitor of Pgp. Extensive in vitro evidence indicates that resistance to natural products can be mediated by Pgp.15 Surveys of human tumors have found very high levels of expression in a majority of adrenocortical carcinomas.16,17 Furthermore, in vitro studies with the ACC cell line, H295, have dem- onstrated high levels of Pgp expression with evidence of a role for this protein in mediating drug efflux.18
The high levels of Pgp expression and its demon- strated function as a drug efflux pump suggests that the treatment of ACC may benefit from the addition of a Pgp antagonist. Previous studies have shown that in vitro, mitotane can increase drug accumulation by decreasing drug efflux, resulting in enhanced cytotox- icity. This effect was observed in cell lines expressing Pgp, not of adrenocortical origin, suggesting that mi- totane interferes with Pgp function.
Because the pharmacokinetics of mitotane re- quire prolonged administration before levels of 10-15 ug/mL are reached, mitotane was administered con- tinuously beginning 2-3 weeks before the initial che- motherapy infusion. However, it is clear that even this lead time in mitotane administration was not suffi- cient to reach the desired blood levels. Thus in only 5 of the 36 patients were levels > 10 µg/mL attained by the time of the first cycle. Furthermore, a level of 10 ug/mL was reached in 17 of 36 patients and 25 of 36 patients, respectively, by the time of the second and third cycles. Although this was not optimal in terms of protocol design, we were reluctant to delay the start of chemotherapy further in these patients with meta- static disease.
Although it is not possible to conclude whether this mitotane/chemotherapy regimen is superior to single-agent mitotane, the results of the current study suggest the infusional therapy administered in this regimen had some activity. This statement is sup- ported by 1) the observation that response to therapy was independent of mitotane level, 2) the fact that in several cases, responses occurred quickly (as early as the first cycle) before substantial mitotane levels had been attained, 3) responses were observed in patients whose tumors had not responded to prior mitotane, and 4) responses were observed in some patients with mitotane levels < 4 µg/mL. An additional observation relates to the activity of mitotane in patients with this disease. Although, as discussed earlier, response rates as high as 60-80% have been reported in the litera- ture, the results of the current study indicate that a more modest response rate is likely. In the current study mitotane was administered very aggressively, to a point of maximal toxicity, and responses were as- sessed using modern imaging techniques. Conse- quently, unless one believes that the coincident ad- ministration of chemotherapy had a negative impact on the activity of mitotane, a response rate higher than that observed herein should not be expected for sin- gle-agent mitotane.
A total of five patients achieved complete or par- tial responses, with minimal responses observed in an additional three patients. However, because we had designed the study to evaluate a multimodality ther-
apy, patients with minimal responses underwent sur- gery, even though they were continuing to respond, because of plans to administer additional chemother- apy postoperatively. Although the overall response rate was not high, patients who responded to therapy appeared to benefit from the treatment regimen, as suggested by their overall survival from a landmark corresponding to the time of response. The extent to which the chemotherapy contributed to the survival cannot be ascertained because it could be argued that the improvement came from the surgical intervention, a modality known to prolong survival in patients with this disease. However, this is unlikely because four of the eight patients in this group did not undergo sur- gical resection. One interpretation is that a combined, aggressive approach to patients with disseminated disease can improve overall survival.
As shown in the current study, even in patients with mitotane levels > 10 µg/mL we were unable to demonstrate in vivo evidence of Pgp inhibition using a surrogate assay. Although it is possible that Pgp inhi- bition still could have occurred in tumor cells, this appears unlikely. Because toxicity from mitotane made the treatment difficult for many patients, we would suggest that a more effective and less toxic Pgp inhibitor be used in a follow-up study. Several agents that currently are under development are potential candidates. Unlike the agents used in initial studies that were not sufficiently potent and dissimilar from subsequent agents such as PSC-833 that had problem- atic pharmacokinetic interactions, these newer agents are more potent, do not appear to be substrates for Pgp, and result in irreversible inhibition. These in- clude XR-9576 (Tariquidar), LY335979, R101933, and OC144-093.32-35 Unlike mitotane, which failed to per- form adequately as a Pgp inhibitor, these agents have been shown to inhibit rhodamine efflux from CD56+ cells completely, and to increase the accumulation of 99mtechnetium-Sestambi in the liver and some tu- mors.
Finally, we would add some observations regard- ing the use of mitotane in the treatment of patients with ACC. Although we have suggested that the re- sponse rate to this agent is far less than previously reported, the literature indicates that it does have some activity as a single agent.7 More important, it can effectively suppress hormonal production or secre- tion, a property that is desirable in patients with man- ifestations of excess hormone production. With regard to the latter, in three patients, we were impressed with the extent to which the symptoms of excess hormone production could be controlled, even as tumor size increased dramatically. These observations make it easy to understand how the antitumor activity of this
agent may have been overestimated in early studies. They also emphasize the need to initiate other ther- apy, such as ketoconazole, as mitotane is withdrawn, lest patients suffer from marked Cushing symptoms as mitotane is eliminated slowly over weeks to months. We also found, as has been noted by others, that after prolonged mitotane administration, the daily dose re- quired to maintain levels of 10-20 µg/mL was reduced substantially, with some patients requiring as little as 500 mg to 1 g. At these reduced doses, tolerability improved substantially, indicating that after an initial difficult period, long-term therapy can be better tol- erated. It also highlights the possibility of significant toxicity developing in patients receiving long-term ad- ministration of mitotane if the dose is not adjusted.
In the current study, we report a single institution trial of 36 patients with ACC who were treated with infusional doxorubicin, vincristine, and etoposide with oral mitotane. Responses were observed in 22% of patients. The lack of an association between re- sponse and maximum mitotane levels, and the obser- vation that responses occurred before mitotane levels reached “therapeutic” doses, suggest that the infu- sional regimen utilized has some activity in this dis- ease. Although mitotane was used as a Pgp antagonist, a surrogate assay using CD56+ cells found that in vivo inhibition of Pgp was not achieved. Given mitotane’s lack of efficacy as an in vivo Pgp antagonist, and the chronic toxicity associated with mitotane therapy, fu- ture efforts in the treatment of ACC will be directed at using more effective and less toxic Pgp antagonists.
REFERENCES
1. Brennan MF. Adrenocortical carcinoma. CA Cancer J Clin. 1987;37:348-365.
2. Wooten MD, King DK. Adrenal cortical carcinoma. Epide- miology and treatment with mitotane and a review of the literature. Cancer. 1993;72:3145-3155.
3. Demeure MJ, Somberg LB. Functioning and nonfunctioning adrenocortical carcinoma: clinical presentation and thera- peutic strategies. Surg Oncol Clin N Am. 1998;7:791-805.
4. Norton JA. Adrenal tumors. In: Devita VT Jr., Hellman S, Rosenberg SA, editors. Cancer: principles and practice of oncology. 5th edition. Philadelphia: Lippincott-Raven, 1997: 1659-1669.
5. Pommier RF, Brennan MF. An eleven-year experience with adrenocortical carcinoma. Surgery. 1992;112:963-970.
6. Bergenstal DM, Hertz R, Lipsett MB. Chemotherapy of ad- renocortical cancer with o’p’DDD. Ann Intern Med. 1960;53: 672-682.
7. Hutter AM Jr., Kayhoe DE. Adrenal cortical carcinoma. Re- sults of treatment with o,p’DDD in 138 patients. Am J Med. 1966;41:581-592.
8. Lubitz JA, Freeman L, Okun R. Mitotane use in inoperable adrenalcortical carcinoma. JAMA. 1973;223:1109-1112.
9. Nader S, Hickey RC, Sellin RV, et al. Adrenal cortical carci- noma. A study of 77 cases. Cancer. 1983;52:707-711.
10. van Slooten H, Moolenaar AJ, van Seters AP, et al. The treatment of adrenocortical carcinoma with o,p’-DDD: prognostic implications of serum level monitoring. Eur J Cancer Clin Oncol. 1984;20:47-53.
11. Vassilopoulou-Sellin R, Guinee VF, Klein MJ, et al. Impact of adjuvant mitotane on the clinical course of patients with adrenocortical cancer. Cancer. 1993;71:3119-3123.
12. Samaan NA, Hickey RC. Adrenal cortical carcinoma. Semin Oncol. 1987;14:292-296.
13. Decker RA, Elson P, Hogan TF, et al. Eastern Cooperative Oncology Group study 1879: mitotane and adriamycin in patients with advanced adrenocortical carcinoma. Surgery. 1991;110:1006-1013.
14. Haq MM, Legha SS, Samaan NA, et al. Cytotoxic chemother- apy in adrenal cortical carcinoma. Cancer Treat Rep. 1980; 64:909-913.
15. Ling V, Gerlach J, Kartner N. Multidrug resistance. Breast Cancer Res Treat. 1984;4:89-94.
16. Flynn SD, Murren JR, Kirby WM, et al. P-glycoprotein ex- pression and multidrug resistance in adrenocortical carci- noma. Surgery. 1992;112:981-986.
17. Haak HR, van Seters AP, Moolenaar AJ, et al. Expression of P-glycoprotein in relation to clinical manifestation, treat- ment and prognosis of adrenocortical cancer. Eur J Cancer. 1993;29A:1036-1038.
18. Bates SE, Shieh CY, Mickley LA, et al. Mitotane enhances cytotoxicity of chemotherapy in cell lines expressing a mul- tidrug resistance gene (mdr-1/P-glycoprotein) which is also expressed by adrenocortical carcinomas. J Clin Endocrinol Metab. 1991;73:18-29.
19. Feller N, Hoekman K, Kuiper CM, et al. A patient with adrenocortical carcinoma: characterization of its biological activity and drug resistance profile. Cancer Res. 1997;3:389- 394.
20. Luton JP, Cerdas S, Billaud L, et al. Clinical features of adrenocortical carcinoma, prognostic factors, and the effect of mitotane therapy. N Engl J Med. 1990;322:1195-1201.
21. Robey R, Bakke S, Stein W, et al. Efflux of rhodamine from CD56+ cells as a surrogate marker for reversal of P-glyco- protein-mediated drug efflux by PSC 833. Blood. 1999;93: 306-314.
22. Anderson JR, Cain KC, Gelber RD, et al. Analysis and inter- pretation of the comparison of survival by treatment out- come variables in cancer clinical trials. Cancer Treat Rep. 1985;69:1139-1146.
23. Icard P, Chapuis Y, Andreassian B, et al. Adrenocortical carcinoma in surgically treated patients: a retrospective
study on 156 cases by the French Association of Endocrine Surgery. Surgery. 1992;112:972-979.
24. Jensen JC, Pass HI, Sindelar WF, et al. Recurrent or meta- static disease in select patients with adrenocortical carci- noma. Aggressive resection vs chemotherapy. Arch Surg. 1991;126:457-461.
25. Bellantone R, Ferrante A, Boscherini M, et al. Role of reop- eration in recurrence of adrenal cortical carcinoma: results from 188 cases collected in the Italian National Registry for Adrenal Cortical Carcinoma. Surgery. 1997;122:1212-1218.
26. Cai W, Counsell RE, Djanegara T, et al. J Metabolic activa- tion and binding of mitotane in adrenal cortex homoge- nates. J Pharm Sci. 1995;84:134-138.
27. Haak HR, Hermans J, van de Velde CJ, et al. Optimal treatment of adrenocortical carcinoma with mitotane: results in a con- secutive series of 96 patients. Br J Cancer. 1994;69:947-951.
28. von Slooten H, van Seters AP, Smeenk D, et al. O,p’-DDD (mitotane) levels in plasma and tissues during chemother- apy and at autopsy. Cancer Chemother Pharmacol. 1982;9: 85-88.
29. Bukowski RM, Wolfe M, Levine HS, et al. Phase II trial of mitotane and cisplatin in patients with adrenal carcinoma: a Southwest Oncology Group study. J Clin Oncol. 1993;11:161- 165.
30. Williamson SK, Lew D, Miller GJ, et al. Phase II evaluation of cisplatin and etoposide followed by mitotane at disease progression in patients with locally advanced or metastatic adrenocortical carcinoma: a Southwest Oncology Group Study. Cancer. 2000;88:1159-1165.
31. Fojo AT. Multidrug resistance in human tumors. Cancer Treat Res. 1989;48:27-36.
32. Stewart A, Steiner J, Mellows G, et al. Phase I trial of XR9576 in healthy volunteers demonstrates modulation of P-glyco- protein in CD56+ lymphocytes after oral and intravenous administration. Clin Cancer Res. 2000;6(11):4186-4191.
33. Dantzig AH, Shepard RL, Law KL, et al. Selectivity of the multidrug resistance modulator, LY335979, for P-glycopro- tein and effect on cytochrome P-450 activities. J Pharmacol Exp Ther. 1999;290:854-862.
34. van Zuylen L, Sparreboom A, van der Gaast A, et al. The orally administered P-glycoprotein inhibitor R101933 does not alter the plasma pharmacokinetics of docetaxel. Clin Cancer Res. 2000;4:1365-1371.
35. Newman MJ, Rodarte JC, Benbatoul KD, et al. Discovery and characterization of OC144-093, a novel inhibitor of P-glyco- protein-mediated multidrug resistance. Cancer Res. 2000;60: 2964-2972.