Translational proof of mechanism (PoM) for sorafenib with bevacizumab: Endpoint analysis and clinical activity

3574

Background: We hypothesized that molecular targeting of tumor and its microenvironment with rational combination therapy would yield improved outcome. Bevacizumab (B) is a monocolonal antibody to vascular endothelial growth factor (VEGFF) and sorafenib (S) is a small molecule inhibitor of the VEGF receptor-2 and RAF-kinase.We have previously reported that S 200mg bid with B 5mg/kg q2wk resulted in partial responses and prolonged disease stabilization (ovarian cancers, renal cell cancer, leiomyosarcoma). Serial tumor biopsies were obtained for translational proof of mechanism analysis. Methods: Percutaneous core biopsies of metastatic sites were obtained at baseline, after 2wk of S (group A) or B (group B), and at 6wks (2 wk into S+B combination therapy). Tissue was used for: lysate array (TLA) proteomic analysis of signaling proteins, IHC for CD31, VEGF, and Ki67, Raf/Ras mutation, and SNP analysis. Results: 18 sets of 3 biopsies were obtained, assessed for morphologic response (MR; decrease to <60% viable tumor seen), and processed for TLA and IHC. 11 pts had morphologic response including all 4 PR and 6/12 SD pts. Greater reduction in change in phospho(p)-ERK was seen at 2wk in group A pts (starting with sorafenib only) (p = 0.003). 9/11 MR pts had inhibition of ERK or AKT activation (p < 0.03). OvCa pts were more likely to have reduction in bRaf (p = 0.02). Pts in group B (starting with bevacizumab only) had lower microvessel density as measured by CD31 IHC (p = 0.05) at 2 wks; there was also a trend to lower VEGF expression by IHC in that group.These effects were muted at time 3, after combination treatment. No relationships to clinical outcomes were found, every pt had at least 1SNP in >1 tested gene; there were no Ras mutations and 1 V600KRaf mutation . Conclusions: These exploratory results demonstrate proof of mechanism in the tumor and its microenvironment for the combination of S and B. Randomization of initial exposure for the first 4 wks allowed interrogation of the role of the independent agents; the Raf/ERK axis was inhibited by S treatment first and there was reduction in tissue angiogenesis markers with B first. The molecularly targeted combination of sorafenib and bevacizumab is biologically active and on target. Phase II studies in multiple tumor types are ongoing.

No significant financial relationships to disclose.

Modulation of erlotinib pharmacokinetics in mice by a novel cytochrome P450 3A4 inhibitor, BAS 100

Administration of BAS 100, a novel mechanism-based CYP3A4 inhibitor isolated from grapefruit juice, resulted in a 2.1-fold increase in erlotinib exposure following oral administration to wild-type and humanised CYP3A4 transgenic mice. This study illustrates the potential of BAS 100 to increase the low and variable oral bioavailability of erlotinib in cancer patients.

Osteonecrosis of the jaw (ONJ) in androgen-independent prostate cancer (AIPC) patients receiving ATTP (bevacizumab, docetaxel, thalidomide, and prednisone)

19594

Background: ONJ has been associated with IV bisphosphonate use in cancer patients (pts) and chemotherapy may be an additional risk factor. ONJ is believed to result from localized vascular insufficiency due to faulty bone remodeling. ONJ incidence in pts not receiving chemotherapy (e.g. Paget's disease) is reportedly only 0.8%; prostate cancer pts have an incidence of 6.5%. Methods: We reviewed data from pts with advanced AIPC who developed ONJ while being treated on a Phase II study of ATTP. Results: Six of 36 (17%) pts treated with ATTP had ONJ confirmed by oral surgery. Four of the 6 pts presented with pain and, in five, the ONJ was mandibular in location. All pts were treated conservatively with either sequestrectomy or oral cleansing with chlorhexidine. All pts had been treated monthly with IV zoledronic acid (ZA). The mean duration of ZA use before diagnosis of ONJ was 20 months (mos). One patient had been treated with oral alendronate for 3 years and then developed ONJ after 5 mos of ZA. Of the 36 pts on-study, previous dental history could be verified in 24 pts. ONJ was diagnosed in 4 of the 5 pts with a prior dental infection or invasive dental procedure. Pts received an average of 11 cycles of ATTP before ONJ was diagnosed. All pts with ONJ had received full doses of bevacizumab and most had received full doses of all medications for all cycles. Conclusions: The possibility exists that the risk for ONJ may be higher with specific chemotherapy regimens, particularly those that include steroids or anti-angiogenic agents. However, at this time prior dental infections or procedures remain the greatest known risk factor. It is also possible that the relatively high incidence of ONJ in prostate cancer patients may reflect increased awareness in this population. Randomized phase III trials in AIPC, specifically addressing the incidence of ONJ, are needed to determine whether specific chemotherapy regimens are associated with an increased risk of this complication.

No significant financial relationships to disclose.

A phase II trial of thalidomide, bevacizumab, and docetaxel in patients (pts) with metastatic androgen-independent prostate cancer (AIPC)

5114

Background: Angiogenesis plays a vital role in the progression of prostate cancer. Antiangiogenic agents thalidomide (T) and bevacizumab (Bv) may enhance docetaxel (Doc) activity in metastatic AIPC. However, T and Bv have different antiangiogenic mechanisms. Since tumor angiogenesis is a complex interplay of multiple angiogenic factors, we reasoned that combination of mechanistically different antiangiogenic agents T and Bv with Doc might be associated with an adequately high and durable PSA response to merit further study. Methods: Pts are required to have progressive metastatic AIPC and no prior chemotherapy for AIPC. Treatment consists of Doc 75 mg/m2 plus Bv 15 mg/kg day 1, q 21 days as a cycle (C), plus T 200 mg qhs and prednisone 10 mg qd. Enoxaparin is used for thrombosis prevention and pegfilgrastim initiated after detection of grade ≥3 neutropenia. PSA is assayed q C and staging studies are done at C 0, C 2, & then q 3 Cs. Results: 39 pts were enrolled, median age 66 [54–79], Gleason score 8 [74% Gs 10∼8, 26% Gs 7∼6], on-study PSA 92 ng/ml [5.9–4399] and pre-study PSA doubling time 1.6 months [0.3–18.2, 87 % <3 months]. Median treatment Cs is 14 [1- 28]. 34 pts (87%) had PSA declines of ≥50%, with median ≥50% PSA-duration 12 Cs [0∼28]. 3 pts have PSA declines around 40% and 2 stable. 26 pts (67%) had >80% PSA declines. 17 pts with measurable disease were evaluable: 1 CR, 9 PR, & 7 SD, with 59% ORR. Significant toxicities: febrile neutropenia (5/39), syncope (4/39), colon perforation or fistula (2/39), grade 3 bleeding (2/39), thrombosis (2/39). Conclusions: This trial is the first study to combine antiangiogenic agents of different mechanisms with Doc in metastatic AIPC. Most of the accrued patients have unfavorable characteristics as evidenced by a high Gleason score and a short PSA doubling time. However, the combination of T and Bv with Doc appears to result in both a high durable PSA decline rate (87%) and a high response in measurable disease (59%) with acceptable toxicities.

No significant financial relationships to disclose.

A pharmacogenetic study of docetaxel and thalidomide in patients with androgen-independent prostate cancer (AIPC) using targeted human DMET genotyping platform

3580

Background: Pharmacogenetic research holds the promise of individualizing cancer therapy by reducing inter-individual variability in drug response, thus enhancing efficacy and reducing toxicity. Past research has been limited due to the lack of a robust genotyping platform that can screen for single nucleotide polymorphisms (SNPs) in the dozens of genes known to be involved in drug disposition. We pilot tested the new Affymetrix Targeted Human Drug Metabolizing Enzymes and Transporter (DMET) 1.0 panel in an exploratory study of docetaxel and thalidomide. The DMET 1.0 panel tests for 1,229 genetic variations in 169 drug disposition genes, including 49 CYP450 genes, 73 non-CYP genes, and 47 transporters. Methods: DNA samples from 47 patients with AIPC enrolled in a randomized phase II trial using docetaxel and thalidomide vs. docetaxel alone were genotyped using the DMET 1.0 panel. Patients’ response was determined using RECIST criteria. Toxicities were graded using the NCI-CTC, and patients were identified if they experienced grade 3 or 4 toxicity. Given the distinct side effect profiles of these two drugs, specific toxicities were assigned as being due to either docetaxel or thalidomide. An association between the SNP parameters and clinical response or toxicity was tested using Mehta’s modification to Fisher’s exact test. Reported results were limited to those where p<0.01. Results: Six SNPs in three genes were associated with response to therapy: PPAR-delta (p=0.0011), SULT1C2 (p=0.0083), and CHST3 (4 SNPs, p=0.0001 to 0.0034). For toxicities associated with docetaxel, five SNPs in three genes were identified: UGT1A1 (2 SNPs, p=0.0009 to 0.0094), UGT1A9 (2 SNPs, p=0.0016 to 0.0096), and CYP2A7 (p=0.0027). SNPs in CYP2B6 (p=0.0033), ABCC1 (p=0.0036), and ABCC6 (p=0.0075) were associated with toxicities from thalidomide. Conclusion: We identified nine genes in which SNPs were potentially significantly associated with clinical response and toxicity to treatment. These results highlight the important role that non-CYP450 and phase II drug metabolizing enzymes may play in the efficacy and disposition of docetaxel and thalidomide. Confirmatory studies are warranted.

No significant financial relationships to disclose.

Pharmacogenetics of organic anion transporting peptides (OATP): An association of a SLCO1B3 polymorphism with the overall survival in white Caucasian patients with androgen independent prostate cancer

5118

Background: Steroid hormone have been implicated in playing a fundamental role in pathogenesis of prostate cancer and several studies have reported an association between clinical outcome and polymorphism of genes involved in steroid metabolism. Recently we observed that polymorphic enzyme CYP17 which influences the rate-limiting steps in androgen biosynthesis was associated with overall survival. However, the effect of genetic variation on the transport of steroid hormones remains unknown. The purpose of this study was to evaluate the association between clinical outcomes in patients with androgen independent prostate cancer, and polymorphism in the SLCO1B3 gene that are potentially important in the differential uptake of steroid hormones. Methods: One hundred seventy nine white Caucasian patients with androgen independent prostate cancer were enrolled in this study. The duration of survival was computed from the date of prostate cancer diagnosis until the date of death or last follow-up. The SLCO1B3 polymorphism (-334 T>G) was analyzed using an automated sequencer. Results: The T allele (TT or TG) and G allele (GG) frequency of SLCO1B3 among all patients was 0.16 and 0.84, respectively. The frequency of the SLCO1B3 genotype was similar in patients and healthy volunteer. No significant differences were observed in the frequencies of the SLCO1B3 genotype in relation to categorized Gleason scores and age at diagnosis. Interestingly, the median survival was significantly longer in 134 patients with the SLCO1B3 G allele (8.4 years) genotype than 45 patients carrying at least one T allele (6.4 years) genotype (hazard ratio=1.58, 95%CI: 1.13 - 2.26, p=0.0105 by Logrank test). Similarly, the survival probability at ten years (22.9% in T allele vs 41.1% in G allele) is observed. Conclusions: Our results indicate that the SLCO1B3 polymorphism is associated with overall survival in white Caucasian patients with androgen independent prostate cancer. Large population-based studies are needed to confirm these results.

No significant financial relationships to disclose.

Variants in the SLCO1B3 Gene: Interethnic Distribution and Association with Paclitaxel Pharmacokinetics

To explore retrospectively the relationships between paclitaxel pharmacokinetics and three known, non-synonymous single-nucleotide polymorphisms (SNPs) in SLCO1B3, the gene encoding organic anion transporting polypeptide (OATP)1B3. Accumulation of [(3)H]paclitaxel was studied in Xenopus laevis oocytes injected with cRNA of Oatp1b2, OATP1A2, OATP1B1, OATP1B3, OAT1, OAT3, OCT1, and NTCP. The 334T>G (Ser112Ala), 699G>A (Met233Ile), and 1564G>T (Gly522Cys) loci of SLCO1B3 were screened in 475 individuals from five ethnic groups and 90 European Caucasian cancer patients treated with paclitaxel. Only OATP1B3 was capable of transporting paclitaxel to a significant extent (P=0.003). The 334T>G and 699G>A SNPs were less common in the African-American and Ghanaian populations (P<0.000001). Paclitaxel pharmacokinetics were not associated with the studied SNPs or haplotypes (P>0.3). The studied SNPs in SLCO1B3 appear to play a limited role in the disposition of paclitaxel, although their clinical significance in other ethnic populations remains to be investigated.

Bony metastatic disease responses to sorafenib (BAY 43–9006) independent of PSA in patients with metastatic androgen independent prostate cancer

4506

Background: Sorafenib is a novel bis-aryl urea, multi-kinase inhibitor, approved for the treatment of advanced renal cell carcinoma. It inhibits b- and c-Raf kinase, PDGFR, c-kit, VEGFR, Flt-3 and p38 and demonstrates anti-proliferative and anti-angiogenic activity. Studies have shown a role for anti-angiogenic therapy for androgen-independent prostate cancer (AIPC). Evidence suggests that the Ras-Raf-MAPK-ERK signaling pathway is dysregulated in AIPC and might be targeted by sorafenib. Methods: 22 patients (pts) with progressive metastatic AIPC enrolled in an open-label, single arm phase II study. The primary objective was to determine if sorafenib is associated with a 50% 4 month probability of progression free survival as determined by clinical, radiographic, and PSA criteria. Sorafenib was given continuously at a dose of 400 mg orally twice daily in 28-day cycles. Clinical assessment and PSA measurement occurred every cycle with radiographic measurements every 2 cycles. Results: Baseline patient characteristics included a median (range) age of 64 (51–78), Gleason 8 (6–9), PSA 226.4 ug/l (2–1905), Alk Phos 108 u/l (48–259), Hb 12.7 g/dl (10.2–15.1). 60% of patients had received one prior chemotherapy regimen. Of the 19 pts with progressive disease, 10 progressed only by PSA criteria in the absence of evidence of clinical and radiographic progression. Two pts were found to have dramatic disappearance of bone metastatic lesions as demonstrated by bone scan, even though they met PSA progression criteria at the time the scans were obtained. Toxicities likely related to treatment include: one grade 3 hypertension and hand-foot syndrome; grade 1/2 toxicities: fatigue, anorexia, hypertension, skin rash, nausea, and diarrhea. Conclusion: Sorafenib in AIPC is relatively well tolerated with 2 patients demonstrating evidence of improved bony metastatic lesions. Interpretation of this study is complicated by discordant radiographic and PSA responses. PSA may not be an adequate biomarker for monitoring sorafenib activity. Further study of sorafenib in metastatic AIPC using clinical and radiographic endpoints is warranted.

No significant financial relationships to disclose.

Interim analysis of phase I trial of 17-DMAG

13148

Background: 17-Dimethylaminoethylamino-17-Demethoxygeldanamycin (17-DMAG) binds specifically to heat-shock protein 90, leading to degradation of oncogene client proteins, inhibiting tumor growth. Based on in vitro/in vivo data, 17-DMAG has potent antitumor activity. Methods: 17-DMAG was administered by 1–2 hour infusion twice weekly for 4 weeks/cycle. Eligibility criteria included: solid tumor or lymphoma, ECOG ≤ 2 and adequate organ function. Exclusion criteria: brain metastases, pregnancy, prolonged QTc, or uncontrolled illness. Accelerated titration escalation design, one patient/cohort enters until a single patient experiences dose-limiting toxicity (DLT) or two patients exhibit grade ≥ 2 (G2) toxicity (except nausea / vomiting) during the first cycle. The maximum tolerable dose (MTD) is the dose level at which no more than 1/6 patients experience DLT and the dose below that at which at least 2/6 patients have DLT. Primary objectives: to determine the toxicity profile, and MTD of twice weekly 17-DMAG. Secondary endpoints: pharmacokinetics (PK) and pharmacodynamics of 17-DMAG. PK were determined by LC-MS on blood samples collected before and serially up to 48 hours after 17-DMAG infusion. Results: 6 patients (one patient per dose level) with the following diagnoses were accrued and treated: renal cell, pancreatic, medullary thyroid, NSCLC, and two colorectal cancers. The DLT and MTD have not yet been achieved. 17-DMAG toxicities have been acceptable, nausea/vomiting being the most common toxicity observed. Three patients developed G1 nausea, one patient had G3 nausea. Two patients had G2 vomiting. Other G1 toxicities included: diarrhea, weight loss, fatigue, elevated transaminases, hair loss, and leukocytopenia. PK results showed significant intra-patient variability as the dose was increased from 1–12 mg/m2 over 6 dose levels; the mean AUC increased from 370–970 ng*hr/mL, and the mean Cmax increased from 36–79 ng/mL. There was a 5-fold range in clearance (range: 6.5–33.1 L/h), but there was no association between dose and clearance. Following a single dose, the terminal phase half-life ranged from 16.4 - 24.3 hours. Conclusions: Since the DLT and MTD have not been reached, accrual will continue.

No significant financial relationships to disclose.

Phase I study of lenalidomide, a novel thalidomide analog, in patients with refractory metastatic cancer

13038

Background: Lenalidomide (L), CC5013, is a thalidomide (T) analog that has shown activity in multiple myeloma and myelodysplastic syndromes (MDS); including activity in pts refractory to T. Preclinical studies suggest L has more potent immunomodulatory and antiangiogenic activity than T. We have conducted a phase I clinical trial of L to determine the maximum-tolerated dose (MTD), characterize the side-effects, characterize pharmacokinetics (PK) in pts with solid tumors. Methods: Pts with refractory metastatic cancer were treated on a modified Fibonacci dose-escalation scheme with an oral daily dose of L. Therapy was continued until MTD or disease progression. Results: Forty-five pts have been enrolled (M/F: 37/8). Median age 68 yrs (range, 25–89 yrs). Tumor types include prostate cancer (36 pts), adrenal carcinoma (3 pt), renal cell carcinoma (1 pt), choleangiocarcinoma (1 pt), small intestine cancer (1 pt), colon cancer (2 pt), and melanoma (1 pt). Dose levels were: 5 mg (n = 3), 10 mg (n = 6), 15 mg (n = 3), 20 mg (n = 3 + 7), 25 mg (n = 6), 30 mg (n = 8), 35 mg (n = 6) and 40 mg (n = 3). Due to side effects the dosing schedule was modified from daily dosing to daily dosing for 21 out of 28 days. Therapy has been well tolerated with most frequent gr 1 and gr 2 toxicities include nausea (43%), myalgia (38%), pruritis/rash (54%), and fatigue (38%), neutropenia (30%) with gr 3 diarrhea (n = 1), hypotension (n = 1), pulmonary effusions (n = 1), thromboses (n = 2) and gr 4 neutropenia (n = 2), arrhythmias (n = 1), hemolysis (n = 1). Dose escalation continues. There were no differences observed between dose levels for either oral clearance values (p = 0.47) or the apparent volume of distribution (Vz) values (p = 0.23). Dose-normalized AUC(0-∞), dose-normalized peak concentrations (Cmax), or half-life (t1/2) did not vary between dose levels (p > 0.12). Conclusion: L has been well tolerated up to 35 mg/d (21 days on; 7 days off), doses higher than the recently FDA-approved doses of 5 and 10 mg/d. In this dose range L exhibits linear PK. Tolerability to myelosuppression the dose-limiting toxicity in pts with MDS may therefore be improved in pts without hematological malignancies.

No significant financial relationships to disclose.

A phase II trial of docetaxel, thalidomide, bevacizumab, and prednisone in patients (pts) with metastatic androgen-independent prostate cancer (AIPC)

13037

Background: Angiogenesis plays a vital role in progression of prostate cancer. Antiangiogenic agents’ thalidomide (Td) and bevacizumab (Bv) have been shown in Phase II trials to enhance docetaxel (Doc) activity in metastatic AIPC. Td plus Doc also improved median overall survival as compared to Doc alone. Td and Bv have different antiangiogenic mechanisms. Td appears to affect bFGF, alter circulating endothelial cells and inhibit TNF expression; yet Td does not affect the target of Bv (VEGF). Since tumor angiogenesis is a complex interplay of multiple angiogenic factors, we reasoned that combination of mechanistically different antiangiogenic agents Td and Bv with Doc might be associated with an adequately high and durable PSA response to merit further study. Methods: Pts have progressive, metastatic AIPC, with no prior chemotherapy for AIPC, or prior Td or Bv. Treatment consists of Doc 75 mg/m2 plus Bv 15 mg/kg day 1, q 21 days as a cycle (C), plus Td 200 mg qhs and prednisone10 mg qd. Enoxaparin 1 mg/kg/d sq is used for thrombosis prevention. PSA is assayed q C and radiographic studies are performed at C 0, C 2, & then q 3 Cs. Results: To date, 22 of planned 60 pts have been enrolled, median age 65 [54–79], Gleason score 8 [Gs 6∼7: 32%, Gs 8∼10: 68%], on-study PSA 111 ng/ml [7.7–4399], & pre-treatment PSA doubling time 1.7 months [0.8–18.2]. I) All 22 pts are actively on trial without progression, with median treatment 6 Cs [1–12]. II) 20 pts treated ≥ 2 Cs, 17 (85%) had PSA declines of >50%, 2 declines of >20%, and 1 stable; median duration of >50% PSA decline is 5 Cs [0∼11]. III) 14 pts (of the 20) treated ≥ 5 Cs, 13 (93%) had >50% PSA declines; 6 of them also had measurable disease: 1 CR, 2 PR, & 3 SD (50% RR). IV) Significant toxicities include grade 4 neutropenia (11/22), grade 3 febrile neutropenia (3/22), syncope (2/22), & colon perforation (1/22). No thrombosis was seen. Conclusions: This trial tests a novel strategy of improving conventional treatment of metastatic AIPC. Our data show that the combination of two mechanistically different antiangiogenic agents Td and Bv with Doc does result in a high durable response in PSA (85∼93%) with acceptable toxicities, suggesting that further studies with the combination are warranted. Accrual is continuing.

No significant financial relationships to disclose.