Bellicum: Data From ASH 2015, Orphan Disease Focus & Sizing the BPX-501 Market

Disclosure: The fund I work for, Aju IB Investment, was a private investor in Bellicum before its IPO and invested further in the IPO. The firm continues to hold all of the shares it has purchased in Bellicum. I wrote this article myself and it expresses my own opinions. It does not reflect the views of my employer. I am not receiving any compensation for this blog post. I have no business relationship with any company whose stock is mentioned in this article. This is not a recommendation to buy any security.

Last week I wrote a piece to try to help anyone interested to get people up to speed on Bellicum and what to look for heading into ASH 2015, given the a major update on the BP-004 trial for BPX-501 in Alpha/Beta depleted (and CD19+ B cell depleted) HSCT. I don’t plan to rehash the background in this post so you may want to read it here. In short, we are looking for improvement in several key outcome measures for these haplo-HSCTs, including (most importantly) non-relapse mortality (NRM, or transplant related mortality, TRM), lower infection rates & infection related mortality, better engraftment rates and potentially control of GvHD (the CaspaCIDe switch could be used in the setting of Grade III or IV GvHD or steroid refractory Grade II GvHD). In malignant patients we also want to be cognizant of the relapse rates, which as I pointed out in my prior post, should be lower than matched HSCT due to the Graft vs. Leukemia (or Lymphoma), or “GvL” effect in haplo-HSCT but that the number of patients was too small.

One poster was presented tonight (the main one) and two more are slated, one for tomorrow on immune reconstitution and one the next day on patients with primary immunodeficiencies with active infections.

Not surprisingly, this poster tonight presented by Franco Locatelli focused on the non-malignant orphan blood disorders. Longer follow-up is needed to assess relapse rates in malignant patients.

I’ll focus on a couple of things in this post:

1. ASH poster review. Did BP-004 update meet our expectations so far (the trial is still ongoing) in the first poster at ASH 2015 (there are 2 more coming in smaller areas of focus)?

2. Abstract focus on orphan blood diseases. There are reasons for this focus, which we will discuss.

3. The addressable market for BPX-501. Adam Feuerstein in his post here says: “BPX-501 could one day change the way haplo-sourced transplants are conducted but the relative rarity of these procedures raises concerns that the commercial market for Bellicum might be too small to matter.” I will make the case that haplo-HSCT could be poised to become the standard for the majority of allogeneic HSCTs. If that is so, the market dwarfs the current market for haploHSCTs. I will also argue that to be a big drug, BPX-501 doesn’t need to address a very large market to be an important drug.

Section 1:ASH poster review

Out of 49 patients enrolled to date, 39 patients have been infused with BPX-501. 37 patients have data 30 days post-HSCT and are included in selected analyses. Let’s look at the key findings on morbidities and mortality: transplant related mortality (non-relapse), hospital stay duration, GvHD, engraftment and infection rates. A great proxy for success is also hospitalization rate. These are all discussed in my prior post as the things to look for in this poster.

The key measure — Transplant Related Mortality.

Honestly, this is the prize picture. Though a small n, this is what you really want. This shows nobody died from the transplant, primarily from infections. This analysis below only looks at non-malignant orphan blood disorders.

Hospital stays were much shorter

These rates with BPX-501 were far shorter in length and re-admission rates also lower. This gets to better outcomes.

GvHD Rates seem on pace with control

I continue to believe that we will see some Grades III or IV at some stage. These patients are getting mature T cells added back in the form of BPX-501. Remember, the control has no mature T cells, and therefore is unlikely to have Grade III or IV GvHD. Nevertheless, I don’t feel the low rates of GvHD changes the message. Also remember, some of these patients were at a lower dose, in the escalation portion and less likely to get Grade III or IV GvHD.

Immune reconstitution and Engraftment

Time to engraftment looks to be the same, though the raw data of number of failures would be nice to see. The immune reconstitution appears to be much better with BPX-501. The next two posters tomorrow and Monday on the trial will add more color to these statistics. I expect to see more about this in tomorrow’s poster as well as, to some degree, Monday’s.

Section 2: Orphan Focus at ASH— Regulatory Advantages

The poster tonight by Locatelli focused largely on BPX-501 enabled haploHSCT for orphan blood diseases. There are several reasons for this. First, the data are excellent and set a high standard that will be exceptionally hard for gene therapy companies to beat in orphan blood diseases. Second, the regulatory path might be a bit easier than for malignant disease. Rather than looking at relapse rates, which may require longer term follow up and a larger sample set, patients with non-malignant disease who engraft and are healthy with immune reconstitution after they are discharged days are essentially cured. FDA will be satisfied these are cures, the regulatory pathway is fast, follow-up is short and the evidence will be strong that BPX-501 enabled HaploHSCT is a safe and effective cure. Of note, follow-up for gene therapies is potentially significantly longer.

Unlike gene therapy in sickle cell or thalassemia, for example, where one hopes there is enough expression to convert a patient to transfusion independence, and the durability of the cure is in question. If a patient is cured by HSCT — that patient is fully cured. Their marrow makes all blood cells with the right hemoglobin. If a patient is sick enough to undergo gene therapy, you can be sure that patient qualifies to undergo a haploHSCT enabled by BPX-501. With 4/4 cures in thalassemia major (β0/β0) patients with no adverse events, Bellicum does not need a special trial in these harder to treat thalassemia major patients, like bluebird has announced.

Here are the data that bluebird are showing LentiGlobin isn’t getting there yet:

BPX-501 is 18 for 18 in safely curing orphan blood disorders, including the 4 difficult thalassemia major patients. Over 60 diseases should be covered under a single regulatory approval for BPX-501 in the cure of these diseases.

This example also gets, in part, to Adam Feuerstein’s comment about relative rarity of the haplo-HSCT procedure and market size (more in Section 3). If you appreciate that this could be preferable to gene therapies for orphan blood disorders, including sickle cell and beta thalassemia (not to mention 60+ other orphan blood diseases), you would have to argue BPX-501 has at least a multiple of the market size of bluebird bio in terms of number of potential procedures in these diseases alone, never mind malignant patients needing HSCT. It begs the question, in a practical world, what doctor or patient would choose a gene therapy over BPX-501 enabled HSCT?

Whatever you project for sales of LentiGlobin, you have to consider BPX-501 enabled HSCT for the same patients.

Section 3: Haploidentical HSCT is Becoming the Superior Choice for HSCT (BPX-501 Market Sizing)

Haploidentical HSCTs are quickly being recognized as not only a potential option for people who don’t have a matched donor but also becoming recognized as the preferred method of HSCT over matched donor for malignant disease and particularly matched unrelated (MUDs) for a variety of reasons.

Feuerstein says “BPX-501 could one day change the way haplo-sourced transplants are conducted.” I want to put some numbers to that. But first, it helps to hear from Dr. Stefan Ciurea (MD Anderson), a leading HSCT KOL in the US, who says the following in this recent review article in Blood:

“Haploidentical hematopoietic stem cell transplantation (HaploHSCT), with progenitor cells from HLA-half-matched first degree related donors (siblings, children and parents), could revolutionize hematopoietic stem cell transplantation as it expands this form of treatment to approximately 40% of patients who do not have an HLA-matched donor. This need is particularly acute in developing countries, which usually do not have an unrelated donor registry and/or cost is a major issue in acquiring unrelated donor progenitor cells. Advantages to Haplo HSCT include almost universal (more than 95% of patients will have a halfmatched related donor) and immediate availability of donor progenitor cells, the opportunity to select the best donor among family members to minimize treatment-related mortality, decrease relapse rate and improve outcomes, and the possibility to collect donor cells for cellular therapy post-transplantation, with the goal to enhance the anti-tumor effects of the graft. Despite its potential advantages, until recently, high donor-recipient HLA-histoincompatibility has proven very difficult to overcome.”

Locatelli highlights the same in the Bellicum today PR here:

“Historically, haplo-transplants had to be given without T cells to avoid GvHD, increasing the risk of deadly infections and delayed immune recovery, and relapse in patients with malignant disease. An approach that addresses these risks without elevating the GvHD risk could shift the standard of care, making haplo-sourced transplants — almost always available from a family member — an attractive option for patients”

The Numbers — Allo HSCT Market is Almost Entirely Available to BPX-501

HaploHSCT could take over most of these categories:

• HaploHSCT is the only option for the No Match group (~14k).
• BPX-501 enabled haploHSCT (if safety proves consistent with data so far) should be a better option for MUDs, as we will discuss (~13k).
• Matched related in a malignant setting (90% of all HSCT) should move to haplo for GvL effect (90% of ~8k).
• BPX-501 would be a natural inclusion in the current haploHSCT paradigm (~2k)

As such, there is a very sound argument that haploidentical HSCT, especially enabled by BPX-501, could become the dominant form of HSCT in the next 3–5 years.

Here are the reasons:

1. No Match. Approximately 14,000 patients who don’t have a match today should now get one with a safer haploidentical HSCT because 95%+ of people have a parent or sibling who can be a haploidentical donor. These people currently are not transplanted. The three analysts covering Bellicum suggest a price of $150k per procedure is reasonable. You will need 7,000 TRxs per year to be a >$1B drug at that price. That means if you only penetrate half of the patients in this “No Match” group, those that currently need but can’t find a match for an HSCT, you reach that threshold.

2. MUDs. 55% of patients who currently undergo an HSCT have MUDs. MUDs have been crucial to saving many lives historically. However, as haploHSCTs are becoming safer, MUD HSCTs are becoming obsolete. MUDs costs much more than a family donor and take a lot more time, up to 3–6 months. That is time patients often don’t have. Also, the chance of finding a MUD for a non-Caucasian individual varies widely among different ethnic minority groups. Critically, most people are not aware that it is illegal to manipulate the marrow of a donor in a MUD HSCT. In other words, you can’t do an Alpha /Beta T Cell depleted MUD HSCT like you can with a haploidentical HSCT. It isn’t allowed. And so you can’t use BPX-501 the same way, if at all.

Given that there is still a high rate of GvHD rate (as high as 41% aGvHD) in MUDs, not to mention other morbidities, it is likely MUDs carry an unacceptable risk vs. BPX-501 enabled haplo HSCT. Given these MUD disadvantages, one can envision a doctor much preferring a quicker, faster, safer BPX-501 enabled haplo HSCT over a MUD HSCT.

3. Matched Related. 37% of allogeneic HSCTs today are performed with matched related HSCs. You might be surprised to hear how high the GvHD (up to 40% aGvHD) and other morbidity rates are in this group are. T cell depleted HSCT, and especially BPX-501 enabled haploHSCT are much better.

However, as I have mentioned before, 90% of allogeneic HSCTs today are for malignant disease (ALL, AML, CLL, NHL etc.). Malignant patients are likely better off with a haploHSCT due to the aforementioned GvL effect. HaploHSCT recipients have lower relapse rates, often over 50% lower than when transplanted from matched related donors. This benefit is important and the case for haploHSCT is made stronger by the the lower transplant related morbidity and mortality in t cell depleted haploidentical HSCT.

So you can see that haploidentical HSCT, in its current safer form, addresses a significant unmet need across nearly every allogeneic HSCT type, including MUDs, donors with no match, and matched related in a malignant setting. There is a good case that the market for haploidentical HSCT, and hence BPX-501, is really almost all allogeneic HSCTs that could be performed, or about 35,000–37,000 per year.

Perhaps worth mentioning is that the market for cures of orphan diseases is as large as the market thinks for bluebird and others, these patients would likely seek BPX-501 enabled haploHSCT. That incremental market is not factored into the above.

I probably have one more most in me on this topic. So if I have time this week I will try to do a valuation post. Several people have asked for it. I’ll include a DCF that people can take and manipulate if they don’t agree with the assumptions.