A lot has been written about the immuno-oncology (cancer immunotherapy) field and how it is expected to revolutionize cancer treatment. In 2013, excitement around immuno-oncology and PD-1 antibodies in particular reached record high levels. In 2014, the trend is expected to continue on several fronts. These include potential approvals, new combination regimens, new indications and new targets.
Below is a review of key catalysts and drivers for immuno- oncology in 2014.
Accelerated approvals for Merck’s and BMS’ PD-1 – potentially in 2014
The most important data readout will be for BMS’ (BMY) nivolumab in squamous non-small cell lung cancer (NSCLC). Results are expected in Q2 2014 and may serve the basis for accelerated approval already in 2014 (assuming BMS submits its BLA in Q2). A response rate of >20% is considered sufficient for approval due to limited options for these patients (few known driver mutations, ineligible for Avastin and Alimta). Good response durability will further increase likelihood of approval.
This is BMS’ most imminent chance to get nivolumab approved. Other pivotal trials will read out only in H2 2014/H1 2015.
Merck (MRK) is pursuing melanoma as a bridgehead to market based on phase I data in Yervoy failures (Response rate >40%). Earlier this year the company surprised the market by announcing a rolling submission in Yervoy failures. The rolling submission mechanism implies Merck plans to submit additional data beyond the phase I results but no specifics were given. These additional data will likely come from a large melanoma phase II comparing MK-3475 to chemotherapy in Yervoy-failures. The trial, which completed enrollment in October 2013, is an open label trial that allows cross over, so Merck is already aware of the data and believes the response and PFS results are sufficient for approval.
PD-1 combination regimens – Yervoy and Avastin
On the heels of stellar activity as monotherapy, companies are trying to differentiate their PD-1 programs by combining them with other drugs. In 2014, companies will present multiple combinations with PD-1 antibodies that may clarify their strategic positioning. The most straightforward combination is with CTLA-4 antibodies like Yervoy, as they employ a similar mode of action.
BMS leads the pack with nivolumab +Yervoy combo – The combination is being pursued in an aggressive development program across many tumor types. Last year, the company presented impressive preliminary results in melanoma demonstrating a high response rate and more importantly very deep responses as most patients experienced near complete eradication of tumors (see figure below). The combination is being evaluated in melanoma, lung, renal, colon and brain cancer.
Still no decision on lung P3 for nivo+Yervoy – At ASCO 2014, all eyes will be on nivo+Yervoy combination data in lung cancer. BMS’ recent earnings call curbed expectations to some extent, as the company did not announce a decision to advance the combination to phase III as some had hoped. In addition, management’s tone during the call was perceived as cautious, implying that so far lung data are not phenomenal. On a positive note, the company announced its decision to start a phase III for nivo+Yervoy in renal cancer.
PD-1 +Avastin as another high priority – Combining PD-1 antibodies with Avastin is practically straightforward as the latter is approved in 2 of PD-1’s imminent indication (NSCLC, renal cancer). In recent years, VEGF has been characterized as an immune suppression factor on top of its role in angiogenesis, making the combination even more logical. Animal models presented by Roche demonstrate synergistic activity for the combination.
PD-L1+Avastin combinations are important for Roche – Roche is evaluating its PD-L1 antibody (MPDL3280A) in combination with its VEGF antibody, Avastin, across several tumor types with an emphasis on NSCLC and renal cancer. Although BMS was the first to pursue Avastin in combination, Roche has a broader program for MPDL3280A+ Avastin and is expected to present top line results at ASCO 2014. As Roche markets Avastin, any Avastin combination regimens may give it an edge over competing PD-1 programs, especially if nivo+Yervoy disappoints. Good combination data may also offset concerns about MPDL3280A’s lower efficacy as monotherapy.
Roche recently started a randomized phase II of MPDL3280A+Avastin vs. Sutent in renal cancer.
New indications for PD-1 – Lymphoma, bladder and breast cancer
PD-1’s core indications are lung, melanoma and renal cancer, where the three leaders (BMS, Merck, Roche) are running neck and neck. Understandably, companies are trying to expand beyond those indications and try to differentiate their programs via new indications.
In lymphoma, 1st mover is BMS – BMS has an ongoing phase I in blood cancers for nivolumab, which may have first results in 2014. The company recently posted 2 phase II trials in DLBCL and follicular lymphoma, respectively, which indicates the phase I demonstrated a sufficient level of efficacy as single agent. Interestingly, it looks like BMS views the DLBCL study as a potential pivotal trial as responses will be assessed by an “Independent radiologic review committee” (IRRC).
Bladder cancer pursued by Roche – Rumors about strong activity in bladder cancer surfaced in the past months. Earlier this month Citi’s analyst, Andrew Baum, issued a report titled “Immunotherapy: Bigger, Broader, Bladder”, in which he predicts compelling data for Roche’s MPDL3280A in bladder cancer at ASCO 2014. According to the report, there may be over 50,000 eligible patients for PD-1 therapy, which translates to a $4B opportunity.
Initial signs in breast cancer – The biggest surprise in 2014 may come from breast cancer, with an emphasis on triple-negative breast cancer (~20% of cases). TNBC is associated with a poor prognosis due to tumor aggressiveness and lack of effective treatments . BMS and Merck both have follow on trials that list TNBC as 1 of 4 relevant tumor types. Both trials started in the middle of 2013 but there are already clear signals of efficacy. After decades of stagnation, there are finally 2 promising classes of drugs for this indication: PARP inhibitors and PD-1 antibodies.
New targets beyond PD-1
So far, PD-1 is the only target with clear efficacy in humans. There are many programs that explore new immunotherapy targets employing various different approaches. To date none have demonstrated anything similar to PD-1 as monotherapy. These programs are being evaluated as monotherapy or as add-on to a PD-1 backbone.
The hottest targets can be divided to co-inhibitory checkpoints like PD-1 (where the goal is to inhibit their function) or immune activators (where the goal is to enhance their function). The first group includes LAG3, IDO, TIM3, KIR and BTLA. The second group includes CD137 (41-BB), OX40 and CD27.
Programs expected to generate data in 2014
Incyte’s IDO inhibitor – There is a huge buzz around Incyte’s (INCY) IDO inhibitor (INCB024360), which in contrast to most programs is a small molecule inhibitor. Monotherapy data for INCB024360 did not demonstrate meaningful activity in patients but Incyte’s management indicated that an ongoing combination trial in melanoma is demonstrating encouraging signs of efficacy. The trial evaluates the drug with Yervoy in melanoma patients and per management’s comments, activity seen with this combination exceeds historic response rate with Yervoy alone (~10%). This was enough to spark investors’ and analysts’ imagination with expectations for a 30-40% response rate. As a result, analysts have plugged in $2-3B of INCB024360 revenues in their models without even seeing data from a single arm 20-patient data set (Expected at ASCO 2014). Incyte and Merck recently announced a combination trial for INCB024360 and MK-3475. The drug is also in phase II for ovarian cancer.
Newlink (NLNK) has 2 IDO inhibitors in development. Results from the 1st generation compound (indoximod) are expected at ASCO 2014. A 2nd generation compound (NLG919) is expected to start Phase I imminently. The interest spurred by Incyte led to a significant step up in the company’s stock price.
Targeting macrophages via CSF-1R – Roche is expected to present a large data set for its CSF-1R antibody alone and in combination with chemotherapy. This antibody is designed to inhibit tumor promoting macrophages that are found in many solid tumors. Conceptually, it is complimentary to T cell activating drugs like PD-1 as it removes an important immune-suppressive component from the tumor microenvironment. Early biomarker data were encouraging but it is still unclear whether this will translate to clinical efficacy.
CD27 as a novel co-stimulatory target – Celldex (CLDX) is expected to present updated results for CDX-1127, an agonist antibody for CD27. The company is enrolling patients with melanoma and renal cancer and should have ~15 patients per tumor type. Phase I results presented last year showed preliminary signs of efficacy in the form of 1 durable complete response (Hodgkin lymphoma) and 2 cases of tumor shrinkage that did not qualify as objective responses (colon cancer and NHL). There were also intriguing immunologic effects such as reduction in regulatory T cells and enhanced T cell activation. Initial results from the melanoma expansion cohort has so far been disappointing.
Preclinically, Celldex presented an additive effect when CDX-1127 was combined with anti-PD-L1 (see figure below).
Anti-KIR combination data from BMS – BMS is running 2 combination trials for lirilumab (Anti-KIR, licensed from Innate Pharma) with Yervoy (CTLA-4) and Nivolumab (PD-1), respectively. As KIR is a family of co-inhibitory checkpoints on NK cells, lirilumab has the potential to complement T cell activation by PD-1/CTLA-4 with NK activation. Lirilumab is the only KIR antibody in clinical development for oncology and these combinations could create an important differentiating factor for its PD-1 franchise.
Awaiting more data for B7-H3 – Macrogenics (MGNX) is expected to report clinical results for MGA271, an anti-B7-H3 antibody in phase I. Although B7-H3’s biology is poorly characterized and the antibody employs additional modes of action, the fact B7-H3 belongs to the same family of PD-L1 results in high expectations for this program. Phase I results reported in 2013 did not include meaningful efficacy but updated results in B7-H3 positive tumors are expected in 2H14.
LAG-3 as a widely pursued target– LAG-3 is one of the better characterized immune checkpoints and is pursued by multiple companies. LAG-3 has been shown to be an important immune checkpoint with resemblance, co-localization and potential cross talk with PD-1. Similarly to PD-1, LAG-3 was initially identified as a general immune suppressor receptor on T cells. Only recently researchers started publishing its role in the context of tumor immunity. In animals, combining PD-1 and LAG-3 antibodies proved synergistic (see figure below).
BMS has the most advanced anti-LAG-3 antibody (BMS-986016), which is currently in phase I. The trial started in October 2013, so it might have preliminary results during 2014. Last week, Novartis (NVS), who until recently did not have a strong presence in the immune checkpoint arena, acquired CoStim which has a preclinical LAG-3 program.
New targets on the horizon
2014 will see a record number of new immunotherapy programs. Several programs are expected to start phase I or to be officially disclosed as late stage preclinical programs.
AstraZeneca leads pursuit of OX40 agonists – OX40 is a co-stimulatory protein which is pursued by at least 3 companies: AstraZeneca (AZN) (licensed a from Agonox), GSK (GSK) (licensed from MD Anderson) and Pfizer (PFE) (internally developed). All programs may enter the clinic during 2014. OX40 is an intriguing target given clinical experience with an old murine antibody developed by Agonox’s founders. Although the antibody was administered only 3 times over a week, anti-tumor activity was observed in 12 of 30 patients. The waterfall chart from the trial clearly shows that some patients had a 20%+ reduction.
There are multiple programs that are either early stage or whose targets are undisclosed.
J&J (JNJ) has a VISTA antibody, which was licensed from Immunext in 2012.
Roche has several preclinical antibodies in development to undisclosed targets. Preclinical results with 2 of those programs demonstrate striking activity alone or with an anti-PD-L1.
New immune checkpoints discovered by Compugen – Through its computational discovery engine, Compugen (CGEN) discovered 9 potential members of the B7 family (of which PD-1 and CTLA-4 are members). For some of the new targets, the company presented mechanistic and animal data which validate their immunologic role in cancer and autoimmune diseases. In contrast to the situation with other targets, Compugen may be the only one with real “proprietary” immune checkpoints. Last year, Bayer and Compugen announced a $540M licensing deal ($10M upfront) for 2 of the novel immune-checkpoints.
Portfolio holdings – March 2nd, 2014