The annual meeting of the American Society of Clinical Oncology held earlier this month in Chicago featured a number of pharmacogenomically guided studies, as well as updates from groups developing genomic tests for cancer diagnosis or prognosis. The following is a roundup of some of the meeting's molecular diagnostics and pharmacogenomics highlights.
Tamoxifen and risk of contralateral breast cancer for BRCA1 and BRCA2 mutation carriers: A combined analysis from the Kathleen Cuningham Foundation Consortium for Research into Familial Breast Cancer (kConFab), the International BRCA1 and BRCA2 Carrier Cohort Study (IBCCS), and the Breast Cancer Family Registry (BCFR).
Researchers from the Australian New Zealand Breast Cancer Trials Group and elsewhere investigated whether tamoxifen is effective as a breast cancer prevention strategy for BRCA1/2 mutation carriers. The researchers looked at women with BRCA1/2 mutations and a history of breast cancer since 1970 who enrolled in a hereditary breast cancer study, and enrolled 1,642 women with BRCA1 and 919 women with BRCA2 mutations. Out of these women, 23 percent of BRCA1 mutation carriers and 48 percent of BRCA2 mutation carriers had taken tamoxifen upon initially being diagnosed with breast cancer.
"During 21,344 person-years of follow-up, 596 contralateral breast cancer [cases] were observed," the researchers, led by Kelly-Anne Phillips of the Peter MacCallum Cancer Centre, reported in the abstract. The overall adjusted hazard ratio estimates were 0.31 (95% CI: 0.22-0.45) and 0.24 (95% CI 0.16-0.35) for BRCA1 and BRCA2 mutation carriers, respectively, Phillips et al. found. Meanwhile, the adjusted hazard ratio estimates for 629 BRCA1 and 412 BRCA2 mutation carriers were 0.52 (95% CI: 0.26-1.04) and 0.39 (95% CI: 0.17-0.89), with 4,869 person-years of follow-up.
"Although biased estimates due to non-random use of tamoxifen cannot be excluded, these results are consistent with tamoxifen reducing [breast cancer] risk for both BRCA1 and BRCA2 mutation carriers," the researchers concluded.
Single nucleotide polymorphisms and expression/amplification of HER2 in patients with breast cancer.
A team of UK researchers investigated the interplay of two SNPs in the HER2 gene in amplification or overexpression of HER2 in breast tumors. The two SNPs of interest in this study, as described in the abstract, are Ile655Val, which is "related to a residue in the trans-membrane domain," and Ala1170Pro, which "induces an amino acid change in the intra-cellular carboxyl-terminal tail."
In the study, Alan Boddy of Newcastle University and colleagues enrolled 293 breast cancer patients with known HER2 status from an outpatient clinic, and analyzed their tumors with FISH or IHC tests. Then, they genotyped these women for the two SNPs using a "fluorogenic 5' nuclease assay using the endogenous 5' nuclease activity of AmpliTaq Gold DNA polymerase."
Boddy et al. reported that out of almost 300 enrolled patients, 47 percent had tumors that were HER2 positive, and 53 percent were HER2 negative. The Ile655Val and Ala1170Pro SNPs were found in 21 percent and 29 percent of patients, respectively.
"There was no correlation between Ile655Val genotype and HER2 status (proportions of HER2-positive tumors among Ile-homozygous patients and carriers of the Val allele were respectively 46 percent and 48 percent, p=0.705)," researchers report in the abstract. "However, when the Ala1170Pro SNP was considered, the proportion of HER2-positive tumors was significantly higher among carriers of the Pro allele than it was among Ala homozygous patients (56 percent vs 38 percent, p=0.002)."
Based on this preliminary analysis, Boddy et al. found that Ala1170Pro may play a role in the development of HER2-positive breast tumors. The researchers said that these findings should be validated in a different cohort, and recommended that an evaluation of allele-specific amplification in tumor samples be conducted.
[ pagebreak ]
Exomic sequencing of pancreatic neuroendocrine tumors: Detection of alterations in chromatin remodeling and mTOR pathway genes.
Luis Diaz of Johns Hopkins University and researchers from other institutions explored the genetic basis for pancreatic neuroendocrine tumors, or PanNETs, a rare form of pancreatic neoplasia. In this study, the researchers attempted to gauge the exomic sequences of ten non-familial PanNETs and screened for commonly mutated genes in 58 additional PanNETs.
The study found that 44 percent of the tumors had somatic inactivating mutations in MEN-1, which encodes menin, and 43 percent had mutations in genes encoding either of the two subunits of a transcription/chromatin remodeling complex consisting of DAXX and ATRX. In multivariate analysis, "mutations in the MEN1 and DAXX/ATRX genes, alone or in combination, were associated with better prognosis," the abstract noted.
However, tumors without mutations in MEN1 or DAXX/ATRX occurred primarily in patients with PanNETs localized to the pancreas. Additionally, in 14 percent of the tumors, the researchers identified mutations in genes in the mTOR pathway, such as PIK3CA, PTEN, and TSC2.
"Taken together, these findings could potentially be used to stratify patients with PanNETs and help determine which patients may benefit from treatment with mTOR inhibitors," the researchers concluded.
RTOG 0525: Molecular correlates from a randomized phase III trial of newly diagnosed glioblastoma.
Kenneth Aldape from MD Anderson Cancer Center, along with researchers from other institutions, evaluated four prognostic indicators— the IDH1 mutation, the glioma-CpG island methylator phenotype (G-CIMP), a microarray-based mRNA panel, and an MGMT promoter methylation assay from MDxHealth — on a training set of 220 glioblastoma tumor samples. There were 36 possible combinations of each of the four biomarker-derived subgroups and researchers compared the subgroup analysis with survival data to delineate the samples into four risk groups.
Researchers then validated their findings from this training set on 763 samples from RTOG 0525, a randomized Phase III trial comparing standard adjuvant temozolomide with a dose-dense schedule of the same drug in 1,153 newly diagnosed glioblastoma patients.
Molecular risk classification of RTOG 0525 samples, obtained at the time of biopsy or surgery, showed a "highly significant" survival association. Compared to the recursive partitioning analysis, "this composite molecular classifier better identified patients with long-term survival and appears to improve resolution by revealing an additional distinct risk group," the researchers noted.
"Four distinct biomarkers or biomarker panels were tested in GBM. These biomarkers were compared with clinical outcome in a training set to optimize a method to combine them into a classifier," Aldape et al. said in the abstract. Based on validation data from the RTOG 0525 samples, "this composite panel may represent an improvement over the existing recursive partitioning analysis with respect to risk stratification of patients for GBM." The researchers believe that this finding could potentially impact future clinical trial designs and provide enhanced opportunities for personalization of therapy for glioblastoma.
Several of the study authors have disclosed receiving research funding or holding an advisory role with industry. MDxHealth provided research funding to several researchers involved in the study.
Validation of a 12-gene colon cancer recurrence score (RS) in patients (pts) with stage II colon cancer (CC) from CALGB 9581.
Researchers from Duke University and elsewhere conducted a validation study in tumor specimens from patients enrolled in the clinical trial CALGB 9581 to show that Genomic Health's 12-gene colon cancer risk score can predict the risk of recurrence for patients with stage II disease, as reported in the QUASAR trial.
The researchers collected samples from patients in CALGB 9581 with available tissue and recurrence, as well as from patients without disease recurrence. Gene expression of these samples was analyzed by Genomic Health's RT-PCR platform using FFPE tissue from patients' primary tumors. Then, study authors assessed mismatch repair protein status (either deficient or intact) by IHC for MLH1 and MSH2.
The primary aim of the study was to gauge the "prognostic value of continuous recurrence score alone and in presence of mismatch repair and traditional clinical/pathologic prognostic variables," the researches, led by Duke's Alan Paul Venook, reported in the abstract. "A weighted Cox proportional hazards model was used to test the association between recurrence score and recurrence-free interval based on a Wald-type test statistic constructed using a weighted partial likelihood estimate and robust variance estimate."
RT-PCR analysis was successful in 690 of 736 patients. The continuous recurrence score was "significantly associated" with recurrence-free interval in univariate analysis. Mismatch repair-deficiency was also associated with recurrence-free interval. "Recurrence risk at five years increased as RS increased and among subgroups defined by T-stage and MMR," Venook et al. reported in the abstract. In patients with T3, mismatch repair intact tumors pre-specified in low, intermediate, and high recurrence score groups had average five-year cancer recurrence risks of 13 percent, 16 percent, and 21 percent, respectively.
The data from CALGB 9581 showed that Genomic Health's recurrence score "improves the ability to discriminate higher from lower recurrence risk stage II colon cancer patients beyond known prognostic factors, particularly in T3, mismatch repair-intact patients where traditional factors like grade and LVI were not prognostic," the researchers concluded.
Several study authors hold advisory roles, stock, and have taken research funding from drug developers, while several others are employed by, own stock in, or have advised Genomic Health.
Use of TMPRSS2-ERG gene rearrangement and quantitative ERG expression to predict clinical recurrence after radical prostatectomy.
Investigators from Cleveland Clinic and Genomic Health conducted a study to gauge whether tumor-derived gene expression profiles, including the presence of TMPRSS2-ERG fusions and ERG gene expression, are associated with clinical recurrence after radical prostatectomy. So far, "the association of TMPRSS2-ERG fusions and ERG expression in prostate cancer with adverse clinical outcomes has been controversial, with mixed results in the literature," the researchers acknowledged in an abstract describing the study results.
Researchers in this study identified 2,600 patients with clinical stage T1/T2 prostate cancer treated with radical prostatectomy at Cleveland Clinic from 1987 to 2004. They then selected 127 patients with clinical recurrence and 374 patients without clinical recurrence after a radical prostatectomy.
"For each patient a primary Gleason pattern (GP) sample, secondary (or highest) GP sample, and an adjacent nontumor (NT) tissue sample were evaluated. Surgical Gleason Score (GS) and clinical data were centrally reviewed," according to the abstract. "RNA was extracted from six manually dissected 10 μm formalin-fixed paraffin-embedded sections obtained from RP specimens and expression of TMPRSS2-ERGa, TMPRSS2-ERGb, ERG and reference genes were quantified using RT-PCR." The researchers, led by Sara Falzarano of the Cleveland Clinic, assessed times to clinical recurrence, PSA recurrence, and prostate cancer death using Cox PH regression.
In the study, 441 patients were evaluable, and from them, 848 tumor samples and 410 nontumor samples were assessed. TMPRSS2-ERGa and/or TMPRSS2-ERGb fusions were present in 51.8 percent of tumor samples and 7.5 percent of nontumor samples. The abstract reports "89 percent concordance for TMPRSS2-ERG fusion status between the two tumor samples for each patient overall, and 82 percent concordance in the subset of patients for whom the two specimens were sampled from 2 distinct foci." Meanwhile, the researchers failed to find an association between TMPRSS2-ERG a/b gene rearrangement or ERG expression with clinical recurrence, PSA recurrence, prostate cancer death, or surgical Gleason score.
"This study was notable for the large number of clinical recurrence events, use of a standardized quantitative assay, and rigorous central review of pathology and clinical data," the researchers concluded. "We did not find an association of TMPRSS2-ERG gene rearrangements or ERG expression with aggressiveness of prostate cancer post RP."
Genomic Health has provided research funding to several researchers in this study. Additionally, all the study authors have either advised, own stock, are employed by Genomic Health.
[ pagebreak ]
The role of activating mutations of KRAS, BRAF, and PIK3CA pathway convergence at the transcriptional level and prediction of treatment response to cetuximab in colorectal cancer.
Researchers from Agendia and several research institutions developed three gene expression profiles characterizing KRAS-, BRAF-, or PIK3CA-activated-like tumors to try to show that a combined molecular profiling method can identify patients with activated EGFR signaling.
The researchers, led by Agendia's Iris Simon, collected fresh frozen tumor samples from 381 patients at four European hospitals and analyzed the samples for mutations in KRAS, BRAF and PIK3CA by direct sequencing. Then, using whole-genome microarray analysis, Simon et al. developed three mutation profiles using a 10-fold cross validation method. The profiles were validated in an independent set of 80 patients, as well as in a dataset from metastatic colorectal cancer tumor samples of cetuximab-treated patients.
Overall, 175, or nearly 46 percent, of the tumors harbored oncogenic mutations in KRAS, BRAF, and PIK3CA. After Simon et al. used the three mutation signatures to develop an integrative three-way classification model, they identified tumors as "mutation-like" tumors or "non-mutation-like" tumors with a sensitivity of 90.3 percent and a specificity of 61.7 percent. They found that the three mutation signatures and the combined model were associated with response to cetuximab treatment in patients with metastatic colorectal cancer.
"The convergence of distinct oncogenic mutations in effector genes downstream of EGFR at the transcriptional level allows the discovery of activated profiles that includes both known and unknown activation in this pathway," the researchers concluded in the abstract. "The three profiles allow the identification of patients with activating mutations on EGFR pathway, which may not benefit from drugs that inhibit the EGFR receptor."
Several study authors in this research are employed by Agendia and own stock in the firm.
Combined use of MammaPrint and molecular subtyping profile (BluePrint) to identify subgroups with marked differences in response to neoadjuvant treatment.
Femke De Snoo and other investigators from Agendia and the Netherlands Cancer Institute attempted to apply two gene expression arrays, Agendia's MammaPrint and BluePrint, to discern specific molecular subtypes within breast cancer tumors and how patients' response to neoadjuvant treatment differ by these subtypes.
The study relied on a cohort of 133 breast cancer patients treated with T/FAC (Taxol, 5-fluoracil, doxorubicin, cyclophosphamide) neoadjuvant chemotherapy. Genome-wide expression data, which was publicly available from an MD Anderson database, was used to determine which patients responded to T/FAC stratified by MammaPrint and molecular subtype via BluePrint. Using MammaPrint and BluePrint scores, the researchers subtyped patients into four groups: MammaPrint low-risk/luminal-type, MammaPrint high-risk/luminal-type, basal-type, and ERBB2-type.
Within this cohort, 20 percent of patients were classified as basal-type, 62 percent as luminal-type, and 18 percent as ERBB2-type. "The overall pathologic complete response [pCR] of this patient cohort was 26 percent and differed substantially among the subgroups," the researchers reported in the abstract. "pCR was observed in 9 percent of all luminal-type samples, in 50 percent of the ERBB2-type samples, and in 56 percent of the basal-type samples." De Snoo et al. reported that the pCR rates for patients in the ERBB2-type and basal-type groups were higher compared to classification based on IHC/CISH assessed ER and HER2 receptor status.
"We observed marked differences in response to neoadjuvant treatment in groups stratified by MammaPrint and BluePrint," researchers concluded. "These findings confirm differences in chemotherapy response among molecular subgroups and indicate that the BluePrint profile described here will help to further establish a clinical correlation between molecular subtyping and treatment response.
Several researchers involved with this study have disclosed they are employed by Agendia and own stock in the company.
Prognostic value of a 46-gene cell cycle progression (CCP) RNA signature for prostate cancer death in a conservatively managed watchful waiting needle biopsy cohort.
Investigators from the Wolfson Institute of Preventative Medicine, St. Bartholomew's Hospital, Thames Cancer Registry, and Myriad Genetics investigated the utility of a 46-gene cell cycle progression score in predicting outcomes in patients diagnosed with prostate cancer by needle biopsy who are monitoring their disease by watchful waiting.
Previously, Myriad had reported data suggesting that the CCP RNA signature predicts biochemical recurrence after radical prostatectomy and of prostate cancer-specific death in patients diagnosed with cancer by transurethral resection of the prostate.
In the study, researchers extracted mRNA from paraffin-embedded needle biopsies from men in the UK who were diagnosed with localized prostate cancer between 1990 and 1996. A CCP score was calculated for each patient. The primary endpoint in the study was death from prostate cancer after a mean follow up of 9.8 years.
In an univariate analysis of 352 patients, the hazard ratio for prostate cancer death was 2.67 for a change from the 25th to 75th percentile of the CCP score. In a multivariate analysis of 333 patients, which factored in patients' Gleason score and PSA, "CCP dominated with only Gleason score providing a significant, additional contribution and baseline PSA was nonsignificant," the researchers state in the abstract.
Based on these findings, the authors concluded that "for patients managed by active surveillance, the CCP score is the strongest predictor of cancer death outcome yet described." Myriad's assay may provide independent prognostic information "valuable in the management of prostate cancer," they wrote.
Myriad has provided research funding to several investigators in this study. Additionally, several researchers are employed by and own stock in the company.