"Marge" Bests Herceptin
The treatment of HER2+ breast cancer was revolutionized by approval of Roche's (OTCQX:RHHBY) Herceptin (trastuzumab) in the late 90s. In the past two decades, Herceptin has saved thousands of lives and registered over $75 billion in global sales.
Since that time, HER2 has become one of the most drugged targets in oncology, with approvals of the antibody Perjeta/pertuzumab (Roche), the antibody-drug-conjugate Kadcyla/ado-trastuzumab emastine (Roche), the HER2/EGFR kinase inhibitor Tykerb/lapatinib (GSK), and Nerlynx/neratinib (PBYI). In addition to the approved drugs targeting HER2, other therapies are currently in clinical development, such as the potent HER2 inhibitor tucatinib (SGEN). Notably, cell therapy approaches to directly target and eradicate HER2 expressing cells using TCR-engineered T cells have been stalled due to fatal toxicity.
Given Herceptin's established supremacy, the crowded nature of the "drug HER2" field, and the potential near-term launch of trastuzumab biosimilars in the USA, it's easy to see why analysts wrote off and may continue to write off MacroGenics' (MGNX) Fc-optimized antibody margetuximab. Indeed, margetuximab is nearly identical to trastuzumab, but importantly harbors three amino acid mutations in the Fc domain, which enhance antibody-dependent cellular cytotoxicity, ADCC.
Relative to trastuzumab, margetuximab's Fc domain displays enhanced affinity to CD16A activating receptor on NK cells and reduced affinity to the CD32B inhibitory receptor. The end result: margetuximab is trastuzumab with much more potent ADCC-capabilities.
As demonstrated in several studies, ADCC is an important mechanism of action for trastuzumab. Unlike the kinase inhibitors Tykerb and lucatinib, trastuzumab causes direct lysis of tumor cells by NK cells, helping to bridge innate and adaptive immunity for further immune activation. Furthermore, it has been demonstrated in multiple clinical trials that patients harboring the low affinity "F" CD16 allele (homozygous or heterozygous) have worse clinical outcomes than patients homozygous for the high affinity "V" allele. (Mnemonic is F for Failure and V for Very good response). In 2017, researchers published a retrospective analysis of 1,251 patients in JAMA Oncology, conclusively demonstrating the importance of ADCC by noting the following:
As hypothesized, patients with genotypes FCB3A-158V/V or FCB3A-158V/F received greater benefit from trastuzumab (HR, 0.31; 95% CI, 0.22-0.43; P < .001) than patients who were homozygous for the low-affinity allele F/F (HR, 0.71; 95% CI, 0.51-1.01; P = .05).
There is little doubt that ADCC plays a profound roll in trastuzumab's efficacy given ample in vitro, in vivo, and clinical data. Thus, it is rather bizarre that sell-side analysts, including Citi's now-infamous Yigal Nochomovitz, downgraded MacroGenics stock just days before the positive margetuximab SOPHIA trial data readout was announced publicly. He was not the only one. On the conference call, a chorus of analysts was apologetic for writing off margetuximab. Even after the positive data readout, certain analysts continue to doubt the efficacy of margetuximab. Denial is not just a river in Egypt...
Analysis of SOPHIA Trial Results
So, what did the data show? In their press release, MGNX was quite terse. Only two hazard ratios were provided, a hazard ratio of 0.74 for the entire cohort which barely achieved statistically significant p=.033 and a hazard ratio of 0.68 for F allele carriers, an estimated 85% of the study population which readily cleared significance (p=0.005). They could not provide OS data as the trial is still in progress and were tight-lipped about response rates and PFS data for competitive reasons. Notably, the better hazard ratio was seen in low-affinity F allele carriers, consistent with drug's mechanism of action. On the conference call, management stated that OS trends were favorable but did not provide any quantitative insight. Assuming a base case PFS of five months for the control group, these hazard ratios likely translate into an approximate two-month gain in PFS, which will likely translate to OS. However, all of these numbers are subject to median statistics. Notably, the FDA will also look at the long-end tail of the Kaplan-Meier survivor curve. To this, MGNX added a significant amount of detail at a medical conference last December, providing detailed vignettes on three heavily-pretreated patients harboring the low affinity allele that progressed on trastuzumab and then had durable multi-year responses to margetuximab. An example of one such patient is provided below. It is these patients who are homozygous for the low affinity F allele that previously failed on trastuzumab who stand to benefit most from margetuximab's approval.
As management (who would face lawsuits if untrue) guided that there were no additional safety signals vs. trastuzumab on the conference call, I have little doubt that margetuximab will get approved in the 3rd line setting. Back-computation ( x*0.15+.85*0.68=0.74 ) of hazard ratios indicates that if the total population had an HR of 0.74 and the low affinity allele group making up 85% of the population had an HR of 0.68, then the implied HR for the 15% homozygous V population was x=1.08, which is almost certainly not statistically significant. Thus, consistent with the drug's mechanism, margetuximab is statistically superior to trastuzumab in 85% of the population carrying the F allele and non-inferior to trastuzumab in the 15% who harbor the natural high-affinity homozygous V/V allele.
It's taken almost two decades since Herceptin's approval, but thanks to MGNX, there is now a HER2 antibody which is conclusively better than Herceptin. This has big implications for the trastuzumab biosimilar market and MGNX's long run prospects.
The 3rd line indication will likely only see ~5,000-10,000 patients treated for a median of six months. Assuming pricing similar to trastuzumab ($70,000/year), this will lead to initial revenues of 175 to 350 million dollars, not bad for a company with a market cap of ~1 billion with 200 million in cash reserves and this is a conservative estimate, which excludes ex-US revs.
Margetuximab to the Frontline? Billions at Stake.
However, the most lucrative position is in the 1st line setting, which is currently over $7 billion/year. MacroGenics does not have the data for 1st, but they should invest in and commence a head-to-head trial comparing margetuximab to trastuzumab. I estimate the trial would take 4-5 years to run, but the prize would be great. If they pre-selected for F allele carriers, trial success would be nearly certain. While pricing power would be diminished due to biosimilars, MGNX would be able to offer a superior product at the same price or slightly higher price carving out significant market share, potentially 50% of the market. While pricing pressure due to biosimilar could be expected, it's not unreasonable to assume to margetuximab could gross $1 billion of revenue when approved in the first line setting and it would have patent protection for a decade or more.
The Rest of Pipeline Adds Significant Value
Given the potential of margetuximab in breast and gastric cancers, MGNX is significantly undervalued after the recent pullback. Importantly, the company's anti-PD1 antibody MGA012 which Incyte (INCY) licensed in exchange for almost $1 billion in potential milestones and generous 15-24% royalties is currently being developed for multiple indications as a standalone, in combination with other in-licensed antibodies from Agenus (AGEN) and in combination with MGNX's own bispecific antibodies.
Notably, MGNX's CD3xCD123 bispecific flotetuzumab has demonstrated complete responses in AML, which can be further improved by combining with MGA012, as planned.
The potential value of MGNX's first-in-class PD1XLAG3 bispecific antibody MGD013 is worth considering. Along with PD1 and TIM3, LAG3 is another major inhibitory surface checkpoint molecule and numerous in vivo studies show increased response rates when combining anti-PD1 and anti-LAG3 antibodies. Given the preclinical data, it's not a stretch to believe that MGD013 will prove clinically superior to blockbusters Opdivo and Keytruda, as margetuximab is to Herceptin.
MGNX has potentially over-focused on B7-H3. It has an antibody (enoblituzumab), a bi-specific (MGD009), and an antibody-drug conjugate, MGC018, all targeting B7-H3. While the trial data for enoblituzumab has been underwhelming, there may be rational for targeting B7-H3 as it is overexpressed in multiple tumor types, including lung cancer and preclinical data for MGC018 has been quite compelling, although there have been safety concerns for MGD009. MGNX's B7-H3 pipeline is a mixed bag, and it's hard to assign much value to it at present. Upcoming data readouts should help clarify any potential long-term value.
The B7-H3 franchise aside, the SOPHIA trial has proven that margetuximab is a best-in-class molecule. The rest of the pipeline (MGA012, MGD013, and flotetuzumab) are quite compelling, with MGD013 having the best-in-class potential for a checkpoint-targeted bispecific.
Assuming that margetuximab was MGNX's only pipeline candidate and was only ever approved in the 3rd line setting, MGNX is likely still a bit undervalued. Revenues of 175-350 million are entirely reasonable. Based on industry comps of 4x revenue, this would value MGNX at ~$800 million to $1.4 billion, including cash on hand.
MacroGenics is Now an Attractive Acquisition Target
Given the current valuation, one essentially gets the rest of the pipeline for free, including revenues and royalties from MGA012, MGD013, flotetuzumab, the B7-H3 franchise, and a few interesting preclinical candidates including a PDL1xCD137 bispecific. Most importantly, there is strong potential for margetuximab to move to the 1st line and eventually rake in billions in revenues per year. The SOPHIA trial shows that it is truly a best-in-class molecule.
As it stands, MGNX appears to be significantly undervalued at ~$22/share and is a compelling buyout candidate for Roche and others that want to buy best-in-class molecules on the cheap. I would expect a buyout in $40+ range.
Disclosure: I am/we are long MGNX. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.