Sangamo BioSciences (NASDAQ:SGMO) was founded in 1995 ostensibly to commercialize zinc finger nuclease technology. Twenty years have passed since its founding and there hasn’t been a single drug approval nor is one imminent. In twenty years, Sangamo hasn’t even advanced a drug past phase 2. What it has done is treat its investors to a steady diet of insider sales, misleading press releases, and dilution.
To find evidence of a Sangamo executive stock purchase you have to go back to August 2010, when CEO Edward Lanphier and CFO Ward Wolfe dug deep to purchase 10,000 and 9,000 shares at $3.00. That pales in comparison to $20.5 million in stock sales from the pair ($17 million for Lanphier and $3.5 million for Wolfe). Sangamo executives have excelled at selling company stock as well with 6 secondary offerings since it became a public company.
Throw in 19 consecutive years of losses since inception and one would think this would be enough to send investors running to the hills. Instead, Sangamo proves P.T. Barnum correct with a market capitalization of over $1.2 billion. Let’s explore why Sangamo’s approach to genetic therapy is fundamentally flawed. First we will examine its efforts in B-thalassemia. Then we will deconstruct Sangamo’s purportedly successful results in HIV with its SB-728 therapy.
Zinc Finger Nuclease in hemoglobinopathies
Sangamo and partner Biogen are developing a zinc finger nuclease (ZFN) based approach to the treatment of sickle cell disease and B-thalassemia. These programs have not yet entered the clinic. Both diseases result from mutations in the proteins that form hemoglobin – the oxygen carrying component of red blood cells. To understand the folly of Sangamo’s approach we’ll need a brief review of hemoglobin synthesis.
Hemoglobin is a tetramer that typically contains two alpha chains and two beta chains. Each individual has two copies of each gene; one from each parent. A mutation that renders one of these proteins defective is generally asymptomatic. This is the so-called heterozygous or carrier state. A mutation in both copies of either gene leads to disease.
Fetuses use a slightly different form of hemoglobin which consists of two gamma chains and 2 alpha chains. Fetal hemoglobin has a higher affinity for oxygen and facilitates the transfer of oxygen from the mother’s blood to the fetus. Fetal hemoglobin is generally present at <1% in adults unless they have a hemoglobinopathy such as sickle cell disease or thalassemia in which case it may be present at greater concentrations. Higher concentrations of fetal hemoglobin may ameliorate some of the symptoms of thalassemia because fetal hemoglobin is not mutated in these patients and it is capable of carrying oxygen. We say “may” because even though some patients with hemoglobinopathies have higher levels of fetal hemoglobin, there has never been a trial to demonstrate this. This is simply an observed correlation.
The fatal flaw:
Assuming that one could introduce DNA into a cell, the most obvious approach to treating thalassemia is to introduce a normal (wild type) copy of the defective gene. This is exactly the approach being taken by Bluebird bio (NASDAQ:BLUE). While there is no guarantee that this approach will work, it is direct and Bluebird is conducting proper studies to test it.
To call Sangamo’s approach to the treatment of thalassemia indirect is an understatement. Sangamo has decided rather than attempting to fix the underlying problem – namely, a mutation in the beta globin gene – it will attempt to ameliorate the symptoms it causes by increasing the amount of fetal hemoglobin produced. Since fetal hemoglobin is normally repressed in adults, Sangamo’s strategy involves “de-repressing” fetal hemoglobin by inactivating its repressor protein BCL11a.
Sangamo has engineered a zinc finger nuclease that can be introduced into red cell precursors. This ZFN is designed to cleave and inactivate the BCL11a gene. Of course, inactivation of one copy of BCL11a is not enough. Instead, both copies of the gene must be inactivated for this to have any chance of working. After confirming that BCL11a has been inactivated, SGMO will re-introduce the modified cells back into the patient.
The barrier to gene therapy is not selecting the right construct. The scientific community understands what gene is mutated and knows how to create a wild type copy for introduction into the cell. The barrier to gene therapy has long been successfully introducing the expression construct into the cells and effecting persistent gene expression. If one solves the problem of introducing DNA into cells, then what possible reason could there be for introducing anything more than a good copy of the mutated gene? Sangamo’s persistence with the zinc finger approach suggests a slavish desire use the technology on which they became a public company 15 years ago in spite of the fact that more direct approaches exist. When it comes to altering someone’s genome, less is more.
Does this sound familiar?
In 2009, Sangamo tried a similar approach to treat peripheral neuropathy. The company hypothesized that Vascular Endothelial Growth Factor also known as VEGF was deficient in patients with diabetic peripheral neuropathy. Rather than injecting the VEGF protein or a gene therapy cassette that caused cells to express VEGF, Sangamo introduced a plasmid (SB-509) that would hopefully express a zinc finger protein that they thought would direct the cell to synthesize VEGF. On October 3, 2011, Sangamo announced that the phase 2 study of SB-509 failed. The parallels between this failed approach and the current approach to treating hemoglobinopathies are striking. In both cases, a more direct approach is available. In both cases, we believe Sangamo uses an indirect route simply to make use of the zinc finger technology.
What could go wrong?
The gene Sangamo plans to inactivate BCL11a, was initially discovered in aggressive B-cell leukemias and is commonly overexpressed in Hodgkin’s disease. In fact, the name BCL comes from B-Cell Lymphoma. Perhaps more troubling is the observation from murine studies that cells lacking BCL11a can induce lymphomas in recipient mice. Will the same be observed in humans? We won’t know until Sangamo exposes enough patients. You have to ask yourself why on Earth someone would knowingly delete a gene that is known to have tumor suppressing activity in mice.
Assuming Sangamo can achieve BCL11a deletions in red blood cell precursors and that these cells when re-introduced into the patient don’t cause cancer, they might be able to increase the levels of fetal hemoglobin and potentially reduce the symptoms of thalassemia. Did we mention that this approach requires that the ZFN modified cells be re-introduced in the context of bone marrow transplant conditioning? Bone marrow transplants are potentially curative for thalassemia and sickle cell disease provided one can find a matched donor so the Sangamo approach, should it meet all of the aforementioned hurdles, does eliminate the need to find a matched donor.
The real question to ask is why anyone would bother with such an indirect approach that at best might reduce symptoms when the gene therapy approach offered by Bluebird promises to effect a cure?
SB-728 in HIV: No evidence of success
We believe Sangamo is making the same mistakes with HIV. Like with B-thalassemia, Sangamo is taking an indirect and circuitous route in its HIV treatment, one that we don’t believe makes scientific sense. Instead of going after the HIV virus directly, Sangamo’s SB-728 works by manipulating the T-cells of HIV patients. The HIV virus works by attaching to receptors on healthy T-cells (a type of white blood cell), entering the cell, and eventually killing the T-cell and impacting the patient’s ability to fight infections. There are several co-receptors on the T-cell; CD4, CCR5, and CXCR4 are three well-known and established ones. Sangamo’s SB-728 supposedly works by causing a mutation in the CCR5 receptor, thus making the cell immune to HIV. It does not target the HIV virus directly. It simply tries to boost a patient’s immunity to HIV.
So far, the results from several phase 1/2 trials for SB-728 have been disappointing, but one wouldn’t be able to tell from Sangamo’s enthusiasm. We believe that Sangamo has cherry-picked and highlighted a handful of patients in every press release and has not mentioned the many patients who have failed therapy. A closer examination shows there is nothing to be excited about. Sangamo likes to point out that 6 patients across their phase 1 and 2 trials have been on extended Treatment Interruption (TI) and remain off highly active antiretroviral therapy (HAART). However, that’s 6 patients across a total of 40 patients treated, which is only 15%. Why should anyone be excited about a 15% response rate in a treatment? If Sangamo bothered to include a control arm, one might be able to objectively assess if SB-728 is doing anything but the last time Sangamo included a control arm was in the phase 2 trial of SB-509 and we know how that worked out (Sangamo BioSciences Announces Phase 2b Trial of SB-509 in Diabetic Neuropathy Did Not Meet Key Study Endpoints). The only thing Sangamo appears more adamant about than using their zinc finger technology is omitting a control arm.
Furthermore, no single patient has been “functionally cured,” a phrase that Sangamo uses liberally in every press release in describing SB-728. The patients that are on extended TI do not have “undetectable viral loads.” Instead, they have demonstrated “continuing control of circulating viral load at low levels.” In fact, each of the 3 subjects that Sangamo has touted as having reached “undetectable viral loads” only reached undetectable viral loads for a transitory period of time. All 3 later saw an increase in HIV viral load to detectable levels. SB-728 did not provide a “cure,” as Sangamo likes to say. We’ll highlight one of the subjects from the SB-728-209 Cohort 5 phase 2 trial to make Sangamo’s sleight of hand clearer.
It’s clear that this patient regressed and viral loads (VL) began to rise significantly before 31 weeks. No updates on the patient have been provided since May of 2014.
The goal of HIV therapy is not to be off drugs and have low levels of disease. The goal of HIV therapy is to have an undetectable HIV viral load and remain so. This has not happened in any of Sangamo’s 40 SB-728 patients. That is a 0% success rate. The patients that remain on treatment interruption have detectable HIV viral loads and are at significant risk of an increase in viral loads. This shows that Sangamo’s SB-728 is quite far from becoming a “functional cure” and may never be one.
Sangamo is now downplaying SB-728 efficacy
As we have seen SB-728’s trial results deteriorate, Sangamo has changed its language around SB-728, downplaying its potential. In earlier press releases, Sangamo described SB-728 as a potential “functional cure.”
In later press releases, Sangamo started describing SB-728 more as a treatment that allows “functional control” of HIV and not a “functional cure.”
Finally, at the Sixth International Workshop on HIV Persistence during Therapy in 2013, Sangamo stated that SB-728 is no longer being developed as a “functional cure” for HIV, which is what the market is so enthusiastic about.
Existing HAART therapy provides undetectable viral loads
It is clear from Sangamo’s own words that Sangamo realizes SB-728 is not a potential “cure” for HIV and now must be positioned as a potential treatment for functional control of HIV. However, there already exists a treatment today that allows for incredible control rates for HIV. In order for SB-728 to achieve any sort of success for the treatment of HIV, it must perform better than today’s standard of care (SOC), which is HAART. Studies of patients on HAART have shown significantly higher rates of response and undetectable viral loads. Well-established drugs like Stribild have shown 94% and 93% of patients reached undetectable rates of viral load2. These types of response and success rates are common with HAART.
In fact, HAART works so well that in SGMO’s phase 1 study, Sangamo had to abandon the portion of the study that targeted HIV patients who had failed HAART because they couldn’t find enough patients. That’s how successful HAART is!
It’s also important to note that the patients who reached undetectable viral loads after receiving SB-728 were all CCR5-Δ32 heterozygous patients, which means they were naturally more resistant to HIV therapy. About 10% of the population is heterozygous, though closer to 20% in people of European ancestry. In Sangamo’s phase 2 portion of the SB-728-902 trial (Cohort 5), Sangamo purposefully chose only CCR5-Δ32 heterozygous patients. Of the 10 patients treated, only 2 ever reached undetectable viral load levels at any point in time, and only 1 remained on TI longer than protocol. Given this population was hand-picked to show the most success, we’re surprised no one achieved successful long-term undetectable viral loads. This is further indication that SB-728 does not work.
Sangamo has been unable to find a partner for SB-728
Lastly, Sangamo has been actively searching for a partner for its HIV program. They have been looking for years, and despite consistent press releases touting the success of its early HIV programs, no partnership has been signed. On the 3Q 2014 earnings call, CEO Ed Lanphier said that they have been in “active discussions” with partners. Yet, on the 4Q 2014 earnings call, in response to a question whether a partnership would happen soon after data from yet another phase 2 study for SB-728 would be announced in late 2015, CEO Ed Lanphier admitted in the Q&A session that no partner was close.
We think this shows that experts have seen the HIV data, have come to the same conclusions we have, and do not want to partner with Sangamo.
1. Sangamo has a 20-year track record of drug development failure. Management has enriched itself at the expense of shareholders. If you bought SGMO stock at $15 per share on its IPO in April, 2000, you would have made 1.3% per year. Unfortunately, the share count increased 5-fold in that same period of time.
2. Sangamo’s approach is indirect and at best can ameliorate symptoms as opposed to gene therapy approaches which simply replace the defective gene and offer the hope of a cure.
3. Sangamo’s purportedly ground-breaking results in HIV with SB-728 do not stand up under close examination. SB-728 has not shown any indication that it can provide a “functional cure” for HIV, does not compare well to today’s standard of care, and cannot find a partner.
4. Sangamo refuses to conduct studies that rigorously assess whether their drugs work. This includes the inclusion of a control arm so assiduously avoided by Sangamo’s management. Rest assured we believe that if Sangamo could demonstrate efficacy in a properly controlled study, they would.
5. The same line of reasoning can be applied to the rest of Sangamo’s pipeline. Stated simply, if one can re-introduce a normal copy of a defective gene, thereby effecting a cure, why would anyone bother with the Sangamo approach? We wonder whether sell-side analysts who cover SGMO would ask this question if they were putting the interests of shareholders above the next secondary offering. Absent this, shareholders need to ask themselves why this time it’s different.
6. We believe Sangamo’s market capitalization of $1.2 billion grossly overestimates the likelihood of success of the HIV program and thalassemia pre-clinical program.
7. We are short Sangamo with a price target of $3.26, which is the value of the cash on the balance sheet. Given Sangamo’s 20-year history of failure to move any asset past phase 2 trials and lack of development with SB-728, we assign no value to its zinc finger platform and merely value Sangamo for the cash on its balance sheet.
According to TipRanks.com, which measures analysts’ and bloggers’ success rate based on how their calls perform, Blogger Alpha Exposure has a total average return of 38.7% and a 74% success rate. Alpha Exposure is Ranked #133 out of 4269 Bloggers