Description

As rising rates have driven two consecutive years of capitulation by investors who bought into biotech names at much richer valuations during the 2020-21 XBI peak, this is creating some asymmetric opportunities for value investors willing to step outside our usual comfort zones and accumulate some attractive option-like positions in biotech names trading below net cash and nearing an upcoming hard catalyst that could result in extreme asymmetric upside.

IO Biotech (IOBT) is another bombed-out microcap clinical-stage biotech currently trading at a deep discount to net cash, and a fraction of its valuation at a very recent secondary equity raise in August, which was fully subscribed by well-regarded biotech specialist investors during an already dismal fundraising environment, and leaves the company fully funded through 2025.

IOBT was previously recommended 18 months ago by VIC member mack855; I think mack's writeup is well worth reading and gives a thorough overview of the previous positive Phase 2 PoC data with IO102-IO103, making a reasonable case for a strong probability of success for the ongoing Phase 3. Both the stock and the writeup appear to have gotten absolutely no love since then.

With IOBT now at a deep discount to net cash and its recent $2/share secondary equity raise, and only six months away from a prespecified Phase 3 interim analysis that will serve as a hard catalyst for the stock, I think a small position at these levels is an increasingly attractive option which could have extremely positive expected value using conservative assumptions. The generous net cash cushion could additionally provide investors with meaningful residual liquidation value in a downside scenario in which the interim analysis is clearly negative.

In looking carefully at disclosures from the currently ongoing Phase 3 study (evaluating IO102-IO3 plus Merck’s PD-1 mAb Keytruda versus Keytruda alone), I think it is particularly interesting that an analysis of the projected time to reach 226 events of disease progression or death is consistent with a substantial prolongation of progression-free survival (PFS) in the study population relative to that previously reported with Keytruda monotherapy. I do not view this as conclusive, but think it provides useful circumstantial evidence further supporting continued observation of a benefit in PFS, making a small position in IOBT particularly interesting and timely ahead of its interim readout.

IO102-IO103 overview

IOBT’s lead asset IO102-IO103 is an interesting first-in-class immuno-oncology combination therapy currently progressing through the later innings of a fully enrolled Phase 3 pivotal study for first-line treatment of advanced metastatic melanoma (MM), which is now just 6 months from a prespecified interim analysis of efficacy that will serve as a significant event-driven catalyst for the stock. IO102-IO103 is a subcutaneously administered bivalent cancer vaccine using peptide sequences designed to engage and activate human T cells specific to PD-L1 (a ligand of PD-1, targeted by blockbuster immune checkpoint inhibitors such as Keytruda and Opdivo) and indoleamine 2,3-dioxygenase (IDO), another checkpoint protein highly upregulated in melanoma. As a novel and well-tolerated first-line MM treatment coadministered with Merck's Keytruda (pembrolizumab), IO102-IO103 has blockbuster potential if this pivotal study is able to replicate the promising efficacy and safety data demonstrated in its previously completed Phase 2.

Keytruda itself notched $21B in sales last year (of course, clearly a different beast with approvals across multiple indications), and the market size for melanoma therapeutics is projected to reach nearly $7.5B annually by 2029. To provide a conservative back-of-the-envelope benchmark estimate for purposes of calculating expected values with a generous margin of safety, we can say that capturing even 10% of this market would result in peak revenues in excess of $750MM and easily support a valuation of $2B or more.

As a relevant direct comp, Opdualag (nivolumab+relatlimab combination therapy) was approved just last year for first-line treatment of advanced melanoma with tumor cell PD-L1 expression of less than 1%, and has already reached $166MM in sales in the latest quarter. BMS is projecting Opdualag revenues of $4 billion annually by 2030. To be honest this seems overly ambitious to me, but the fact that Opdualag itself is already nearing a $750MM run rate so soon after launch supports that $750MM in peak revenues would be an extremely conservative estimate for a novel first-line treatment for MM, particularly one with the efficacy and tolerability advantages over Opdualag and PD-1 monotherapy implied by IOBT’s completed Ph2.

To temper this optimism a bit, IOBT’s positive Ph2 data showed overall response rate (ORR) and progression-free survival (PFS) outcomes that would be an impressive improvement on current standards of care if confirmed in Ph3, but these data originated from a single n=30 open-label study, and there is of course considerable remaining uncertainty around the outcome of the confirmatory Ph3 that will only be resolved through completing the study. Drug development is a hard and relentlessly empirical game; there's certainly been a history of promising Ph2 results from a variety of immuno-oncology and cancer vaccine approaches that have failed to pan out in subsequent Phase 3 studies, sometimes for reasons that are never fully understood. Keeping in mind both the irreducible uncertainty and very large potential upside, a straightforward expected value approach using a series of conservative evidence-based assumptions is a reasonable way to objectively identify attractive and mispriced bets.

Net cash, burn rate, and downside case

In the current situation, what makes IOBT a particularly interesting and timely event-driven play is the combination of the deep discount to net cash and the rapidly approaching prespecified efficacy interim analysis in Phase 3.

This effectively creates a near-term hard catalyst where in the success case, any confirmatory signal of efficacy could create massive near-term upside for the stock given the scale of the commercial opportunity in first-line MM, and conversely in the failure case, a clear signal of a *lack* of efficacy has at least some potential to drive early termination for futility, and a significant return of capital to shareholders while much of the current net cash cushion remains.

If the prespecified ORR analysis in 6 months (which should provide highly meaningful evidence of a signal in either direction given a full year of data on 255 patients) shows that the efficacy seen in Phase 2 clearly is not replicating and the active arm is not separating from placebo, this would provide a clear no-go signal that (I would argue) should motivate any rational sponsor or data monitoring committee to recommend stopping the study for futility.

On August 9 of this year, the company completed a private placement of 37MM shares and warrants (struck at $2.47), priced at $2.025 per unit for net proceeds of $71.9MM. This was fully subscribed by biotech specialist and HF investors including Lundbeckfonden BioCapital, Kurma, Vivo Capital, Armistice Capital, Marshall Wace, Samsara BioCapital, Novo Holdings, and Pivotal Life Sciences. The company has not sold any shares under an ATM adopted in Feb 2023, opting instead to complete the August secondary at $2.

As of the 9/30/2023 10-Q, IOBT had current assets of $168MM ($165.5MM of cash and equivalents + $2.7MM prepaid expenses) against $15MM total liabilities, for $153MM of net cash. TTM cash burn rate is $66MM and TTM operating loss is $72MM. With the interim readout in June 2024, let's charge them a full $90MM or five quarters of TTM operating losses for severance and wind-down costs through liquidation. This gets us to $63MM, or two thirds of the current market cap in a downside case where a failed interim analysis winds up the company.

Of course, some biotech managements are notorious for resisting liquidation and returns of capital where this makes sense for shareholders. In the case of IOBT, reasonably sophisticated biotech VC and crossover investors collectively own a large stake (bought at much higher levels) relative to management, and Vivo Capital and Lundbeckfonden are both on the Board. Together with independent Board members and an independent DMC, there is enough adult supervision here to add appropriate discipline in scenarios where the interim data clearly does not support continued development.

There's also certainly scope for scenarios where the interim analysis is equivocal and the Board is persuaded to run the study out to the primary endpoint of PFS on the full population, projected in 2H2025. In some of these scenarios patience may be rewarded with success in the final analyses, and in some scenarios it may be rewarded with a terminal value of zero. For simplicity and conservatism, I’ll present expected value scenarios exploring a range of probabilities of the success, liquidation, and zero-value outcomes.

Event-driven value investing in biotechnology – an expected value framework

In the not too distant past, biotech specialist investors piled onto IOBT and many similar names at vastly richer valuations than today’s. Following the subsequent extreme dislocation in the XBI, there are interesting opportunities to accumulate positions in some of these names that represent mispriced options with potentially extreme asymmetric upside. Traditional value investors will understandably be put off by the cash burn and irreducible uncertainty, resulting in many net cash biotechs being almost universally neglected in the current environment.

How to usefully think about valuing an event-driven situation like this, where you can reasonably develop an informed intuition that some outcomes contribute highly positive asymmetric expected value that make a small position worth a look, but the irredicible uncertainty is real enough that coming up with exact probabilities and values for each case amounts to false precision?

Well I'm taking a page from our pal Cathie Wood's open-source model of TSLA, and making available a simple expected value model where you can plug in the assumptions you're comfortable with underwriting and build a reasonable understanding of whether you think this setup delivers attractive asymmetric expected value.

Valuation estimates for the success case and liquidation case are based on the conservative assumptions described in the previous section.

As an objective starting point for establishing the base rate probability of a successful outcome, the following Biotechnology Innovation Organization (BIO) whitepaper presents empirical findings on success rates and ultimate likelihood of approval by study phase and disease area, generated from a decade of longitudinal data on 9704 drug development programs.

The overall likelihood of approval for oncology programs currently in Ph3 stands at 43.9%; further breaking down the data puts the Ph3 likelihood of approval at 39.8% in solid tumor indications and at 48.2% for immuno-oncology products.

It’s certainly possible to make a case for a higher than average probability of success. The BIO whitepaper itself goes on to present results from a machine learning algorithm trained on the dataset (pg20), where breakthrough therapy designation and a positive Ph2 outcome (both of which IO102-IO103 has) further increment the base rate probability of approval of a Ph3 oncology asset by +20.6% and +12.8% respectively. But the point here is to be conservative, and recognize the Moneyball / behavioral finance axiom that investors are highly likely to pay insufficient attention to base rates of probabilities in either direction, often leading to significantly mispriced bets when implied probabilities are far off from the objective data on base rates of success. Let’s use the lowest real-world base rate above of 39.8% in solid tumor indications (call it 0.4) and present a range of expected values for probabilities anchored at, above, and below this empirical rate.

Looking over the scenarios below, at a high level, I’d make the case that these support an extremely attractive expected value using empirical data on rates of historical Ph3 success as a benchmark, and also supports an expected value above the current share price across a range of fairly draconian assumptions, including a probability of Ph3 success as low as 5 percent.

As would be expected, the upside case contributes the vast majority of expected value, but possibility of $0.95/sh liquidation recovery enables overall expected value to be positive down to very low probabilities of success. Scenarios assigning low or zero probability of cash liquidation proceeds after a failed interim still show positive expected value for probabilities of success as low as 0.1. This is enough to convince me that this is an attractive option-like opportunity with positive expected value.

IO102-IO103 drug development program

The previous July 2022 recommendation by mack885 does a good job in presenting detailed background on the IO102-IO103 program and results of the prior Ph2 trial; the company’s 10-K and investor decks include a comprehensive overview of the Ph2 results with informative comparisons to the topline data from other recent MM trials. I’ll aim to provide a focused non-technical summary of the key data and subsequent developments relevant to underwriting the probability of success of the ongoing Ph3 clinical trial.

The ongoing Ph3 study is designed as a fairly straightforward 2-arm randomized controlled parallel-group trial, with patients randomized 1:1 to receive IO102-IO103 plus Keytruda versus Keytruda alone. The primary endpoint is progression free survival (PFS); key secondary endpoints include overall response rate (ORR) and overall survival (OS). The protocol includes a prespecified interim analysis of ORR which is scheduled to occur one year after 225 patients have been randomized. IOBT had randomized 225 patients as of June 2023 (implying availability of interim results in June-July 2024), and fully enrolled the trial reaching 380 patients in November 2023.

The IO102-IO103 program has received the coveted Breakthrough Therapy designation from FDA based on its positive Ph2 results. This enables frequent communications from FDA to ensure data being collected can support approval, and also includes all features of the FDA’s Fast Track program such as eligibility for the Accelerated Approval pathway and Priority Review of a future BLA filing. The company has stated that discussions with FDA have included the potential for pursuing accelerated approval as early as next year if the interim analysis of ORR is positive, followed by full FDA approval and EMA approvals based on the primary endpoint of PFS, projected to read out in 2H2025.

For full transparency, I should mention the statistical threshold for formally declaring an early success supporting accelerated approval on this interim is very stringent at p<0.005, and I don’t necessarily expect ORR results to reach this very high level of significance. The interim look is designed this way because of a quirk of clinical trial statistics called alpha spending, which entails that an early interim analysis using an extremely stringent threshold like this will enable subsequent statistical analyses to be done at no meaningful penalty, using a traditional threshold near p<0.05. The bottom line is that a homerun early success on ORR at p<0.005 will support accelerated approval in 2H2024, while a less resounding success that shows clear benefit in ORR at p>0.005 will support filing via the traditional full approval pathway using the full analysis of the PFS primary endpoint in 2H2025.

Analysis of emerging data on PFS relative to pembrolizumab monotherapy

Interestingly, on June 14 of this year IOBT announced an increase in the total number of patients to be enrolled in the trial from 300 to 380. This was done to accelerate the projected time to reach the primary endpoint of PFS (which is prespecified to be assessed when 226 events of disease progression or death are registered in the trial). Because unblinded data on differences between the active and placebo arms cannot be reviewed at any point prior to the interim analysis, the fact pattern suggests that the emerging blinded data on PFS in the overall study population may have been longer than previously assumed. In other words, this is circumstantial evidence that as the study has progressed, overall rates of disease progression or death are continuing to show a decrease relative to historical rates on pembrolizumab monotherapy.

In the KEYNOTE-006 pivotal trial of pembrolizumab versus ipilimumab in advanced MM, 6-month PFS rates in the pembrolizumab arms were 46.4% and 47.3%, and 12-month PFS rates on pembrolizumab monotherapy were each below 20% as shown in the Kaplan-Meier plots below. In the IOBT Phase 3, 225 patients were enrolled as of June 14 of this year, but the primary PFS endpoint assessing 226 events of progression or death is currently projected to be reached in the second half of 2025, over two years from having enrolled 225 patients (many of whom had enrolled far earlier than that 6/14/23 milestone date). It appears increasingly likely that these PFS endpoint projections are consistent with a substantial prolongation of PFS relative to that previously reported with pembrolizumab alone.

If rates of 6-month PFS in the overall study population were similar to the 47% reported for pembrolizumab monotherapy in KEYNOTE-006, more than half of the 225 patients enrolled as of the June 2023 milestone should have already progressed. By the June 2024 interim analysis of ORR, over 80% or 180 of these patients should have progressed.

Having reached full enrollment of N=380 last month, it is apparent that 380-225=155 new patients were enrolled between June and November of this year. If 6-month PFS rates were similar to pembrolizumab data in KEYNOTE-006, more than half of these patients (155*0.5=78) should also have progressed by the June interim. These progression rates would be consistent with at least 180+78=258 patients already having experienced disease progression or death by the June 2024 interim, far in advance of the company’s projection for 226 events to be reached in the second half of 2025.

I do not view this at entirely conclusive (we don’t know all details of the enrolled study population relative to KEYNOTE-006), but in addition to follow-up data from the completed Ph2 further described below, this does provide interesting and meaningful circumstantial confirmatory evidence further supporting an observation of incremental benefit over historical rates of PFS in the ongoing Ph3.

IO102-IO103 Phase 2 follow-up data

Results from a new January 5, 2023 data cut with an additional 15 months of patient follow-up for the Ph2 study of IO102-IO103 + nivolumab have been reported, and continue to be encouraging. With minimum follow-up time of of 49.8 months, median overall survival was not reached, median PFS was 25.5 months, 50% of patients (15/30) achieved a complete response, and the ORR was 80%.

Data on PFS and overall survival compare favorably to current standards of care and the results of recent MM clinical trials as outlined below.

Safety data specific to IO102-IO103 have been fairly benign relative to the indication, with the most common treatment-emergent adverse events being local reactions at the site of injection. In the Ph2 trial where IO102-IO103 was coadministered with BMS’ PD-1 agent nivolumab, rates of CTCAE grade 3+ adverse events were broadly similar to rates with PD-1 agents alone in the KEYNOTE-006 and CHECKMATE-066 studies, and significantly lower than PD-1/CTLA-4 combination therapy with nivolumab+ipilimumab in CHECKMATE-067.

Of note, non-first-line MM patients who were anti-PD-1 antibody refractory and enrolled in a separate cohort of the earlier Ph2 had no significant response to therapy, which the company attributes to the vaccination approach being suited to first-line rather than late-line/refractory treatment. This could be plausible given the mechanism (if stimulating T cell activation versus PD-L1 is less relevant for tumors that have evolved resistance to anti-PD-1), but does mean that future use is likely appropriate in first or early line settings. A separate ongoing Phase 2 basket trial in several non-melanoma solid tumor indications has released interim data showing several partial responses in non-small cell lung cancer and squamous cell carcinoma of the head and neck (ORR ~50-60%); these data appear too early to be conclusive. The company is also investigating IO102-IO103 for melanoma and head & neck cancers in the neo-adjuvant (preoperative) setting, and has two preclinical-stage solid tumor vaccine programs targeting arginase-1 and TGF beta 1. I’m assigning zero value to these additional programs and indications for simplicity. At a high level, a clear failure for IO102-IO103 in the large ongoing MM Phase 3 should effectively invalidate the rationale for further development of these indications and candidates and motivate liquidation if the Board is rational, while a success could imply additional value above the MM-specific upside case valuations above.

Conclusion

Given the scale of the commercial opportunity for a new first-line treatment for MM, attributing even a small probability of success to IOBT’s ongoing Ph3 program implies an expected value far in excess of the current share price, with some additional protection in the downside case provided by IOBT’s net cash cushion if the upcoming interim analysis shows a clear signal of futility. An analysis of the projected time to reach 226 events of disease progression or death in the ongoing Ph3 appears consistent with a substantial prolongation of PFS in the study population relative to historical data with Keytruda monotherapy, providing additional interesting circumstantial evidence supporting continued observation of a potential benefit on this endpoint. Overall, the data support a small position in IOBT is a timely event-driven value investment with positive expected value heading into its upcoming Ph3 interim analysis.

Catalyst

Upcoming prespecified interim analysis of ORR in early 3Q2024