The Cytomx paradox

Failure of the company’s in-house lead raises questions about the viability of its extensively partnered probody technology.

During its 14 years of existence Cytomx has generated little in the way of convincing data from in-house projects, a fact underlined by yesterday’s failure of the group’s lead, CX-2009, in breast cancer.

Yet the probody technology on which Cytomx was built did attract considerable interest from pharma companies, and Cytomx scored deals with Pfizer, Bristol Myers Squibb, Abbvie, Amgen and Astellas. The second of those best illustrates the paradox, with Bristol effectively betting its post-Yervoy immuno-oncology strategy on a tech that appears unable to deliver what it promises to.

Probodies have an antibody structure, but add masking domains that render them inactive until cleaved by proteases in the tumour microenvironment. The fundamental aim is to avoid off-tumour activity, and they can be used to build naked MAbs, bispecific T-cell engagers and antibody-drug conjugates.

Still, based on Cytomx’s own work there is little to support the idea behind probodies, and since excitement about the technology hit a peak in March 2018 Cytomx’s stock has lost 95% of its value, including a 30% fall at market open this morning.

Breast cancer disappointment

This latest crash was sparked by data from a study of praluzatamab ravtansine (CX-2009), an anti-CD166 probody-drug conjugate, in breast cancer. In ER-positive/Her2-negative patients the remission rate with praluzatamab was at least 10%, crossing an efficacy benchmark; however, in a triple-negative cohort this bar was missed.

The trial also included a third cohort, TNBC patients given praluzatamab plus Cytomx’s anti-PD-L1 probody pacmilimab (CX-072). The company said nothing about how this performed, but scrapped both TNBC cohorts anyway, adding that it would now not advance praluzatamab without a partner.

This is not the first time praluzatamab has disappointed. The latest setback adds lack of efficacy – in a space being shaken up by Astrazeneca/Daiichi Sankyo’s Enhertu – to a profile in which ocular toxicity has stood out, both at Asco 2020 and in the latest dataset.

Mizuho analysts said they did not expect praluzatamab to be partnered, and removed it and pacmilimab from their Cytomx model. A separate probody ADC, the anti-CD71 asset CX-2029, is partnered with Abbvie, but also disappointed at Asco 2020, with injection site reactions and anaemia spooking investors.

Buying into probodies: Cytomx's deals with pharma
Date Licensee Asset(s) involved Up-front
Mar 2020 Astellas xCD3 probody discovery $80m
Oct 2017 Amgen EGFRxCD3 probody, CX-904 $60m*
Mar 2017 Bristol Myers Squibb (Deal extension) $200m
Apr 2016 Abbvie Anti-CD71 probody-drug conjugate, CX-2029 $30m
May 2014 Bristol Myers Squibb Anti-CTLA-4 probodies & others $50m
Jun 2013 Pfizer** Anti-EGFR probody & others $25m
Notes: *included $20m equity; **deal terminated in Mar 2018. Source: company filings.

But if probodies, designed specifically to widen the therapeutic window of standard MAbs, are actually themselves beset with toxicities, where does this leave their development?

Bristol might today be asking itself that question; two clinical-stage anti-CTLA-4 probodies, BMS-986249 and the non-fucosylated BMS-986288, comprise the US pharma company’s efforts to develop a less toxic version of Yervoy. They were originated under a Cytomx alliance involving a $50m up-front fee in 2014, and a further $200m three years later.

In total Cytomx received $445m from partners, in addition to over $400m of investor cash comprising VC funding, IPO proceeds and secondaries. Yet today its market cap stands at $90m, which even in the bear market suggests an astonishing display of value destruction.

Share This Article