Originally published June 7.
By Turna Ray
CHICAGO — Pfizer presented new data on several investigational targeted drugs at the American Society of Clinical Oncology's annual meeting here this week, making good on its previously expressed desire to increasingly apply pharmacogenomic strategies to its drug development process.
The focus for Pfizer at the meeting this year is personalized medicine, said Mace Rothenberg, Pfizer Oncology's senior VP of clinical development and medical affairs, at a briefing this week. According to the drug giant, the abstracts presented this year display the company's “focused approach to cancer drug development through the identification and validation of molecular targets.”
In particular, non-small cell lung cancer is a big focus for Pfizer at ASCO, where the company presented several abstracts from ongoing studies on two personalized NSCLC drugs: the ALK-inhibitor crizotinib (PF-02341066) and the pan-HER inhibitor PF-00299804.
The briefing, which Pfizer held to discuss crizotinib and its overall personalized medicine strategy in oncology, featured company officials alongside six academic collaborators. They all agreed that as drug development moves from a one-size-fits-all strategy to increasing application of pharmacogenomic strategies, industry, academia, and the government will need to collaborate.
Pfizer has embraced a collaborative approach to drug development. For example, through the formation of the Asian Cancer Research Group earlier this year, Lilly, Pfizer, and Merck agreed to share pre-competitive genomics data to advance research on lung and gastric cancers (PGx Reporter 2/24/2010).
Several researchers assisting Pfizer with the development of crizotinib were involved in the Tumor Sequencing Project Consortium, a project to sequence the exonic regions of 1,000 genes in 200 specimens of adenocarcinoma of the lung. The consortium, a collaboration among multiple academic centers, revealed molecular targets in lung cancer that many drug firms, including Pfizer, have applied to current drug development.
Paul Bunn, professor of medicine at the University of Colorado who worked with Pfizer in studying crizotinib, added that plans are underway to develop a lung cancer mutation consortium that will develop a panel of 10 genes linked to the disease. Patients can then go to a CLIA lab that is using this panel, and can get have their lung tumors tested for stratification into ongoing clinical trials. “This represents an interesting way of proceeding with clinical trials,” Bunn said.
Collaboration and advancement of genomic tools are closing the "translational gap" between discovery research and commercialization, Pfizer's Rothenberg suggested. He noted that it took 40 years from the discovery of the defective BCR-ABL protein and the development of Novartis' Gleevec as a treatment for chronic myelogenous leukemia and gastrointestinal stromal tumors.
In the case of crizotinib, the EML4-AKL fusion oncogene was discovered in December 2007, and the drug is now in late-stage clinical trials. “So, hopefully, we're closing that [translational] gap,” Rothenberg said.
In addition to crizotinib and PF-00299804, Pfizer also released pharmacogenomic data at ASCO on bosutinib, a dual SRC and BCR-ABL kinase inhibitor in chronic myelogenous leukemia; Sutent in renal cancer; the glioblastoma treatment PF-04948568, and the previously discontinued melanoma drug tremelimumab.
The studies discussed below were all financially supported by Pfizer. In addition to the involvement of academic researchers, Pfizer also provided research leadership in many of the studies.
NSCLC: Crizotinib, PF-00299804
At ASCO, Pfizer presented data from two preliminary studies that provided justification for advancing investigational agents in late-stage trials in molecularly defined patient populations. These include a Phase I/II study evaluating crizotinib in patients carrying the EML4-ALK-fusion oncogene, and a Phase II trial of the oral pan-HER inhibitor PF-00299804 in patients with certain EGFR mutations.
[ pagebreak ]
In the early-stage study looking at crizotinib, researchers recruited patients with ALK fusions as part of an expanded cohort in a Phase II, first-in-patient monotherapy trial of the drug. Patients with ALK fusions were determined by FISH testing. In the study to date, 76 ALK-positive, previously treated NSCLC patients have been treated in the trial with crizotinib, and 50 patients are evaluable for response.
Interim results show the overall response rate to be 64 percent and the disease control rate at 90 percent. Although the median progression-free survival, or PFS, data is not yet mature, “this study supports the concept of molecular selection of NSCLC patients for appropriately designed treatment,” the researchers added.
In the study, patients on crizotinib experienced gastrointestinal toxicities, including nausea (55 percent) and vomiting (39 percent), most frequently. “The oral ALK inhibitor [crizotinib] demonstrated a high response rate in patients selected for ALK fusions and was associated with a good safety profile," the researchers reported, adding that based on these results a phase III study for crizotinib has been initiated.
Pfizer estimates that ALK gene alterations are considered a key driver of lung tumorigenesis in around 45,000 newly diagnosed NSCLC patients annually. According to estimates, approximately 3 percent of NSCLC patients have ALK fusions.
Since the occurrence of this mutation in the lung cancer population is low, Pfizer needed to have a hypothesis and a companion test before proceeding onto clinical trials. “You would have never seen this activity without a test,” Bruce Johnson, director of the Lung Cancer Program at the Dana-Farber Cancer Institute, said during the briefing.
Abbott Molecular developed the FISH test that is being used in clinical trials for crizotinib, but Pfizer and its collaborators are also exploring the use of RT-PCR and immunohistochemistry approaches to detecting ALK fusions (PGx Reporter 09/02/09). Yung-Jue Bang, a professor at the Seoul National University College of Medicine and a Pfizer collaborator, said at the briefing that the FISH test had nearly 100 percent concordance with IHC tests conducted in his laboratory.
At the briefing on crizotinib, Pfizer and academic researchers were enthusiastic about these interim results. David Johnson, director of hematology and medical oncology at Vanderbilt University Medical School, put the results in the context of Herceptin's efficacy in HER2-positive patients.
HER2 is overexpressed in one-in-five women with breast cancer in the US, who usually have a response rate of between 30 percent and 40 percent. “We're talking about a doubling of that” with regard to crizotinib, Johnson said. No cost-effectiveness analysis has been done with regard to crizotinib as survival data is not yet available, but researchers believe that the drug, administered to patients most likely to benefit with the help of a companion test, will be cost effective in terms of cost per life gained.
In a Phase II trial involving the investigational pan-HER inhibitor PF-00299804, researchers enrolled 39 patients with EGFR-mutated NSCLC and treated them with the investigational drug. Researchers are planning to ultimately enroll around 80 patients, including 30 Asian and 30 non-Asian participants.
Endpoints in this study include: progression-free survival at four months, tumor response, safety, serial pharmacokinetics, and tissue- and blood-based EGFR biomarkers, including T790M mutations, which has shown to cause secondary resistance in 50 percent of NSCLC patients after they see some benefit from EGFR-tyrosine kinase inhibitors.
The researchers reported that 29 patients out of 39 have been evaluated so far. Of these, one patient has experienced a complete response, six have had partial responses, and 16 have stable disease after treatment for six weeks or more. Preliminary PFS rates at three, four, and six months were 90 percent, 79 percent, and 79 percent, respectively.
Mutation status was obtained in 31 out of 39 patients. “All evaluable patients with known EGFR mutant NSCLC [14 patients] showed tumor shrinkage,” the researchers reported in the abstract.
The most commonly associated adverse reactions were diarrhea (79 percent), dermatitis acneiform (49 percent), stomatitis (42 percent), acne (24 percent), rash (21 percent), and anorexia (18 percent). One patient discontinued the trial due to adverse reactions and 10 participants required dose reductions while on the study.
These early results suggest that “PF-00299804 demonstrates encouraging efficacy as first-line treatment of selected Asian and non-Asian patients with NSCLC, as reflected by PFS rates,” the researchers said. “Of note, tumor shrinkage was seen in all evaluable patients with EGFR mutant disease."
Although most of the adverse events were not severe and were manageable, “to ensure long-term tolerability, a lower starting dose is being considered,” the researchers concluded.
At ASCO, Pfizer presented a total of six abstracts on PF-00299804, including one comparing the investigational agent to Genentech/OSI's Tarceva (erlotinib). At the 12-week median follow-up, PF299804 showed significantly longer progression-free survival in patients in the overall population compared to Tarceva, according to the abstract. This trend was consistent across several subgroups, including EGFR wild-type patients. Common adverse events, such as diarrhea and acne, were higher in the PF-00299804 arm.
PF-00299804 will be studied in a double-blind, placebo-controlled, randomized Phase III trial in NSCLC patients previously treated with standard chemotherapy and EGFR inhibitor treatment.
Renal Cancer: Sutent
Genomic Health announced in 2008 that it was developing a companion test to personalize treatment with Pfizer's Sutent in renal cancer. The test will be based on Genomic Health's Oncotype DX platform (PGx Reporter 01/09/08). In a presentation at ASCO this week, researchers from Pfizer, Cleveland Clinic, and Genomic Health discussed data from an exploratory study to identify pharmacogenomic markers to predict disease recurrence in patients with clear-cell renal cancer.
[ pagebreak ]
In the retrospective investigation, researchers identified archived paraffin-embedded samples from patients with stage I-III clear-cell renal cell carcinoma who underwent nephrectomy at Cleveland Clinic between 1985 and 2003, and re-reviewed and verified these patients' pathology and RCC recurrence status. Then, researchers extracted RNA from dissected tumor sections and analyzed expression of RNA using RT-PCR.
Researchers had complete clinical pathology data and tissue blocks from 931 patients, the majority of whom were male with a median age of 61 with stage I/II/III renal cancer. The primary endpoint was recurrence-free interval, or RFI, defined as the time from nephrectomy to first recurrence or death due to RCC.
Researchers identified 732 genes of interest, including five reference genes. Of these, 448 genes were significantly associated with RFI and 300 genes were significantly associated with at least four of five covariates; but these findings were not adjusted for false discovery.
After adjusting for clinical and pathological covariates and false discovery, 16 genes remained “significantly and strongly associated with RFI” in a multivariate model.
“Among these 16 genes, increased expression was associated with lower risk of recurrence for angiogenesis-related (including EMCN and NOS3) and immune-related (including CCL5 and CXCL9) genes,” abstract states.
Since the genomic factors that were identified seemed to “strongly associate” with risk of recurrence and clinical/pathological covariates, they “can be used to create a multi-gene algorithm to predict recurrence of ccRCC,” the researchers concluded.
Brian Rini, Associate Professor of Medicine at the Cleveland Clinic and lead author of the trial, discussed the nature of approximately 72 “potentially significant genes” identified in the exploratory study.
According to Rini, the most predominant gene group that was discovered was angiogenesis. “Increased expression of angiogenesis was associated with a lower risk of recurrence,” Rini said.
Other identified groups include immune-response genes, extracellular matrix disallocation genes, and cell-cycle genes. For some immune-response genes, such as IL-8 and IL-6, "increased expression was associated with worse outcome, in terms of higher risk of recurrence, whereas other immune-response genes were associated with lower risk of recurrence,” he said. Cell-adhesion and cell cycle genes tended to be associated with a higher risk of renal cancer recurrence, Rini added.
According to researchers, the project is the largest genomic study of localized clear-cell renal cell carcinoma ever performed.
“The magnitude of hazard ratios [associated with identified genes in renal cancer] are comparable to those of other clinically used prognostic markers, such as ER and HER2 in breast cancer,” the abstract notes.
Next steps for this data will be to create a multi-gene algorithm to predict recurrence of renal cancer after nephrectomy. Pfizer is planning to validate these markers in two different cohorts and one of these trials is currently ongoing. “We hope that these [markers] will move us toward a more molecular stratification [in treating] renal cancer that moves beyond histological, pathologic, and clinical stratification, into low, intermediate, and high risk of recurrence, based on these molecular features,” Rini said.
Finally, the company aims to develop a predictive test to identify patients with clear-cell renal cell cancer who are most likely to benefit from treatment with Sutent.
Although genomically targeted drugs may demand premium pricing, Pfizer officials and researchers asserted that the overall cost of care to the patient and the healthcare system will be lowered by reducing drug toxicities and improved efficacy of such treatments. “The vast majority of cancer patients who get treatment today see no benefit,” said Bunn. “So, it's better for [payors] if patients respond to the drugs they're taking.”
In an abstract presented at the meeting, Pfizer presented data on the efficacy of tremelimumab, a drug previously discontinued in late-stage studies due to low efficacy. The abstract presented this week showed it to have benefit in patients with low c-reactive protein (CRP) levels.
[ pagebreak ]
According to the abstract, researchers conducted exploratory analysis of a randomized Phase III study, A3671009, comparing the anti-CTLA4 monoclonal antibody tremelimumab to dacarbazine or temozolomide in patients with unresectable melanoma. In this retrospective analysis, researchers evaluated whether baseline serum CRP levels could serve as a predictive marker for treatment benefit with tremelimumab.
Earlier attempts to revive the drug — which Pfizer had stopped developing in 2008 — using pharmacogenomic markers proved unsuccessful. In the Feb. 1 issue of Clinical Cancer Research, researchers from Pfizer and other research institutions published results from a Phase II study on patients treated with tremelimumab, which included pharmacogenetic assessment of polymorphisms in the CTLA4, FcγRIIa, and IgG2 genes. However, no clinically meaningful associations emerged from this investigation (PGx Reporter 04/28/10).
In the latest ASCO abstract, researchers seem to have identified a protein marker that would allow targeted administration of the drug in patients most likely to benefit.
In the study, those patients identified as having low baseline CRP, around 326 out of 525, had an overall survival of 19.1 months for the tremelimumab arm versus 12.7 months for the chemotherapy arm (p = 0.0037). Patients with high baseline CRP had a 0.86 hazard ratio for survival, favoring the chemotherapy arm, “indicating that this group of patients did not apparently benefit from tremelimumab compared to chemotherapy,” the abstract notes.
Similarly, the investigation found that the tumor response rate to tremelimumab but not to chemotherapy was higher for patients with low CRP. Rates of diarrhea and serious adverse events were similar between the two groups based on CRP levels. “The results were robust for alternative definitions of baseline CRP,” the researcher reported, concluding that “low baseline CRP identifies a subset of patients with significant survival benefit from tremelimumab compared to chemotherapy.”
Based on these results, the researchers recommended that Pfizer conduct a prospective clinical trial of tremelimumab compared to chemotherapy in patients with low serum CRP.
Pfizer announced earlier this year that it would work with Qiagen subsidiary DxS to develop a companion test for PF-04948568 (PGx Reporter 02/10/10), a vaccine comprising 13 amino acids unique to EGFRvIII for the treatment of glioblastoma.
Researchers from Pfizer, Celldex Therapeutics, and several academic centers reported data from a multi-center, randomized, open-label Phase IIb/III study that enrolled patients with resected, newly diagnosed, EGFRvIII+ve glioblastoma after they had radiation therapy and were treated with temozolomide.
PF-04948568 was administered every other week three times prior to starting maintenance treatment with temozolomide until treatment progressed. The primary endpoint of the study was the progression-free rate at 5.5 months from first vaccination.
Researchers enrolled 81 patients, with 65 receiving PF-04948568. The trial had to be amended to a single-arm design after 14 out of 16 patients randomized to the standard care alone arm refused to participate after being informed of their randomization status. As such, researchers reported in the abstract on the interim efficacy data on the first 40 patients treated with PF-04948568.
In this interim analysis, 28 out of 40 patients were alive and progression-free at 5.5 months. Local injection site reactions were the most common treatment-related adverse events and were reported in the majority of patients. Reversible hypersensitivity reactions (all grades) occurred in 5 percent of patients, and one patient reported a serious reaction and discontinued treatment with PF-04948568.
[ pagebreak ]
“Vaccination with PF-04948568 was well tolerated in combination with maintenance temozolomide in patients with newly diagnosed, fully resected glioblastoma multiforme,” the researchers concluded. The progression-free rate of 70 percent at 5.5 months reported from this multicenter experience is similar to that reported in a previous study conducted by Duke University and MD Anderson Cancer Center, the researchers added.
“Difficulties retaining control-arm patients in this open label US trial suggest that additional study of the vaccine would require a placebo-controlled trial,” the abstract notes.
Chronic Myeloid Leukemia: Bosutinib
Researchers from Pfizer, MD Anderson, and institutes in Korea, Russia, and Italy reported results from a Phase I/II study investigating the efficacy and safety of bosutinib in patients with chronic phase chronic myeloid leukemia who failed treatment with Gleevec (imatinib).
Bosutinib is an orally bioavailable dual SRC/ABL tyrosine kinase inhibitor, with some inhibitory activity against the oncogene platelet-derived growth factor receptor, or c-kit. The abstract reports preliminary data on 299 patients, 72 percent of whom were imatinib-resistant, 28 percent were imatinib-intolerant. Patients also had prior therapy with agents other than imatinib, including interferon and stem cell transplant.
The most common adverse events experienced by bosutinib-treated patients were nausea, vomiting, and diarrhea that was usually grade 1-2 and manageable. Adverse events improved spontaneously after three to four weeks. Grade 3-4 nonhematologic toxicities were rash (9 percent) and diarrhea (8 percent). The most common grade 3-4 hematologic abnormalities were: thrombocytopenia (23 percent), neutropenia (14 percent), and anemia (9 percent).
Researchers also reported other common grade 3-4 laboratory abnormalities, such as hypermagnesemia (11 percent) and increased alanine aminotransferase (10 percent). In the study, 56 (19 percent) patients discontinued due to toxicity.
Of 132 patients evaluable for hematological response, 78 percent had a complete response; 58 percent of evaluable patients achieved a major cytogenetic response, 46 percent of which were complete.
Among evaluable patients, 156 were analyzed for a molecular response. Of these, 76 (49 percent) achieved a major molecular response, and 47 (30 percent) of these had a complete molecular response.
Researchers identified 20 different mutations at baseline in 45 out of 99 patients tested. Complete response occurred in 78 percent of patients with mutations and in 89 percent of patients with no mutations. Major cytogenetic response occurred in 60 percent of patients with mutations and 54 percent of patients with no mutations.
Based on these results, researchers concluded that “bosutinib is an active agent for patients with CP CML who failed prior imatinib, including patients with a wide variety of BCR-ABL mutations.” Additionally, researchers noted that the drug “demonstrated a favorable toxicity profile with minimal hematologic toxicity.”