Below is my ASCO 2014 preview (better late than never…). I tried to make this recap as comprehensive as possible but it is practically impossible to cover all the interesting stuff (let me know if I missed anything dramatic). Unlike last year, I decided to group interesting abstracts based on mechanism of action rather than companies in order to provide a more holistic perspective. On top of attending the conference itself, I will try to attend as many analyst events as possible (this year I have Clovis, Roche, BMS and Incyte on my list) and include them in my post-ASCO write-up.
PD-1 antibodies – still promising but challenges start to emerge
PD-1 antibodies continue to dominate the scene and this ASCO will have clinical data sets for 6 antibodies from BMS, Merck, Roche, AstraZeneca, Merck-Serono and Curetech. So far, updated results in the abstracts for the 3 primary indications (lung, renal, melanoma) fell short of expectations and raise 4 significant issues:
1. Most patients do not experience long term remissions
2. Combination data with Yervoy in lung cancer suggest limited to no added benefit
3. Combination with approved and investigational agents may be too toxic
4. Contradicting results on the use of PD-L1 as a predictive marker
On a more positive note, PD-1 antibodies are still the most promising class of drugs in development for cancer and in some patients they lead to phenomenal long term remissions. In addition, there appears to be good activity across a wide variety of other tumor types including bladder, head and neck and ovarian cancer.
PD-1 for Lung cancer
BMS’ (BMY) Nivolumab + Yervoy results were uninspiring with a 17% response rate for nivo’s higher dose. This is lower than response rate seen to date with monotherapy PD-1 antibodies (20-30%) and there was no correlation between response and PD-L1 expression. The safety profile appears problematic with a discontinuation rate of 35% and 3 (6.5%) treatment-related deaths. If updated results are not materially different, it would be hard to justify advancing this combination in lung cancer, despite BMS’ intentions to do so and promising activity in melanoma.
As 1st line treatment, nivolumab demonstrated a 30% response rate and a median PFS of 7 months. Response rate in PD-L1+ patients was 67% vs. 0% for PD-L1- patients. When Nivolumab was added to standard chemotherapy regimens, response rate was 33-50%, which appears numerically better than that expected with chemotherapy alone (30%+) but the increase in response rate is not dramatic and the small sample size makes it hard to interpret the data.
Updated results in pre-treated patients demonstrated median survival of 10 months, which is encouraging but not dramatically better than what is expected in this population with other salvage therapies. The 3 mg/kg arm had better overall survival (15 months) but it is unclear whether this effect is real or just a statistical fluke. Surprisingly, PD-L1+ patients had lower survival.
Merck (MRK) will present data for MK-3475 in 1st line (all PD-L1+) and pre-treated NSCLC patients. The trials demonstrated activity of 21%-36% but this includes also unconfirmed responses. Interestingly, in pre-treated patients, PD-L1+ patients had a higher response rate than PD-L1- (24% vs. 8%).
PD-1 for Renal cancer
Nivolumab will have monotherapy and combination data. Interestingly, BMS issued a press release that included more updated data than what appears in the abstracts.
The monotherapy response rate of 20-22% is lower than that of previous reports (30%). PFS of 4 months for the 2 higher doses implies the drug is active but again, the extent of benefit over standard of care is limited. The overall survival data (25 months) is encouraging and appears significantly higher than typical survival in this patient population (predominantly 3rd line).
BMS will also present results for combination with Yervoy and approved agents (Sutent, Votrient). In both cases, response rate climbs to 43%-52% . The nivo+ Yervoy abstract reports a PFS of 9 months, which appears higher than nivo monotherapy but can also be explained (at least partially) by the less pre-treated nature of the population (23% were 1st line). Combining Nivolumab with Sutent or Votrient had toxicity issues with 60-73% Gr 3/4 rate and a 20-24% discontinuation rate.
PD-1 for Melanoma
BMS will present updated 3-year survival rate for nivolumab monotherapy and the highly anticipated combination data with Yervoy. The abstract for the monotherapy study includes an impressive survival rate of 41% at 3 years vs. 48% at 2 years. This may represent a “tail” (flattening of the curve) driven by durable responses. PD-L1 expression was associated with better survival and PFS. The combination data will be disclosed at the conference and expectations are for a 50%+ response rate.
MK-3475 will have monotherapy data, which appears in-line with Nivolumab. One abstract reports a 41% response rate and a PFS of 7 months. PD-L1 expression correlated with improved response rate (51% vs. 6%) and PFS (12 vs. 3 months). Another presentation titled “Efficacy and safety of the anti-PD-1 monoclonal antibody MK-3475 in 411 patients with melanoma” will be presented at the conference.
Additional indications for PD-1 antibodies
This year’s meeting will shed light on applicability of PD-1 antibodies in additional tumors. The most promising data will be for Roche’s MPDL3280A (Anti-PD-L1) in bladder cancer, where 10 (50%) of 20 patients with PD-L1+ tumors achieved a response. Nivolumab will have preliminary data in ovarian cancer. The abstract reports a response rate of 23% in platinum-resistant patients. Merck will present results in head and neck cancer, where activity has been observed but the abstract does not disclose any efficacy readouts. On its ASCO press release, BMS announced nivolumab received breakthrough designation for Hodkin’s lymphoma which suggests there is good activity in this tumor type as well.
New immuno-oncology programs
2 combination studies evaluating new immunotherapies with Yervoy in melanoma patients will be presented. The significance of the data is in establishing proof of concept for the synergy with an immune checkpoint. Going forward both programs will likely be pursued with PD-1 antibodies.
After months of speculations, Incyte (INCY) will report combination data for its IDO inhibitor (INCB024360) with Yervoy in melanoma patients. The abstract included information for the 300mg dose (too toxic) and for 8 patients in the 25mg dose, where a 38% response rate is reported. This response rate is numerically higher than Yervoy’s typical response rate of 10-15% but durability of response appears short. At the conference, Incyte will provide updated results for the 25mg cohort and report response rate with a higher dose (50 mg). In order to maintain the positive sentiment around this program, response rate needs to stay above 30% with evidence of durable responses, a hallmark of effective immunotherapy.
Amgen (AMGN) will also present for the first time results from a study evaluating it cancer vaccine Talimogene laherparepvec (T-VEC) with Yervoy in melanoma patients. The abstract reports 41% response rate including an impressive 24% CR rate. Similarly to the IDO data set, investor attention at the conference will be on durability of responses.
Next-gen EGFR inhibitors – Clovis vs. AstraZeneca
Investors are closely watching the face-off between AstraZeneca (AZN) and Clovis (CLVS), which will present data for their respective EGFR inhibitors. According to the abstracts, both agents have promising efficacy in NSCLC patients who failed approved EGFR inhibitors (Tarceva, Iressa) and harbor the T790M EGFR mutation. Both drugs recently received breakthrough designation from the FDA.
AstraZeneca is expected to report improved activity for AZD9291, with a response rate of 64% in 89 patients with T790M+ tumors according to the abstract. Durability seems strong as 97% of patients who had confirmed responses were still on treatment as of the cutoff date. Clovis’ abstract for CO-1686 does not provide new meaningful data and includes the previously presented response rate of 67% in just 9 patients. At the conference, investigators will have efficacy data for more patients (40+), which are expected to be positive based on the breakthrough designation CO-1686 received last week.
The two drugs appear comparable in term of efficacy (with the caveats of cross-trial comparison) but their safety profiles are differentiated. AZD9291 leads to more classic EGFR toxicities (rash, GI) and more importantly, 5 cases of interstitial lung disease-like symptoms were noted. For CO-1686, investors’ primary concern is QTc prolongation and whether it becomes clinically meaningful with prolonged treatment. Going forward and given the similar efficacy profile, a favorable safety profile may become an important advantage in this patient population, especially if the drugs reach first line EGFR+ NSCLC.
FGFR inhibitors – Clovis still in the lead
This year’s meeting will have several presentations of FGFR inhibitors. Despite strong scientific evidence for the implication of the FGF pathway in many tumor types, FGFR inhibitors have demonstrated limited activity as monotherapy. The only exception to date is Clovis’ lucitanib, a dual FGFR/VEGFR inhibitor that demonstrated the promising results in a small data set of breast cancer patients with “FGF-aberrant” tumors. Lucitanib’s efficacy is attributed to the parallel inhibition of the VEGF and FGF pathways.
At ASCO, Clovis will present updated results which may corroborate the initial efficacy signal in breast cancer and potentially other tumor types. The abstract discloses the previously reported 50% response rate in 12 heavily pre-treated breast cancer patients.
Other companies will have data for their selective FGFR programs. AstraZeneca will report data for AZD4547 in 2 trials (gastric cancer and solid tumors). The drug generated responses in 1/38 and 1/7 patients with FGF aberrations, respectively. Novartis (NVS) will report phase I data for BGJ398 in squamous NSCLC patients with FGFR1 amplification. The abstract reports encouraging preliminary signs of activity with 4 responses among 21 patients. J&J (JNJ) will present phase I data for JNJ-42756493, including a subset of FGF pathway aberrations. According to the abstract, of 8 biomarker positive patients, two urothelial cancer patients achieved a response.
Hutchison MediPharma will report data for its dual FGFR/VEGFR inhibitor, which should be viewed as a direct competitor to lucitanib. Of 17 patients who received an improved formulation of the drug, 4 achieved a response. It is not disclosed whether these patients had FGF aberrations so it is unclear whether their responses are mediated by FGFR inhibition.
CD38 antibodies – Sanofi emerges as a worthy opponent
This year’s meeting will hae updated results for the 2 leading CD38 antibodies for multiple myeloma.
Genamb’s (GEN)/J&J’s daratumumab will have 2 oral presentations as monotherapy and in combination with Revlimid. The abstract for the monotherapy study reports mixed efficacy data for 2 expansion cohorts (8 and 16 mg/kg). The combined response rate for the two cohorts was only 19% (compared to 33% for ≥4 mg/kg at ASH 2012). Nevertheless, the 16 mg/kg arm had a response rate of 46%, which suggests a dose dependent response but sample size is small (13 patients). The daratumumab+Revlimid combination continues to look active with responses in 8 (72%) of 11 patients.
Sanofi’s (SNY) /Immunogen (IMGN) SAR650984 will also have data as monotherapy or in combination with Revlimid. The abstract for the monotherapy study reports a 24% response rate including a 33% response rate for doses ≥ 10mg/kg. Of note, 2 patients achieved a complete response whereas in daratumumab’s data set there were none. Efficacy in combination with Revlimid was 58% (7/12 patients). This is numerically lower than daratumumab+Revlimid but the small sample size makes it hard to compare the 2 drugs. In addition, patients in the SAR650984 trials appear more heavily pre-treated.
Although both agents are active in multiple myeloma, daratumumab has a clear development lead with 2 phase III trials (still not active) and a pivotal phase II trial in relapsed/refractory patients.
Antibody-drug conjugates- strong data in lymphoma and ovarian cancer
This year’s ASCO will have a record number of ADC presentations, predominantly from Roche. Most of the ADCs to be presented this year are powered by Seattle Genetics’ (SGEN) and Immunogen’s technologies.
Roche will present results from the ROMULUS trial, which compares its CD22 ADC (pinotuzumab) vs. its CD79b ADC (polatuzumab) when added to Rituxan in 2 subtypes of non-hodgkin’s lymphoma (DLBCL and follicular lymphoma). Both regimens demonstrated strong response rate of 54%/51% in DLBCL and 67%/60% in FL. However, both agents have tolerability issues with 36% of patients experiencing serious side effects and a high discontinuation rate. There were also 2 deaths related to pinotuzumab. Before pursuing phase III, Roche will probably explore regimens to minimize toxicity, especially peripheral neuropathy.
2 ADCs targeting CD19 will report data in NHL, both are expected to report a 40+% response rate. Sanofi’s SAR3419 (utilizes Immunogen’s technology) will have phase II data in DLBCL, which appears surprisingly good based on the abstract. Of 41 patients, 18 (43.9%) achieved a response including a CR rate of 12.2%. Although patients were relatively less heavily-pretreated (only 31.7% had received ≥ 3 prior regimens), this response rate is impressive and superior to previous updates. Seattle Genetics will report phase I results for SGN-CD19A in NHL. Of 20 evaluable patients the abstract discloses 8 (40%) responders, 6 of whom achieved a CR. Updated results may include stronger efficacy with further dose escalation.
Roche will present phase I results for its anti-NaPi2b (DNIB0600A) and anti-mesothelin (DMOT4039A) ADCs. Both agents utilize Seattle Genetics’ technology and appear active in ovarian cancer. The abstract for DNIB0600A reports a strong response rate of 41% at the highest dose in 17 patients with NaPi2b+ tumors. Evidence of anti-tumor activity was seen in 7/21 NaPi2b+ lung cancer patients. Roche recently started a phase II in platinum resistant ovarian cancer that compares DNIB0600A versus Doxil. As Doxil has a typical response rate of ~10% and PFS OF ~3 months, updated results at ASCO will help to evaluate likelihood of success. The abstract for DMOT4039A reports an encouraging 30% response rate in platinum-resistant ovarian cancer as well as additional signs of anti-tumor activity. There was also 1 PR in 19 pancreatic cancer patients. Bayer has a mesothelin ADC (BAY 94-9343 ,utilizes Immunogen’s technology) which demonstrated activity in mesothelioma but no responses were noted for in small cohort of ovarian cancer patients. BAY 94-9343 is being tested in an expansion cohort of ovarian cancer patients.
Immunogen will present updated phase I results for 2 proprietary programs. Although the focus will be on ways to overcome toxicity, data may include efficacy updates. For IMGN853 (anti-FRa), Immunogen will present a modified dosing regimen designed to minimize ocular toxicity thereby allowing maximal exposure. According to the abstract, higher doses led to clinical benefit in 10/24 but actual response rate is not disclosed. IMGN529 (anti-CD37) will have phase I results in NHL. Although signs of activity were seen, the dosing protocol had to be amended to include steroids in order to prevent neurtopenia associated with cytokine release.
Additional ADCs which utilize Seattle Genetics’ technology will have data this year: ABT-414 (anti-EGFRvIII) in glioblastoma, DSTP3086S( anti-STEAP-1) in prostate cancer and MLN0264 (anti-GCC) in colon cancer.
Oral agents for CLL – Pharmacyclics under pressure
The biggest event in CLL this year is phase III data for Pharmacyclics’ (PCYC) Imbruvica (partnered with J&J). The trial, which compares Imbruvica to Arzerra in relapsed/refractory patients was already stopped after a positive interim analysis. While the ASCO abstract is not published, the EHA abstract for this trial is available and reveals that Imbruvica generated dramatically better response rate (43% vs 4%), PFS (not reached vs. 8 months), and overall survival (HR=0.43). Despite this phenomenal data set, the market is growing increasingly skeptical about Imbruvica’s commercial potential following a weak launch. It remains to be seen whether ASCO will be the turning point for Imbruvica’s sales trajectory.
Roche will present updated results for ABT-199, which is viewed as Imbruvica’s strongest competitor, as monotherapy and in combination with Rituxan. The monotherapy abstract reports an impressive response rate of 79% (22% CR), which appears superior to Imbruvica. The drug’s safety profile was improved thanks to the new dosing regimen that appears to have resolved TLS (tumor lysis syndrome) concerns.
Next-gen sequencing (NGS) gaining momentum – Foundation Medicine is the clear leader
Multiple groups will report experience with NGS as a tool to identify actionable mutations in order to guide treatment decisions. Many reports include Foundation Medicine’s (FMI) technology and exemplify its disruptive potential. Importantly, NGS appears more sensitive and accurate than other technologies, and can identify effective treatments that are otherwise not commonly used for a given tumor type. The abstracts include cases in which FMI’s system identified ALK mutations in tumors originally characterized as ALK-negative. These patients were then treated with ALK inhibitors and most responded. Another group reported the identification of a novel BRAF mutation in pancreatic cancer that may confer sensitivity to BRAF/MEK inhibitors. A third abstract describes the identification actionable mutations in tumors that were considered negative for all commonly used diagnostic tests.
Jakafi in pancreatic cancer and PV
Incyte will present phase II and phase III data for its Jakafi (Jak1/2 inhibitor) in pancreatic cancer and PV (polycythemia vera), respectively.
The pancreatic cancer trial did not show an overall survival benefit for the entire patient population, but a subset analysis in patients with high CRP levels (n=60) demonstrated a survival benefit. At ASCO, investor focus will be on the reliability and robustness of CRP levels as a selection criterion for the upcoming phase III.
In the PV trial, Jakafi achieved the primary endpoint in a highly significant manner, leading to a response in 21% of patients vs. 1% in the control arm. Although this response rate is lower than expected, responses were highly durable and most patients derived some sort of clinical benefit with Jakafi. These results will likely lead to approval in PV, which should translate to accelerated growth in Jakafi sales (already factored in by investors).
Additional notable abstracts
MEK162 – Novartis will present multiple data sets for, MEK162, a MEK inhibitor licensed from Array (ARRY) Biopharma. The most intriguing abstract is for a combination with LEE001 (CDK4/6 inhibitor) in NRAS mutated melanoma. MEK162, is currently in phase III in NRAS melanoma based on a 20% response rate. Preliminary results for the combination include a response rate of 43% but most responses were not confirmed so more follow up is needed. Another study evaluated MEK162 with BYL719 (PI3K-alpha inhibitor) and demonstrated preliminary signs of activity in KRAS-mutated ovarian cancer (3 of 4 patients had a response). Attempts to combine MEK inhibitors with PI3K inhibitors encountered limited success to date due to a narrow therapeutic window. Using an isoform-selective PI3K inhibitor may finally allow simultaneous inhibition of the 2 pathways.
CSF1R inhibitors – Roche and Plexxikon (Daiichi Sankyo) will present impressive efficacy for their respective CSF1R inhibitors in a rare joint tumor (PVNS) characterized by constant activation of CSF1R. Interestingly, although Roche targets CSF1R via an antibody and Plexxikon utilizes a small molecule kinase inhibitor, the clinical effect appears comparable between the two agents. Roche’s antibody (RG7155) led to responses in 7/10 (70%) patients, who were not amenable to surgery. Responses were associated with symptomatic improvement and appear durable (up to 17 months). Plexxikon’s PLX3397 led to responses in 7/11 (64%) of patients, which were also associated with symptomatic improvement.
MEK/BRAF combination in melanoma – GSK will present phase III results from COMBI-d, which evaluated Tafinlar/Mekinist vs. Talifnar in 1st line BRAF+ melanoma. Although the PFS benefit was marginal (2 weeks), preliminary survival analysis provides reason for optimism with a HR of 0.63 (p=0.023). If the survival benefit is maintained in the final analysis, MEK/BRAF combination will likely be the standard of care for 1st line BRAF+ melanoma. Later in the year, Roche will report phase III results for its MEK/BRAF combination in the same setting.
Combination regimens for BRAF+ colorectal cancer – At ASCO investigators will present data for combination regimens in BRAF+ colorectal (CRC) cancer. In contrast to experience in melanoma, BRAF+ CRC is not sensitive to BRAF inhibitors (5% response rate). Three abstracts report intriguing preliminary activity in combination with other agents. One abstract reports that 4/5 patients achieved durable responses when treated with Zelboraf combined with Erbitux (anti-EGFR) and chemotherapy. Another study evaluates GSK’s MEK/BRAF combo (Tafinlar/Mekinist) with Erbitux and the abstract reports responses in 4/6 evaluable patients. A third study evaluates LGX818 (Novartis’ BRAF inhibitor) and Erbitux with or without BYL719 (PI3K-alpha inhibitor). Of 18 patients, there were 3 PRs and 2 very durable cases of stable disease (1 year). At ASCO, investigators will present updated results with the triple arm.
MET inhibitors – Despite failures of Roche’s Metmab and Arqule’s (ARQL) tivatinib, the industry is still very interested in Met inhibitors based on the large amount of active programs. Although some abstracts report limited activity even in biomarker-defined tumors, there are preliminary encouraging signs in some of the trials that could help defining relevant niche populations. Amgen’s AMG 337 achieved 5 responses among 7 gastric cancer patients with MET amplification. Pfizer’s Xalkori, which is also a MET inhibitor, generated a response rate of 33% among 12 lung cancer patients with MET-amplified tumors with a median duration of response of 8 months.