Astra I/O master plan hits a bump in head and neck cancer

The 5% fall in Astrazeneca’s valuation yesterday following the disclosure that two studies of its checkpoint inhibitors in head and neck cancer have been placed on partial clinical hold may not, for once, be an over-reaction. And this is despite the fact the bleeding events that prompted the hold are unlikely to be linked to treatment with the anti-PD-L1 antibody durvalumab and anti-CTLA4 antibody tremelimumab, which are being tested in the phase III studies.

The share price fall does, however, highlight the nervousness investors have about the strategic nature of this indication, more specifically termed squamous cell carcinoma of the head and neck (SCCHN), for Astra. Although SCCHN affects many fewer people than the fiercely competitive field of non-small cell lung cancer, where Astra is also in phase III with its checkpoint inhibitors, the UK-based group’s stronger position relative to rivals makes SCCHN likely to be the more commercially valuable indication.

Bleeding ‘eck

Following reports of a number of bleeding events, Astra voluntarily suspended enrolment into its two SCCHN studies, Kestrel in first line and Eagle in second line, in mid-September and informed regulators of its action.

The FDA subsequently formally placed the two studies on partial clinical hold – meaning new recruitment is suspended, but patients already enrolled continue on drug – pending completion of an analysis of these events. It seems Astra then updated its listing and announced the hold. 

Astra said its analysis of the bleeding events is now complete and has been submitted to FDA for review. It believes the bleeding episodes are not connected with drug therapy and hopes to receive a response and resume recruitment soon. Bleeding is a known complication in SCCHN primarily due to the proximity of tumours to major blood vessels and the use of other therapies, which can include surgery and radiation.


Prior to the partial hold, Kestrel and Eagle had been due to read out in late 2017 or early 2018 and the best case scenario now is likely to be a few months of delay to those timelines. Astra does have two phase II studies underway in SCCHN, Hawk and Condor, that are both due to render results before this time. Both studies are already fully recruited so are unaffected by the regulatory action.

Hawk tests durvalumab in PD-L1-positive patients, using a cut-off of 25% expression, while Condor tests durvalumab and tremelimumab as single agents and in combination in PD-L1-negative patients.

Hawk may potentially read out in the next two months, so could provide an important injection of confidence. The company has previously suggested Hawk may have registrational potential, but in the light of the clinical hold, this now seems optimistic.

Phase II/III studies of checkpoint inhibitors in SCCHN
Study Setting N Design Trial ID Data
 Hawk*  2L, PD-L1 positive 112 Durvalumab NCT02207530 Dec-16
 Condor*  2L, PD-L1 negative 240 3 arms: durva, tremelimumab,  durva + treme    NCT02319044 Oct-17
Kestrel 1L 760 durva +/- treme vs SoC   NCT02551159 Nov-17  
Eagle 2L   720 durva +/- treme vs SoC NCT02369874 Feb-18
Keynote-040 2L 466 Keytruda vs SoC NCT02252042 Jan-17
Keynote-147* >2L 74 Keytruda +/- acalabrutinib NCT02454179 Sept-17
Keynote-048 1L 825 Keytruda +/- chemo vs Erbitux + chemo NCT02358031 Nov-17
Checkmate-141 2L 360 Opdivo vs Investigator's choice NCT02105636 Efficacy stop
Checkmate- 651 1L 490 Opdivo + Yervoy vs SoC NCT02741570 Jul-18  
Note: All studies are Phase III except where indicated with an asterisk.

... while competitors are getting to market first

Astra has made a greater investment in SCCHN than its I/O rivals, but it is not without competition. Indeed, it has been beaten to market by Merck, which obtained an accelerated approval for Keytruda as a monotherapy in second line SCCHN in August. This was based on data from a phase I study, Keynote-12, which showed an objective response rate of 16%. Merck has a confirmatory phase III study, Keynote-040, underway with a readout scheduled for early next year.

And Bristol-Myers Squibb’s Opdivo is currently under regulatory review in SCCHN, based on the data from the Checkmate-141 study, which was stopped early for efficacy in January this year. Indeed, with breakthrough therapy status, Opdivo may gain formal US approval in this indication in the next few weeks.

Checkmate-141 data were presented at AACR and showed Opdivo conferring a 2.4 month improvement in median overall survival (7.5 vs 5.1 months), equivalent to a 30% reduction in the risk of death. This will provide a benchmark for comparison for Keynote-040 and later Eagle, when these data become available.

These are the key competitors for Astra, as Roche and Pfizer/Merck KGaA do not yet have any mid- or late-stage trials with their PD-L1 antibodies, Tecentriq and avelumab, in SCCHN. Opdivo and Keytruda are or are likely to be approved as single agents, but the future looks likely to be combinations – especially in PD-L1-negative patients – and thus Astra has a potential advantage. 

Bristol-Myers is trialling the combination of Opdivo and Yervoy in the first line setting in Checkmate-651, and these data, when available, will no doubt be compared with those of the combination arm from the Kestrel study, which should report first. Merck, which of course lacks an anti-CTLA4 drug, does not have an dual I/O combination in phase III, although is testing the combination of Keytruda with chemotherapy in Keynote-048. 

One competitor’s SCCHN study that will be closely watched by Astra at least is likely to be the Keynote-147 study of Keytruda with the UK group’s own BTK inhibitor, acalabrutinib. If successful, it seems likely that Astra would likely attempt copy this, substituting durvalumab – or even advance a triple durva/treme/acalabrutinib combo into trials.

While SCCHN has a lower profile than lung cancer and melanoma as a teting ground for checkpoint inhibitors, the field is still moving fast and is capable of generating the same surprises.

To contact the writer of this story email Robin Davison at or follow @RobinDavison2 on Twitter

Related Topics

Share This Article