TIL Therapy Elicits Encouraging Activity in Advanced NSCLC

Adoptive cell transfer with tumor-infiltrating lymphocytes was found to be feasible, have a manageable toxicity profile, and expanded in 95% of patients with metastatic non–small cell lung cancer who had disease progression on nivolumab (Opdivo).

Ben Creelan, MD

Adoptive cell transfer with tumor-infiltrating lymphocytes (TILs) was found to be feasible, have a manageable toxicity profile, and expanded in 95% of patients with metastatic non—small cell lung cancer (NSCLC) who had disease progression on nivolumab (Opdivo), according to results of a small phase 1 trial presented at the 2020 AACR Virtual Annual Meeting I.1

Specifically, the treatment also led to tumor shrinkage in the 20-patient trial, with a best objective response rate (BORR) of 25%; this comprised 2 ongoing complete responses (CRs), 1 partial response (PR), 2 unconfirmed PRs, and 1 unconfirmed PR that is ongoing and awaiting a patient’s next CT scan. If the unconfirmed PR is confirmed, the BORR will be 33%, explained lead study author Ben Creelan, MD, in a virtual presentation during the meeting.

“Despite progression on nivolumab, we did see a decrease in tumor size for many patients, and the ORR was in around one-quarter of patients, and perhaps in a one-third of patients if our unconfirmed PR is confirmed,” said Creelan, of the Thoracic Oncology Program at Moffitt Cancer Center. “This [includes] several patients who had clinical benefit, had escape lesions after local ablative therapy, or a patient who had a response on biopsy despite radiographic pseudoprogression. We’re excited about carrying TILs further in lung cancer.”

TIL adoptive cell transfer therapy has previously demonstrated activity in patients with metastatic melanoma, in which the durable CR rate ranged from 10% to 15%.2

To be eligible for enrollment on the phase 1 study, patients with NSCLC must have had safely accessible metastasis plus at least 1 measurable lesion, could not have received prior treatment with PD-1 inhibitors, no prior use of corticosteroids, could not have had untreated brain metastases, a lung diffusion capacity greater than 50%, cardiac left ventricular ejection fraction greater than 50%, and a negative cardiac stress test.

Patients underwent resection for any accessible metastatic lesions for TIL manufacturing and were given nivolumab intravenously at 240 mg every 2 weeks for 4 doses. If PRs or clinical benefit were observed with the PD-1 inhibitor, then patients continued on nivolumab and did not undergo TIL therapy.

However, if patients experienced disease progression or stable disease with enlarging or new lesions after 2 serial CT scans, then they underwent lymphodepletion with cyclophosphamide at 60 mg/kg plus sodium 2-mercaptoethanesulfonate on days —7, –6, followed by fludarabine at 25 mg/m2 on days —7 to –3.

TILs were infused at ≥20 x 109 to 1010 CD3+ cells, depending on growth, on day 0. On days 1 to 6, an intermediate dose of interleukin-2 (IL-2) was given at 18 miU/m2 over 6, 12, and 24 hours, and then at 4.5 miU/m2 over 24 hours for 3 days. Investigators also administered apheresis for functional studies, as well as maintenance nivolumab at 480 mg every 4 weeks at day +29 up until 1 year.

The primary end points of the trial were safety and tolerability of cyclophosphamide/fludarabine/TIL/IL-2 in this patient population, as well as success of TIL expansion and BORR.

Of the 31 patients who were screened, 20 were enrolled, had their TILs harvested, and received nivolumab. Twelve patients were ineligible after screening due to active brain metastases (n = 4), their lung diffusion capacity was less than 50% (n = 3), they did not have lung cancer (n = 2), had TKI-naïve, EGFR-mutant disease (n = 2), and had prior second cancer (n = 1). Of those 20 patients, 4 could not receive TIL therapy due because of nivolumab response, insufficient TIL growth, decline in performance status, or lost insurance and transportation (n = 1 each).

Of the 16 patients who did receive TIL therapy, 12 were evaluable at day +28, 1 patient was still awaiting a CT scan at day +2, 2 patients discontinued because of toxicity and were unevaluable at day +28, and 1 patient did not have a measurable tumor prior to TIL therapy and were not RECIST evaluable.

The majority of patients had low or no PD-L1 expression, as well as low tumor mutational burden (TMB). The median age was 54 years, the median PD-L1 tumor proportion score was 6%, and the median TMB level was 4.5 mutations/megabase. The median number of smoking pack-years was 9, and the median number of prior therapies was 0 to 1.

Additional results showed that patients experienced clinical responses before and after TIL therapy. The median overall survival both from the start of enrollment and from post-TIL therapy was not reached. One case of apparent pathological response was reported, despite radiographic pseudo-progression. Furthermore, 3 patients with the emergence of “escape lesions” were effectively treated with local ablative treatment.

“Patients mostly had either new lesions or progressive disease with nivolumab monotherapy, although 2 patients had a PR. They subsequently had new lesions and enlargement of non-target lesions, allowing them to enter them into the cyclophosphamide/fludarabine/TIL/IL-2 arm,” Creelan explained. “The majority of patients were treated with TIL/IL-2, and they actually had a resting of progression with the subsequent CT scans.”

Moreover, at week 16, investigators noted persistence of infused TRV-β clonotypes that correlated with benefit (P =.018).

Creelan also noted that that TILs are reactive to autologous tumor and predicted neoepitope, and that TCR clonotype and phenotype analyses suggest good persistence of transferred TILs. The frequency of T cells with specificity for tumor antigens may be between 10% and 15%. Of the 2 patients with CRs, it was found that both had T cells that recognized multiple antigens.

“The TIL [therapy] was reactive to the autologous tumor for many patients and predicted neoepitopes for many of them,” said Creelan. “Clonotype and phenotype analyses suggested good persistence of the transferred TILs—going out to several months. The frequency of T cells showed us a specificity of antigens in about 10% of the clonotypes for tested TIL.”

Regarding safety, TIL therapy was found to have a manageable toxicity profile comprised of expected adverse events (AEs), which included cytopenia, cytokine capillary leak—which is seen traditionally with IL-2—fever, and chills. Additionally, all AEs occurred early and resolved within 2 weeks, according to Creelan.

The pivotal next steps for continuing TIL development include optimizing the treatment for more patients, making the approach more accessible to patients by finding ways to produce the therapy faster and with fewer AEs, Creelan concluded.


  1. Creelan B, Wang C, Teer J, et al. Durable complete responses to adoptive cell transfer using tumor infiltrating lymphocytes (TIL) in non-small cell lung cancer (NSCLC): a phase I trial. Presented at: 2020 AACR Virtual Annual Meeting I; April 27-28, 2020. Abstract CT056.
  2. Rosenberg SA, Restifo NP. Adoptive cell transfer as personalized immunotherapy for human cancer. Science. 2015;348(6230):62-68. doi:10.1126/science.aaa4967.

<<< 2020 AACR Annual Meeting