Curagen – Positive Results at ASCO 2009

/* Style Definitions */ table.MsoNormalTable {mso-style-name:”Table Normal”; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-priority:99; mso-style-qformat:yes; mso-style-parent:””; mso-padding-alt:0cm 5.4pt 0cm 5.4pt; mso-para-margin:0cm; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:11.0pt; font-family:”Calibri”,”sans-serif”; mso-ascii-font-family:Calibri; mso-ascii-theme-font:minor-latin; mso-fareast-font-family:”Times New Roman”; mso-fareast-theme-font:minor-fareast; mso-hansi-font-family:Calibri; mso-hansi-theme-font:minor-latin; mso-bidi-font-family:Arial; mso-bidi-theme-font:minor-bidi;}

Last month at the ASCO meeting, Curagen (CRGN) presented results for its lead drug, CR-011, in breast cancer and melanoma patients. CR011 had activity in both indications, however, most of the drug’s value should be ascribed to the breast cancer program, which represents a huge commercial opportunity and better chances of approval.

As I previously wrote, the significance of the breast cancer trial is not only in the clinical activity of CR-011, but more importantly, the ability to identify patients who are likely to respond to the drug. By defining the right target population, Curagen could substantially improve chances of approval, shorten development time and enjoy high market acceptance.

The poster child for this strategy is Roche’s (RHHBY.PK) Herceptin, which is approved only for breast cancer patients whose tumors express Herceptin’s target, a protein called Her2. Although only ~25% of breast cancer patients are eligible for Herceptin, it has become one of the top selling oncology drugs, bringing in $4.4B in sales last year. Had Herceptin been tested in the entire breast cancer population, it would probably never have gained approval.

At ASCO, the strategy of selecting the right patients for Herceptin paid off once again, this time in gastric cancer. As part of a phase III trial, gastric cancer patients were first screened for Her2 expression and Herceptin was given only to “Her2 positive” patients. Herceptin increased survival by 2.7 month, an improvement which will likely get the drug approved for “HER2 positive” gastric cancer. This entirely new market could boost Herceptin’s annual sales by over $1B.

As an antibody based drug, CR-011 should follow the footsteps of Herceptin, because it also binds a specific protein on tumors – GPNMB. Similarly to Herceptin, CR-011’s target is expressed only in ~25% of breast cancer patients. However, CR-011 is an antibody-drug conjugate (ADC) while Herceptin is a conventional “naked” antibody. As a result, unlike Herceptin, which works by a plethora of effects (signal modulation and recruiting the immune system), CR011 exerts its activity by a single mechanism of action. This should make the process of picking the right patients even more straightforward: The expression of GPNMB on tumors.

Based on the update at ASCO, CR-011 clearly has activity in breast cancer. In addition, there are preliminary signs of a correlation between response and GPNMB expression. To get a clearer answer, investors would have to wait until year end.

Good data but small numbers

Curagen presented top line results for 18 breast cancer patients.  The patient population was particularly tough to treat, with a median of four prior chemotherapy regimens in the metastatic setting. This figure excludes biologic agents and earlier stage treatments such as post or pre-operative therapies, therefore, the actual number of prior therapies patients had received was higher, probably five or six.

In terms of clinical activity, one (6%) patient achieved a confirmed partial response (a sustained reduction of at least 30% in tumor burden) and two additional patients achieved an unconfirmed partial response (a transient tumor reduction of 30% or more). Overall, some level of tumor shrinkage was observed in 9 (50%) patients.

A confirmed response rate of 6% in a phase I trial is nothing to write home about, but it is important to remember that patients were enrolled to the study regardless of whether their tumors express high levels of GPNMB. Assuming that the patients in the trial represented the overall population, most of them expressed low levels of GPNMB, or none at all, and were therefore unlikely to respond to CR-011 in the first place. In addition, the first two patients who were enrolled in the study had pre-existing nerve side effects, which deteriorated after the first treatment cycle with CR-011. These patients were taken off the study immediately so they did not have the opportunity to respond to the drug.

Looking for correlation

At the time of presentation, there were GPNMB expression data for only five of the patients: One patient had high levels of GPNMB, one had low levels of GPNMB and three had “GPNMB negative” tumors. Crossing these findings with the efficacy data resulted in some interesting observations. The “GPNMB positive” patient was one of the unconfirmed PRs, with a 51% reduction after 6 weeks. The patient with the low GPNMB expression had a mixed response (shrinkage of some lesions in parallel to growth in other lesions), coupled with marked improvement in disease symptoms. The patient remained on study for almost 4 months, illustrating the benefit she was deriving from CR-011. Of the 3 patients with negative staining, 2 did not respond to the drug at all while one had stable disease after the first cycle and progressed shortly after.


/* Style Definitions */ table.MsoNormalTable {mso-style-name:”Table Normal”; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-priority:99; mso-style-qformat:yes; mso-style-parent:””; mso-padding-alt:0cm 5.4pt 0cm 5.4pt; mso-para-margin:0cm; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:11.0pt; font-family:”Calibri”,”sans-serif”; mso-ascii-font-family:Calibri; mso-ascii-theme-font:minor-latin; mso-fareast-font-family:”Times New Roman”; mso-fareast-theme-font:minor-fareast; mso-hansi-font-family:Calibri; mso-hansi-theme-font:minor-latin; mso-bidi-font-family:Arial; mso-bidi-theme-font:minor-bidi;}

It is very hard to interpret these findings due to the small sample size, but the overall trend is encouraging, with a possible correlation between GPNMB expression and clinical response.

The trial is still enrolling patients and investigators are trying to obtain tissue samples for as many patients as possible. According to Ronit Simantov, Curagen’s Chief Medical Officer, the recent data made investigators more inclined to do a biopsy and obtain tissue sample. The company expects to have response data for approximately 40 patients by year end, of whom, about 20 are expected to have tumor samples for GPNMB expression.

T-DM1 as a benchmark

The best benchmark for assessing CR-011 is Roche’s T-DM1, which is also an antibody drug conjugate (comprises of Herceptin and a drug payload). T-DM1, currently in two pivotal breast cancer studies, is one of the most important cancer drugs in development at the moment.

The two agents have a lot in common. Both utilize validated ADC technologies; CR-011 employs Seattle Genetics’ (SGEN) technology whereas T-DM1 employs Immunogen’s (IMGN) technology. Furthermore, the targets for CR-011 and T-DM1 (GPNMB and Her2, respectively) are found on a minority of breast cancer patients and are associated with disease aggressiveness and poor outcome. Lastly, according to Timothy Shannon, Curagen’s CEO, the company patented GPNMB as a target, similarly to what Genentech/Roche did with Her2. This strategy enabled Roche to be the sole company who can sell anti-Her2 antibodies. If Curagen succeeds in defending its intellectual property, it will also be able to block other companies from selling anti-GPNMB antibodies, which could translate into a huge advantage.  

In contrast to T-DM1, CR-011 binds a fairly new and less validated target. Thus, the burden of proof lies on Curagen to show that it can identify the right patients for CR-011 in advance. The critical event for CR-011 is expected to occur later this year, when there will be more tissue samples for assessing whether CR-011 has preferential activity in “GPNMB positive” patients. If so, a high single digit response rate for the entire trial would position CR-011 as an extremely promising drug. To put things in perspective, in its first trial, T-DM1 had a response rate of 37.5% in patients who express high levels of T-DM1’s target, Her2. These patients represent ~25% of the general breast cancer population, so if T-DM1 were evaluated in unselected patients, it would probably achieve a response rate of 9-10%.

The Deal with Celldex

Shortly before ASCO, Curagen announced it would be acquired by Celldex (CLDX) in return for approximately $94.5 million worth of Celldex shares. Curagen comes with a dowry of $54.5 million of cash, putting a value of $40 million on CR-011 and the company’s preclinical antibody pipeline.

The rationale for some sort of a deal is clear at this stage, as the CR-011 program is becoming more complex and expensive.  Nevertheless, the deal puts a surprisingly low price tag on CR-011. In a market where good oncology products are licensed for an upfront payment of tens of millions, $40 million for the full ownership of CR-011 seems quite low.

During a call with Curagen’s management team last month, I asked Shannon why such a promising agent is sold for such a price.  According to Shannon, the company has been evaluating other alternatives, but the deal with Celldex represented the maximal value they could find in the market. He emphasized that the company contacted a lot of potential partners, but none of them offered a more attractive deal.

It is unclear to me how so many pharma companies, who are eager to fill their pipelines with clinical stage compounds, overlooked CR-011’s huge potential. CR-011 is one of the most advanced ADCs in the market, with clear activity in breast cancer, as well as in melanoma, thus representing a very large commercial opportunity. Unlike many other compounds, where investigators are struggling to find reliable biomarkers of response, response to CR-011 might be easily predicted owing to CR-011’s mechanism of action as an ADC. Lastly, unlike the case in most targeted therapies, Curagen has a strong patent position which could block other companies from developing direct competitors to CR-011.

Looking on the bright side, since Celldex is a small publicly traded company, it should be very sensitive to CR-011’s performance, so investors could still get exposure to CR-011. If the deal is consummated, Curagen investors might want to keep the Celldex shares they receive.

In summary, in my opinion, there is a clear discrepancy between CR-011’s potential and the acquisition price. Was CR-011 undervalued by the entire industry or is there something I am missing? We should get the answer in 6 months, with the updated results from the phase II trial in breast cancer. If CR-011 manages to demonstrate a high single digit response rate coupled with a correlation between response and GPNMB expression, it would position it as one of the top oncology drugs in development.

              The Biotech Portfolio as of July 5th, 2009



25 thoughts on “Curagen – Positive Results at ASCO 2009

  1. Hi Ohad,

    Thanks for the CRGN update.
    Me too a shareholder of CRGN and thinking of going thru’ the CLDX deal. Seriously, do you think there will be another offer for CRGN?
    Once again, thanks.


  2. Pingback: Breast cancer stage 2 prognosis? | Lobular breast cancer

  3. Hi Ohad,

    What new money do you think we should add to in your biotech portfolio?
    When do you plan on having a new recommendation?


  4. Hi Sam

    It’s hard to pick one or two stocks. In a risky field like biotech, I would spread that across several companies.
    There is one company we are seriously considering at the moment, we might add it during the coming weeks.

    I hope the R788 data are as good as TASKi 1, but large trials tend to be less positive than smaller ones.



  5. Ohad
    What is your current opinion on GENZ and it’s manufacturing troubles , and what do you think of it’s potential competitor for Gauche disease , Protalix.

    Can’t you comment on this , why ?


  6. I own a small stake in incyte but why such a big deal that incyte and fda agree on the drug trial design?

    It is not like they produced positive results from phase 3 trials.

    Also, what scares me is that incyte was once over a 140 per and has lost more than 95 percent from that high.


  7. INCY owns 100% of a potential blockbuster with a very high likelihood of approval. At $550M, it isn’t cheap but not too expensive either. The only issue is the debt.

    Sorry, I don’t follow YMI



  8. Ohad,

    What do you think about NPSP? They have 2 drugs in phase 3: GATTEX™ (teduglutide), NPSP558 (parathyroid hormone 1-84 [rDNA origin] injection)



  9. Sorry denny, I don’t follow NPSP.

    Regarding the MITI rumor, I also heard that but I have no confitrmation. If MITI manages to get full rights to blinatumomab, it will put it in a very attractive position. The $6M figure sounds rather low, though.



  10. Based on the title of the abstract, there are still no signs of efficacy, which is not surprising. In order to remain on track, MT110 must have a good safety profile that will enable dose escalation to therapeutic levels.

    I liked OSI’ new IGF1R inhibitor. Looks like they are the first to suceed with that target, probably have a better chance than all the antibodies in the clinic against the same target. However I am not too optimistic on Tarceva’s ability to grow, following the recent data at ASCO as a maintenance therapy.

    I don’e have a concrete opinion on ONTY.



  11. Ohad,

    A teaching fellow at the Mayo Clinic posted these remarks about CLDX’s Rindopepimut. Do you agree? What would be your rejoinder?

    “I’ve not really looked into this, so I don’t have anything meaningful to add. I will note, however, that they’re using additional methods to stimulate T cells. I think this highlights the difference in approach between companies who simply believe this pathway will take care of itself (a la rindo) versus companies that believe direct stimulation of this pathway is a necessary part of improving the success of immunotherapeutics (ipi, blinatimumab).


    Not to belabour the point, but I think one important point of all these vaccine immunotherapeutics against self antigens is to pay attention to the immune response related side effects (if any). Since we’ve not had much success with vaccine-type immunotherapeutics, new approaches will most likely show outright failure (we see little to no immunological adverse events coupled with little efficacy – theratope and various other vaccine immunotherapeutics) or a degree of success that goes hand in hand with a serious immunological adverse event profile (ipilimumab; remember TGN1412?).

    I don’t think we’re at the point technologically where we can manage to break self-tolerance with these vaccination strategies while demonstrating zero immunological adverse events. In my opinion, anyone currently selling this utopian view is selling a vaccine strategy that doesn’t work.”

    My comments were directed towards rindopepimut, which I thought was the larger focus of the piece. So in regards to A, i would note that the mechanism of action of provenge and ipilimumab are very different that rindopepimut.

    Rindopepimut is simply a 13 amino acid peptide attached to a carrier protein (KLH in this case). So the body has to ramp up an immune response against this peptide. This has historically been a relatively failing proposition in producing an adaptive immune response: you get low titre B-cell derived antigen specific antibodies and a weak T cell response. This is very similar to BLP25.

    MAbs like ipilimumab skip the step of requiring the body to produce high affinity antibodies against the desired target. I think this is key. I do not debate ipilimumab’s efficacy, but its mechanism is very different than what rindo asks of the immune system.

    As for provenge, the earlier papers by Vuk-Pavlovic showed very weak T cell responses. And as we know, the response was largely against the fusion point of the constructed peptide cassette. I won’t deny that provenge has demonstrated some efficacy, but I think we can agree that the mechanism through which that efficacy is manifest depends on a pathway very different than rindo.

    As per usual, I do not believe that such peptide immunotherapeutics elicit a sufficient immune response. I would personally stay away from cldx even if that means I miss out on a phase iii success.


  12. There are plenty of arguments in order to oppose peptide based vaccines. To me CLDX’s rindo could be different because:

    1) it targets a mutated protein for which it might be easier to break tolerance.
    2) They show that EGFRVIII positive tumors turn into EGFRVIII negative tumors – implies the vaccine has biologic activity.

    Other than that it’s a highly risky program.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s