`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Urologic Oncology 6 (2001) 111–115
`
`Review article
`Randomized phase 2 trial of ketoconazole and ketoconazole/doxorubicin
`in androgen independent prostate cancer
`a,
`b
`c
`d
`Randall Millikan, M.D.
`*, Luis Baez, M.D.
`, Tarit Banerjee, M.D.
`, James Wade, M.D.
`e
`e
`f
`, Roger Winn, M.D.
`, Terry L. Smith, B.S.
`,
`Kimberly Edwards, R.N.
`a
`Christopher Logothetis, M.D.
`a
`Center for Genitourinary Oncology, Department of Genitourinary Medical Oncology, University of Texas M.D. Anderson Cancer Center, Houston, TX, USA
`b
`San Juan City Community Oncology Program, San Juan, TX, USA
`c
`Marshfield Community Oncology Program, Marshfield, WI, USA
`d
`Central Illinois Community Oncology Program, Peoria, IL, USA
`e
`Department of Clinical Investigation, Section of Community Oncology, University of Texas M.D. Anderson Cancer Center, Houston, TX, USA
`f
`Department of Biostatistics, University of Texas M.D. Anderson Cancer Center, Houston, TX, USA
`Received 23 June 2000; accepted 3 October 2000
`
`,
`
`Abstract
`
`Eighty-nine patients with progressive prostate cancer despite suppression of testosterone and withdrawal of anti-androgens were stud-
`ied. This was a relatively advanced population, with 63 of 89 having either osseous metastases (mets) beyond the axial skeleton or visceral
`mets. Patients were randomly assigned to receive either ketoconazole alone, or ketoconazole with weekly doxorubicin. All patients re-
`>
` 80% main-
`ceived replacement hydrocortisone. The primary endpoints were response and survival. Based on PSA reduction criteria (
`tained for at least 8 weeks), 14 of 45 patients (31%) in the single-agent ketoconazole arm responded. Sixteen of 44 patients (36%) in the
`combination ketoconazole/doxorubicin arm responded. There were no important differences between the two treatments in any outcome
`measure. The median overall survival for all patients was 12.5 months; median time to progression was 3.3 months. Toxicity was signifi-
`cant with both regimens, and more severe in the doxorubicin arm. Fully 20% of patients in each arm discontinued therapy due to intoler-
`able side effects.Each of these regimens is toxic, and produced responses in fewer than half of treated patients. Although the observed
`median survival does compare favorably with reports from similar cohorts treated in the community, the potential benefit is only modest.
`In our view, neither of these regimens is sufficiently promising to justify phase 3 evaluation. © 2001 Elsevier Science Inc. All rights re-
`served.
`
`
`
`Keywords: Prostate; Ketoconazole; Doxorubicin; Randomized; Community; Phase II
`
`1. Introduction
`
`Hormone refractory prostate cancer continues to be a
`major therapeutic challenge and to claim more than 40,000
`lives annually. Despite the increasing recognition that com-
`bination chemotherapy has clinically relevant anti-tumor
`activity [1,2], there is still no standard therapy for the thou-
`sands of men who will pass the benchmark of androgen in-
`dependent progression this year. Recently, there has been
`renewed interest in the use of ketoconazole in this setting
`
`* Corresponding author. Randall E. Millikan, Ph.D., M.D., The Uni-
`versity of Texas M.D. Anderson Cancer Center, 1515 Holcombe Blvd.,
`1
`1
`Box 13, Houston, TX, 77030, USA. Tel.:
`1-713-792-2830; Fax:
`1-713-
`745-1625.
`
`
`E-mail address: rmillika@notes.mdacc.tmc.edu (R. Millikan).
`
`[3,4]. In fact, some recent reports suggest that ketoconazole
`is among the most active agents available, with high PSA-
`defined response rates reported [5,6]. Renewed interest in
`ketoconazole derives in part from the recognition that bio-
`availability is critically dependent on gastric acidity, which
`may account for some of the variability in previous reports
`using this agent. In addition, ketoconazole may exert effects
`beyond suppression of steroid hormone synthesis [7–9].
`For several years, ketoconazole in combination with
`doxorubicin has been an integral part of our approach to ad-
`vanced prostate cancer at the University of Texas M.D.
`Anderson Cancer Center [910]. We have used weekly doxo-
`rubicin in these studies in keeping with the observation that
`in general, chemotherapeutic agents have more impact in
`prostate cancer when they are given frequently in modest
`doses, as opposed to dosing at or near the MTD every 3 to 4
`
`1078-1439/01/$ – see front matter © 2001 Elsevier Science Inc. All rights reserved.
`PII: S1078-1439(00)00123-X
`
`
`
`
`JANSSEN EXHIBIT 2064
`Mylan v. Janssen IPR2016-01332
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`112
`
`R.A. Millikan et al. / Urologic Oncology 6 (2001) 111–115
`
`weeks as is typical for more chemosensitive cancers [1].
`Others have investigated ketoconazole in combination with
`oral cyclophosphamide [11].
`So far, most reports of ketoconazole use in androgen in-
`dependent prostate cancer (either alone or in combination)
`have come from referral centers, and the suitability of this
`treatment for the community setting is uncertain. In addi-
`tion, we thought that it was of substantial interest to see if
`either (or both) of these regimens appeared sufficiently ac-
`tive to be tested in phase 3 evaluation. Accordingly, we
`have performed a randomized phase 2 trial of ketoconazole
`alone vs. ketoconazole/doxorubicin in community (CCOP)
`affiliates of the M. D. Anderson Cancer Center. The pri-
`mary goals of this study were to define the response rates
`and toxicities observed with these regimens in the commu-
`nity setting. A randomized phase 2 design provides several
`advantages in this context. First, the impact of PSA testing
`continues to be reflected in significant stage migration in
`patients with prostate cancer, making comparison with his-
`torical benchmarks especially problematic. Second, selec-
`tion of one regimen over another for further development is
`an important, well-established use of this design [12]. Fi-
`nally, especially in the context of a common disease in
`which a two-arm trial does not seriously prolong accrual
`time, it is of interest to have some basis for recognizing
`large differences between therapies. We have previously
`used this design to conclude that neither 5-fluorouracil nor
`5-fluorouracil in combination with interferon were promis-
`ing [13] and that vinblastine in combination with estramus-
`tine was likely to be more active than vinblastine alone [14].
`
`2. Materials and methods
`
`2.1. Patients
`
`Eligible patients had histological proof of adenocarci-
`,
`noma, “castrate” testosterone levels (defined as
`50 ng/dl
`for the purposes of this study), and evidence of androgen in-
`dependent progression within the six months prior to regis-
`tration. No prior exposure to the study agents was allowed.
`Patients had no more than one previous exposure to cyto-
`toxic agents. Patients were off antiandrogens for at least 4
`weeks or had progression since antiandrogen withdrawal.
`All patients had serum PSA greater than 4 ng/ml at entry.
`Adequate physiologic reserve was demonstrated by an esti-
`mated creatinine clearance of at least 35 ml/min; serum lev-
`els of hepatic transaminases less than twice the upper limit
`of normal; and no evidence of bifasicular block by electro-
`cardiography (EKG). All patients had either a negative car-
`diac system review with a normal EKG, or a measured left
`ventricular ejection fraction of at least 45%.
`Exclusion criteria included taking supraphysiologic
`doses of corticosteroids, active peptic ulcer disease or a re-
`quirement for antacid use, an active second malignancy or
`uncontrolled lesions within the central nervous system.
`Patients were prospectively stratified according to extent
`
`of disease into four categories: local/regional only (i.e. nor-
`mal bone scan at entry); bone disease limited to the axial
`skeleton; bone disease involving long bones; or visceral in-
`volvement. This stratification was done in order to help in-
`sure prognostic balance between the two groups; no sub-set
`analysis of these groups was planned or performed.
`
`2.2. Treatment
`
`All patients were treated with ketoconazole, 400 mg p.o.
`tid. Care was taken to instruct patients to take this apart
`from meals, either one hour before or two hours after eating.
`In addition, in order to insure an acidic gastric environment,
`patients were instructed to take the ketoconazole with 250
`mg of ascorbic acid. In the ketoconazole/doxorubicin arm,
`patients were treated with ketoconazole as above, and then,
`starting with the second week, were given weekly doxorubi-
`2
`cin, 20 mg/m
` i.v., either as a brief “piggyback” or as a 24 h
`infusion, at the discretion of the treating physician. In the
`absence of toxicity, doxorubicin was continued to a cumula-
`2
`tive dose of 400 mg/m
`. In general, therapy was continued
`until progression or “maximum benefit” as judged by the
`treating physician.
`Patients unable to tolerate ketoconazole due to GI dis-
`tress were offered sucralfate suspension. This was taken 1 h
`before or 2 h after a dose of ketoconazole. If symptoms per-
`sisted, then ketoconazole was reduced to 200 mg tid. All pa-
`tients received hydrocortisone, 20 mg q AM and 10 mg q
`PM. This is necessary to avoid symptomatic corticosteroid
`deficiency at these adrenolytic doses of ketoconazole.
`
`2.3. Criteria for response and progression
`
`For patients without measurable disease, response was
`defined as a PSA decrease of at least 80% sustained for at
`least 8 weeks, with at least stable symptoms. In addition, pa-
`tients with conventionally measurable disease were required
`to show at least 50% reduction in the product of the greatest
`dimension and its perpendicular. Progression was defined
`by any of the following events: 25% increase in PSA above
`the nadir level; new lesions by bone scan or other imaging
`modality; worsening symptoms attributed to prostate can-
`cer. In order to avoid bias introduced by infrequent PSA
`measurements, all progression times were derived by linear
`interpolation of the PSA measurements on either side of the
`threshold for progression (i.e., 25% increase over PSA na-
`dir). This procedure always provides a conservative esti-
`mate for the date of progression.
`
`2.4. Statistical considerations
`
`The historical response rate with ketoconazole is about
`40%. According to the selection design of Simon et al. [12]
`a two arm randomized phase 2 trial with 37 patients per arm
`provides a 90% chance of correct selection if the true re-
`sponse rates are 40% and 55% respectively. Thus the ac-
`crual goal of 45 patients per arm provided ample power for
`correct selection. Although formal hypothesis testing was
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`R.A. Millikan et al. / Urologic Oncology 6 (2001) 111–115
`
`113
`
`not an objective of this small trial, large differences could
`certainly be detected. (This trial had an 80% power to reject
`a null hypothesis if the median survival was 9 months in one
`arm and 17 months in the other.) The primary end-points of
`the trial were response and survival. We also report time to
`progression (TTP). Although “response duration” is of in-
`terest, there is considerable uncertainty in defining both the
`time of “response” and the time of “progression.” Thus, we
`report TTP (which is uncertain on only one end of the inter-
`val) stratified by response. We believe this provides the
`most objective and informative way to describe the durabil-
`ity of the observed responses.
`
`3. Results
`
`Between July 1995 and October 1996, 90 patients were
`registered from 22 CCOP affiliates. No patients were enrolled
`from the University of Texas M. D. Anderson Cancer Center.
`One patient was subsequently found to be ineligible due to
`concurrent therapy with strontium-89. All other patients are
`reported for response, survival and toxicity, irrespective of
`how much therapy they actually received (intent to treat anal-
`ysis). Patient characteristics are summarized in Table 1.
`As of this report (median follow-up of 3 years), all but
`one patient has progressed; 83 of 89 (92%) have died. In the
`ketoconazole only group, 14 of 45 (31%; 95% c.i. 18–47%)
`responded. The median survival was 12.5 months, with 7
`patients (15%) surviving at 2 years. In the ketoconazole/
`doxorubicin group, 16 of 44 (36%; 95% c.i. 22–52%) re-
`sponded. The median survival was 12.5 months, with 3 pa-
`tients (7%) surviving at 2 years. Since rates of PSA decline
`>
`of
`50% are also commonly reported, we relate that 18 of
`45 (40%) in the ketoconazole arm and 22 of 44 (50%) in the
`combination arm met this criterion. Only 14 (16%) of our
`patients had conventionally “measurable” disease at study
`entry. Of the twelve patients for whom follow-up imaging
`studies were obtained, 2 demonstrated a conventional PR,
`.
`and both had PSA reduction of
`80%. Of the other 10,
`there was only one patient with an 80% PSA reduction, but
`
`Table 1
`Patient characteristics at entry
`
`Feature
`
`Local/regional involvement
`Axial skeleton only
`Long bone involvement
`Visceral involvement
`Age at registration
`(range)
`Age at diagnosis
`(range)
`PSA at entry
`(range)
`Time, hormone therapy to registration
`(range)
`
`KETO
`
`KETO/DOX
`
`2
`11
`25
`7
`70.9 yrs
`(51–82)
`65.3 yrs
`(47–78)
`98 mg/ml
`(5–3,255)
`3.8 yrs
`(0.5–9.3)
`
`1
`12
`23
`8
`70.1 yrs
`(48–85)
`67.3 yrs
`(47–81)
`200 ng/ml
`(11–2,335)
`3.0 yrs
`(0.3–11.3)
`
`3
` 6
`
`no objective response in his measurable disease—a 6
`cm liver mass.
`The median TTP for all patients was 3.3 months. TTP by
`response status is shown in Fig. 1. Of those patients who
`have died, all but two died of prostate cancer. Thus, overall
`survival is essentially identical to cause-specific survival in
`this cohort. Kaplan-Meir survival curves for the two cohorts
`are shown in Fig. 2. Extent of disease at study entry had a
`significant impact on outcome. As shown in Fig. 3, patients
`with involvement of long bones or viscera did significantly
`worse than patients with disease confined to the axial skele-
`ton [15]. As is universally reported, patients demonstrating
`a major PSA response had a much better survival (Fig. 4).
`Toxicity was significant, and dominated by gastrointesti-
`nal complaints such as bloating, nausea, heartburn and anor-
`exia. One patient on the single agent ketoconazole arm had
`grade 4 hepatotoxicity requiring hospitalization; he had
`consumed 1 to 2 beers per day for several years, but was not
`known to have established liver disease. There was one
`death from myocardial infarction in each group. A striking
`finding in our study, in agreement with Mahler [16] and our
`previous anecdotal experience, but in contrast to the report
`of Small et al. [5], was the very high rate of patient with-
`drawal for intolerance of therapy. Overall toxicity results
`are summarized in Table 2. Although it is clear that many
`more grade 3 adverse events were observed in the combina-
`tion arm, it is interesting that an equal number of patients
`(fully 20%) in each arm discontinued therapy for intolerable
`side effects.
`All patients participating in this study were free to have
`any other treatment at the time of disease progression. Four
`patients on single agent ketoconazole were subsequently
`treated with ketoconazole/doxorubicin; all responded. Inter-
`estingly, five other patients were treated with mitoxantrone/
`prednisone following progression on single agent ketocona-
`zole and none responded. Three patients initially treated on
`
`Fig. 1. Actuarial survival based on all-cause mortality. All eligible patients
`included. Kaplan-Meier estimates of TTP in non-responders (n561; lower
`curve) and responders (n528; upper curve).
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`114
`
`R.A. Millikan et al. / Urologic Oncology 6 (2001) 111–115
`
`Fig. 2. Kaplan-Meier estimates of overall according to treatment. Keto5
`Ketoconazole; Dox5Doxorubicin.
`
`Fig. 4. Kaplan-Meier estimates of survival by response status.
`
`the combination arm were subsequently treated with either
`single agent estramustine or mitoxantrone/prednisone; none
`responded.
`
`Discussion
`
`The optimal treatment of advanced prostate cancer is not
`known. Recent studies have confirmed a measurable, clini-
`cally relevant palliative impact of some therapies. It is be-
`coming clear that “secondary” hormonal responses in ad-
`vanced prostate cancer are not as rare as once believed. The
`recent report [17] of a 55% response rate and obvious clinical
`palliation achieved with dexamethasone is striking in this re-
`gard. Although physiologic replacement doses of hydrocorti-
`sone were given in this trial, it is unlikely that the observed
`effects were attributable to hydrocortisone. The available data
`on single-agent prednisone [18], does not suggest that the re-
`sponses seen in this trial are associated with physiologic re-
`placement doses of corticosteroids. The effect of dexametha-
`
`Fig. 3. Kaplan-Meier estimates of survival for all patients by extent of disease
`at entry. Patients with disease confined to axial skeleton (n526) had a much
`better outcome than patients with extra-axial or visceral metastases (n563).
`
`sone seems to be a peculiar property of this agent given at the
`0.75 mg bid dose. Remarkably however, there are no options
`for adrenal suppression that have a favorable toxicity profile.
`Exogenous glucocorticoids, ketoconazole, aminoglutethi-
`mide, mitotane, etc. are all fraught with considerable morbid-
`ity. With respect to more conventional cytotoxic therapy,
`many investigators have recently reported combination regi-
`mens in advanced prostate cancer. Although significant alter-
`ation of the natural history of the disease has not yet been
`convincingly demonstrated, meaningful symptom palliation
`by use of cytotoxic therapy is now widely recognized. In fact,
`the National Comprehensive Cancer Network guidelines [19]
`have endorsed two sequential trials of systemic cytotoxic
`therapy (including Sr-89 as an option in this category) as ap-
`propriate in the standard, palliative management of patients
`with androgen independent prostate cancer.
`In our view, the present trial does confirm biologic activ-
`ity of ketoconazole-based therapy, with many patients hav-
`ing significant PSA declines and clinically relevant periods
`of freedom from progression. Further, the median survival
`we observed in the community setting in a cohort with far-
`advanced, progressive disease compares favorably with the
`community experience recently disclosed for mitoxantrone/
`
`Table 2
`Summary of observed toxicities according to NCI Common
`Toxicity Criteria v. 2.0.
`
`Grade
`
`Anemia
`Anorexia
`Bilirubin increase
`Diarrhea
`Fatigue
`Granulocytopenia
`Nausea alone
`Stomatitis
`Thrombocytopenia
`Vomiting
`
`KETO
`<
`2
`
`8
`2
`0
`5
`3
`0
`14
`1
`5
`8
`
`3
`
`0
`1
`1
`2
`1
`0
`2
`0
`0
`1
`
`4
`
`0
`0
`1
`0
`0
`0
`0
`0
`0
`0
`
`KETO/DOX
`<
`2
`
`3
`
`15
`7
`2
`10
`10
`9
`22
`6
`12
`12
`
`3
`1
`0
`0
`4
`8
`3
`6
`1
`3
`
`4
`
`1
`0
`0
`0
`0
`1
`1
`0
`0
`1
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`
`R.A. Millikan et al. / Urologic Oncology 6 (2001) 111–115
`
`115
`
`prednisone (median survival of 12.5 months vs. 7.8 months)
`[18]. Despite this, it is clear that a truly therapeutic thresh-
`old has not yet been crossed.
`In conclusion, we found no evidence that either of these
`regimens is likely to substantially improve the survival of
`patients with androgen independent prostate cancer. How-
`ever, the confirmation of a markedly better outcome in the
`setting of significant PSA declines, and many patients with
`symptom palliation, does confirm some biologic activity for
`this treatment, and suggests to us that further work on such
`regimens is reasonable. In our view, additional studies with
`novel combinations are more appropriate than taking either
`of these regimens on to phase 3 evaluation.
`
`Acknowledgments
`
`We thank Ms. Sherryl Smith for secretarial assistance.
`This work was supported in part by grant CA45809-10 from
`the National Cancer Institute, Bethesda, MD. REM is
`pleased to acknowledge support from the American Cancer
`Society in the form of a Career Development Award.
`
`References
`
`[1] Millikan RE. Chemotherapy of advanced prostatic carcinoma. Sem
`Oncol 1999;25:185–91.
`[2] Oh WK, Kantoff PW. Docetaxel (Taxotere)-based chemotherapy for
`hormone-refractory and locally advanced prostate cancer. Sem Oncol
`1999;26:49–54.
`[3] Jubelirer SJ, Hogan T. High dose ketoconazole for the treatment of
`hormone refractory metastatic prostate carcinoma: 16 cases and re-
`view of the literature. J Urol 1989;142:89–91.
`[4] Muscato JJ, Ahmann TA, Johnson KM, Wilding W, Monaghan G,
`Schlossman DM. Optimal dosing of ketoconazole (KETO) and hy-
`drocortisone (HC) leads to long responses in hormone refractory
`prostate cancer. Proc Annu Meet Am Soc Clin Oncol 1994;13:A701.
`[5] Small EJ, Baron AD, Fippin L, Apodaca D. Ketoconazole retains ac-
`tivity in advanced prostate cancer patients with progression despite
`flutamide withdrawal. J Urol 1997;157:1204–7.
`[6] Pavlick AC, Pecora AL, Scheuch J, et al. Treatment of hormone re-
`
`fractory prostate cancer with ketoconazole, hydrocortisone and cyclo-
`phosphamide. Proc Annu Meet Am Soc Clin Oncol 1996;15:A698.
`[7] Eichenberger T, Trachtenberg J, Toor P, Keating A. Ketoconazole. A
`possible direct cytotoxic effect on prostate carcinoma cells. J Urol
`1989;141:190–91.
`[8] Siegsmund MJ, Cardarelli C, Aksentijevich I, Sugimoto Y, Pastan I,
`Gottesman MM. Ketoconazole effectively reverses multidrug resis-
`tance in highly resistant KB cells. J Urol 1994;151:485–91.
`[9] Sella A, Kilbourn R, Amato R, et al. Phase II study of ketoconazole
`combined with weekly doxorubicin in patients with androgen-inde-
`pendent prostate cancer. J Clin Oncol 1994;12:683–88.
`[10] Ellerhorst JA, Tu SM, Amato RJ, et al. Phase II trial of alternating
`weekly chemohormonal therapy for patients with androgen-indepen-
`dent prostate cancer. Clin Cancer Res 1997;3:2371–76.
`[11] Maulard-Durdux C, Dufour B, Hennequin C, et al. Phase II study of the
`oral cyclophosphamide and oral etoposide combination in hormone-re-
`fractory prostate carcinoma patients. Cancer 1996;77:1144–48.
`[12] Simon R, Wittes RE, Ellenberg SS. Randomized phase II clinical tri-
`als. Cancer Treat Reports 1985;69:1375–81.
`[13] Daliani D, Eisenberg PD, Weems J, Lord R, Fueger R, Logothetis CJ.
`The results of a phase II randomized trial comparing 5-fluorouracil
`and 5-fluorouracil plus alpha-interferon. Observations on the design
`of clinical trials for androgen-independent prostate cancer. J Urol
`1995;153:1587–91.
`[14] Amato RJ, Logothetis CJ, Dexeus FH, Sella A, Kilbourn RG, Fitz K.
`Preliminary results of a phase II trial of estramustine (EMCYT) and
`vinblastine (VLB) for patients (pts) with progressive hormone refrac-
`tory prostrate carcinoma (HRPC). Proc Amer Assoc Cancer Res
`1991;32;A1111:186.
`[15] Logothetis CJ, Hoosein NM, Hsieh JT. The clinical and biological
`study of androgen independent prostate cancer (AI PCa). Sem Oncol
`1994;21:620–9.
`[16] Mahler C, Verhelst J, Denis L. Ketoconazole and liarozole in the
`treatment of advanced prostatic cancer. Cancer 1993;71:1068–73.
`[17] Storlie JA, Buckner JC, Wiseman GA, Burch PA, Hartmann LC, Ri-
`chardson RL. Prostrate specific antigen levels and clinical response to
`low dose dexamethasone for hormone-refractory metastatic prostate
`carcinoma. Cancer 1995;76:96–100.
`[18] Dowling AJ, Czaykowski PM, Krahn MD, Moore MJ, Tannock IF.
`Prostate specific antigen response to mitoxantrone and prednisone in
`patients with refractory prostate cancer. Prognostic factors and gener-
`alizability of a multicenter trial to clinical practice. J Urology 2000;
`163:1481–5.
`[19] Millikan RE, Logothetis C. Update of the NCCN guidelines for treat-
`ment of prostate cancer. Oncology 1997;11:180–93.
`
`