April 2019 Edition Vol.12, Issue 4

Targeted Therapy, Vaccines, CAR-T, and a Practice Changer at AACR 2019

By Neil Canavan

“As a pancreatic cancer clinical trialist, this is not a waterfall plot I expect to see…” Mark O’Hara, MD, Abramson Cancer Center, U. Penn.

Clinicians who specialize in the treatment of pancreatic cancer are unaccustomed to delivering good news, yet, for at least one clinical trial reported at AACR 2019 – specifically, the waterfall plot that clearly shows the efficacy of a new cancer immunotherapy agent called APX005M – this was very good news indeed.

These data, as well as results from investigations of other immunotherapy approaches (5 in all), and one targeted agent, are highlighted herein.


A diagnosis of metastatic pancreatic cancer is almost invariably a death sentence. “Even with the therapeutic advances of FOLFIRINOX (drug combination), and gemcitabine/nab-paclitaxel, responses are still limited to less than 30% of patients,” said Dr. O’Hara, who presented the data for APX005M, “And the median overall survival (OS) is still less than one year.” Further, despite the recent Nobel Prize accolades for the efficacy of so-called checkpoint inhibitors, nothing has changed. “No responses to single agent checkpoints have been noted,” said O’Hara, adding that a recent investigation of checkpoints combined with chemotherapy achieved only, “modest results.”

The target of APX005M is a molecule called CD40, which is not an immunologic brake pedal, as are the approved checkpoint agents to date, but rather, an immune system stimulant, which, among other activities, helps to prime and activate antigen presenting cells, the critical command posts for the mobilization of tumor-killing T cells.

In the present phase 1b study, Dr. O’Hara investigated the use of two different doses of the CD40 agonist, APX005M, combined with gemcitabine/nab-paclitaxel, +/- nivolumab (an anti-PD1 inhibitor) in four cohorts of treatment-naïve patients with metastatic pancreatic ductal adenocarcinoma. The primary and secondary endpoints were safety and efficacy, respectively (N=30).

Results: For safety, the combination of chemotherapy with not one, but two immunotherapy agents resulted in only two grade 3/4 adverse events (febrile neutropenia), which were transient. For the 23 patients evaluable for efficacy, 11 achieved a partial response (PR, with 2 PRs unconfirmed), 9 patients demonstrated stable disease (SD), and one patient had disease progression (PD). Overall, the response rate in this very difficult-to-treat patient population was 46.7%, with durations of response for some patients lasting more than 10 months.

These very encouraging data have already had their impact and a phase 2 investigation of APX005M is currently underway.

Solid Tumor CAR T Cells: IcasM28z and the other (2)

“These data are potentially the most compelling CAR-T data that we’ve ever seen in solid tumors to date.” Nilofer Azad, MD, Johns Hopkins

The efficacy of chimeric antigen receptor (CAR) T cells, a technology whereby a patient’s own T cells are genetically reprogrammed to fight their cancer, has been without precedent in the hematologic setting. However, translation of that success into the language of solid tumors has resulted in, what are at best, mixed messages.

Enter the mesothelin and HER2-targeting CARs.

Regarding the former: “The novelty of our study is that the CAR-T cells are delivered directly to the tumor site using regional delivery techniques,” said Prasad Adusumilli, MD, Memorial Sloan Kettering Cancer Center (MSKCC). The intent of regional administration is to avoid trafficking issues that have arisen with CARs delivered systemically. “If this approach is successful, 2 million patients with mesothelin-expressing solid tumors per year in the United States will be eligible for this treatment.”

The construct, IcasM28z, is a fully humanized CAR – the first of its kind – which includes a “suicide” switch to eradicate the cells should unexpected toxicity occur (a major concern given the past safety record of solid tumor-targeting CARs).

For this phase 1 study, Adusumilli and colleagues investigated the use of IcasM28z in a cohort of patients with malignant pleural disease. After cyclophosphamide preconditioning, IcasM28z CAR T cells were injected directly into the pleural cavity (N=21). After CAR-T administration, 14 of 21 patients went on to receive anti-PD1 checkpoint blockade agents, an addition justified by preclinical studies. (Note: this tumor type is poorly responsive to monotherapy with checkpoint agents.)

Results: Of the 11 patients who received CAR-T + anti-PD-1, 2 patients experienced a CR (complete response) 6 had a PR, and 3 experienced disease progression. The overall response rate (ORR) was 72%. The treatment was well tolerated and no on-target, off-tissue toxicities were observed.

With these data in hand, the next step is to program the observed anti-PD-1 synergy directly into the IcasM28z CAR, via a decoy receptor. “We designed a PD-1 dominant negative receptor,” Adusumilli explained, which renders the receptor inactive. “Cancer cells can keep pressing the brake on the CAR cell, but the CAR is no longer inhibited.” The reasoning here is that since so many immunologic regulatory pathways are redundant, “…if we took out one (immunologic) brake, another would arise.” Thus, a decoy.

A further tweak to this technology, much of which emanates from the lab of Michel Sadelain (MSKCC and co-founder of Juno), the IcasM28z CAR will have its activity modulated in order to avoid T cell exhaustion, a major concern in CAR-T treatment. Further still, Atara Biotherapeutics, the licensee of all this whiz-bang, is also looking to park their CAR in a “universal” allogenic platform.

HER2, as Well

The molecule HER2 (human epidermal growth factor receptor 2) has long been a therapeutic target in certain types of breast cancers, and the HER2-targeting agent, traztuzumab, is one of the most effective oncology drugs approved. However, there are other tumor types that overexpress HER2.

“HER2 can be expressed in a variety of sarcomas, including osteosarcoma, the most common malignant bone tumor,” said Shoba Navai, MD, Baylor College of Medicine. “However, unlike HER2 positive breast cancers, the amount of HER2 on the cell surface of sarcoma is much less. This lower level of expression is not sufficient for HER2 targeting drugs, such as traztuzumab to have efficacy.” Thus, an investigation into the more sensitive HER2-targeting CARs.

For this phase 1 study, patients with HER2 positive sarcomas were lympho-depleted with the standard flu/cy regimen, and then treated with one of two dose levels of HER2 CARs. Patients were followed for six weeks and assessed for safety and efficacy (N=10).

Results: For safety, “…no serious cardiac or pulmonary toxicities were observed. All patients maintained normal heart function throughout the study period.” Which is critically important, given the target. “There was a previously reported patient who died after being treated with a traztuzumab-based CAR (at the NCI),” said Navai. “Our construct uses a different scFv, and in the eight years we’ve been using it we’ve not observed any cardio-toxicity.” Further, as is common to CARs, there was no grade 3/4 cytokine release syndrome (CRS).

As for efficacy, 3 patients achieved CR, 4 had SD, and 4 patients progressed (all but one patient was metastatic at study entry). The survival rate at one year was 60%.

Giving these positive signals, this construct will follow a similar development pathway as the Meso CAR, with the incorporation of an anti-PD-1 moiety. Future studies will include other HER2 expressing solid tumors.

Gamma Delta T Cells, Prophylactic Vaccine

Wrapping up the immunotherapy coverage is one novel approach to treat cancer, and another to prevent it. The novel approach uses a different type of T cell, the γδ (pronounced, gamma delta) T cell, which functions much more like an NK cell of the innate arm of the immune system. As stated by study author, Lawrence Lamb, PhD, co-founder, and CSO of Incysus Therapeutics, “These γδ T cells function as innate immune effectors. They recognize and immediately kill malignant cells experiencing bioenergetic, oxidative or chemically-induced stress, so they don’t require activation by antigen-presenting cells, or recognition of classical major histocompatibility antigens.” Meaning, as with NK cells, you do not need to identify an antigen target for therapeutic efficacy. For γδ T cells to activate all you just need is a stressed cancer cell.

Such stress, as in the setting of glioblastoma (GMB) is provided by the standard-of-care drug temozolimide, and the secret sauce to leveraging this stress is to genetically render the γδ cells chemo-resistant, thereby allowing them to survive in an environment hostile to immune cells.

Coinciding with the proof of principle data reported at AACR, the FDA just approved the investigational new drug application for this approach for the clinical study of genetically modified γδ cells in combination with chemotherapy for the treatment of patients with newly diagnosed GBM.


“The Holy Grail is not curing cancer – the Holy Grail it’s preventing it.” Louis Weiner, MD, Georgetown University

In brief, there was a tantalizing presentation by Steven Lipkin, MD, PhD, Weill Cornell Medicine looking at a way to prevent cancer arising from Lynch Syndrome. “Lynch is a genetic disease which confers approximately 70%-80% lifetime risk of developing colorectal cancer,” said Lipkin. Thanks to revelations provided by what is now widely available genetic testing, “what we thought was an orphan disease we now know is a risk to 1 of every 250 people in the United States.”

In considering a vaccine strategy, Lipkin and colleagues looked deep into the genetics of Lynch and found that there were molecular commonalities caused by typical Lynch frame shift mutations. This enabled the creation of a vaccine with an array of novel components, so-called neoantigens, which in this rare circumstance are shared between (potential) patients.

Using a mouse model of Lynch Syndrome, Lipkin was able to demonstrate extended survival with his vaccine, with results promising enough that human trials will commence in the near future.

“Here you see the very first example I’ve ever heard of a mutated neoantigen that is shared among people with this particular syndrome,” said Dr. Weiner, adding, “Which raises the possibility of being able to create a single type of vaccine with a pool of peptides that could potentially work in millions of people.”

Targeted Therapy: FLT3

“These results are truly practice changing.” Alexander Perl, MD, Abramson Cancer Center, U. Pennsylvania

The target here is the mutated form of the cytokine receptor FLT (pronounced, “flit”) 3. “Because FLT3 is a tyrosine kinase, and its mutated form activates the kinase in cancer, there have been a lot of study in drugs that inhibit FLT3,” said study presenter Alexander Perl, MD, Abramson Cancer Center, U. Penn. However, previously developed agents were only moderately active, too toxic, or quickly led to resistance. A new drug, the FLT3-inhibitor, gilteritinib, was designed to avoid all three drawbacks.

A phase 3 investigation compared gilteritinib vs dealer’s choice salvage therapy in a population of FLT3-mutant positive, relapsed/refractory AML patients (N=371).

“Even though AML is a highly chemotherapy sensitive disease,” explained Perl, “There is a 40 to 70% relapse rate after achieving remission with induction chemo, and the eventual outcomes for those patients are dismal.” The FLT3 mutation is a marker of poor outcome and is present in roughly 30% of patients with AML.

Results: “Response rates were more than doubled with gilteritinib therapy,” said Perl. The rate of CR was 21% for gilteritinib vs 11% for standard-of-care (SOC), and the ORR was 68% vs 26%. The duration of response was also substantially longer — 11.0 months for gilteritinib vs 1.8 months for SOC — leading to an overall survival advantage of 9.3 months vs 5.6 months.




Post a Comment

OBR Archives

To view previous issues of OBR green you can visit our archives. The entire library of OBR green articles is searchable.