In February 2013, AVEO Oncology and Astellas Pharma Inc. announced results from their controlled, multi-center, open-label, TIVO-1 study that compared the investigational VEGF inhibitor, tivozanib, head-to-head with the VEGF inhibitor, sorafenib [Nexavar; BayerHealthcare, Onyx Pharmaceuticals] in recurring or metastatic renal cell carcinoma. Patients (N=517) were randomized to tivozanib or sorafenib and must have undergone prior nephrectomy (complete or partial) for excision of the primary tumor.
On May 2, 2013, the FDA's Oncologic Drug Advisory Committee (ODAC) is meeting to discuss the efficacy and safety of tivozanib and will decide its fate. Seemingly, this is business as usual until a review of the dynamics in play makes ODAC’s yea or nay vote on tivozanib an interesting one to follow.
In the final analysis of TIVO-1, tivozanib demonstrated a statistically significant improvement in progression-free survival (PFS), the primary endpoint of the study, when compared with sorafenib: median PFS 11.9 months in the tivozanib arm vs 9.1 months for sorafenib. However, in the secondary endpoint that measured overall survival (OS), results showed that there was no statistical difference between the two arms. The final OS analysis shows a median OS of 28.8 months for tivozanib versus a median OS of 29.3 months for the sorafenib arm.
At first look, given the OS results, it would seem as if tivozanib did not perform any better (or worse) than its comparator; yet what is confounding the OS results is the fact that the TIVO-1 trial design allowed patients in the sorafenib arm to switch over to the tivozanib arm once their disease progressed. Roughly 74% of patients in the sorafenib arm received a subsequent therapy in the trial versus 36% in the tivozanib arm.
How ODAC will respond to a drug whose OS fared about the same as its comparator, but showed superiority in PFS brings the question of PFS vs OS as a gold standard under scrutiny once again. The validity of PFS and OS as endpoints has been debated before. Furthermore, crossover of patients in cancer clinical trials is a recognized phenomenon, confounding results, and casting shadows of doubt on trial endpoints.
In The Street, Adam Feuerstein is predicting a negative vote by the ODAC panel for recommendation to the FDA. He states, “Admittedly, my conviction level for this prediction isn't all that strong. Today's FDA seems addicted to approving all drugs, so the agency may be unwilling to subject tivozanib to the scrutiny it deserves. If the FDA takes it easy on tivozanib, the advisory panel may do the same.”
He is basing his prediction on the analysis of the OS data over a 2-year period in which the relative risk of death was 25% higher in the tivozanib arm vs the control arm. The greater risk of death between the 2 groups was not statistically significant. “That might work in Aveo's favor, but a hazard ratio of 1.245 against tivozanib is alarming -- even if not statistically significant -- and raises serious questions about the drug or the conduct of the study,” he states.
To read more click here: http://www.thestreet.com/story/11905026/1/why-aveos-kidney-cancer-drug-will-be-turned-down-by-next-weeks-fda-panel.html
Supporting the tivozanib recommendation by the ODAC committee is Aafia Chaudhry, MD. In an online article in PropThink, Dr. Chaudhry writes, “we see very little to suggest that a panel would vote against approval of this drug. While some may argue that the RCC market is crowded, we see tivozanib being used as an alternative to other oral therapies in the frontline setting. The agent’s beneficial dosing cycle and superior tolerability profile make this an exciting addition to the current armamentarium. We believe oncologists would want to have this available in the clinic to offer patients. It also adds a useful alternative to more toxic agents used in sequencing for subsequent lines of therapy.”
To read more click here: http://propthink.com/aveos-inflection-point-tivozanib-likely-to-receive-positive-advisory-vote-fda-approval/5898
Kimryn Rathmell, MD, PhD, Associate professor, The University of North Carolina, Chapel Hill, is a translational scientist and a medical oncologist dedicated to the care of patients with kidney cancer. According to Dr. Rathmell, “Going before ODAC, we have no clear OS advantage with tivozanib as compared with sorafinib, yet PFS is positively impacted at 8.4 months in the investigational arm.”
In patients who were treatment naïve (70% of total study population), tivozanib demonstrated a statistically significant improvement in PFS with a median of 12.7 months vs 9.1 months for sorafenib. According to the AVEO Oncology press release, “This is the longest median PFS reported to date in treatment-naïve advanced RCC patients in a pivotal study.”
In addition to meeting its endpoints, tivozanib demonstrated a good safety profile, with patients tolerating the drug well. The most common adverse event reported was hypertension which is consistent with the safety profile of the Phase 2 study. When asked about the possibility of ODAC not recommending the drug for FDA approval, Rathmell said, “This is a drug that has a strong PFS and a good side effect profile. It would be disappointing to go back to a Phase 3 trial because of a crossover.”
In the United States, kidney cancer is expected to occur in over 65,000 people in 2013, with over 13,000 deaths being attributed to the disease. The death rate is fairly steady annually, and most people don’t know they have it until it is in its advanced stage. “It’s a sneaky disease,” says Rathmell, “about one-third of tumors will reoccur after patients undergo resection.”
To treat the disease, the mainstay of available therapies is tyrosine kinase inhibitors (TKIs) that target the VEGF receptor. “Chemo doesn’t really work,” says Rathmell, “because the disease is very vascular. TKI drugs, [such as tivozanib and sorafanib], target the vasculature; not really the cancer cells.”
Targeting the VEGF receptor is just one corner of the market. “Standard of care used to be interferon even though it was never FDA approved,” says Rathwell. “It’s toxic, hard to give, and only works for a very small margin of patients. With the TKIs most people can take them and get some kind of response.”
In the TKI era, Rathmell assures, the survival time per patient is increasing. “Patients are living longer by a good margin than they were in the interferon era; and collectively, this can’t be pointed to just one drug in the renal cell armamentarium of drugs targeting cancers.”
Every drug opens up a new therapy choice for some patients. If approved, tivozanib will provide another option for patients with RCC as there’s always a patient that can match up better with a new drug than an already existing drug. “The goal is to achieve a long term of disease control without completely destroying quality of life,” says Rathmell.
Is PFS a clinically meaningful endpoint in cancer clinical trials, especially when crossover is part of the study design? Stay tuned - it will probably be a nail biter. First we'll see the FDA's review of the file and then we have ODAC on May 2nd.
By John McCleery, Managing Editor, OBR
New entrants into the oncology marketplace can distinguish themselves in a number of ways. One approach is to be first-in-class with a novel mechanism of action (MOA) – this is true for over one-half of the pipeline agents in pivotal development. Another development strategy is to target an indication with high unmet need, offering a valuable new treatment approach for patients with few options. Of course, all cancer indications have rather high unmet need compared with non-oncology indications due to the life-threatening nature of the disease; however, among the numerous cancer indications, the level of need may vary considerably. Among 69 pipeline agents that are in pivotal clinical development in oncology, only one-third of ongoing pivotal studies with pipeline agents are being conducted in indications with relatively high unmet need1 (Figure 1). These indications include relapsed refractory acute leukemias; second-line therapy for gastric cancer, pancreatic cancer and hepatocellular carcinoma; and fourth-line therapy for multiple myeloma and renal cell carcinoma.
Areas of unmet need may be growing as clinical research continues to identify new biomarkers that identify mechanisms that drive a particular patient’s tumor or that confer a worse prognosis. Patients in these categories may have treatment options in the available armamentarium, but those options may be less effective than a targeted agent directed at their tumor’s specific molecular profile, making this a new area of unmet need. However, despite the oncology community’s growing focus on personalized medicine, only 22% of pipeline agents currently in pivotal trials are being developed in a biomarker-defined patient population. Novel biomarker populations that are currently being studied in pivotal trials in oncology are summarized in Table 1. Many more are being evaluated in early-stage exploratory studies and could begin to affect the market in the next five years.
Through extensive analysis based on a combination of product and indication attributes,2 Kantar Health has evaluated and rated these 69 pipeline agents to identify those with the greatest likelihood of clinical impact in the next three years. Five pipeline agents have emerged as having significant “breakthrough” potential, based primarily on the merits of their clinical data and independent of target indication attributes. We believe these five agents have the potential to dramatically improve efficacy outcomes and may transform the standard of care in their target indications.
1. Dabrafenib + Trametinib
GlaxoSmithKline is developing the BRAF inhibitor dabrafenib and the MEK inhibitor trametinib in melanoma patients with the BRAF-mutation. As monotherapies, neither dabrafenib nor trametinib raises eyebrows. In combination, however, they may be a force to be reckoned with. In a randomized Phase II trial, the combination significantly prolonged progression-free survival (PFS) compared with dabrafenib alone while also reducing the incidence of key toxicities associated with this class of agents. These data are currently being explored in Phase III trials, but Kantar Health anticipates the National Comprehensive Cancer Network (NCCN) will recommend use of the combination regimen based on the highly positive randomized Phase II data, giving this combination a near-term impact on the melanoma market (approval of each agent as a monotherapy is expected in 2013).
The Bruton's tyrosine kinase (Btk) pathway is a downstream mediator of the B-cell receptor pathway, making ibrutinib a novel mechanistic approach to the treatment of B-cell malignancies. Ibrutinib has demonstrated robust activity in heavily pretreated chronic lymphocytic leukemia (CLL) and mantle cell lymphoma (MCL). Furthermore, ibrutinib is amazingly well-tolerated, with very few serious adverse events. The lack of significant toxicity opens the door for it to be combined with current standards of care. Capitalizing on the early efficacy signals, Pharmacyclics and Johnson & Johnson are pursuing multiple pivotal trials of ibrutinib in relapsed/refractory CLL and MCL.
The PI3K pathway undoubtedly represents one of the hottest drug targets in oncology. Idelalisib is the most advanced of the isoform-specific PI3K inhibitors, and it made a big splash when it produced high response rates and PFS in heavily pretreated CLL and non-Hodgkin’s lymphoma (NHL) patients in early trials. Although it’s not without toxicity, the observed toxicities don’t overlap with those associated with current standards of care in CLL and NHL, leaving open the possibility for idelalisib to be combined with existing regimens. Following its 2011 acquisition of Calistoga, Gilead has charged ahead with the development of idelalisib, initiating four Phase III trials in relapsed/refractory CLL and indolent NHL.
Nivolumab (anti-PD-1 monoclonal antibody, BMS) demonstrated impressive activity in a large, first-in-man Phase I trial, with durable objective responses observed in several solid tumors. On top of these efficacy outcomes, nivolumab impressed with its rather tolerable safety profile. From this single (albeit large) trial, Bristol-Myers Squibb plowed ahead, initiating five Phase III trials in three tumor types (melanoma, NSCLC and renal cell carcinoma). The strong Phase I data, good safety profile, broad applicability across multiple tumors and the possibility for a predictive biomarker all contribute to nivolumab’s potential breakthrough status.
5. Palbociclib (PD-0332991)
PD-0332991 (palbociclib; Pfizer) wowed the audience at the 2012 San Antonio Breast Cancer Symposium (SABCS) annual meeting with an 18+-month improvement in median PFS in a randomized Phase II trial of letrozole with or without PD-0332991 as first-line therapy for ER+ metastatic breast cancer. As a CDK 4/6 inhibitor, it will be first-in-class if it hits the market, although inhibitors of other CDKs are also in late-stage development. Although currently in pivotal development only for metastatic breast cancer, the eventual development of this agent in earlier-stage breast cancer can be foreseen, and Pfizer is also investigating the drug in a number of other solid tumors. With broad applicability and stunning Phase II data, PD-0332991 has the ability to be a breakthrough agent in the treatment of cancer.
These five agents are just the tip of the iceberg. The oncology pipeline continues to grow, and the number of novel MOAs in development offer significant promise toward improving outcomes in a variety of indications. The past couple of years have seen some significant breakthroughs in cancer management, and the next few years have great potential to similarly transform care and outcomes.
1. The level of unmet needs was calculated based on the number of existing treatment options in the target indication (with fewer treatment options correlating with higher need), the rate of treatment (with high non-treatment rates correlating with higher need), and efficacy outcomes achieved with current options (with short survival times correlating with higher need). High overall scores for the sum of these three attributes were considered to be a high unmet need, whereas low overall scores were considered to be a low unmet need.
2. Attributes that were scored included mechanistic innovation, strength of prior published data, target indication population size, and level of competition. For more in-depth information on Kantar Health’s scoring methodology, please e-mail email@example.com.
By: Stephanie Hawthorne, PhD, Director, Kantar Health
After AACR 2012 in a piece we called “A Science Geek’s Thoughts on the 2012 AACR Annual Scientific Meeting”, we wrote about three larger-picture, emerging themes that we felt worth discussing: epigenetics, tumor cell metabolism, and cancer stem cell biology (science that leads to the exploitation of the stem cell for novel drug targets for therapeutic interdiction, not stem cells themselves as therapies). While these themes certainly repeated themselves in 2013, epigenetics and cancer stem cell research have matured, while tumor cell metabolism seems to still be searching for its killer app (although we continue to hear that Agios is very close to putting a molecule into the clinic after having recently published data on two inhibitors of IDH1 and IDH2 in the April 4 edition of Science). We also would add another concept that has caught our eye in the past, and that is the promise of harnessing the body’s immune system to fight cancer and produce long-lasting benefit. While this concept is hardly new as Bristol-Myers’ (BMY, NC) Yervoy (ipilimumab,) has been approved since May of 2011; however, Yervoy has not been, per se, a biotech pure-play. Further, the drug’s tolerability issues will likely limit its uptake and duration of use. This year, Bristol’s nivolumab has reset the bar for long-term benefit along with tolerability, even though its first-in-human study was published in the Journal of Clinical Oncology in 2010; likewise, nivolumab was the subject of two articles in the New England Journal of Medicine in 2012. We believe that future hindsight will mark the AACR 2013 plenary session as a harbinger for future ASCO plenary sessions. Inspiring talks were offered by Katie Couric regarding her crusade against colorectal cancer and role in creating Stand Up to Cancer (SU2C) as well as Siddhartha Mukherjee of Columbia University and author of The Emperor of All Maladies (a must-read, in our view). On a personal note, the AACR program organizers are to be commended for putting together a plenary agenda that was decidedly a non-snoozer.
Immunotherapy: When the Body Cures the Body
The common theme amongst the various presentations at the AACR plenary (as well as others throughout the conference) concerned complete eradication of tumors and/or durable remission of effect. Perhaps most emblematic of this was the talk given by Suzanne Topalian of The Kimmel Cancer Center at Johns Hopkins University. Dr. Topalian, the lead author of the 2012 NEJM article, discussed the biology of the PD-1/PD-L1 axis and the differences between it and the CTLA-4 axis, the target of Yervoy. Nivolumab blocks the binding of both PD-L1 and PD-L2 to the PD-1 receptor. She then discussed findings of both the JCO and NEJM articles, which examined the use of nivolumab in non-small cell lung cancer (NSCLC), melanoma, kidney, colon or prostate cancer. What most impressed us, and immediately caused us to crystallize the impact on patients of Dr. Topalian’s data, was the slide she showed on the durability of response of three patients who has ceased treatment with nivolumab. Dr. Topalian provided vignettes of three patients-one with colorectal cancer (CRC), one with renal (kidney) cancer (RCC), and one with melanoma-showing that patients can remain in complete or partial response 4-5 years into their therapy. The patient with CRC remains in complete response just short of four years into follow up after stopping treatment shy of one year. The RCC patient received therapy for approximately six months had a partial response (PR) as best response shortly thereafter, then went on to produce a complete response (CR) at roughly 3.5 years, and at nearly 4.5 years remains in CR. Finally, Dr. Topalian’s example of the melanoma patient was a bit more complicated. The patient achieved a PR just shy of two years on therapy, was diagnosed with a new metastasis just over three years after diagnosis, was retreated to a PR at year four, and was still in PR at roughly 4.5 years. Dr. Topalian also showed a slide that she showed at ESMO 2012. The key data point on this slide is a bullet that indicated 28 of 54 responses lasted more than one year in patients who had more than one year of follow-up. Grade 3-4 adverse events were a very reasonable 15%, and treatment-related mortality was only 1%. Perhaps most exciting was the slide shown by Dr. Topalian that described a patient with squamous non-small cell lung cancer, a tumor type that is widely thought to be non-immunogenic. This particular patient responded to anti-PD-1 antibody therapy after an immediately previous treatment with epigenetic therapy of 5-azacitidine (Vidaza, CELG, MO) and entinostat (Syndax, private, NC), an inhibitor of HDAC 6. Dr. Topalian indicated that this patient has recently completed two years of anti-PD-1 therapy. Moreover, Dr. Topalian indicated that there are five more patients like the one in the example she cited who have achieved durable responses after prior treatment with epigenetic therapy. While the number of patients is small, the 100% response rate in this population is impressive. We eagerly await the presentation of these data at an upcoming scientific conference. In terms of biomarkers and other correlates of response, Dr. Topalian showed a number of slides that spoke to the level of PD-L1 in the tissue as the most highly predictive of response in patients. In her findings, 13/31 patients with detectable levels (>5% of tumor cells) of PD-L1 achieved a PR or CR, whereas 0/18 patients with undetectable levels of PD-L1 did so. This has intuitive appeal to us, as it likely means that those patients without detectable PD-L1 have some other immune mediator of their disease. She also showed that patients whose disease had evidence of tumor infiltrating lymphocytes (TILs) experience improvements in overall survival. Dr. Topalian closed with suggestions as to how best to improve the therapeutic profile of these remarkable antibodies. These involved the blockade of lymphocyte activation gene 3, or LAG-3, and the blockade of interleukin-10 (IL-10), a cytokine known to tamp down the immune response to cancer cells, among other functions. We look forward to future developments with not only anti-PD-1 and PDL-1 therapy, as well as others that will enhance their effects.
Apart from the very cool science that was on display during Dr. Topalian’s talk, we were struck by the thought of the effect on patients as the knowledge of the durable effect of PD-1 blockade make its way across patient blogs and discussion groups. What patient would not want to be cancer-free for three, four, five years or more? Anti-PD-1 therapy could make this possible. Even if the duration of response diminishes over time, as it inevitably will, we believe the pull of durable response could eventually make drugs in the class the largest selling oncology drugs extant. The robust effect, broad applicability, and acceptable tolerability could result in sales of the drugs that would put sales of Avastin (~$6 billion) to shame. Were Vidaza, Dacogen, or SGI-110 to hitch their respective wagons to the PD-1 bandwagon, it would likely be an unmitigated (and currently unanticipated) boon to the top line for these drugs. We take a deeper dive into the science later in this report.
Engaging Scientists in a Discussion about Stem Cells with Slick Graphics
The JMP Securities Award for the Sexiest Scientific Graphics goes to Hans (a/k/a “Very”) Clevers of the Hubrecht Institute in Utrecht, The Netherlands. Dr. Clevers talk was peppered with some of the slickest scientific animations and graphics we have ever seen this side of the SciFi Channel. If every science teacher in America had access to graphics like this we would end the crisis in science education virtually overnight. Sadly, we speculate that Dr. Clevers presentation must have been hideously expensive and that he is the fortunate recipient of gobs of grant money. Having said that, Dr. Clevers scientific findings were as informative as his graphics were elegant. His work focuses on Lgr5 cells in the intestinal crypt, where stem cells demonstrate significant proliferation; a gut stem cell can transform itself into an enterocyte or goblet cell in the space of roughly five days then die within days, a huge turnover rate compared to other cells in the body. Most of Dr. Clevers’ talk concerned his work on the creation of a gut model he and his research team derived from Lgr5 cells, which they can use in vitro and in vivo; however, we were encouraged by his further validation of some of our favorite cancer stem cell targets, including wnt, frizzled, LRP, Lgr and Rspo. All of these components comprise a sophisticated signaling complex on the exterior of the cells of the colon. In colorectal cancer, the wnt signaling pathway is locked in the “on” position, while Rspo binding to Lgr activates Lgr and, in turn, Lgr then binds to LRP and amplifies wnt-mediated signaling. The Lgr5 cells can be readily located as they are in contact with so-call Paneth cells, which also are found in the intestinal crypt, and appear to nurture the Lgr5 cells. In Dr. Clevers’ words, every cell that touches a Paneth cell is a stem cell, and Paneth cells secrete gobs of wnt3 as part of their support role to the Lgr5 cell. Dr. Clevers also showed that inhibition of the protein porcupine, can block the secretion of wnt. Novartis (NVS, NC) is developing a small molecule inhibitor of porcupine, referred to as LGK974. In addition, privately-held OncoMed Pharmaceuticals is developing OMP-54F28, an antagonist of the wnt signaling pathway.
Session on Epigenetics Packs the House
The Sunday morning plenary session on epigenetics, hosted by Robert Copeland, PhD, Chief Scientific Officer of privately-held Epizyme, played to a packed house. While there were multiple sessions held during the five days of AACR on epigenetics, none was as forward-looking nor as absolutely packed as this one (an estimate of the number of attendees to the session exceeded 10,000). Peter Laird of the USC Epigenome Center in Los Angeles led the speaker’s roster and provided a fascinating discussion of how the epigenetic signature of cancer cells is vastly different than that of normal cells in terms of both the number of genes silenced as well as turned on (or “de-silenced”). Dr. Laird showed that there is massive (“focal”) hypermethylation at CpG islands (the start sites for gene transcription) across a wide variety of tumor types. These regions of hypermethylation are in areas of gene expression that would normally be associated with genes that would either prevent the outbreak of cancer or provide signals to the cancer cell that would normally tell cancer cells to die. What was particularly fascinating about Dr. Laird’s talk was not only did he provide us with optimism towards finding additional validated drug targets in these hypomethylated regions, but also tied mutations in kinase signaling pathways (such as BRAF) to changes in the epigenome. Fearless prediction: more of these integrative, or “systems biology” studies, will continue to be done (indeed, must be done) as we advance our assault against cancer. In our opinion, not only will mutations or changes in the regulation of the epigenome be tied to mutations or up-regulation of cellular growth pathways and vice versa, but simultaneous changes in aspects like cellular metabolism, immune evasion, and stem cell-like properties will be tied together in a more holistic picture going forward. During the same talk, James “Jay” Bradner from the Dana-Farber Cancer Center gave an outstanding talk (though similar to the one we had seen at ASH 2012) about his work on the probe molecule, JQ-1, and its role in the regulation of the MYC gene, one of the most commonly over-expressed genes in all of cancer (behind only p53). JQ-1 (and presumably chemical matter from privately-held Constellation Pharmaceuticals) blocks a key regulator of MYC activation, bromodomain 4 (brd4), a key regulator of MYC gene transcription. Dr. Bradner’s slides showed the pattern of benefit that is now becoming familiar to those of us groupies of epigenetics-that of a flat-line pattern of tumor growth (better yet, death). Dr. Bradner showed some amazing CT scans of a patient with KRAS-mutated lung cancer who experienced a near-total regression of a massive tumor. Dr. Copeland finished the session with a comprehensive discussion of the two targets of focus for Epizyme, DOT1L, mutated in mixed-lineage leukemia (MLL), and EZH2, a very hot gene target that has generated a lot of interest due to its dysregulation in a variety of tumor types (NHL, breast, prostate, colon, gastric and bladder, among others) and which has drawn the attention of a number of large pharmaceutical and biotechnology companies. Again, the shape of Dr. Copeland’s Kaplan Meier curves was virtually identical to those of the other speakers-flat-to-downward sloping, rather than right-shifted as has been seen with so many other interventional therapies in oncology. We will be producing a fuller report on the epigenetics space in the near-term.
Early Signs that Epigenetics Holds the Keys to Immunotherapy & Vice Versa
In addition to the separate plenary sessions on immunotherapy and epigenetics, we noted several poster presentations that hinted strongly at the ability to boost anti-tumor immune responses through the epigenetic reprogramming, primarily through the use of hypothmethylating agents. Some of these presentations were in prior notes by us relation to Astex (ASTX, MO) in which tumor specific antigen expression, namely of NY-ESO1 and MAGE-A, in peripheral blood mononuclear cells (PBMCs) was found to increase in response to global gene demethylation in AML/MDS patients treated with SGI-110. In a separate poster, Dacogen induced demethylation in AML cell lines correlated with increased NY-ESO1 and MAGE-A protein expression, further supporting the likelihood that T cell activation plays a role in the mediation of tumor responses by hypomethylating agents (HMAs). Future experiments from Griffiths et al. will measure if tumor antigen specific CD4+/CD8+ T cell responses are in fact elicited by HMA therapy. Meanwhile, a forthcoming prospective study will test the hypothesis in vivo, assessing the combined effect of Dacogen plus the NY-ESO1 vaccine CDX-1401 (CLDX, NC) in patients with AML/MDS. One has to take the good with the bad when it comes to HMA induced gene expression, however. This was evidenced in part in a poster presented by Edward Kadel et al of Genentech (Roche, NC) showing that differential expression of PD-L1 mRNA in NSCLC cell lines was regulated by the level of the promoter by DNA methylation. Induction of PD-L1 expression was most responsive to treatment with Dacogen in cell lines where baseline methylation was high and gene expression low. One initial interpretation of these data might be that DNA demethylation and immune checkpoint activation are antagonistic and that without PD-1/PD-L1 blockade, HMA activity is actually undermined by immnosuppressive mechanisms. Such was the conclusion reached by Garcia-Manero et al in a poster presented at last ASH last December in which the expression of PD-1, PD-L1, PD-L2 and CTLA4 were all induced in the PBMCs of AML/MDS patients treated with low dose Dacogen, or Vidaza in combination with Revlimid (CELG) or vorinostat. On the other hand, it is possible that the induction of PD-L1 expression though HMA exposure acts as primer or homing mechanism for activated T cells. Recall in relation to the plenary talk by Dr. Topalian that tumors responding to nivolumab exhibited relatively higher levels of PD-L1 expression compared with non-responders. Taken together, we believe that future clinical studies of epigenetic therapy used combination with PD-1/PD-L1 antibodies (given perhaps in sequence) will indeed show these regimens to be synergistic in their activities.
In our view, AACR is an exceptional meeting in that it provides the scientific and business community with an opportunity to view the up-and-coming drugs and drug targets of the future. To use a sports analogy, it is akin to being able to see all the college football or basketball teams and their players over the course of just a few days. While there is a lot of information for investors to digest, it also provides incredible opportunities to select the stars of the future, without the distraction of running to numerous sessions to see clinical data. The latter purpose is best served by ASCO, coming in June. However, there is a certain pride (and profit) to be had when one can say “I saw that drug when it was still in Phase II trials” like a promising rookie. We hope the character of the AACR conference never changes.
by Mike King, Managing Director and Senior Biotechnology Analyst, JMP Securities
On Sunday at the American Association for Cancer Research (AACR) 2013 annual meeting, Jose Baselga, M.D., Ph.D., physician-in-chief at the Memorial Sloan-Kettering Cancer Center, presented new data from the Phase III EMILIA trial (Abstract LB-63) that suggested a relationship between tumor biomarkers and efficacy of Kadcyla® (TDM-1, ado-trastuzumab emtansine; Roche/Genentech). Kadcyla is a novel antibody-drug conjugate that comprises a potent microtubule polymerization inhibitor conjugated to the Herceptin® (trastuzumab; Roche / Genentech) monoclonal antibody via a highly stable linker. Kadcyla is designed to take advantage of the targeted nature of the antibody to selectively deliver the cytotoxic agent to HER2+ breast cancer cells. The EMILIA trial evaluated Kadcyla monotherapy versus Xeloda® (capecitabine, Roche) plus Tykerb® (lapatinib, GlaxoSmithKline) (XT) in relapsed HER2+ metastatic breast cancer patients. The positive results led to its regulatory approval in the United States in February 2013. Roche’s Herceptin, currently dominates the HER2+ market, with about 85% utilization in front-line HER2+ metastatic breast cancer patients. Although Tykerb is approved in the second-line in combination with Xeloda, about half of U.S. patients are rechallenged with Herceptin plus chemotherapy, leaving Tykerb for later-lines of therapy.
Baselga and colleagues conducted a biomarker analysis as part of the EMILIA study to determine whether variable expression of HER2, HER3, EGFR, PTEN and mutational status of PIK3CA gene impact efficacy of Kadcyla or Tykerb plus Xeloda. Patients in both treatment arms were evaluated according to intratumoral HER2 mRNA levels (at least median HER2 mRNA level versus less than median HER2 mRNA level). Patients with tumors expressing higher levels of HER2 mRNA (median or higher) derived greater overall survival benefit from Kadcyla (34.1 months) versus patients with tumors expressing lower levels of HER2 (26.5 months). In contrast, there was no difference in overall survival benefit in patients treated with Tykerb/Xeloda when stratified by HER2 mRNA levels (24.8 months for high levels versus 23.7 months for low levels). Similar analysis based on mRNA expression levels of HER3, EGFR and PTEN revealed no relationship between these biomarkers and efficacies of Kadcyla or Tykerb/Xeloda regimens. However, the mutational status of PIK3CA was discovered to be associated with improved outcomes when treated with Kadcyla but not Tykerb/Xeloda.
The PIK3CA gene encodes the alpha isoform of PI3K catalytic subunit, which is frequently mutated in HER2+ breast cancers and is thought to represent one of the resistance mechanisms to HER2-targeted therapy. PIK3CA-activating mutations may impact efficacy of anti-HER2 therapies as PIK3CA lies downstream from HER2 in the signal transduction pathway. Therefore, efficacies of these regimens were examined in patients according to PIK3CA-activating mutational status. Indeed, for patients in the Tykerb/Xeloda arm, overall survival of patients with PIK3CA activating mutations was worse compared to patients with PIK3CA wild-type tumors (17.3 versus 27.8 months). In contrast, overall survival appeared unaffected by PIK3CA mutational status in patients in the Kadcyla arm: in patients with mutated PIK3CA Kadcyla improved median progression-free survival (mPFS) from 4.3 months to 10.9 months, and in patients with wild-type PIK3CA Kadcyla improved mPFS from 6.4 to 9.8 months. These data suggest that Kadcyla may be able to bypass a common HER2 resistance pathway that has confounded anti-HER2 therapies and offers a new treatment option to women who would have otherwise not fared as well from other HER2-targeted therapies. Prior data has suggested no significant association between PIK3CA mutation and Herceptin efficacy, similar to what is observed here for Kadcyla, suggesting that this resistance mechanism may primarily have implications for use of Tykerb or other HER2 family tyrosine kinase inhibitors.
These results are thought-provoking and set the stage for future biomarker-driven analyses to guide choice of therapy upon relapse, although it’s premature to begin integrating into clinical practice at this time. However, these data could support tighter definitions of patients eligible for therapy with these agents, possibly by raising the threshold for HER2 positivity or by using PIK3CA status to help physicians make an informed decision on whether Tykerb or Kadcyla is a better second-line option for individual patients. The field has grown adept at incorporating biomarker analyses into initial patient diagnosis to guide treatment algorithms, but we are only beginning to understand the potential importance of re-testing patients throughout the progression of disease to further optimize and individualize care.
By: Stephanie Hawthorne, PhD, Director, Kantar Health; Tatiana Spicakova, PhD, Consultant, Kantar Health; and Greg Wolfe, PhD, Senior Consultant, Kantar Health