We assign no value to Sangamo Biosciences SB-728-T in its expected indication; i.e., functional cure for HIV. We are establishing coverage of Sangamo Biosciences Inc. with a Sell rating. Sangamo’s $575 million equity capitalization is supported in part by (1) an adoptive T-cell therapy, SB-728-T, that Sangamo believes may provide a functional cure of HIV and (2) the Berlin Patient, an individual who Sangamo inexorably reminds investors is ‘HIV-free without HAART following hematopoietic stem cell transplant from an allogeneic, HLA matched, CCR5 delta-32 homozygous donor’. Although the mechanisms behind this individual’s apparent cure are unknown and the case has yet to be replicated, the donor’s CCR5 genotype (CCR5 delta-32/32) appears to have been recognized by some as causative.
Sangamo identified HIV as a potential therapeutic target for its zinc finger technology in 1996, two years before its IPO. Thirteen years later, in February 2009, the first phase 1 study of SB-728-T was initiated. To date, three phase 1 studies that enrolled 42 patients have been completed. Although data for all 42 patients is available based on the date of the primary endpoint evaluation, Sangamo has selectively reported data from two studies that enrolled 21 patients. Data from the third study, which enrolled 21 treatment-naïve/off-treatment individuals and was designed to assess the safety and antiviral effect of SB-728-T, has been withheld. Currently, two studies of SB-728-T are ongoing, one of which Sangamo has officially designated a phase 1/2 study and another, an expansion cohort, that has been designated a phase 2 study. More recently, Sangamo has begun referring to both studies as phase 2 trials. Sangamo will present preliminary results from these two studies at a medical conference on May 15, 2013 and complete data by year end. The data ostensibly will form the basis for what Sangamo Biosciences’ management has signaled will be a phase 3 ready and partnerable HIV program by the end of 2013 and early 2014, respectively.
HIV enters cells by attaching to the CD4 receptor and subsequent binding to surface co-receptors. R5-tropic and X4-tropic HIV bind to the CCR5 and CXCR4 co-receptors, respectively while some HIV isolates utilize CXCR4 in addition to CCR5 (dual-tropic HIV).
The CCR5 co-receptor is used by most viral isolates found in new or early infection and approximately 50% in antiretroviral experienced individuals. In the AIDS Clinical Trial Group A5211, a phase 2b study of the CCR5 antagonist vicriviroc in ART-experienced individuals, 50% of subjects had virus that used the CCR5 coreceptor, 46% had dual-tropic populations that used both the CCR5 and CXCR4 coreceptor and 4% had virus that used the CXCR4 coreceptor exclusively. SB-728-T’s targeted population is R5-tropic HIV.
CCR5-Targeted Gene Therapy
CCR5 was identified as a potential anti-HIV target based on a genetic association study of individuals who, despite repeated high-risk sexual behavior, remained resistant to infection by HIV. Subsequent studies of these individuals led to the discovery of a mutation in the CCR5 gene. The mutation was found to be a 32-nucleotide deletion, referred to as CCR5 delta-32. The deletion mutation results in a non-functional CCR5 receptor that prevents R5-tropic HIV cell entry. Individuals homozygous for the mutation have resistance to HIV infection. In addition, population based studies have found an enrichment of individuals heterozygous for the mutation (delta-32 heterozygotes) among HIV+ individuals with slower disease progression, suggesting a protective effect in these individuals.
SB-728-T is designed to genetically recreate the mutation. To accomplish this, the CCR5 gene in patient-derived CD4+ T cells is disrupted using Sangamo’s gene editing technology. The cell preparation and treatment protocol involves the delivery and transient expression of SB-728 into patient-derived CD4 T-cells using adenoviral vectors. Prior to re-infusion, the modified cells are expanded and stimulated using a protocol developed at the University of Pennsylvania that employs anti-CD3 and anti-CD28 monoclonal antibodies. The hypothesis is that the re-infused CD4 T-cells will be protected from infection and destruction and will expand in vivo, producing an HIV-resistant CD4+ T cell population.
The Berlin Patient, Elite Controllers and Non-Progressors Exploited by Sangamo
In February 2008, a poster abstract was presented at the 15th Conference on Retroviruses and Opportunistic (CROI 2008) . The abstract reported the case of a delta-32 heterozygous, HIV-infected individual who underwent allogeneic stem cell transplant with a donor selected to be homozygous for the CCR5 delta-32 mutation to treat relapsed acute myeloid leukemia. The day prior to transplant, antiretroviral therapy was discontinued. Two months after the transplant, the patient’s delta-32 heterozygous genotype had changed to a delta-32 homozygous genotype. At the time of the report, the patient had maintained undetectable viral load for >285 days in the absence of ART. Approximately 45 days after the report at CROI 2008, the patient experienced a second relapse of AML and subsequently underwent a second SCT from the same donor. The individual has reportedly maintained undetectable viral load in the absence of ART ever since.
On February 12, 2009, details of the case were reported in the New England Journal of Medicine and the patient has since been described as the first individual to have been cured of HIV.
Although the mechanisms behind the individual’s cure remain unknown, Sangamo Biosciences has implicated the transplant donor’s homozygosity for the CCR5 delta-32 mutation in the ‘apparent eradication of HIV’ and ‘functional cure’. Sangamo has provided further support for this theory by redefining groups of individuals, broadly referred to as controllers, who are capable of spontaneously controlling HIV infection. Among the latter are elite controllers, which Sangamo defines as individuals ‘that have only one of the two copies of their CCR5 gene mutated’ and long-term non-progressors, which Sangamo defines as individuals ‘who carry the CCR5 delta-32 mutation on only one of their two CCR5 gene copies (heterozygotes)’.
The foregoing definitions are undoubtedly helpful investor relations tools and lend credibility to the SB-728-T approach, but they are false.
Platform-Validating, Game-Changing, Paradigm-Shifting Data Supporting Functional Cure Expected Soon
Preliminary and complete data from two studies of SB-728-T in HIV positive individuals is expected to be released on May 15th at ASGCT and the second half of 2013, respectively. In preparation, bullish analysts have established outlandish price targets and published speculative research arguing that these two small, uncontrolled studies could establish ‘strong evidence for a functional cure for HIV’; could herald ‘a paradigm shift’ in the treatment of HIV; and could validate Sangamo’s entire zinc finger technology platform. Others have speculated that ‘positive’ data could trigger a takeout. One analyst, reflecting on the data release, opined that the ‘potential for a cure for this incurable disease should be game-changing’. Although the meaning of the statement is not clear, it appears that the analyst is telling investors that ‘there’s a chance’ and the chance should be capitalized at an obscene multiple. If history is a guide, the table-pounding and blind faith is not necessarily a harbinger of positive data or a platform-validating event, but underwriter anticipation of a financing.
In recently published reports, arguments have been made that SB-728-T is potentially worth between $400 million (~$6 per share) and $710 million ($13 per share) and at least one analyst is modeling approval of SB-728-T as early as 2015. Although it is likely that only a fraction of the sane rely on these estimates, SB-728-T is Sangamo’s most advanced development compound and inflated expectations are clearly baked into the shares.
These estimates are per se absurd. SB-728-T is a potential first-in-class, phase 1 gene therapy expected to functionally cure HIV that has been evaluated in 42 individuals, data from 21 of whom Sangamo has not released.
Unprecedented Use of the Word Unprecedented
Results reported for subsets of individuals (some as small as n=1) in open label phase 1 safety studies of SB-728-T have been described as ‘unprecedented’, ‘dramatic’ and ‘ground-breaking’, Sangamo’s preferred terms for describing almost all reported data [9,10,11,12,13,7,14,15,16,17,18,19,20,21].
For example, Sangamo has reported ‘unprecedented and durable increase in CD4+ cells’, ‘unprecedented improvement in immune system health’, ‘unprecedented antiviral and immunological activities’, ‘unprecedented durable engraftment’, and ‘unprecedented improvements in overall CD4+ T-cell counts and CD4 counts and normalization of CD8+ T-cell ratios’.
More recently, an individual whose laboratory performed a post-hoc statistical analysis that led to the observation that certain cell surface marker and gene expression profiles may predict superior response to SB-728-T was quoted in a Sangamo press release as saying: ‘The ability of SB-728-T to durably reconstitute the immune system in HIV-infected subjects after a single treatment has never been observed before with any other therapeutic approach’. Sangamo’s chief medical officer has declared the same, saying ‘There is no other drug that has been shown to have the same dramatic effect on the immune system in this setting’ .
These statements significantly overvalue the data, if not plainly misrepresent it, in our view.
SB-728-T a Phase 3 Ready Asset by End of 2013?
Although no controlled studies are in progress, nor have any been conducted, Sangamo has told investors that SB-728-T will be a phase 3 ready asset by the end of 2013 , which suggests that Sangamo expects to have completed end of phase 2 meetings with the FDA within the next several months and has a phase 3 ready regulatory endpoint and pathway in mind.
Neither the endpoint nor the pathway has been adequately described, which is not unexpected since there are no established surrogate endpoints for ‘functional cure’, there is no consensus on how (types of assays) or where (body compart-ments) to measure HIV in the context of clinical endpoints in cure research and because Sangamo has not described the nature of the functional cure with specificity (e.g., aviremia in the absence of ART for ‘x years’, no disease progression, no CD4 loss, lack of HIV transmission or a combination of two or more of these objectives).
What Sangamo has said is the following: the data it currently has suggests that ‘with sufficiently high levels of circulating biallelically modified cells and good engraftment, SB-728-T can provide a ‘functional cure’ for HIV’. The claim is wildly optimistic and the data suggest no such thing. It is supported by a post-hoc analysis that found a statistically significant correlation between numbers of ZFN-modified cells and viral load at the end of a 12-week structured treatment interruption in one cohort of a phase 1 study. The statistical analysis incorporated data from 5 individuals in this six patient cohort with R5-tropic virus and intact immune function as measured by CD4.
It is unclear how the correlation, which informed the design of an uncontrolled study that has been designated a phase 2 trial, has increased the probability of success or, in the multiple-inflating language of the Street, de-risked the development program. At best, the correlation confirms the plausibility of the original hypothesis and should be surprising only to those who expected that there might be an inverse correlation be-tween levels of ZFN-modified cells and viral control. At worst, it is useless statistical noise. The suggestion that the data is evidence of a dose-response is questionable.
Since Sangamo is promising to deliver a phase 3 ready asset by year end, it is perplexing that this phase 2 study has no control arm and that Sangamo has yet to disclose when it will conduct a study with a control arm; for example, unmodified CD3/CD28 costimulated autologous CD4 T-cells.
The safety of CD3/CD28 costimulated autologous CD4 T-cells has been established in this setting and patients might be expected to benefit [22–27]. Levine and colleagues conducted a phase 1 adoptive immunotherapy trial that enrolled eight HIV-infected individuals who were treated with autologous, ex-vivo expanded CD4+ T cells costimulated with anti-CD3/anti-CD28 antibodies. Immune reconstitution, as measured by increases in CD4% and CD4:CD8 ratios, was observed in all eight subjects at the end of the study. At CROI 2006, follow-up results were reported. Improvement in CD4 counts, CD4%, and CD4:CD8 ratios were maintained for a median of 3.3 years.
SB-728-T is Not a Phase 3 Ready Asset
If the primary endpoint of the anticipated pivotal program is to be some unknown surrogate measure of functional cure and if functional cure is defined as control of HIV replication in the absence of ART, it seems implausible that SB-728-T will be in pivotal ‘cure’ studies any time soon. The NDA for maraviroc, the first anti-HIV agent to target a host protein (rather than a viral target), was supported by results from 28 phase 1 studies in >600 healthy volunteers/HIV-1 infected individuals, one phase 2 study in 917 R5-tropic, ARV-naïve HIV+ individuals and one phase 2 study in 186 non-CCR5 tropic ARV-experienced HIV+ individuals.
It is difficult to believe that a gene therapy would undergo any less rigorous evaluation prior to entering phase 3 clinical trials. And although Sangamo may have a different notion of what constitutes a phase 3 ready asset, most would agree that an uncontrolled, short-term study of a putative cure for HIV that will have been evaluated in ~10 HIV+ individuals does not meet the threshold of a phase 3 ready asset.
Persistence of latent infection is an obstacle in the pursuit of a cure. Even if there were strong evidence of antiviral activity from phase 1 studies of SB-728-T, which there is not, the threshold for safely discontinuing ART is not known. What is known is that HIV can establish a state of latent infection at the level of individual T cells . A paper published in AIDS in 2010 (Rebound of plasma viremia following cessation of antiretroviral therapy despite profoundly low levels of HIV reservoir: implications for eradication ) highlights this stark reality.
Chun et al. reported results from a study of 44 individuals that had been on ART for a median of 7.7 years and had achieved prolonged viral suppression. The authors highlighted the case of an individual who had been well-controlled on ART for 10.5 years ; according to the authors, this individual had the lowest HIV reservoir ever measured in their lab or elsewhere (1 infected cell per 1.7 billion CD4+ cells). ART was discontinued to determine if viral rebound would occur and what the kinetics of rebound would be if it occurred. 50 days after cessation of HAART, this individual’s viral load rebounded to 1593 copies/ml (3.2 log10 copies), spontaneously became undetectable, and then rebounded on day 143 to 8684 copies/ml (3.9 log10 copies) – see Figure 1 – at which point HAART was reinstituted.
Ongoing Studies of SB-728-T
SB-728-T currently is being evaluated in two small, uncontrolled studies (SB-728-902 Cohort 531 and SB-728-110132). Both studies are designed to maximize engraftment of SB-728-T, one by means of patient selection, and one by means of a pre-infusion conditioning regimen. The rationale for the studies is based on a post-hoc analysis that showed a statistically significant correlation between engrafted biallelically modified T-cells and reduction in viral load at the end of a 12-week treatment interruption.
Although Sangamo’s ability to transduce T-cells at therapeutically-relevant levels in the setting of the clinical studies conducted to date has not been an issue, engraftment levels observed in phase 1, despite assertions to the contrary17, were not encouraging [33,32] (one of six individuals in one phase 1 study did not engraft).
Cohort 5, designated a phase 2 study, is evaluating SB-728-T in an aviremic population of HIV-infected individuals heterozygous for the CCR5 delta-32 mutation (see Summary below). The ‘1101 study is evaluating the safety and effect of escalating doses of a non-myeloablative pre-infusion conditioning regimen on engraftment and is expected to enroll at least 9 (and up to 18) aviremic individuals in three dose cohorts.
Prior Phase 1 Studies of SB-728-T
SB-728-T has been evaluated in three phase 1 studies. The first phase 1 study [RAC Protocol #0704-843 or the UPENN study] enrolled 12 treatment-experienced, aviremic individuals in two cohorts (6 immunologic responders and 6 immunologic non-responders). Immunologic non-response (i.e., impaired CD4+ T cell restoration on HAART) was defined as CD4+ T cell count <500 cells/mm3 after 2 years of stable HAART and >200 cells/mm3 at screen. No individuals were enrolled on Cohort 1.
The second phase 1 study, SB-728-902, enrolled 9 treatment-experienced, aviremic individuals (immunologic non-responders) in three dose cohorts (Cohorts 1 to 3) [34,35].
Treatment Naive/Off Treatment Study - SB-728-1002
In October 2010 Sangamo opened a third study , discontinued for as yet undisclosed reasons, that enrolled 21 treatment-naïve or off-treatment HIV-infected individuals through October 2011 (SB-728-1002/NCT01252641). Sangamo has never reported data from these 21 patients, despite promises that it would.
February 2, 2011 - Q4 2010 Earnings Conference Call
“Our initial clinical observations of the pharmacology of the modified T cells have been encouraging which is why we have now moved into the viremic population at both ends of the spectrum with expansion of our Phase 1 trial into heart failures and initiation of the Phase 1/2 trial, SB-728-T-1002 in treatment naive subjects. Both of these populations will allow us to not only look at changes in the immune system, but also at the impact of the virus on the modified T cell product.” - Edward Lanphier
April 26, 2011 - Q1 2011 Earnings Conference Call
“Also in the second half of the year we expect to present additional data from our Phase I HIV studies, as well as preliminary data from our Phase I-II study in treatment-naïve subjects.” - Edward Lanphier
October 25, 2011 - Q3-2011 Conference Call – - Q&A Session
Analyst Question: ”Thank you. Just to confirm the data that you will be showing in the treatment naive patients, that was about 13 patients that were enrolled, and you will report that by year end?
Answer: “No, the 1002 trial that Geoff discussed says we have accrued 21 subjects on that trial, and we expect to present the full cohort in 2012. - Edward Lanphier
Analyst Question: “Okay, pardon me. That 21 does not include the new cohort 5, is that correct?”
Answer: “That’s correct.What I was answering there,Ted, the subjects on the 1002 trial, which is the treatment naive study.” - Edward Lanphier
October 21, 2011 – BioCentury Publications Inc. News Makers in the Biotech Industry Conference
“We also have an exploratory study which is ongoing in treatment naive patients. We will be discontinuing that study to concentrate our resources on the proof of concept studies in the aviremic HAART-treated patients.” - Edward Lanphier
Tying CCR5 Genotype to Functional Cure
Functional cure is broadly defined as host-mediated control of HIV replication in the absence of antiviral therapy, although stricter criteria have been proposed. Examples of naturally occurring functional cure include subsets of untreated HIV-infected individuals who do not progress to AIDs or death, referred to broadly as controllers. These individuals have been defined variably but controllers with undetectable viral load have been referred to as elite controllers and those with low, but detectable viral load, have been referred to as viremic controllers . Controllers are the subject of intense study as they have the potential to accelerate and guide HIV cure research.
The existence of controllers was first discovered during the early (pre-ART) years of the HIV epidemic when the research community began identifying HIV infected individuals who did not progress and were able to maintain stable CD4+ cell counts. It was subsequently observed that certain individuals had undetectable viral loads (referred to at the time as ‘controllers’) while others had low, but detectable, viral loads; the latter group fits the classical definition of ‘long-term non-progressors’, or LTNPs, who are recognized as a distinct population from elite controllers [38,39,40,41]. Elite controllers are estimated to represent between 0.5% to <1% of all HIV-infected individuals [41,38]. Given the genetic heterogeneity of elite controllers , their situation is not likely to be recapitulated with any gene therapy.
Although numerous mechanisms to explain elite control and non-progression have been suggested, they remain theoretical  and the subject of study. Increasing numbers of what have been referred to as ‘atypical’ elite-controllers are being identified. Mendoza et al. reported on four such cases that were distinct from conventional LTNPs/ECs (extraordinarily low HIV burdens and comparatively weak immune responses) . One individual was heterozygous for the CCR5 delta-32 mutation and none were homozygous.
Sangamo BioSciences Redefines Elite Control
Sangamo, and only Sangamo, has defined elite controllers according to CCR5 genotype; for example, management has defined elite controllers as individuals ‘that have only one of the two copies of their CCR5 gene mutated’ [45,46] and has stated that ‘those with only one copy of the delta-32 mutation are identified as “Elite Controllers” for their immune systems’ ability to resist progression of HIV without the need for HAART.’ 
Sangamo has also created a novel definition of long term non-progressors; ‘Individuals who carry the CCR5 delta-32 mutation on only one of their two CCR5 gene copies (heterozygotes), so-called “long term non-progressors” tend to take longer to develop AIDS’.
Sangamo’s definitions, although they lend credence to a CCR5-centric cure doctrine, might be identified by some as lacking ‘truthiness’. Those with only one copy of the delta-32 mutation are identified as delta-32 heterozygotes and so-called long-term non-progressors have never been classified solely on the basis of presence of the delta-32 mutation.
The definition is more than academic. When the first phase 1 study of SB-728-T was evaluated by the Recombinant DNA Advisory Committee, two studies were cited suggesting enrichment of delta-32 heterozygotes in the population of long-term non-progressors (Cohen et al., 1997 and Eugen-Olsen et al., 1997). In the former study, there was actually no evidence of slower disease progression among CCR5 delta-32 heterozygotes while the latter study found a 3-year delay in progression to AIDs. More recent studies have failed to find enrichment of CCR5 delta-32 heterozygotes in the LTNP population. In addition, in some studies CCR5 delta-32 heterozygosity has been strongly associated with lower viral loads and/or higher CD4+ T-cell counts, but most are silent on the topic of disease progression.
A less obvious form of the questionable definition is now making its way into the literature. Wilen and colleagues from Sangamo, UPENN and Tulane 47 stated the following in a research paper titled Engineering HIV-Resistant Human CD4+ T Cells with CXCR4-Specific Zinc-Finger Nucleases [PLoS pathogens 7, no. 4 (2011): e1002020];
The apparent eradication of HIV resulting from a ccr5 delta-32 homozygous allogeneic bone marrow transplant into an HIV-infected patient represents the first reported “cure” of HIV.
Although there is no evidence that the donor’s CCR5 genotype did not contribute to this patient’s cure (the Berlin Patient), there is likewise no debate that the mechanisms behind the Berlin Patient’s apparent cure are not completely understood.
Inaccurate Generalizations – Controllers
In addition to inaccurately defining controllers, Sangamo has made what are largely irrelevant and inaccurate generalizations about the proportion of controllers harboring the CCR5 delta-32 mutation; e.g., Sangamo says ‘A population of individuals that is immune to HIV infection, despite multiple exposures to the virus, has been identified and extensively studied. The majority of these individuals have a natural mutation, CCR5 delta-32’. The claim is debatable based on research from other studies of elite controllers; e.g., the Elite Controller Collaborative Study .
Pereyra and colleagues reported on a cohort of 126 individuals who had achieved control of HIV infection in the absence of ART . The study participants were classified into 2 controller subgroups: (1) those that maintained viral loads below the level of detection as measured by then available assays – elite controllers; and (2) a group that maintained viral loads between 50-2,000 copies/ mL (defined as viremic controllers). Data from these study participants were compared to untreated persons with progressive infection (chronic progressors) in an attempt to identify immunologic and host genetic factors associated with control.
A total of 66 elite controllers and 60 viremic controllers were compared to 30 chronic progressors. The presence of the CCR5 delta-32 mutation was determined for all HIV controllers (elite and viremic); none were homozygous for the CCR5 delta-32 mutation. 14% of elite controllers were heterozygous for the delta-32 mutation compared to 22% of viremic controllers. As observed by Pereyra et al., this frequency is approximately the same as the frequency in individuals not infected with HIV .
Quantitative Immune Reconstitution
Sangamo’s claims of unprecedented immune reconstitution are supported by data of questionable rigor; specifically, increases in CD4 counts and normalized CD4:CD8 ratios observed in subsets of individuals who participated in two phase 1 studies. In order to contemplate the validity of Sangamo’s conclusion that SB-728-T has demonstrated a unique ability to ‘durably reconstitute the immune system’ one must accept CD4 count as a reliable surrogate marker of immune restoration in the setting in which SB-728-T was studied as well as the validity of Sangamo’s historical controls, some of which are known to exist only because of Sangamo’s interminable use of the word ‘unprecedented’. The literature suggests that the surrogate marker is not reliable and the historical controls should be rejected.
Absolute change in CD4 count is an imperfect marker of immune reconstitution as was demonstrated in the SILCAAT  and ESPRIT  studies (ART vs. ART + IL-2) ; in both cases, cytokine-mediated increases in CD4 T-cell counts were not predictive of clinical benefit. Sangamo had previously speculated that an early cytokine burst (IL-2, IL-7 and IL-15) following infusion of SB-728-T played a key role in the ‘dramatic’ post-infusion expansion of CD4+T-cells.
Sangamo has not explained why SB-728-T cytokine-mediated increases in CD4 count are qualitatively different than IL-2 mediated increases, but management is now touting quantitatively superior increases in CD4 count. However, there is no evidence that the nebulous but ‘significantly’ greater increases in CD4 count cited by Sangamo have any clinical significance nor is there any evidence that the increases were related to ZFN modification. Sangamo’s ‘trust us, this wouldn’t have happened in the absence of ZFN modification’ stance is ridiculous.
In March 2013, a presentation at CROI identified three historical controls  that purportedly back Sangamo’s use of the term unprecedented to describe its phase 1 data (‘Previous t-cell infusion studies manipulated in a similar without CCR5 modifications did not result in sustained CD4 reconstitution’). The controls meet none of the well-accepted criteria for validity and can be rejected on purely academic grounds.
Criteria for Valid Historical Controls (Pocock, 1976) 
- A recent study with the same treatment
- Identical eligibility criteria, workup, and evaluations
- Prognostic factors are known and are the same in both treatment groups
Although not defined, we interpret ‘sustained’, as used by Sangamo, to be the 12-month time frame for the outcome measure. One of the three historical controls was a six month study, making a comparison to Sangamo’s 12-month study difficult. This study evaluated CD4-zeta gene–modified CD4 and CD8 T-cells expanded ex-vivo using an anti-CD3 and anti-CD28 stimulation protocol. 40 aviremic patients on stable HAART were enrolled. This study incorporated a control arm that received unmodified T-cells, because ‘adoptive transfer of ex vivo–activated T cells alone may have anti-HIV efficacy’ (Tebas et al.).
The antiviral efficacy of the activation/stimulation protocol was first reported by Levine and colleagues in 1996  and was described in more detail in 2008 (‘crosslinking of CD3 and CD28 with bead-immobilized antibody renders CD4+ T lymphocytes highly resistant to HIV infection’, Powell and Levine, 2008) . So, even if ‘quantitative immune reconstitution’ were an informative clinical endpoint, increased CD4 count is a known effect of the infusion of CD3/CD28 activated CD4 T-cells and has been observed in other anti-HIV gene therapy trials using the same cell preparation protocol [56,24].
CD3/CD28 Stimulated CD4+ T-Cells in HIV – Examples in the Literature
- We describe a procedure for large-scale enrichment, growth, and harvesting CD4+ T cells. This method may be effective for HIV-1 immunotherapy, as the mode of stimulation, with anti-CD3 plus anti-CD28 coated beads (CD3/CD28 beads) induces a potent antiviral effect (Large-Scale Production of CD4+ T Cells from HIV-1-Infected Donors After CD3/CD28 Costimulation  (Levine, 2009).
- We conducted an adoptive immunotherapy trial infusing purified autologous CD4+ T cells from HIV-infected sub-jects, which were expanded in the presence of magnetic bead-bound anti-CD3 and anti-CD28 (CD3/28). This method of costimulation has been shown to increase the expression of beta chemokines and down-regulate the ex-pression of CCR5, conferring resistance to HIV infection by CCR5-tropic isolates (Bernstein, 2004).
- Adoptive transfer of costimulated CD4+ T cells induces expansion of peripheral T cells and decreased CCR5 ex-pression in HIV infection 
- Immune reconstitution following autologous transfers of CD3/CD28 stimulated CD4+ T cells to HIV-infected persons 
- Intrinsic resistance to T cell infection with HIV type 1 induced by CD28 costimulation 
- Differential regulation of HIV-1 fusion cofactor expression by CD28 costimulation of CD4+ T cells 
- Antiviral effect and ex vivo CD4+ T cell proliferation in HIV-positive patients as a result of CD28 costimulation 
Sangamo’s historical controls also include a study of gene-modified autologous T-cells in ten HIV-infected individuals with multidrug-resistant virus, eight of whom had full blown AIDs. Median CD4 count at baseline was 93 cells/mm3 and median viral load was 4.96 log10 RNA copies/ml.  Thus, even if the cross-study comparisons were valid, which they are not, this is not the patient population in which SB-728-T has been studied.
SB-728-T Antiviral Activity
Although allegedly on the brink of phase 3 studies, virtually nothing is known about SB-728-T’s antiviral activity. What can be discerned is limited to results from two open label phase 1 studies, one of which enrolled 6 individuals who participated in a 12-week structured treatment interruption (clinicaltrials.gov, NCT00842634 – Cohorts 2 and 3) and another that enrolled 9 individuals who did not participate in a structured treatment interruption (clinicaltrials.gov; NCT01044654); which is to say, nothing.
A third phase 1 study, commenced in October 2010, enrolled 21 treatment-naïve or off-treatment HIV-infected individuals (clinicaltrials.gov, NCT01252641). The data from this study has not been presented despite repeated statements by management that it would. In a press release dated March 3, 2011, Sangamo announced that it would present data from the ‘1002 study in late 201160. During an investor conference call on April 26, 2011, Sangamo reported that it would present preliminary data from the study at a scientific or medical meeting in the second half of 2011. On July 27, 2011, Sangamo confirmed this plan, saying that the preliminary data would be reported ‘later this year’61. On October 21, 2011  Sangamo announced that the study would be discontinued to concentrate resources on aviremic HAART-treated patients. On October 26, 2011  Sangamo announced that it had just completed accrual of the ‘1002 study and stated that the data would be presented at an ‘appropriate meeting in 2012 when we have more prolonged follow up data on the entire cohort.’
The primary and secondary outcomes were assessed at 12 months. The last assessment (for the secondary outcome measures) would therefore have taken place over six months ago. Sangamo has not reported the data; last updated in May 2012, the clinicaltrials.gov record indicates that study is ongoing, but not recruiting .
This study might have provided some evidence of antiviral activity (or inactivity). SB-728-T’s potential in this treatment naïve setting was characterized by one of the principal investigators  as the ‘whole shooting match’; Don’t Say Cure: Questions for Jay Lalezari, MD. Cure Blog (2011).
This could all eventually lead to nothing. But we’ve proven the loop can be completed. The next step is to test the therapy in treatment-naïve patients. What will those cells do in the presence of virus? Presumably they will expand even further. In the presence of HIV, they obviously have a selective growth advantage. We should see even more expansion. The whole shooting match is what hap-pens to those viral loads. If those viral loads come down, you and I are living in a different universe. If they don’t, I don’t think there’s any gray area – Jay Lalezari, MD.
At the 51st Interscience Conference on Antimicrobial Agents and Chemotherapy (9/2011), detailed results from an open label, single arm proof of concept trial were presented (Cohort 2 – Protocol #0704-843). Cohort 2 enrolled 6 immunologic responders (defined by CD4 count >450 cells/ml); Cohort 3 of this study enrolled 6 immunologic non-responders (CD4 > 200 and < 500 cells/ml).
In addition to data from Cohort 2, results from a phase 1 dose escalation study were presented (clinicaltrials.gov, SB-728-902). This study enrolled 9 individuals (3 per cohort) fully-suppressed on HAART with suboptimal CD4 count (<500 cells/mm3) 66 after a median of 21 years of treatment . Data from one individual who participated in an apparently off-protocol, 28-week treatment interruption beginning 12 months post infusion was presented. Of the six individuals enrolled on Cohort 2 who participated in the structured treatment interruption, four completed the STI.
In a press release announcing the presentation at ICAAC, Sangamo highlighted data from one individual that had undetectable viral load during the 12-week STI after an initial rebound; the individual was heterozygous for the delta-32 mutation (Subject 205). Expansion Cohort 5 (i.e., the ongoing ‘phase 2’ study) is enrolling CCR5 delta-32 heterozygotes  ‘to confirm the occurrence of aviremia during treatment interruption’10 as observed in this one individual.
In the same press release, Sangamo highlighted (1) a ‘0.8 to >2.0-log reduction in viral load from peak’ during STI that was observed in 3 of 6 subjects with the highest estimated circulating levels of cells with both copies of the CCR5 gene modified and (2) the ‘confirmation and extension of previous observations of unprecedented improvements in overall CD4+ T-cell counts and CD4+: CD8+ T-cell ratios’ and ‘engraftment, expansion, trafficking and persistence of the ZFN modified cells’.
At the 19th Conference on Retroviruses and Opportunistic Infections (CROI 2012), viral load decreases (from peak during STI) in two other individuals enrolled on Cohort 2 of this study were highlighted  (study participants 201 and 203) – Figure 3; additional data on study participant 205 was also presented – Figure 4).
This Data is Not Unprecedented or Informative
The response observed in the SB-728-T treated individual heterozygous for the CCR5 delta-32 mutation, while interesting, cannot be interpreted in the absence of a control arm. Examples in the literature are sparse and conflicting but one study suggested that CCR5 delta-32 heterozygotes are more likely to control viremia during STI [69,70] while another found no correlation between CCR5 genotype and viral rebound during STI in well-controlled, chronically infected individuals .
Although the antiviral evidence gleaned from this single patient might not meet Sangamo’s threshold for a issuing a press release sprinkled with terms such as ‘cure’ and ‘unprecedented’, it met the investigators’ threshold for invoking the term ‘dramatic’73. Complete results were published ahead of print in December 2012 in Blood . A prior phase 1 study of VRX496 enrolled 5 individuals who had failed > 2 HAART regimens. 4 of 5 individuals had stable or increased CD4 counts and viral loads were stable in 4 of 5 (in one patient there was 1.7 log10 change in viral load).
Conducted at UPENN, Jacobi Medical Center, and the MacGregor Clinic, the protocol and entry criteria were similar to Cohort 2 of the UPENN SB-728-T study. 17 aviremic individuals stably controlled on ART with CD4 cell count >450 cells/mL at screening were enrolled; 4 individuals were screened out. Primary end points were safety, effects of infusion on viral load before STI, time to viral recrudescence and set point during STI, and change in CD4 T-cell counts from baseline to STI (cohort-specific outcomes in Cohort 2 of the SB-728-T phase 1 study conducted at UPENN were time to viral recrudescence to >500 copies/ml viral load and viral set point following STI.