乌鸦传媒

Expert Critique

FROM THE ASCO Reading Room

Hardeep Phull, MD Hematologist/Oncologist Sharp Healthcare & Sharp Rees Stealy

Chimeric antigen receptor T cells (CARTs) are a novel cell-based therapy designed to direct host-derived T-cells against cancer cells through tumor-specific surface receptors. Our preliminary experience with CARTs has been primarily against a variety of relapsed, refractory B-cell malignancies including ALL, CLL, and NHL. However, this approach is also being explored against a wide variety of solid malignancies as well.

Most recently, data for the efficacy of the JCAR17 platform for relapsed/refractory diffuse large B-cell lymphoma (DLBCL), the most common lymphoma in the United States, were presented by Dr. Jeremy Abramson and his colleagues at the 2017 International Conference on Malignant Lymphoma. This specific construct consists of equal quantities of CD4+/CD8+ CARTs which are transduced separately after apheresis and directed against CD-19 on tumor cells subsequently in a 1:1 infusion. There are unique features of this platform, including a 4-1BB co-stimulatory domain and an EGFR 鈥渢racking鈥� domain. These preclinical design details incorporated into well-defined JCAR17 constructs have been purported to result in better expansion, improved efficacy against tumors, and more predictable and consistent pharmacodynamics.

In this ongoing Phase II multicenter study, 55 heavily pretreated patients with relapsed/refractory DLBCL, including almost 50% without a response to autologous stem cell transplantation (ASCT), were treated with fludarabine/cyclophosphamide lymphodepletion, followed by a dose of JCAR17. At 3 months, the overall response rate (ORR) was found to be 51%, with a complete response (CR) rate of 39%. Remarkably, this high degree of response was preserved in high-risk patients, including those who were refractory to chemotherapy after transplantation, or had double/triple hit lymphomas. Also, patients with higher doses of JCAR17 had an improved ORR/CR, associated with higher CD8+ cell expansion, suggesting a durable remission. Regarding safety, about one-third of the patients did develop cytokine release syndrome (CRS) as expected, requiring treatment with tocilizumab and steroids, although only one case was grade 3-4. The rate of neurologic toxicity was 22%, with 16% being grade 3-4 events. Nevertheless, the adverse events were completely reversible in all cases, except for in one patient with neurotoxicity. Interestingly, in one patient who relapsed after obtaining a CR following JCAR17 infusion, administration of a PD-1 inhibitor resulted in remission within 3 months, suggesting a potential resistance mechanism of immune escape which can be abrogated by PD-1 blockade.

We are finding that CD-19 directed CARTs for relapsed/refractory DLBCL can induce remarkably durable remissions and cures in patients who previously had limited treatment options after being chemotherapy refractory, with an associated low chance of achieving remission and poor prognosis. Meanwhile, we are getting better at predicting and treating CRS and neurotoxicity, although further investigation is essential to better understand and mitigate these life-threatening adverse events. Nevertheless, as the CART repertoire expands to other B-cell malignancies and solid cancers, we can expect to see further studies investigating potential variables regarding their function in various settings. The ongoing study on DLBCL with JCAR17 noted above seems to suggest that unique features incorporated into the design of CART constructs can result in consistent performance and improved efficacy, with possible mechanisms for resistance that can also be overcome. Whether CART therapy could be used at an earlier line of therapy, combined with novel agents such as checkpoint inhibitors, or replace ASCT altogether remain possibilities, as suggested by Dr. Abramson.

In clinical practice, we know that further variables will need to be investigated to further fine-tune CARTs in B-cell malignancies including CART factors (i.e., alterations in various CART domains, the selection of the tumor surface marker, the 鈥渄ose鈥� of cells infused, ideal CD4:CD8 ratios, in vivo tracking and anergy, resistance mechanisms, combination treatments) as well as patient factors (prior therapies, native immune response, tumor expression profiles, and disease burden).

Full Critique

Chimeric antigen receptor (CAR) T-cell therapy can induce durable remissions and possibly cure patients with relapsed/refractory diffuse large B-cell lymphoma (DLBCL) with poor prognosis and no standard therapeutic options.

"This completely novel therapy makes a dent in the treatment of the most common lymphoma subtype, with predictable and manageable toxicity," Jeremy Abramson, MD, clinical director of the Center for Lymphoma at Massachusetts General Hospital Cancer Center in Boston, told 乌鸦传媒. "For oncologists who treat refractory DLBCL, this type of treatment looks like a game changer."

He noted that several small molecules and immunotherapy approaches have been tried for DLBCL patients who have had at least two prior lines of therapy, but none of the treatments have induced remission. There are no standard treatment options for this population, which has an unmet medical need, Abramson said.

"We are targeting the most difficult-to-treat patients in the most common lymphoma in the U.S., with 70,000 new cases each year. This is the ideal patient population to study, because nothing else works, including novel therapies. The majority of patients who have relapsed after prior lines of therapy are usually chemotherapy refractory. Their , with a low chance of remission."

Abramson presented data at the 2017 International Conference on Malignant Lymphoma in Lugano, Switzerland, about encouraging efficacy and safety with the CAR T cell JCAR17, which is one of a number of anti-CD19 CAR T cells in development. JCAR17 is a CD19-directed CAR T cell product administered in equal quantities of CD4-positive and CD8-positive CAR-positive T cells.

Preclinical data show that administration of a leads to better expansion and anti-tumor activity. Flat dosing and defined composition allow for administration of a consistent dose of CD4-positive and CD8-positive CAR T cells, he explained.

"This CAR T cell has unique features, including a 41BB co-stimulatory domain and a truncated EGFR receptor that allows tracking of the compound." As opposed to other anti-CD19 CAR T cells, in JCAR17 after apheresis for lymphocytes, T cells are selected out, transduced separately, encoding the receptor, and then re-infused in a 1:1 ratio.

In general, T cells are transduced and expanded, and each patient is given a different quantity of T cells. "We know exactly what dose each patient is getting. This gives us a better sense of how the product is working."

Abramson reported results from an . Of the 55 evaluable patients, 40 had DLBCL not otherwise specified, 14 had transformed DLBCL, and one patient had grade 3B follicular lymphoma. Fifteen patients had double-hit or triple-hit lymphoma.

The 55 patients, median age of 61, had a median of three prior treatments; three-quarters of the patients were chemotherapy refractory, and about half had not responded to autologous stem cell transplantation (ASCT). Patients received one of two doses of JCAR17 after fludarabine/cyclophosphamide lymphodepletion.

Overall, the evaluable toxicity was manageable, Abramson said, with about one-third of patients developing cytokine release syndrome (CRS), which was grade 1 or 2 for all except one patient who had grade 3-4 CRS. No increase in CRS was seen from the first to second dose levels. The rate of neurologic toxicity was low (22%), with 16% grade 3-4 and 6% grade 1-2.

All the cases if CRS and all but one neurotoxicity were completely reversible, he said, adding that "about 60% of patients had no toxicity." Rescue medications included tocilizumab in 11% of patients and dexamethasone in 24%.

Abramson called the response rate "encouraging," with the best overall response rate (ORR) of 76%, with 52% complete responses (CRs). At 3 months, ORR was 51% and CR was 39%. Subsets of patients who were chemotherapy refractory post-transplant, or had double-hit or triple-hit lymphoma, all showed high-quality responses.

"If we look across the dose levels, we see some suggestion of durable remission in the higher dose level at 3 months, with a 64% ORR and 46% CR," said Abramson. He noted there was higher CD8 expansion at the higher dose level, and that a more refined product might further optimize the CR rate.

Many DLBCL patients relapse with detectable CAR T cells in their blood. The vast majority still have the CD19 antigen, which is different from in acute lymphoblastic leukemia, he said. "This suggests an immune evasion mechanism in relapse. An inflammatory response may induce immune expression of checkpoints on T cells. It's appealing to think this mechanism of resistance may be due to programmed death-1 (PD-1) interaction, because we could block those responses."

One chemotherapy-refractory patient who achieved CR relapsed with detectable CAR T cells at low levels and was re-treated with CAR T cells with no response. The patient received a PD-1 inhibitor and went back into remission 3 months later. Abramson said he believes that immune escape appears to be the dominant mechanism of resistance for a significant proportion of patients.

Assuming JCAR17 receives FDA approval, the best candidates for therapy may be those who had two lines of prior therapy and for whom ASCT was not effective. "Ideally, we could use CAR T cell therapy earlier. Putting patients on three-plus lines of standard therapy is probably too late," said Abramson. "After R-CHOP, less than half of patients are cured with high-dose chemotherapy and ASCT. Perhaps we should do CAR T cell therapy instead of ASCT."

Results from a planned randomized trial will be needed to prove that CAR T cell therapy can replace ASCT, he added.

An ongoing JCAR17 study with pivotal dose expansion is accruing patients, and a randomized DLBCL trial is studying a combination of the CAR T cell therapy with or without checkpoint inhibitors.