T-cell Changes: Why Only Some Respond to Ipilimumab

Janis C. Kelly

May 28, 2014

The immunotherapy ipilimumab (Yervoy, Bristol-Myers Squibb Company) works amazingly well in some patients, hardly at all in others. A groundbreaking study that used deep sequencing techniques offers some clues as to why.

Ipilimumab, which is marketed for melanoma but is being explored in several other cancer types, including prostate cancer, acts as a checkpoint blocker by inhibiting cytotoxic T lymphocyte– associated antigen–4 (CTLA-4).

Immune repertoire sequencing has confirmed that blocking CTLA-4 increased turnover and diversity of the T-cell repertoire in some patients with advanced prostate cancer or metastatic melanoma, but also showed that patients who survived longest maintained clones of high-frequency T-cells they had developed before starting treatment.

The new finding was reported by Lawrence Fong, MD, professor in residence, Division of Hematology/Oncology, University of California, San Francisco, and colleagues in Science Translational Medicine.

"We demonstrated how anti-CTLA-4 changes the T-cell clones circulating in a patient," Dr. Fong told Medscape Medical News. "This 'scrambling of the deck' could allow a patient to generate T cells that target the cancer and could explain why this treatment takes time to work. This is consistent with what people suspected might be going on, but our results are the first demonstration of this."

"The surprising finding was that better clinical outcomes were associated with maintenance of a preexisting immune response," Dr. Fong added.

Monitoring changes in the T-cell repertoire might eventually help clinicians meet 2 of the challenges of treating cancer with targeted biological agents: selecting patients who are likely to benefit from treatment (while avoiding unnecessary expense and toxicity risk for patients unlikely to benefit), and avoiding premature discontinuation of treatment in patients with delayed responses.

For example, only around 10% to 15% of patients with malignant melanoma respond to ipilimumab, but long-term data reported in 2013 showed that 17% to 25% of responders survived more than 3 to 10 years, vs average survival of 10 to 11 months in similar patients treated with conventional chemotherapy.

T-cell clonotype monitoring, as described by Dr. Fong and colleagues, might help physicians select the subgroup of patients most likely to respond to anti-CTLA-4 treatment. However, Dr. Fong said that the deep sequencing techniques used in this study are less likely to be useful in initial patient selection than in monitoring response after beginning treatment.

CTLA-4 is a T-cell surface protein that acts as a brake on killing of tumor cells. Anti-CTLA-4 antibodies block the brake and allow T cells to attack and kill cancer cells. Dr. Fong's group examined the underlying mechanism by using the LymphoSIGHT (Sequenta, Inc.) immune repertoire sequencing platform to measure the frequency of individual, rearranged T-cell receptor β (TCRβ) genes as a way of characterizing the diversity of rearrangements, known as T-cell clonotypes.

The study was done using cryopreserved peripheral blood mononuclear cells (PBMCs) from 25 castration-resistant prostate cancer (CRPC) patients concurrently enrolled in a phase 1/2 clinical trial of ipilimumab (up to 4 doses ranging from 1.5 to 10 mg/kg and GM-CSF 250 µg/kg/day), and from from 21 patients concurrently enrolled in a phase 1 trial of single-agent tremelimumab (a CTLA-4 blocker that had been under development by Pfizer) at 15 mg/kg given every 3 months. The clonotype repertoires and survival times in these 2 groups of cancer patients were compared with those of a cohort of 9 untreated control participants, which had been obtained from Cellular Technology Limited.

This study showed in that in the CRPC patients treated with ipilumumab and GM-CSF and in the metastatic melanoma patients treated with tremelimumab, the immunotherapy treatment increased TCR diversity (seen in the number of unique TCR clonotypes) and that the clonotype repertoire continued to evolve during months of treatment. However, the number of clonotypes that increased with treatment did not have a straightforward association with clinical outcome. Instead, improved overall survival was seen in patients who maintained high-frequency clones that they had at baseline, prior to treatment. Patients who had decreases in their baseline highest-frequency clonotypes also had shorter survival.

The important clonotypes associated with better survival included T cells with high-avidity T-cell receptors, such as virus-reactive T cells. The authors speculate that these clones might represent preexisting high-avidity T cells.

"We propose that for most long-term survivors, high-avidity clones present at baseline are readily available after checkpoint inhibition to recognize relevant tumor antigens and mediate durable responses. For others, an increasingly diverse and evolving TCR repertoire allows for discovery and expansion of subdominant tumor-reactive clones. Because only a proportion of patients obtain long-term benefit, identification of these preferential clonotypes could improve patient selection for patients receiving single-agent CTLA-4 blocking antibodies," the authors wrote.

The researchers also suggested that determining antigen specificity for persistent clonotypes could inform potential vaccination strategies in combination or in sequence with CTLA-4 blockade to increase efficacy in the majority of patients who will not benefit from CTLA-4 blockade alone.

Dr. Fong told Medscape Medical News that the study data did not align neatly with either the "threshold" model of CTLA-4 blockade or the "attenuation" model, but suggested that preexisting T-cell responses might hold the key to better outcomes after anti-CTLA-4 blockade.

Dr. Fong said that patients appear to develop individual-specific immune responses to their own tumors rather than responses to shared antigens. "I think we showed that people do not neatly conform to a model based on animal models. If anything, we see both models in action: an increase in the diversity of the repertoire (threshold model), but a maintenance of the high-avidity T cells (attenuation model)," he commented.

Dr. Fong reported no relevant financial interests. Coauthors Mark Klinger, Craig Cummings, and Malek Faham are employees of and own stock in Sequenta. Sequenta has pending patent applications on the TCR sequencing platform.

Sci Transl Med. Published online May 28, 2014. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.