Biomarker-driven approach in cancer therapy





1Richard D. BAIRD, MA, MRCP, PhD

1,2Carlos CALDAS, MD, FMedSci
1. Department of Oncology
University of Cambridge
Addenbrooke’s Hospital
Cambridge, UK
2. Cancer Research UK Cambridge
Research Institute
Li Ka Shing Centre
Cambridge, UK

Biomarker-driven approach in cancer therapy


by R. D. Baird and C. Caldas, United Kingdom



There is great variability in tumor biology among cancer patients, determining response to treatments and clinical outcomes. Much of this variability can now be explained by differences in the biomarker profiles between tumors. This review uses breast cancer as a focus for discussing contemporary biomarker-driven approaches in cancer therapy. The most well-established biomarkers in breast cancer are tumor expression of the estrogen receptor (ER) and receptor tyrosine-protein kinase erbB-2 (HER2), high expression levels of which open up the possibility of ER- and HER2-directed therapies, respectively. Additional cancer drivers have been identified in recent decades, including aberrations in the PI3K-AKT-mTOR pathway (phosphatidylinositol 3-kinase; protein kinase B; mammalian target of rapamycin), fibroblast growth factor receptor, cell-cycle control, tumor angiogenesis, and immune control. To tackle these cancer drivers, a large number of novel molecular targeted therapies in breast cancer are in clinical development. It is increasingly apparent that most molecular targeted therapies will only work in biomarker-defined subsets of patients. This has required the development of new clinical trial methodologies, including “basket” and “umbrella” trial designs to match molecular therapy to the patient’s biomarker profile, as well as novel Bayesian adaptive approaches. Key to the success of this approach will be repeated sampling of patient tumors over time, which can be difficult if invasive tumor biopsies are required, but may be transformed by new “liquid biopsy” technologies to measure circulating free DNA and circulating tumor cells.

Medicographia. 2015;37:287-295 (see French abstract on page 295)



No 2 cancer patients are exactly the same. Even patients with cancers that are the same size and stage, and that are very similar at the microscopic level using standard pathological and immunohistochemical criteria, can have dramatically different natural histories and responses to therapy. Clinicians know this only too well. Indeed, one of the central questions of oncology is why 2 similar patients treated in the same way can have such different clinical outcomes. It is now well established that this variation can be largely explained by differences in the molecular or biomarker profiles between different cancers. This review focuses on breast cancer as a paradigm for contemporary biomarker-driven approaches in cancer therapy.

Breast cancer is not one but many diseases, each defined by its biomarker profile

The best established breast cancer biomarkers are estrogen receptor (ER) expression and receptor tyrosine-protein kinase erbB-2 (HER2) expression. It is striking that with all the advances in molecular biology and genomic technologies in the last 2 years, ER and HER2 are the only universally accepted biomarkers used in clinical practice for the management of breast cancer patients. However, there remain large differences in clinical outcomes between patients with the same ER and HER2 status. With the development of novel genetic technologies over the last 20 years, a number of genomic classifiers have been developed to help explain this clinical variability. The first generation of breast cancer classifiers examined the correlation of gene expression with clinical outcomes, using complementary DNA (cDNA) microarray and polymerase chain reaction (PCR)-based techniques. This work was pioneered by the Stanford group, who used unsupervised hierarchical clustering of gene-expression profiles to discover 5 new “intrinsic subtypes” of breast cancer: luminal A, luminal B, normal breast like, HER2-positive, and basal-like cancers.1





Figure 1
Figure 1. Survival curves for ten subtypes of breast cancer identified by
METABRIC researchers.

For each integrative cluster (IntClust), based on joint copy number and gene expression data, the
number of samples at risk and total number of deaths (in parentheses) are indicated.
Copied from reference 5: Dawson et al. EMBO J. 2013;32:617-628. © 2013, European Molecular
Biology Organization.



There have since been a number of efforts to standardize and commercialize gene-expression–based breast cancer assays. The MammaPrint test is a 70-gene expression signature originally developed from a series of patients who had undergone definitive surgery, but no systemic therapy, and for whom long-term follow-up was available2; it was approved by the US Food and Drug Administration (FDA) for prognostic prediction in 2007. Oncotype DX is a 21-gene, real-time PCRbased assay which was initially developed to predict the risk of cancer recurrence for women treated for stage I and II hormone- receptor–positive, lymph-node–negative, invasive breast cancer by surgery and 5 years of adjuvant tamoxifen.3

The largest breast cancer genomic study to date has been performed by the METABRIC group (Molecular Taxonomy of Breast Cancer International Consortium), combining gene copy number and expression from 2000 primary breast cancers, with long-term clinical follow-up.4 Analysis of the combined DNA-RNA profiles revealed that breast cancer can be classified into 10 different subgroups, each with different clinical outcomes (Figure 1).5 This new taxonomy of breast cancer has now been robustly validated in 7500 external samples.6 In contrast to the early gene-expression–based studies discussed above, this included 7 distinct subgroups of ERpositive breast cancer, and separation of triple-negative can- cers (breast cancer negative for estrogen, progesterone, and HER2 receptors) into at least 2 subtypes. The METABRIC classifier may be the only breast cancer genomic classifier that may help direct targeted-therapy selection, with enrichment for particular drivers in different iClusters (Integrative Clustering of Multiple Genomic Data Types).5

Tumor heterogeneity in space and time – the need for repeat tumor sampling
In the past, comprehensive studies of tumor heterogeneity at the molecular level have been constrained by the lack of available biopsy specimens. Clinicians naturally wanted to avoid unnecessary invasive biopsy procedures, particularly if this didn’t change patient management.

There is now, however, a recognition that repeat tumor biopsy at the time of metastatic relapse can be crucial to optimization of patient care, since ER and HER2 status can change over time. In one meta-analysis, for example, it was found overall that 14% of patients were reclassified from ER-negative to ER-positive and 5% of patients were reclassified from HER2- negative to HER2-positive, opening up the possibility of new endocrine therapy and anti-HER2–therapy options, respectively.7 While ER- and HER2-directed therapies have without question transformed the clinical outcomes for patients with breast cancer, painstaking translational research over recent years has uncovered additional key drivers of malignant progression. These include nodes in the PI3K-AKT-mTOR pathway (phosphatidylinositol 3-kinase; AKT or protein kinase B; mammalian target of rapamycin); receptor tyrosine kinases (eg, fibroblast growth factor receptor [FGFR]; cell-cycle control (eg, cyclindependent kinases 4 and 6 [CDK4/ 6]; DNA damage repair pathways (eg, poly(ADP-ribose) polymerase [PARP]); and immune checkpoints.

Molecular targeted therapies for breast cancer

The discovery and validation of novel drug targets in breast cancer has led to a large number of corresponding drug-discovery programs. Selected molecular targeted therapies for breast cancer in clinical development or recently approved are shown in Table I.8-17

HER2-directed therapies
In the last 15 years, trastuzumab has transformed the outcomes for patients with HER2-positive breast cancer. Patients in this subgroup used to have some of the worst survival statistics out of all breast cancer patients, until the addition of trastuzumab to standard chemotherapy brought these survival curves back up to the rest of the group.18


Table I
Table I. Selected molecular targeted
therapies in clinical development or
recently approved for the treatment of
breast cancer.

Abbreviations: AKT, also known as protein
kinase B; CDK, cyclin-dependent kinase; ER,
estrogen receptor; HER2, receptor tyrosine-protein
kinase erbB-2; FGFR, fibroblast growth factor
receptor; mTOR, mammalian target of rapamycin;
mTORC, mTOR complex; oral SERDs, oral selective
estrogen receptor degraders; PARP, poly-
(ADP-ribose) polymerase; PD-1, programmed
cell death protein 1; PD-L1, programmed deathligand
1; PGM, Personal Genome Machine;
PI3K, phosphatidylinositol 3-kinase; TKI, tyrosine
kinase inhibitor.



In recent years, the addition of pertuzumab, another HER2- targeting monoclonal antibody, to standard therapy, has also yielded a further dramatic step change in overall survival for patients with HER2-positive metastatic breast cancer. The CLEOPATRA study (CLinical Evaluation Of Pertuzumab And TRAstuzumab) showed that the median overall survival for these patients could be improved from 40.8 months to 56.5 months, with minimal additional toxicity,8 leading to regula- tory approval for pertuzumab in 2012. Another drug approved in 2013 for the treatment of HER2-positive metastatic breast cancer was the antibody drug conjugate, trastuzumab emtansine (T-DM1). T-DM1 binds the trastuzumab antibody to a cytotoxic maytansine derivative (DM1), using a stable linker. This specifically targets cytotoxic drug delivery to HER2-overexpressing cells, thereby minimizing cytotoxic exposure to normal tissues and improving therapeutic index. FDA approval for T-DM1 was granted following the phase 3 EMILIA trial (NCT00829166; not an acronym), which randomized patients with HER2-positive advanced breast cancer, who had received prior taxane- and trastuzumab-based therapy, to either T-DM1 or lapatinib + capecitabine.9 In this second/thirdline population, T-DM1 demonstrated improved efficacy (median progression-free survival [PFS],9.4 months vs 6.4 months, respectively), and reduced toxicity.

PARP inhibitors
A further important group of molecular targeted therapies are those that target cancer-cell defects in DNA repair. Preeminent in this category are inhibitors of the enzyme PARP. Ten years ago, preclinical data suggested that cancer cells with impaired DNA repair via the homologous recombination pathway were exquisitely sensitive to PARP inhibition,19 via socalled “synthetic lethality.” Early phase clinical trials of PARP inhibitors suggested a high degree of antitumor activity for these agents, particularly against tumors with breast cancer gene (BRCA) mutations.20 Hopes were high that these drugs might reach a wider group of patients than merely those with BRCA mutations, including high-grade serous ovarian cancer, triple-negative breast cancer, and other sporadic cancers with a “BRCAness” phenotype.21 However, the clinical development programs of PARP inhibitors encountered a number of serious challenges. First, the PARP-inhibitor field was damaged by the results from the iniparib development program; partly by a surprising negative phase 3 trial when iniparib was added to standard chemotherapy, and subsequently by the revelation that iniparib wasn’t a PARP inhibitor after all.22,23 Second, the level of antitumor activity for non–BRCA-mutant tumors was disappointing, including data in triple-negative breast cancer.24 However, in the last few years, the PARP-inhibitor field has experienced a resurgence, with clinical development focused on BRCA-mutant tumors. In early 2015, regulatory approval for olaparib was granted for the maintenance treatment of BRCA-mutant ovarian cancer. Ongoing clinical research strategies are focused on the development of predictive biomarkers for defective homologous recombination and the use of PARP inhibitors in combination with other agents, for example PI3K pathway inhibitors.25

VEGF
It is now well understood that a nascent tumor cannot grow beyond a tiny size without creating a network of blood vessels to supply it with nutrients and to take away waste products, a process called angiogenesis. This hallmark trait of cancer is mediated by a number of factors including vascular endothelial growth factor (VEGF) and its receptor. A number of drugs have been developed to inhibit VEGF, including the monoclonal antibody bevacizumab, which gained conditional FDA approval following the results of the E2100 trial. In this study, the investigators found that the addition of bevacizumab to paclitaxel for the first-line treatment of metastatic breast cancer yielded an improved overall response rate [36.9% vs 21.2%; P<0.001] and increased median PFS (11.8 vs 5.9 months; P<0.001).26 However, there proved to be no overall survival difference for the addition of bevacizumab and, in fact, treatment was associated with an increased risk of potentially dangerous toxicities, including hypertension, bleeding, and gastrointestinal perforation. Thus in December 2010, the FDA recommended removing the first-line metastatic breast cancer indication from the approval for bevacizumab. In the last few years, this conclusion has been bolstered by negative clinical trial results for bevacizumab in the adjuvant breast cancer setting.27 Current clinical research efforts include those focused on identifying predictive biomarkers for maximum bevacizumab benefit, for example the ARTemis trial (Avastin Randomized Trial with neo-adjuvant chemotherapy for patients with early breast cancer [NCT01093235]).

PI3K-AKT-mTOR inhibitors
The PI3K-AKT-mTOR pathway is one of the most commonly deranged in human cancer. For example, up to 40% of breast cancers are found to bear PIK3CA mutations. Activation of the pathway in clinical cancer samples is associated with poor prognosis in multiple tumor types, and functional studies in preclinical models suggest that the pathway’s activation can promote malignant progression through increased proliferation, increased cell survival, and promotion of invasion and metastasis. Furthermore, the nodes in the pathway are druggable targets, which has led to a huge number of inhibitors in clinical development. Selected PI3K, AKT, and mTOR inhibitors are listed in Table I and are reviewed more comprehensively elsewhere.28

First-generation “rapalog” mTOR complex 1 (mTORC1) inhibitors were the first agents in this group to reach the clinic and in 2012, everolimus was approved for the treatment of postmenopausal, hormone-receptor–positive metastatic breast cancer, in combination with the aromatase inhibitor exemestane.16 This approval was based on the BOLERO-2 study (Breast cancer trials of OraL EveROlimus-2), which demonstrated an improved median PFS for patients treated on the everolimus-exemestane arm (6.9 months), compared with exemestane alone (2.8 months). However, this PFS benefit came at the cost of increased side effects, including stomatitis (any grade: 56% vs 11%), skin rash (36% vs 6%), diarrhea (30% vs 16%), hyperglycemia (13% vs 2%), and pneumonitis (12% vs 0%). So while the approval of everolimus was an important step forward, there remained a pressing need to develop PI3K-pathway agents with an improved therapeutic index.

Consequently, there have been great hopes for direct inhibitors of the catalytic subunit of PIK3, which is encoded by the PIK3CA gene. Preclinical models of PIK3CA-driven cancers suggested that PIK3CA-mutant cell lines were considerably more sensitive to early PI3K inhibitors such as buparlisib (BKM120) and pictilisib (GDC-0941).29

Early clinical trials showed that these agents were generally well tolerated, but showed somewhat disappointing single agent antitumor activity, even in patients with PIK3CA-mutant tumors.10,11

However, it has since become clear that different PI3K isoforms have distinct biological functions,30 and targeting these with isoform-selective PI3K inhibitors may yield an improved therapeutic index. Indeed, the early clinical trial results for &alpha-isoform selective inhibitors of PI3K are more encouraging, and seem to have increased activity in patients with PIK3CA-mutant tumors.12,13

FGFR inhibitors
Mutations and amplifications in the genes encoding FGFR are among the most common abnormalities found in breast cancer. A range of FGFR inhibitors are in development, including drugs that target FGFR, but also other kinases (eg, VEGF, kinase insert domain receptor [KDR], etc), and drugs that more selectively target FGFR in isolation (Table I). While responses to these agents have been reported, they do not predictably occur in patients bearing FGFR molecular abnormalities. Current development strategies in breast cancer are focused on combination therapies with endocrine and cytotoxic chemotherapies.

CDK4/6 inhibitors
Disruption of the cell cycle is a well-defined hallmark of cancers in general, and many genetic alterations in cell-cycle control have been described in breast cancer specifically.4 The CDKs bind with cyclin partners to help control progression through the cell cycle, and a number of CDK-targeted drugs have been developed to interfere with this. Until recently, the clinical experience with CDK inhibitors was characterized by a lack of clinical activity and the presence of significant toxicities.31 Palbociclib was subsequently developed as a reversible, oral, small-molecule inhibitor of CDK4/6, which has an important role in regulating the G1/S phase transition via the control of retinoblastoma protein (Rb) phosphorylation. In February 2015, palbociclib was granted accelerated approval for the treatment of hormone-receptor–positive metastatic breast cancer, in combination with letrozole, based on the results from the PALOMA-1 trial (not an acronym).17 This showed a dramatic increase in median PFS from 10.2 months for patients treated with letrozole alone, to 20.2 months for patients treated with letrozole in combination with palbociclib. Confirmatory phase 3 trials are underway, and a number of other CDK4/6 inhibitors are now in clinical development.

Immunotherapies
Researchers have tried over many decades to enhance the patient immune response to treat cancers, unfortunately with very limited success. However, in the last few years, the development of novel immune-checkpoint inhibitors has reinvigorated the immunotherapy field. These checkpoint inhibitors block some of the inhibitory signals between effector T-cells and tumor cells, including cytotoxic T-lymphocyte antigen 4 (CTLA-4), programmed cell death protein 1 (PD-1), and programmed death-ligand 1 (PD-L1), thus “taking the brakes off” the immune response. Clinical trial results in melanoma and non–small cell lung cancer have been particularly exciting, with a subset of patients achieving long-term disease control, in some cases after only a few doses of treatment.32,33 Encouraging activity has also been seen in other tumor types including bladder, head and neck, and triple-negative breast cancer. Once again, significant efforts are underway to identify biomarkers that will predict response to immunotherapies. The early data on tumor PD-L1 expression appears promising, with an enriched response rate for PD-L1–positive cancers; however, many patients with PD-L1–negative cancers also respond to therapy.34 Interestingly, the subset of breast cancer patients in the METABRIC series, which is characterized by a dearth of genomic aberrations (iCluster 4),4 is also characterized by increased immune infiltrates and may be a rational group to target with immune-checkpoint therapies. More recently, a large immunohistochemical study has revealed that approximately 20% of basal-like breast cancers express the PD-L1 protein, mostly in tumor-infiltrating immune cells.35 This subgroup could gain particular benefit from anti– PD-L1 directed therapies.

The need for robust biomarkers to predict clinical outcome
One common theme arching across all these clinical drug development programs is the recognition that different patients can respond very differently to any given drug. For these molecular targeted therapies, it is imperative that robust biomarkers are identified to predict which subgroups of patients are likely to get the greatest benefit from these new drugs, which are frequently very expensive and sometimes associated with significant toxicity. There are a number of distinct categories for biomarkers correlated with drug effect; these are defined in Box 1 (page 292).10,36

Biomarker-driven clinical trial design

Trials investigating novel molecular therapies and biomarker profiles have required significant changes to the traditional paradigm for clinical drug development.37 In phase 1 clinical trials, the recommended phase 2 dose is now selected not only on the basis of toxicity, but is also informed by pharmacokinetic- pharmacodynamic relationships. There is now a blurring between the old division between phase 1, phase 2, and phase 3 trials, with many early phase clinical trials having both dose-escalation and large expansion phases.





Alongside this is a trend away from large, unselected phase 3 trials, to smaller, often phase 2 trials seeking to identify larger effect sizes in biomarker-defined patient subgroups.38 Furthermore, there are now major efforts in cancer centers across the world to establish routine molecular profiling services for cancer patients. These efforts help to identify groups of patients with particular molecular aberrations required for clinical trial entry, and also permit researchers to study correlations between molecular profiles and clinical outcomes (Table II).39-46

“Umbrella” and “basket” trials
Umbrella trials tend to be multi-arm studies focused on particular tumor types. Patients are assigned to specific drugs based on the molecular profile of their cancer (Figure 2).


Figure 2
Figure 2. Umbrella clinical trial design: biomarker profiling for
“matched molecular therapy.”

Abbreviations: FGFR, fibroblast growth factor receptor; PI3K, phosphatidylinositol
3-kinase; PIK3CA, phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic
subunit α; PD-L1, programmed death-ligand 1.



In a basket trial, a single drug is tested across cancer types, in patients whose tumors have been found to have a particular abnormality, often a specific gene mutation (Figure 2). One example of such a study is the VE-BASKET trial (NCT01524978); which is testing the serine/threonine-protein kinase BRAF inhibitor vemurafenib in patients whose cancers bear a mutation in the BRAF gene, also referred to as v-Raf murine sarcoma viral oncogene homolog B1. This trial provides the opportunity to quickly test a novel agent across multiple tumor types. However, one of the disadvantages of this approach is that for uncommon gene mutations, a large number of patients have to be tested (in some cases after an additional research tumor biopsy) in order to identify the small group eligible for the study. This can be unsatisfactory for both patients and investigators.

A further challenge to the basket trial, multi–tumor type approach, is that the function of particular gene mutations is often highly context dependent. For example, BRAF inhibitors work much better in BRAF-mutant melanoma than they do in BRAF-mutant colorectal cancer. Recent work from the Bernards group has suggested that this may be due to feedback activation of epidermal growth factor receptor (EGFR), which allows the cancer cells to continue proliferating in the presence of the BRAF inhibitor. This feedback loop is not seen in melanoma cells because they express low levels of EGFR.47



Table II (right page). Selected clinical trial programs
matching biomarker profiles to targeted therapies.

Abbreviations: aCGH, array comparative genomic hybridization; BRAF, v-Raf
murine sarcoma viral oncogene homolog B1; CCND1, cyclin D1; CNV, copy
number variation; EGFR, epidermal growth factor receptor; FISH, fluorescence
in situ hybridization; IHC, immunohistochemistry; KRAS, Kirsten rat sarcoma viral
oncogene homolog; MSKCC, Memorial Sloan Kettering Cancer Center;
NGS, next-generation sequencing; NSCLC, non–small cell lung cancer; PCR,
polymerase chain reaction; RPPA, reverse phase protein arrays; RXRs, retinoid
X receptors; US NCI, US National Cancer Institute; VEGF, vascular endothelial
growth factor; VEGFR2, vascular endothelial growth factor receptor 2; VHIO,
Vall d’Hebron Institute of Oncology; WIN, Worldwide Innovative Network.



This work has led to combinations of BRAF and EGFR inhibitors, with chemotherapy being tested in BRAF-mutant colorectal cancer with encouraging preliminary results.48

Contemporary clinical trial designs that aim to identify the best fit between biomarker profile and molecular targeted therapy also increasingly use multi-arm, multi-stage designs (eg, FOCUS4, [Molecular selection of therapy in metastatic colorectal cancer: a molecularly stratified randomized controlled trial program]), and Bayesian adaptive approaches (eg, BATTLE- 1 [Biomarker-integrated Approaches of Targeted Therapy for Lung Cancer Elimination], I-SPY 2 Trial [Investigation of Serial studies to Predict Your Therapeutic Response with Imaging and molecular AnaLysis 2]).

“Liquid biopsy” – a great new opportunity for biomarker-driven clinical drug development

This review would not be complete without considering the transformative potential of new “liquid biopsy” technologies, which provide a method to characterize tumor biomarkers over time, with repeated sampling, which is much less invasive than traditional solid tumor biopsies. At the current time, 2 of the most high-profile technologies are those to assay circulating tumor cells (CTCs) and circulating tumor DNA.49 In this fast-moving field, multiple research groups are investigating the potential for liquid biopsies to yield clinically useful predictive biomarkers, early-response biomarkers,36 and biomarkers of acquired drug resistance.50

Conclusions

This review has considered how recent advances in our understanding of the molecular biology of breast cancer have led to the development of novel molecular targeted therapies. It is clear that such therapies will work best in subsets of patients defined by their biomarker profile. Novel clinical trial methodologies are being employed to help match these biomarker profiles for individual patients to the best molecular therapy, using basket, umbrella, and adaptive approaches. Current experimental therapeutics trials are translationally intensive, with repeat tumor sampling over time to capture changing molecular profiles, which occur as a result of cancer clonal evolution. These kinds of trials may be facilitated by new liquid biopsy technologies for monitoring cell-free DNA and circulating tumor cells.

References
1. Perou CM, Sørlie T, Eisen MB, et al. Molecular portraits of human breast tumours. Nature. 2000;406:747-752.
2. van’t Veer LJ, Dai H, van de Vijver MJ, et al. Gene expression profiling predicts clinical outcome of breast cancer. Nature. 2002;415:530-536.
3. Paik S, Shak S, Tang G, et al. A multigene assay to predict recurrence of tamoxifen- treated, node-negative breast cancer. N Engl J Med. 2004;351:2817- 2826.
4. Curtis C, Shah SP, Chin SF, et al. The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature. 2012;486:346- 352.
5. Dawson SJ, Rueda OM, Aparicio S, Caldas C. A new genome-driven integrated classification of breast cancer and its implications. EMBO J. 2013;32:617- 628.
6. Ali HR, Rueda OM, Chin SF, et al. Genome-driven integrated classification of breast cancer validated in over 7,500 samples. Genome Biol. 2014;15:431.
7. Aurilio G, Disalvatore D, Pruneri G, et al. A meta-analysis of oestrogen receptor, progesterone receptor and human epidermal growth factor receptor 2 discordance between primary breast cancer and metastases. Eur J Cancer. 2014;50: 277-289.
8. Swain SM, Baselga J, Kim SB, et al; CLEOPATRA Study Group. Pertuzumab, trastuzumab, and docetaxel in HER2-positive metastatic breast cancer. N Engl J Med. 2015;372:724-734.
9. Verma S, Miles D, Gianni L, et al; EMILIA Study Group. Trastuzumab emtansine for HER2-positive advanced breast cancer. N Engl J Med. 2012;367:1783- 1791.
10. Sarker D, Ang JE, Baird R, et al. First-in-human phase I study of pictilisib (GDC- 0941), a potent pan-class I phosphatidylinositol-3-kinase (PI3K) inhibitor, in patients with advanced solid tumors. Clin Cancer Res. 2014;21:77-86.
11. Bendell JC, Rodon J, Burris HA, et al. Phase I, dose-escalation study of BKM120, an oral pan-Class I PI3K inhibitor, in patients with advanced solid tumors. J Clin Oncol. 2012;30:282-290.
12. Juric D, Rodin J, Gonzalez-Angulo AM, et al. BYL719, a next generation PI3K alpha specific inhibitor: preliminary safety, PK, and efficacy results from the firstin- human study. Cancer Res. 2012;72:CT-01. doi:10.1158/1538-7445.AM2012- CT-01.
13. Juric D, Krop I, Ramanathan RK, et al. GDC-0032, a beta isoform-sparing PI3K inhibitor: results of a first-in-human phase Ia dose escalation study. Cancer Res. 2013;73:LB-64. doi:10.1158/1538-7445.AM2013-LB-64.
14. Banerji U, Ranson M, Schellens JHM, et al. Results of two phase I multicenter trials of AZD5363, an inhibitor of AKT1, 2 and 3: biomarker and early clinical evaluation in Western and Japanese patients with advanced solid tumors. Cancer Res. 2013;73:LB-66. doi:10.1158/1538-7445.AM2013-LB-66.
15. Yap TA, Yan L, Patnaik A, et al. First-in-man clinical trial of the oral pan-AKT inhibitor MK-2206 in patients with advanced solid tumors. J Clin Oncol. 2011; 29:4688-4695.
16. Baselga J, Campone M, Piccart M, et al. Everolimus in postmenopausal hormone- receptor-positive advanced breast cancer. N Engl J Med. 2012;366: 520-529.
17. Finn RS, Crown JP, Lang I, et al. The cyclin-dependent kinase 4/6 inhibitor palbociclib in combination with letrozole versus letrozole alone as first-line treatment of oestrogen receptor-positive, HER2-negative, advanced breast cancer (PALOMA-1/TRIO-18): a randomised phase 2 study. Lancet Oncol. 2015;16: 25-35.
18. Vici P, Pizzuti L, Natoli C, et al. Outcomes of HER2-positive early breast cancer patients in the pre-trastuzumab and trastuzumab eras: a real-world multicenter observational analysis. The RETROHER study. Breast Cancer Res Treat. 2014;147:599-607.
19. Farmer H, McCabe N, Lord CJ, et al. Targeting the DNA repair defect in BRCA mutant cells as a therapeutic strategy. Nature. 2005;434:917-921.
20. Fong PC, Boss DS, Yap TA, et al. Inhibition of poly(ADP-ribose) polymerase in tumors from BRCA mutation carriers. N Engl J Med. 2009;361:123-134.
21. Turner N, Tutt A, Ashworth A. Opinion: Hallmarks of ‘BRCAness’ in sporadic cancers. Nat Rev Cancer. 2004;4:814-819.
22. O’Shaughnessy JA, Osborne C, Pippen JE, et al. Iniparib plus chemotherapy in metastatic triple-negative breast cancer. N Engl J Med. 2011;364:205-214.
23. Sinha G. Downfall of iniparib: a PARP inhibitor that doesn’t inhibit PARP after all. J Natl Cancer Inst. 2014;106:djt447.
24. Gelmon K, Tischkowitz M, Mackay H, et al. Olaparib in patients with recurrent high-grade serous or poorly differentiated ovarian carcinoma or triple-negative breast cancer: a phase 2, multicentre, open-label, non-randomised study. Lancet Oncol. 2011;12:852-861.
25. Ibrahim YH, García-García C, Serra V, et al. PI3K inhibition impairs BRCA1/2 expression and sensitizes BRCA-proficient triple-negative breast cancer to PARP inhibition. Cancer Discov. 2012;2:1036-1047.
26. Miller KD, Wang M, Gralow J, et al. Paclitaxel plus bevacizumab versus paclitaxel alone for metastatic breast cancer. N Engl J Med. 2007;357:2666-2676.
27. Cameron D, Brown J, Dent R, et al. Adjuvant bevacizumab-containing therapy in triple-negative breast cancer (BEATRICE): Primary results of a randomised, phase 3 trial. Lancet Oncol. 2013;14:933-942.
28. Rodón J, Dienstmann R, Serra V, Tabernero JM. Development of PI3K inhib itors: lessons learned from early clinical trials. Nat Rev Clin Oncol. 2013;10: 143-153.
29. O’Brien C, Wallin JJ, Sampath D, et al. Predictive biomarkers of sensitivity to the phosphatidylinositol 3’ kinase inhibitor GDC-0941 in breast cancer preclinical models. Clin Cancer Res. 2010;16:3670-3683.
30. Vanhaesebroeck B, Guillermet-Guibert J, Graupera M, Bilanges B. The emerging mechanisms of isoform-specific PI3K signalling. Nat Rev Mol Cell Biol. 2010;11:329-341.
31. Stone A, Sutherland RL, Musgrove EA. Inhibitors of cell cycle kinases: recent advances and future prospects as cancer therapeutics. Crit Rev Oncog. 2012; 17:175-198.
32. Prieto PA, Yang JC, Sherry RM, et al. CTLA-4 blockade with ipilimumab: longterm follow-up of 177 patients with metastatic melanoma. Clin Cancer Res. 2012;18:2039-2047.
33. Gettinger SN, Horn L, Gandhi L, et al. Overall survival and long-term safety of nivolumab ( anti–programmed death 1 antibody, BMS-936558, ONO-4538) in patients with previously treated advanced non–small-cell lung cancer. J Clin Oncol. 2015 Apr 20. Epub ahead of print. pii:10.1200/JCO.2014.58.3708.
34. Patel SP, Kurzrock R. PD-L1 Expression as a predictive biomarker in cancer immunotherapy. Mol Cancer Ther. 2015;14:847-856.
35. Ali HR, Glont SE, Blows FM, et al. PD-L1 protein expression in breast cancer is rare, enriched in basal-like tumours and associated with infiltrating lymphocytes. Ann Oncol. 2015 Apr 20. Epub ahead of print.
36. Dawson SJ, Tsui DW, Murtaza M, et al. Analysis of circulating tumor DNA to monitor metastatic breast cancer. N Engl J Med. 2013;368:1199-1209.
37. Yap TA, Sandhu SK, Workman P, de Bono JS. Envisioning the future of early anticancer drug development. Nat Rev Cancer. 2010;10:514-523.
38. Stewart DJ, Kurzrock R. Cancer: the road to Amiens. J Clin Oncol. 2009;27: 328-233.
39. Kim ES, Herbst RS, Wistuba II, et al. The BATTLE trial: personalizing therapy for lung cancer. Cancer Discov. 2011;1:44-53.
40. Le Tourneau C, Paoletti X, Servant N, et al. Randomised proof-of-concept phase II trial comparing targeted therapy based on tumour molecular profiling vs conventional therapy in patients with refractory cancer: results of the feasibility part of the SHIVA trial. Br J Cancer. 2014;111:17-24.
41. Rodon J, Soria JC, Berger R, et al. Challenges in initiating and conducting personalized cancer therapy trials: perspectives from WINTHER, a Worldwide Innovative Network (WIN) Consortium Trial. Ann Oncol. 2015 Apr 23. Epub ahead of print.
42. André F, Bachelot T, Commo F, et al. Comparative genomic hybridisation array and DNA sequencing to direct treatment of metastatic breast cancer: a multicentre, prospective trial (SAFIR01/UNICANCER). Lancet Oncol. 2014;15:267-274.
43. Rollins BJ, MacConaill LE, Wagle N, et al. PROFILE: broadly based genomic testing for all patients at a major cancer center. J Clin Oncol (Meeting Abstracts). 2013;31(15 suppl):1531. http://meetinglibrary.asco.org/content/116795-132.
44. Sequist LV, Heist RS, Shaw AT, et al. Implementing multiplexed genotyping of non-small-cell lung cancers into routine clinical practice. Ann Oncol. 2011;22: 2616-2624.
45. Voest EE. Delivering stratified medicine: a strategic overview. EACR Mol Pathol Course. 2012. http://www.eacr.org/mpathcourse2012/presentations/E Voest_ presentation.pdf.
46. Dienstmann R, Serpico D, Rodon J, et al. Molecular profiling of patients with colorectal cancer and matched targeted therapy in phase I clinical trials. Mol Cancer Ther. 2012;11:2062-2071.
47. Prahallad A, Sun C, Huang S, et al. Unresponsiveness of colon cancer to BRAF (V600E) inhibition through feedback activation of EGFR. Nature. 2012:483;100- 103.
48. Hong DS, Morris VK, Fu S, et al. Phase 1B study of vemurafenib in combination with irinotecan and cetuximab in patients with BRAF-mutated advanced cancers and metastatic colorectal cancer. J Clin Oncol (Meeting Abstracts). 2014;32(15 suppl):3516. http://meetinglibrary.asco.org/content/132589-144.
49. Haber D, Velculescu VE. Blood-based analyses of cancer: circulating tumor cells and circulating tumor DNA. Cancer Discov. 2014;4:650-661.
50. Murtaza M, Dawson SJ, Tsui DW, et al. Non-invasive analysis of acquired resistance to cancer therapy by sequencing of plasma DNA. Nature. 2013;497:108- 112.



Keywords: basket/umbrella trial; biomarker profile; breast cancer; estrogen receptor, HER2, liquid biopsy; targeted therapy