It’s early days yet, but checkpoint immunotherapies have the potential to jump-start the high-risk, high reward glioblastoma drug development space, which has been prone to many a setback, the latest casualty being Celldex Therapeutics Inc.’s Rintega vaccine.

Checkpoint inhibitors, namely Bristol-Myers Squibb Co.’s Yervoy and Opdivo, Merck & Co. Inc.’s Keytruda and Roche’s newly approved Tecentriq (atezolizumab), have dramatically shaken up the oncology landscape.

Among many other studies of its checkpoint inhibitors in various tumor types to be presented at the American Society of Clinical Oncology annual meeting in June, Bristol will be presenting updated data for GBM patients enrolled in a Phase I cohort of the Phase III CheckMate 143 trial (see sidebar).

[Sidebar#00160523014]

It’s unclear how far the checkpoint inhibitors will go in terms of breadth of indications, and whether they can score another breakthrough in treating glioblastoma multiforme patients, who have a terribly poor prognosis. The research literature suggests a median survival of one year for GBM and a five-year survival rate of less than 10%.

Datamonitor Healthcare estimates that the number of GBM cases diagnosed in the US, Japan and five major EU markets will rise from 27,230 in 2013 to 35,420 in 2033, based on demographic trends and other factors.

Merck’s alkylating chemotherapy agent Temodar (temozolomide), now generic, is commonly used with radiation therapy for first-line patients and may also be used as a monotherapy in the second-line setting.

Roche’s Avastin (bevacizumab) is an option for recurrent GBM. The drug received accelerated approval for this indication in 2009 from FDA, but was not approved in Europe. Two Phase III studies in newly diagnosed GBM yielded mixed results – AVAglio showed an improved quality of life but an RTOG study showed the opposite, and neither study showed a benefit for an improvement in overall survival Also see "Avastin QoL Data In First-Line Glioblastoma Muddy Expansion Plans" - Pink Sheet, 3 March, 2014..

Still, many clinicians still find value in using Avastin in relieving symptoms, such as edema, and minimizing the need for steroids, and would like to see it retain labeling for glioblastoma, Patrick Wen, director of the Dana Farber's Center for Neuro-Oncology, commented in an interview.

As with other cancer types, interest in using checkpoint inhibitors has been rising in brain cancer. However, as Johns Hopkins specialist Michael Lim and colleagues pointed out in an August 11, 2015 article in Nature Reviews Neurology, they are relying on data extrapolated from other tumor types, and “at present, the exact involvement of checkpoint pathways in brain tumor pathogenesis is unknown.”

Bristol's Early Start

Of the leading four sponsors of PD-1/L1 checkpoint inhibitors, Bristol is by far the most invested in GBM. Early-stage, investigator-sponsored studies of Merck’s Keytruda (pembrolizumab) and AstraZeneca PLC’s PD-L1 inhibitor durvalumab (MEDI4736) are ongoing, and Roche presented preliminary safety results from a GBM cohort of atezolizumab in a Phase I study in 2015, but the data are very early and in a small group of patients.

Bristol got an early start with Opdivo (nivolumab) in glioblastoma, having initiated the Phase III cohort of the CheckMate 143 study of Opdivo with or without the CTLA-4 inhibitor Yervoy (ipilimumab) against Avastin for recurrent disease back in 2014. That study is event-driven and results could be available later this year.

A second Phase III study – CheckMate 498 – tests Opdivo against Merck’s standard-of care alkylating agent Temodar, both on top of radiation therapy in newly diagnosed GBM with unmethylated MGMT. Methylation of the MGMT gene is associated with better response to chemo and a better prognosis, whereas those with unmethylated MGMT have a poorer prognosis and limited treatment options.

With that study, Bristol “strategically positioned Opdivo head-to-head with Temodar," in MGMT unmethylated type GBM, Datamonitor analyst Brandon Goode wrote in a recent report on the GBM pipeline.

“Positive data from this study would propel Opdivo into the market as the only therapy dedicated specifically to the treatment of this subtype of disease that affects a large share of patients,” Goode noted.

The primary completion date on CheckMate 498 is March 2019.

Tough Nut To Crack

There are many challenges in developing a drug for glioblastoma, including difficulty getting over the blood-brain barrier, tumor resistance mechanisms, heterogeneity of disease and the fact that microscopic disease is invisible on imaging studies, Mark Gilbert, senior investigator and chief of the National Cancer Institute’s neuro-oncology branch at the National Cancer Institute, said in an interview.

That has made it tough to get through Phase III and approval Also see "R&D Realities Disturb Dreams About Orphan Brain Cancer Blockbusters" - Pink Sheet, 1 October, 2010..

EMD Serono Inc.'s cilengitide and AstraZeneca PLC's Recentin (cediranib) are among the drugs that failed in the past. Exelixis Inc. was bullish on taking its multikinase inhibitor XL184, now marketed as Cometriq (cabozantinib), into Phase III, but the indication dropped from its pipeline in 2011 ([A#06150100005_02]).

Celldex’s Rintega (rindopepimut), a cancer vaccine designed to induce immunity against EGFRvIII, a permanently activated mutation on tumor cells, is the latest failure. The drug had breakthrough therapy status with FDA for EGFRvIII-positive glioblastoma, which accounts for about 30% of GBM cases.

Rintega: Lessons Learned

The company announced the failure of the Phase III ACT IV study in EGFRvIII+ newly diagnosed GBM in March. The trial was terminated early when it became clear the drug was unlikely to improve overall survival, but it was shown to be safe ([A#14160307003]).

The company said that it was at a loss to explain the outcome because performance in mid-stage single arm studies of newly diagnosed patients had looked promising in comparison with historical controls with Temodar. But in the pivotal trial, performance of the comparator Temodar was far better than expected, according to the company.

Overall survival in an interim analysis was 20.4 months for Rintega vs. 21.1 months for the comparator Temodar (both drugs were given on top of radiation).

Researchers interviewed for this article said that they would need to see the full results to understand what happened in the study.

Some question whether the right historical controls were used for comparison in the mid-stage studies. Celldex noted that the OS for temozolomide ranged from 12 months to 18 months across trials.

But there are other historical studies that show similar results for temozolomide as the Rintega pivotal trial specifically in EGFRvIII+ newly diagnosed patients with stable disease, Datamonitor's Goode pointed out. In a trial published by Nicola Montano and colleagues in the journal Neoplasia in December 2011, the OS for EGFRvIII+ patients treated with Temodar was 19.3 months, which is close to ACT IV results, he noted.

H.C. Wainwright analyst Swayampakula Ramakanth also highlighted the same study in a March 8 note, observing that “the majority of the patients in the Montano study had received gross total resection (59 out of 73 patients) and received standard-of-care treatment consisting of radiotherapy and temozolomide, similar to the patients in the control arm of the ACT IV study.”

The Montano data suggest EGFRvIII expression was a positive, not a negative prognostic factor in the pivotal study of Rintega, Ramakanth said.

Rindopepimut did show a survival benefit over Avastin in a Phase II study of recurrent GBM, but Avastin is not a terribly effective drug and that wasn’t the target population for the pivotal trial, Goode added.

The Rintega failure emphasizes the need for more randomized mid-stage work, because the use of single arm studies with historical controls can be misleading, Lim, who is director of brain tumor immunotherapy at Johns Hopkins, told the Pink Sheet.

NCI's Gilbert suggested that the optimal study might be an integrated randomized Phase II/Phase III study, where efficacy is examined early and researchers move quickly to a fully powered Phase III study if they see a signal. Gilbert is the principal investigator on a multi-arm study of checkpoint inhibitors and Temodar that the NCI is running with such a design (NCT02311920). The first cohort has enrolled and once the safety evaluation of the patients is complete, the next one will open.

Gilbert commented that the EGFRvIII target was interesting, with some promising earlier stage results, and that following the failure, "we need to always, always remember that there are other cancers that were in the same boat we are in, like melanoma, which now is a phenomenally successful area."

Bristol Undeterred

BioMedTracker analyst Edny Inui acknowledges the Rintega failure was disappointing but does not expect it will keep pharmas from the space, which offers a large market share for those who can make even a small impact.

Inui does expect, however, that interest will shift to more targeted immunotherapy as opposed to global immunotherapy.

No vaccine has succeeded in 50 years of ongoing effort, but that “will not detract enthusiasm for immuno-oncology approaches for glioblastoma,” Bristol’s Jean Viallet, global clinical research lead in oncology, asserted in an interview.

Johns Hopkins' Lim sees a silver lining in the Rintega trial failure. The study has hopefully established that Rintega is safe and the door is still open to research of well-tolerated vaccines as part of combinations, including with PD-1 and IDO inhibitors, he said.

In January, a Duke University investigator started a study (NCT02529072) of a dendritic cell vaccine developed in-house in combination with Opdivo in 66 patients.

What’s In The Pipeline

The industry-sponsored pipeline is diverse and includes PD-1 inhibitors, gene therapies and cancer vaccines, Goode noted in his GBM pipeline report.

According to the BioMedTracker database, for brain cancers overall, including malignant gliomas, anaplastic astrocytomas and GBM, there are now nine drugs in Phase III, 39 in Phase II and 24 in Phase I. There are also 73 suspended programs in the pipeline.

Six drugs are in late-stage development for GBM. In addition to Opdivo, they include two cellular therapies – Northwest Biotherapeutics Inc.’s DCVax-L vaccine and ImmunoCellular Therapeutics Ltd.’s ICT-107 autologous dendritic vaccine ICT-107 – and VBL Therapeutics’ gene therapy VBL-111, which is being developed in combination with Avastin (see chart below).

In their Nature Reviews Neurology paper, Lim and colleagues noted that “the traditional assumption has been that immune responses in the CNS were limited,” because of the blood-brain barrier, among other reasons, but “this view has recently been challenged, as it has become clear the CNS [central nervous system] actively communicates with the immune system.”

Regarding checkpoint inhibitors, the authors were enthusiastic, while cautioning of challenges in glioblastoma studies, including the lack of assays for immune response, the need for different measures of response for checkpoint inhibitors, and management of immune-related CNS adverse events, they noted.

Bristol's Viallet said that the field is clearly focusing on reversing mechanisms of immunosuppression in the tumor microenvironment as the most promising way to bring immuno-oncology to patients with glioblastoma.

The rate of PD-L1 expression is very high in GBM at 88% of first-line tumors and 72.2% in recurrent disease, and PD-1 therapy improved overall survival in preclinical studies, Goode noted in his report. Furthermore, there are some clinical data; results for 10 patients in a safety cohort of one study suggested biologic activity of checkpoint inhibition in GBM.

Goode cautioned in his pipeline report that excitement aside, the supporting clinical data are lacking for checkpoint inhibitors.

Inui noted that along with increased interest in exploring checkpoint inhibitors will come a rise in development work to tackle the challenge of how to sequence and use immunotherapies together and with standard-of-care treatments. A range of checkpoint inhibitors and other types of immunotherapies could potentially play a role in treatment.

“That’s something you will see more of as immunotherapy fever takes over in the glioblastoma space,” Inui said.

Glioblastoma tumors are highly immunosuppressive, which suggests a combination approach is needed, and more research is also needed to develop biomarkers to help predict response, Lim said.

Personalized Approaches

The trend in recent years has been for therapies to be personalized, based on mutation and MGMT methylation status, the Datamonitor report notes.

Immunocellular Therapeutics’ ICT-107 autologous dendritic cell vaccine targets the human leukocyte antigen A2 (HLA-A2), which is present in about half of patients. This candidate targets six different antigens associated with GBM and by “targeting multiple biomarkers, ICT-107 overcomes the tumor escape mechanisms commonly associated with single antigen targeting,” Goode noted.

“Additionally, the therapy targets biomarkers found on cancer stem cells, which are believed to be a key contributor to the proliferation and recurrence of tumors," the analyst added.

However, while ICT-107 improved progression-free survival significantly in a Phase II study, it failed to improve OS, he noted.

A pivotal study will stratify outcomes based on MGMT methylation status, which could enhance positioning with payers, in Datamonitor’s view.

On the other hand, ICT-107 development trails behind NorthWest Biotherapeutics' DCVax-L, which Datamonitor points out has a shot at being the first brain cancer vaccine on the market, with approval possible as early as the second quarter in 2017 in the US (versus a third-quarter 2020 US entry for ICT-107).

DCVax-L still needs to prove itself clinically, however. Early stage data looked promising but were derived from very small trials that were not peer-reviewed, Goode noted in his report.

Interim Phase III results were expected in 2015 but never materialized. A partial clinical hold was placed on the trial in December, stopping enrollment of new patients. In a May 2 statement, the company said it has enrolled over 300 out of 348 patients in the pivotal trial and hopes for a resolution soon.

The company also said it was in talks to run three different Phase II studies of its vaccine in three different combinations. It did not disclose the tumor types or trial partners, aside from saying the deals involved "leading participants in the immunotherapy field."

The partial hold is unlikely to have a big effect on the study considering 300 were already enrolled and continued, Inui commented, but the hold is not a good sign for development generally if robust results aren’t observed in GBM.

Glioblastoma: Late-stage Pipeline

Sponsor

Drug

Target+

Drug Type

Pros

Cons

Northwest Biotherapeutics

DCVax-L

Immune system

Dendritic cell vaccine

Could be first brain cancer vaccine to market. Compassionate use in Germany lends credibility. Early data suggest long-term survival benefit. Safe, infrequent dosing after initial infusions.

Personalized and complex manufacturing. Expected to be expensive. Limited trial data for small number of patients, not peer reviewed.

ImmunoCellular Therapeutics

ICT-107

AIM-2, gp100, HER-2, Il-13ra2, MAGE-1, TRP-2

Dendritic cell vaccine

Targets multiple biomarkers to overcome tumor escape. Safe, infrequent dosing. Works in unmethyated and methylated GBM.

Limited to HLA-A2 positive disease. Likely to be expensive. Personalized and complex, Likely to reach market long after competing DCVax-L.

Bristol-Myers Squibb

Opdivo

PD-1

Monoclonal antibody

Bristol's brand-success in numerous cancers. Well-tolerated. Tested in high unmet need subtype: unmethylated GBM.

Lack of GBM-specific data. Expensive when used with Yervoy. Biweekly dosing is more frequent than vaccines.

VBL Therapeutics

VB-111, developed in combination with Roche’s Avastin

TNF-alfa

Gene therapy biologic

Novel mechanism involving gene therapy. Avoids problems with treatment resistance by targeting tumor vasculature. Less costly to make than dendritic vaccines.

Use with Avastin means more expensive and more side effects than other pipeline drugs. May not enter EU as Avastin not approved there.

Tocagen

Toca-511 and Toca Fc

DNA

Gene therapy biologic and small molecule combination

Novel mechanism of action involving gene therapy elicits antitumor autoimmune response and kills cancers directly. Less costly to make than dendritic vaccines.

Limited to patients undergoing planned surgery. Limited to recurrent GBM. Company lacks experience in market.

+AIM-2 = absent in melanoma 2; gp100 = glycoprotein 100; HER-2 = human epidermal growth factor receptor 2; IL-13Ra2 = interleukin-13 receptor alpha 2; MAGE-1 = melanoma antigen-encoding gene 1; PD-1 = programmed cell death protein 1; TNF-alpha = tumor necrosis factor-alpha; TRP-2 = tyrosinase-related protein 2

Sources: Datamonitor Healthcare, BioMedTracker. Examples of strengths and weaknesses based on Datamonitor assessments.

Can Checkpoint Inhibitors Jump-Start Glioblastoma Drug Development?

More from Clinical Trials

More from R&D