Cellular Immunotherapy & Unum Therapeutics: Out of Many, One

Posted in Biotech financing, Portfolio news | 2 Comments

More exciting news has emerged from the immuno-oncology field over the past couple weeks from both Penn/Novartis and NCI/Kite.  The former released new CAR-T data, widely heralded as “unprecedented” in its efficacy: in a trial of relapsed, refractory ALL patients, 90% had complete remissions after their infusions (27/30), and 78% were still alive after treatment, including the first ALL patient treated with the therapy over 2 years ago. Further, three patients used CAR-T as a bridge to curative allogenic stem cell transplant while in remission. Additional news came from the NCI/Kite team, who published a Lancet paper this month on their ALL/NHL Phase 1 trial, which also reported strong response rates on the order of 70%+ (here), intriguingly without the prolonged B-cell aplasia seen with other CD19 CAR-Ts.  These new data highlight the truly transformational potential of cellular immunotherapy.

Although the data is new, the field is not – and it’s worth remembering how long breakthroughs take in this business.  The first CAR-T construct was described in a 1989 paper by immunotherapy pioneer Zelig Eshhar in PNAS (here).  Early first-generation constructs (more here), typically featuring a single signaling domain most commonly derived from the CD3z component of the TCR/CD3 complex, were explored with different antigens in different indications, but for most of the next two decades they had very minimal efficacy, and little to no sustained responses. Over that time, more advances were made in terms of signaling complexes and transmembrane chimeras.

The first reports of compelling clinical efficacy were in 2010, nearly 21 years after CAR-T constructs were first described, and Penn’s August 2011 NEJM patient report really catalyzed the field (here). Since that time, other labs have shown great data – like MSK and the NCI – and there’s been an explosion of commercialization interest in the field.

What’s unusual and worth noting about the explosion in the CAR-T space is that it wasn’t the traditional venture community that dove in first – unlike a typical new modality such as RNAi or antibodies, where VC-backed companies have often led the charge (like Atlas’ investments early in Alnylam or  Morphosys, respectively). While many VCs had previously dabbled in cellular immunotherapy for years, nothing had worked well so there was little enthusiasm for the space, and little venture funding flowing into it.

But it was Novartis, one of the largest of the Pharma’s, who boldly moved more quickly and more aggressively into the field than others, jumping into the dual risks of both cell therapy and gene therapy. Mark Fishman and his team at Novartis deserve real credit for opening up the commercial realization of this field. Their significant and landmark deal with the University of Pennsylvania around Carl June’s work (here, here) put the field under the spotlight and established them squarely as the leading large biopharma player in the space. Although it was still a very limited clinical story (e.g., the deal’s press release talks about three patients with responses), this big move into the space triggered a wave of renewed interest in cellular immunotherapy – which has continued to be supported by accumulation of great clinical data.

Several CD19-directed CAR-T constructs had shown early compelling efficacy by the time Novartis had done their deal with Penn – and these other groups (MSK, NCI, Hutchinson, etc) provided great academic partners for venture groups to pour into the field (e.g., Kite Pharma, Juno Therapeutics, etc).

But it is interesting to me that it required both compelling early clinical data and a Big Pharma jumping into such a “risky” field in a major way before the venture community got conviction. In some ways, the major clinical risk – do these constructs work – was eliminated before the venture and investment communities got fully engaged in this modality. The big risks now lay with late stage clinical studies, speed-to-approval, manufacturing and process optimization, marketing and distribution of a process-as-the-product therapeutic regimen – all of which require big capital infusions to wrestle to the ground, and are typically not the risks that venture capitalists fund to overcome. The big Series A and B financings at Juno, the IPO at Kite, the Celgene deal at bluebird – and many others – are a reflection of this significant capital requirement in the CD19 race today.

Atlas was very close to jumping into that race in the months that followed the Penn-Novartis announcement, but we failed to seize the opportunity fast enough for a variety of reasons. However, given the valuations in the field today, I suspect we’ll regret letting that one get away; chalk it up onto our version of the Bessemer “anti-portfolio” – awesome deals that got away.

About a year ago, sensing the CD19 race wasn’t one where we could compete to win as early stage venture investors, we began to shift our focus to “next generation” approaches in the CAR-T space. These include things like suicide switches to turn them off, alternative solid tumor antigen approaches, and allogenic off-the-shelf concepts. All of those are interesting, but the one area that got us really excited was Dario Campana’s work on a “universal” next generation engineered chimera aimed at augmenting the efficacy of antibody therapeutics (here, covered in SciBX here)– which has formed the basis of a new startup called Unum.

Unum Therapeutics

Unum is a new cellular immunotherapy company dedicated to antibody-coupled T-cell receptor (ACTR) therapeutics.  Today Unum announced its Series A financing, co-led by Atlas and Fidelity Biosciences, with participation from Sanofi-Genzyme BioVentures (here). This financing should advance Unum’s lead candidate through initial proof-of-concept clinical studies.

In short, ACTR technology enables genetic-programming of T-cells to attack tumor cells in an antibody-directed manner.  Unum’s ACTR constructs are not restricted to a single antigen or narrow set of tumors, but instead can be universally applied to augment any antibody-directed anti-cancer therapy with a cell-surface antigen.  Hence the name, derived from e pluribus unum – “out of many, one”.

Unum’s ACTR approach leverages the huge portfolio of mAb programs directed against cancer antigens across the industry. This could be the “one CAR construct to optimize them all”. Pharma has been optimizing the mAb arsenal for decades – tuning PK, potency, epitope selection, etc; Unum’s platform doesn’t reinvent those features but leverages them. Furthermore, most cancer-directed mAbs have also been optimized for ADCC effects through CD16 binding – making them ideal pairings for Unum’s technology. Finding the right set of antibody partners for the Unum platform, whether it be approved drugs or development-stage mAbs, is a big focus for the company.

In addition, ACTRs leverage years of experience in CAR-T cells and construct optimization (i.e., 3rd generation protein chimeras), but adds features like dose titration (via administered antibody dosing regimens) and cessation of therapy (by no longer dosing the antibody, these ACTR cells will not be effective). These attributes are very unique among CAR-T stories today. Most CAR-T’s expand in vivo so dose titration in a conventional sense is difficult, and in many cases the engineered T-cells appear to be active “forever” – i.e., the on-going B-cell aplasia seen in most of the CD19 CAR-T clinical studies to date. These unique features of ACTRs should also open up solid tumor targets in particular.

In addition to a compelling scientific and clinical approach, Unum has a superb founding team:

  • As mentioned above, Dario Campana is the company’s Scientific Founder and a real thought leader in the field of cancer cell therapies. At the National University of Singapore (NUS), he developed the ACTR technology that forms the basis for Unum. Before NUS, while at St. Jude Children’s Research Hospital, he created the CD19-CART construct that Carl June took into the clinic and demonstrated such profound activity.
  • Chuck Wilson is Unum’s founder and CEO, and has been both a Pharma BD executive and a serial Atlas-backed entrepreneur. Chuck drove the BD aspects of the landmark Novartis-Penn deal, as head of partnering for research and early development. Before Novartis, Chuck was a founder and biotech executive at Archemix, a prior Atlas portfolio company in the aptamer field. We’ve worked with Chuck for the past thirteen years either as investors or potential collaborators.
  • Lastly, Unum’s Chief Scientific Officer is cancer biologist Seth Ettenberg, who was most recently Cambridge site head for Novartis Oncology Biotherapeutics. Seth has extensive immuno-oncology expertise, leading the scientific aspects of the Novartis-Penn CAR partnership, as well as helping with the acquisition and integration of checkpoint-focused CoStim Pharma.

We expect early patient data in 2015 at Unum. Like most product engines, Unum has been and will continue to be discussing with pharma partners around the opportunity to expand the number of antibody-directed programs we’re exploring. We expect to have advanced those dialogues significantly over the coming months.

It’s exciting to be a part of such a transformational field, and a unique approach within it – looking forward to working with Chuck, Dario, Seth, Michael Gladstone at Atlas, Ben Auspitz from Fidelity, and Bernard Davitian at Sanofi – to help build Unum into a great new immunotherapy story.


Fueling Up To Attack Cancer: Two Big Startup Financings

Posted in Atlas Venture, Biotech financing, Portfolio news | Leave a comment

Immuno-oncology, antibody-drug conjugates, and cancer metabolism are three of the hottest and potentially most transformative approaches in cancer research today, and have occupied center stage at recent oncology meetings like ASCO and AACR (here, here).

The immuno-oncology (I/O) field has certainly hit its stride with impressive late stage clinical data and approvals around PD1/PD-L1 interventions, as well as exciting early data on engineered T-cell therapies in a variety of hematological malignancies.  Antibody-drug conjugates (ADCs) have also been at the forefront of cancer R&D: over 30 ADC programs are in clinical testing today (here), building on the exciting data and approvals for both Kadcyla in metastatic breast cancer and Adcetris in NHL.  And in cancer metabolism, Agios’ early but impressive data on AG-221 in IDH2 mutated cancers validates the approach of targeting key glycolytic pathway liabilities (here).

Like most early stage biotech investors, we’ve been active in all three areas.  Earlier this year, Novartis acquired CoStim Pharmaceuticals, an immuno-oncology checkpoint inhibitor company backed by MPM and Atlas (here); we’ve also been supporting bispecific antibody programs in the checkpoint and I/O space at F-star (here), and have two additional soon-to-be announced I/O investments (more later!).  In cancer metabolism, we’re pushing a lipid metabolism target (ACC) into the clinic over the next few months at Nimbus Discovery with a broad range of applications, including metabolic disease/NASH, but also in specific cancers dependent on de novo lipid synthesis (preclinical AACR data here).  Further, Navitor is working in nutrient sensing and metabolism around the mTORC1 pathway, with clear implications in cancer research, among other applications (here).

Earlier today we announced two new financings around drug discovery platform companies targeting these cancer pathways – Raze Therapeutics and Bicycle Therapeutics – and thought (re)introducing their stories and our approach would be of interest.  These were both Atlas’ seed investments that have met their initial milestones and are beginning to scale.

Raze Therapeutics

Cancer anabolic metabolism – specifically mitochondrial one-carbon energetics – has emerged recently as a major driver of cancer proliferation, survival, and biomass accumulation.  Raze, originally seeded by Atlas and led by my partner Peter Barrett as “Project C”, has launched publicly today with a $24M financing to advance a number of therapies targeting key nodes in these pathways.

Raze’s scientific founders – Vamsi Mootha (HHMI, MGH), Josh Rabinowitz (Princeton), and David Sabatini (Whitehead/MIT) – bring extensive mitochondrial proteome, systems biology, metabolic flux, and cancer biology expertise.  Their labs provided novel insights into several key nodes/targets in serine biosynthesis, glycine uptake, and related anabolic pathways.

Raze’s initial seed hypothesis was aimed at addressing the question of whether the specific set of first-in-class foundational targets were drug-able with novel chemical matter of sufficient quality to warrant a broader drug discovery campaign.  After close to eighteen months, the team has more than delivered against those early hurdles, expanding our biologic insights into these pathways and establishing a compelling platform for new therapeutics creation.

Raze’s approach of antagonizing these one-carbon metabolism pathways is supported by clinical experience with anti-folates, thymidylate synthase inhibitors, and ornithine decarboxylase inhibitors that target single downstream arms of the one carbon pathway.  Raze’s programs differentiate from these prior agents in two important ways: (i) potential for broader and more profound therapeutic efficacy, as Raze’s inhibitors would simultaneously inhibit all three critical one-carbon utilizing pathways, and (ii) significantly better safety properties than current metabolism-directed agents, since thymidylate synthase, dihydrofolate reductase and other enzymes are generally elevated in “normal” proliferating cells, whereas Raze’s targets are specifically upregulated only in transformed cells.  Lastly, Raze and its founders have identified specific patient populations, including genetically-defined subgroups, in both solid and liquid tumors for targeting with one-carbon metabolism-based therapeutics.

In addition to Raze’s great scientific foundation, the company has built a great emerging team.  In addition to our three scientific co-founders, both Keith Flaherty (MGH/Harvard) and Mike Gilman (Atlas EIR and CEO of Padlock Therapeutics) have joined the SAB.  Mike was also the Acting CEO for part of the company’s formative “seed” phase.  Vipin Suri and Adam Friedman, both Atlas EIRs, are currently driving the tumor biology and corporate development functions, respectively.  Lastly, as of only a week ago, our newest partner Jason Rhodes has jumped in and taken the helm of Raze as the Acting CEO going forward as we ramp up lead optimization activities against a set of targets.

With today’s announcement, we’ve brought forward a $24M financing supported by great co-investors at MPM, MerckSerono Ventures, Partners Innovation Fund, Astellas Ventures, and Novartis (NIBR); this should fuel the company’s programs into development.  Exciting times for Raze and cancer anabolism.

Bicycle Therapeutics

After several years of exploring different applications of their bicyclic-macrocycle platform, and perfecting the “art” around their phage-based discovery screening system, Bicycle Therapeutics has powered up with a $32M financing to focus its attention on the drug-conjugate space – aptly called our “Bicycle-Drug-Conjugate” approach (BDC).

The quick background: Atlas and Novartis Ventures co-founded Bicycle several years ago with distinguished scientific founders Sir Gregory Winter (MRC) and Christian Heinis (EPFL) around their phage-based macrocycle discovery approach (here); over the past few years, with the further support of solid co-investors SVLS, SR One, and more recently Astellas, the company has explored a variety of applications for bi-cyclic therapeutics in capital efficient manner. As described previously (here), the key with any new macrocycle platform is figuring out the true path to differentiated offerings relative to existing approaches (i.e., versus oral, cell-penetrant NCEs and parenteral, extracellular antibodies); applying a good technology to the wrong applications results in the same outcome as backing a failed technology.  This is where Bicycle has been both thoughtful and disciplined.

Bicycle’s dual-loop chemical-scaffold constrained macrocycles have several key characteristics: high potency (sub-nM), robust screening fidelity, species-tunable selectivity, very high protease and metabolic stability, ease of manufacture, as well as simple and controlled conjugation of payloads.

But the real excitement comes from turning an apparent liability – their very short half-life – into a differentiated feature.  One of the big problems with ADCs is that antibodies have long half-lives and expose normal tissues to toxic payloads over days if not weeks in circulation; long after dosing, ADCs can be detected in most vascularized tissues.  Bystander damage to host tissues is one of the dose-limiting toxicities of these approaches.

BDCs, by contrast, rapidly extravasate with high volumes of distribution after parenteral dosing, quickly binding to cell surface tumor antigens and internalizing, and any unbound BDC is quickly cleared by the kidney within minutes.  Plasma half-lives for bicycles in patients are predicted to be on the order of a few hours, not days or weeks, and yet even with these short exposures we see very potent anti-tumor activity in our in vivo xenograft models.  Rather than being an Achilles heel of the platform, these short half-lives, when coupled with rapid extravasation and internalization, have the potential to be major differentiators vs conventional ADC approaches.

This financing should fuel the Bicycle platform for several years, bringing its lead program through clinical proof-of-concept studies in cancer patients (undisclosed first-in-class target), as well as advance several other programs into development.  As with all drug discovery platforms, we also anticipate partnering extensively with Pharma oncology groups that have expertise in specific cell surface antigens that could be targeted by this BDC modality.


These are two oncology-focused drug discovery engines launched or seeded by Atlas over the past few years – one targeting the “big biology” around mitochondrial energetics in cancer, the other applying a novel modality to tumor-targeting of cytotoxic payloads.  We have high hopes for them, and their aspirations of impacting cancer care.