Who won the immuno-oncology battle?
Clearly there are limitations in assessing such a contest at this stage of development, but it is pertinent to note that since the ASCO annual meeting abstracts were released last month, Bristol-Myers Squibb has seen approximately $5.2 billion shaved off its market cap, while Merck & Co. has seen its grow by around $4.5 billion and Roche by $3.8 billion.
According to post-ASCO investor notes, analysts don't believe that Bristol-Myers Squibb has lost its overall lead in the immuno-oncology (IO) race, but the gap is clearly closing. One could argue that while Bristol-Myers Squibb has failed to meet high expectations, both Merck and Roche have delivered better-than-expected progress.
There are many moving parts driving this trend, but in short – and despite some highly impressive data showcased by Bristol-Myers Squibb – concerns have grown over the toxicity of combinations and the role of biomarkers; two points of focus that have hit sentiment in Bristol-Myers Squibb disproportionately due to aspirations for combining nivolumab and Yervoy and its strategies towards targeting PD-L1-positive patients (see below).
The flip-side to market-leading revenue expectations for Yervoy and nivolumab is that Bristol-Myers Squibb is now increasingly leveraged to success in IO. Current consensus forecasts for 2020 apportion approximately a third of total revenue to these two franchises.
Investors were jittery again on Wednesday when Bristol-Myers Squibb's share price closed down 3 percent despite the lack of a notable catalyst for this sell-off. International Strategy & Investment analyst Mark Schoenebaum postulated that some shareholders are concerned that expectations for ongoing Phase III studies of nivolumab versus Taxotere in second-line non-small-cell lung cancer (NSCLC) are too high, particularly as enrolled patients have not been screened for being PD-L1 positive (a decision that based on recent study data could diminish the result) and the primary endpoint is overall survival rather than response rate.
Furthermore, pre-ASCO confirmation that the FDA should approve Merck's pembrolizumab on or before a PDUFA date of October 28 for second-line, Yervoy-resistant melanoma looks to be an increasingly important event and gives Merck a shot at becoming the first company to secure approval for a PD-1/PD-L1 antibody.
Bernstein analyst Tim Anderson recently suggested that investors are under-appreciating the potential loss of first-line usage to off-label use of pembrolizumab; an assertion that oncologists confirmed in our recent pre-ASCO Physician Views poll – see Physician Views Poll Results: First-to-market status for Merck & Co.'s MK-3475 in melanoma could open substantial opportunity for off-label front-line usage.
Where next for immuno-therapies?
Immuno-oncology products dominated ASCO as expected, but there was little in the way of surprise. Instead key themes were the accumulation of clinical data across indications where development is now more advanced (NSCLC, melanoma and renal cancer) and the continued expansion of development into new indications. In this respect, Roche scored a notable victory with its impressive data for the PD-L1 inhibitor MPDL3280A in bladder cancer.
The data – which demonstrated an overall response rate (ORR) of 43 percent in high PD-L1 expressing patients – underscored not only the applicability of IO therapies across multiple tumour types, but further evidence that use of biomarkers will play a role in both the development and commercialisation of these products, noted analysts (MPDL3280A delivered a notably lower ORR of 11 percent in low/non PD-L1 expressers).
Similarly, Merck demonstrated robust data for pembrolizumab in head and neck cancer patients, while there was further evidence of a broadening in IO portfolios, notably in the case of Roche who confirmed the progression of a humanised OX-40-targeting monoclonal antibody into clinical studies.
Further data for ABT199 in chronic lymphocytic leukaemia (CLL) and a handful of other early-stage disclosures prompted a chorus of approval from analysts who were left impressed by the Swiss company's showing at ASCO. Society Generale's Stephen McGarry noted "the impressive breadth and detail of the Roche data, from drugs that work by multiple mechanisms of action in different indications, suggests that Roche is likely to deliver on its protect-and-grow strategy for its oncology franchises – despite a competitive environment."
Did AstraZeneca deliver?
Fresh from defending itself against Pfizer's takeover attempts with a strategy indebted to bullish projections for its oncology pipeline, AstraZeneca duly attracted notable attention at ASCO.
Speaking to Forbes this week, CEO Pascal Soriot continues to talk a good game, and suggested that rather than putting the company under pressure, its decision not to entertain Pfizer's bids has AstraZeneca employees fired up to succeed.
In terms of AstraZeneca's prospects in IO, future commercial opportunity continues to be driven by the promise of its portfolio, rather than substantial data. At the company's ASCO analyst event, for example, management did present for the first time Phase I dose-escalation data for its PD-L1 inhibitor MEDI-4736 with the CTLA4 inhibitor tremelimumab. However, discussion around this presentation was still firmly centred on the possibility that both products may be better and safer than Bristol-Myers Squibb's nivolumab and Yervoy. As JP Morgan's James Gordon noted "the meeting highlighted Astra’s confidence in their early-stage oncology pipeline, but provided limited clinical data to substantiate this confidence."
Soriot noted to Forbes' Matthew Herper that he was impressed with the volume of clinical data presented by AstraZeneca at this year's ASCO versus last year's event, where he conceded that the company's output was low versus his previous visits as a Genentech employee. He will be aiming to make a further step up in 2015 with MEDI-4736 expected to have a more prominent role.
Where AstraZeneca did deliver was with data for its various small-molecule therapeutics. Suggestions that it may be able to file AZD9291 for T790M-positive, second-line NSCLC in Q1 of next year have boosted the possibility that having caught up with development of Clovis Oncology's CO-1686, AstraZeneca may now overtake its smaller rival. The race to market between these two drugs is the most competitive to emerge from ASCO 2014, with both Soriot and his counterpart at Clovis Pat Mahaffay claiming superiority for their products.
Who 'lost' at ASCO 2014?
Beyond the creeping suggestion that Merck and Roche are continuing to gain ground on Bristol-Myers Squibb in the IO race, a handful of other companies arguably experienced a disappointing ASCO, although in each instance there appears to be scope for reprieve.
In Clovis' case not only did data at ASCO confirm that AstraZeneca's AZD9291 is now level pegging both in terms of time-to-market and efficacy data, but investors were spooked by Mahaffay's disclosure that some patients who received CO-1686 had developed hypoglycaemia and were subsequently being treated with insulin.
There has been considerable debate as to whether Clovis' 30 percent share price slump since the beginning of ASCO has been driven by the commercial implication of a hypoglycaemia side-effect or the way in which it was disclosed by the company, particularly Mahaffay's admission in an interview with CNBC that the drugmaker did not initially know how to treat the side effect.
In the same interview, Mahaffay stressed that CO-1686 "does not cause diabetes" and that no patients had discontinued the study due to hypoglycaemia. He also suggested that CO-1686 would deliver superior progression-free survival (PFS) data versus AstraZeneca's drug, while Clovis bulls will reassure themselves with the fact that treatment-related hypoglycaemia is fairly common with cancer therapies. That said, ASCO may have provided AstraZeneca the necessary momentum in this particular race.
Eli Lilly's efforts to secure a commercially appealing approval for its anti-VEGF antibody ramucirumab in second-line NSCLC also prompted a mixed response from industry observers at ASCO. Having secured approval for gastric cancer, but failed to demonstrate sufficient efficacy in breast cancer, ramucirumab has been an inconsistent pipeline 'asset' for Eli Lilly. However, top-line data in NSCLC had raised expectations ahead of ASCO when published in February.
Full data from the REVEL study demonstrated, however, that while a combination of ramucirumab and docetaxel had delivered a statistically significant hazard ratio of 0.86 versus docetaxel alone, it only provided an overall survival benefit (OS) of 1.4 months. With developers of IO drugs also targeting the NSCLC market, analysts had suggested that a two month to three month OS benefit would be necessary if REVEL was to deliver a commercially successful result. That said, ramucirumab's benefit was broadly consistent across both squamous and non-squamous NSCLC patients; with Roche's Avastin not approved or used in the squamous setting – and with refractory patients having access to little in the way of effective therapy – Eli Lilly may find some traction here, noted ISI analyst Mark Schoenebaum and Goldman Sachs' Jami Rubin.
Another notably disappointing result at ASCO concerned the ALTTO study of GlaxoSmithKline/Novartis' Tykerb plus Roche's Herceptin versus Herceptin alone in adjuvant HER2 breast cancer patients – a result that prompted widespread debate as to how the treatment of adjuvant and neoadjuvant breast cancer is treated and whether partial complete response (pCR) can be used as a reliable surrogate endpoint; a previous neo-adjuvant study using pCR as its endpoint demonstrated that adding Tykerb to Herceptin did generate a statistically significant reduction in tumour size.
Puma Biotechnology – developing neratinib, a similar product to Tykerb, which utilises pCR in its study design – has seen its share price fall 25 percent over the past five days with investors concerned that approval in the adjuvant setting is now much less likely. Responding to suggestions that pCR is a potentially flawed surrogate endpoint, Donald Berry of M.D. Anderson Cancer Centre – who designed the I-SPY2 trial into which neratinib has been enrolled – told Forbes' Matthew Herper that reaction to the ALTTO data was overblown. He suggests that the study was not big enough to reach statistical significance. The jury remains out, with the head of the FDA's oncology division also suggesting that his team don't yet know how to use data from neoadjuvant studies and at present is considering the totality of data for submitted drugs based on these trials.
What role for biomarkers in immuno-oncology?
The role of biomarkers (particularly PD-L1) in shaping development, approval and usage of IO therapies remained a key area of debate at ASCO.
Monotherapy PD-1/PD-L1 inhibitor data presented in a number of new solid tumour indications – such as bladder cancer and head and neck cancer – demonstrated significantly different response rates for high expressers of PD-L1 and low/non-expressers.
One particular challenge (or opportunity) is that while some regimens deliver notably higher responses in patients who are high expressers of PD-L1, more muted responses in low or non-expressers of PD-L1 are still comparable or better than current standards of care. There may be some disconnect between physicians and payers on this front; while numerous oncologists at ASCO were cited as saying that they would not want to deprive low/non-expressers of therapy with an IO product, some analysts have suggested that payers may be more cautious towards stratifying patients based on biomarker status given the potential price that these drugs will command.
Roche – recognised as the global leader in biomarker-led pharmaceuticals – appears to be advocating a win-win approach for developers; use of biomarkers to expedite development timelines followed by stratification of patients once products are approved in order to assess who should be targeted with more aggressive regimens, such as combination versus monotherapy.
To read more Friday Five articles, click here.