Subbiah Explores Treatment of Non–Small Cell Lung Cancer With a KIF5B-RET Fusion

April 9, 2021
Targeted Oncology Staff

Case-Based Roundtable Meetings Spotlight, Case-Based Roundtable Meetings Spotlight March 2021: Solid Tumors,
Pages: 40

Vivek Subbiah, MD, explored the treatment options for RET-fusion positive non–small cell lung cancer during a virtual Targeted Oncology Case-Based Roundtable event.

Vivek Subbiah, MD, associate Professor, Department of Investigational Cancer Therapeutics, Division of Cancer Medicine, and medical director, Clinical Center for Targeted Therapy, Cancer Medicine Division The University of Texas MD Anderson Cancer Center, explored the treatment of RET-fusion positive non–small cell lung cancer during a virtual Targeted Oncology Case-Based Roundtable event.

Targeted OncologyTM: What is the frequency of distribution of RET fusions and mutations in lung cancer?

SUBBIAH: Lung cancer has become a poster child for precision oncology, beyond EGFR mutations, ALK fusions, ROS1 fusions, and, recently, NTRK fusions. We see RET fusions in non–small cell lung cancer [NSCLC] with a frequency of 1% to 2%.1 In major cancer centers, the more we test, the more we find. As long as we do broad-based panel NGS testing in these patients, we are able to identify these [mutations]. It’s seen in 10% to 20% of papillary thyroid cancers. In medullary thyroid cancer [MTC], almost 100% of the patients with a hereditary form of thyroid cancer have a RET mutation, and anywhere from 60% to 90% of patients with MTC have a RET mutation. The most common RET mutation in MTC is M918, and in lung cancer, the most common fusion partner is the KIF5B.

How do the National Comprehensive Cancer Network (NCCN) guidelines suggest treating patients with NSCLC and RET mutations?

The latest NCCN guidelines for RET-rearranged positive NSCLC [say for] RET rearrangement discovered prior to first-line systemic therapy, the preferred first-line therapy is selpercatinib [Retevmo] or pralsetinib [Gavreto].2

Vandetanib [Caprelsa] and cabozantinib [Cabometyx] are category 2B and useful under certain circumstances. [For] the second line on progression from frontline [treatment], either selpercatinib or pralsetinib is recommended. [The NCCN guidelines] suggest initiating systemic therapy for NSCLC based on adenocarcinoma guidelines.

If the RET rearrangement is discovered during the first line of chemotherapy, [the guidelines] suggest completing planned systemic therapy, including maintenance therapy, or interrupting therapy followed by selpercatinib or pralsetinib and, in some rare cases, cabozantinib or vandetanib. Again, cabozantinib and vandetanib are category 2B.

Which studies have looked at selpercatinib as treatment in this setting?

LIBRETTO-001 [NCT03157128] is an ongoing, global, multicenter clinical trial.3 As of the latest update, there have been 702 patients enrolled in other cohorts—that includes RET fusion–positive NSCLC, RET fusion–positive thyroid cancers, and RET mutation–positive MTC.4 We are focusing on NSCLC. There were 39 patients, and 105 [were] previously treated with platinum chemotherapy.3

At that time, there were 531 patients on the study, and among those patients there were 253 with RET fusion–positive NSCLC enrolled on the study. Per agreement with the FDA, the primary analysis set for defined criteria consisted of the first 105 consecutively treated RET fusion–positive platinum chemotherapy refractory patients with NSCLC, and the first 55 patients with RET-mutant MTC previously treated with vandetanib or cabozantinib.

For lung cancer, we are going to focus on the first 105 previously treated patients and then the 39 RET fusion–positive treatment-naive patients [from a study] published in the New England Journal of Medicine.

[For the 105 previously treated patients], the median age was the same, 61 years. The male to female ratio was 62 female and 43 male patients. Seventy-one percent of the patients were never smokers. Similar to ALK and other fusion-driven tumors, these patients tend to be light smokers or never smokers, and a few patients were current smokers.

Most of the patients had an ECOG performance status of 1, and almost 90% of the patients that were enrolled on both the arms had adenocarcinoma. The previous platinum chemotherapy treated cohort range of therapy was 3 lines of therapy. Apparently in Japan there was a patient treated with 15 lines of therapy. I don’t know what 15 lines of therapy you would have in cancer, but we have a patient on the study who went through 15 lines of therapy for NSCLC.

All 105 previously treated patients received platinum therapy. There were 55% who received anti–PD-L1 or anti–PD-1 therapy. Almost 50% of patients received another multitargeted tyrosine kinase inhibitor [TKI] for the RET fusion. There were 36% in the 105 data set [of patients who were not treatment naive] who had brain metastasis at enrollment. There were 18% who were treatment naive who had brain metastasis.

The most common fusion partner for the RET fusion was KIF5B in 56% [of patients with previous chemotherapy treatment] and 67% in the untreated cohort.

What was the efficacy in LIBRETTO-001?

The objective response rate [ORR] in the previously treated chemotherapy cohort, based on investigative assessment—that means the local site investigative assessment—was 70%. By independent blinded central review, the ORR was 64%. In the previously untreated patients, the investigative assessed cohort ORR was 90%, and the independent review ORR was 85%. There were 2 complete responses [CRs] in both the independent review and investigator assessment in the previously treated population.

The median duration of response [DOR] was 17.5 months by independent review and 20.3 months by investigator assessment [for patients who were previously treated]. In the previously untreated cohort, the median DOR was not estimable because the patients were still on study at the time of publication.

The median time to response was 1.8 months [overall]. We did scans every 2 months. I think this was the first time they did scans, so the median time to response was 1.8 months. When we treated these patients, and especially if they were clinically symptomatic, they became better within the first couple of weeks once we started treatment.

Responses were observed regardless of prior therapy with anti–PD-L1 or a multitargeted TKI, and regardless of the RET fusion partner.

Most importantly, 1 of our [end] points is central nervous system [CNS] data. There were 38 of 105 patients who had a CNS metastasis at baseline. Among those 38 patients, 11 patients had measurable disease, [meaning it] could be measured by a central radiologist. Ten of those 11 patients with measurable lesions responded for a total of 91% objective intracranial response rate, a CR rate of 27%, and a partial response rate of 7%. The median intracranial DOR was 10.1 months in the data set that was published.

Have there been any updates to these data?

There was an updated poster that we presented at the 2020 American Society of Clinical Oncology [ASCO] annual meeting.5 Of the 253 patients, 80 patients had CNS disease at baseline. Among those 80 patients, 22 patients were deemed measurable by central evaluation. The number of patients with nonmeasurable disease was 58, and most of them were heavily pretreated before enrolling on this study.

Among the 22 patients with measurable CNS disease at baseline, the CNS ORR was 81.8%. There was an 85.7% ORR in the patients who [did not receive prior] radiation. There was 75% ORR in patients with prior CNS radiotherapy. Almost all patients in this data set with measurable disease had an intracranial response. The CNS DOR in the data set was 9.4 months, and the median follow-up in the patients without prior CNS radiotherapy was 15.7 months.

Can you discuss the progression-free survival (PFS) and safety data from LIBRETTO-001?

In the previously treated therapy cohort, the 1-year PFS rate was 66% by independent review. The 1-year PFS rate was 75% in the previously untreated cohort.3

Adverse events [AEs] included hypertension—grade 3 hypertension was seen in 20 patients, which is 14%. We also saw grade 3 increased aspartate transaminase [AST] and alanine transaminase [ALT] levels in these patients. Eight percent of patients had an increased AST level, and 10% of patients had an increased ALT level.

In the entire data set, among all tumor types of 702 patients, 5% of the patients had a permanent discontinuation due to an AE, including ALT elevation, sepsis, increased AST, and drug hypersensitivity action in 0.3 to 0.4% of patients.6 Interruptions due to AEs were seen in 42% of patients, and dose reductions due to an AE were reported in 31% of patients.

On May 8, 2020, the FDA approved selpercatinib for [patients with] lung and thyroid cancer with RET gene mutations or fusions.7

What other studies looked at therapies for RET-positive NSCLC?

The ARROW study [NCT03037385] for pralsetinib was in advanced RET fusion–positive NSCLC.8 The study has not been published, but the drug is FDA approved.9 This was presented at the ASCO meeting, so we have that data set.

This was a phase 2 study of 400 mg once-daily dosing of pralsetinib, and the phase 2 dose expansion is still ongoing. The primary end point was ORR. The key secondary end points were DOR and disease control rate.

In this study, 87 patients [had] received prior platinum therapy, and 27 patients…were treatment naive. Almost all of the patients had metastatic disease. Even in this study, the KIF5B fusions were 75% of the previously treated cohort and 70% of the treatment naive cohort. All of them were heavily pretreated before enrolling on the study in the prior platinum cohort.

There were 43% of patients with CNS disease in the prior platinum cohort, and 37% of the patients had CNS metastasis in the treatment-naive cohort.

How did patients do on the ARROW trial?

The ORR in this study in the treatment-naive cohort was 70%, and the overall response rate was 57% in the closed platinum treated cohort.

The data show 96% of the evaluable patients had tumor reduction, and 100% of the patients who were treatment naive responded. Interestingly, 6% of the evaluable patients had a CR, and 12% of CRs were in treatment-naive patients. Patients responded to pralsetinib regardless of prior treatment and regardless of the fusion partner.

Among the 87 patients with RET fusion–positive NSCLC, 8 had measurable CNS metastasis.10 This drug also has CNS activity. Half of the patients, 4 out of the 8 patients, had responses in the intracranial lesions, 2 patients had a CNS CR in this data set, and 75% of those responders had a DOR of greater than 6 months.

AEs seen in greater than 15% of patients with pralsetinib included grade 3 fatigue in 2.3% of patients and grade 3 hypertension in 14% of patients. An interesting AE is pneumonitis. Pneumonitis was reported in 10% at any grade, and 2.7% at grade 3/4, and 1.8% of patients discontinued permanently due to pneumonitis, pneumonia, or sepsis. Interruptions due to an AE were reported in 60% and dose reduction due to an AE was reported in 36% of the data set.

In terms of hematology parameters, there was a decrease in neutrophils reported, and 10% had grade 3/4 toxicity.

Based on these data, the FDA on September 4, 2020, approved pralsetinib for lung cancer with RET gene fusions.9 Interestingly, they also approved an Oncomine [Dx Target] Test as a companion diagnostic for pralsetinib.

How do you feel about the evolution of this landscape over the course of 2020?

We had not 1 but 2 drugs approved in 2020. [It was] a tough and challenging year for all of us, with the COVID-19 [coronavirus disease 2019] pandemic and so many different stressors in the world and practice. But for precision oncology, we have 2 selective RET inhibitors approved in NSCLC.

The phase 1 studies [of these drugs] were started in 2017; 3 years from the start of a phase 1 study to an FDA approval attests to the fact that both these agents are highly active in a high area of unmet need.

Would immunotherapy, despite a PD-L1 expression of 80%, work in a patient such as this?

I think that’s a great question. We published the data; I can talk about my own experience. [At The University of Texas] MD Anderson Cancer Center, we published the data showing lack of responsiveness of patients with NSCLC who harbored RET fusions to immunotherapy. [Memorial] Sloan Kettering Cancer Center also published similar data.

We don’t know why these patients are PD-L1 positive. I would think PD-L1 positivity here is a red herring, and most of these patients with RET fusions have a very low tumor mutation burden. And, in fact, when we sequence these patients, they are devoid of immune-infiltrating T cells.

[As with] other oncogenic driver lung cancers, such as ALK and ROS1, RET fusions also do not respond well to immunotherapy. If possible, if we have a patient with RET-altered NSCLC, I think we would prioritize a selective TKI as a preferred option over immunotherapy. The NCCN guidelines do not say anything about PD-L1; as long as it’s RET fusion–positive disease, the preferred first line is a TKI.

We are still learning [about] immune responsiveness and oncogene drivers because, regarding immunotherapy and oncogenic responses, most of the immunotherapy trials excluded patients with EGFR and ALK fusions. We don’t know how the patients responded with RET fusions, ROS1 fusions, even NTRK fusions on immunotherapy studies.

Only recently, with broad-based commercial planners, we’ve started testing for RET fusions. It remains to be seen if patients post RET inhibitor respond to an immunotherapy or a chemoimmunotherapy. If an RET TKI wasn’t available for a patient for financial reasons, or a patient does not tolerate it, or if you want to administer immunotherapy, I think we would go with a chemoimmunotherapy option rather than just immunotherapy alone.

But there are no absolutes in medicine. There might be a handful of patients with RET fusion lung cancer who might respond to immunotherapy, but we’ll wait [and see].

So today, if I had a patient who was RET fusion positive, I would offer one of the selective TKIs, if possible.


1. Subbiah V, Cote GJ. Advances in targeting RET-dependent cancers. Cancer Discov. 2020;10(4):498-505. doi:10.1158/2159-8290.CD-19-1116

2. NCCN. Clinical Practice Guidelines in Oncology. Non-small cell lung cancer, version 2.2021. Accessed March 4, 2021.

3. Drilon A, Oxnard GR, Tan DSW, et al. Efficacy of selpercatinib in RET fusion-positive nonsmall- cell lung cancer. N Engl J Med. 2020;383(9):813-824. doi:10.1056/NEJMoa2005653

4. Goto K, Oxnard GR, Tan DSW, et al. Selpercatinib (LOXO-292) in patients with RET-fusion+ non-small cell lung cancer. J Clin Oncol. 2020;38(suppl 15):3584. doi:10.1200/JCO.2020.38.15_suppl.3584

5. Subbiah V, Gainor JF, Oxnard GR, et al. Intracranial activity of selpercatinib (LOXO-292) in RET fusion-positive non-small cell lung cancer (NSCLC) patients on the LIBRETTO-001 trial. J Clin Oncol. 2020;38(suppl 15):9516. doi:10.1200/JCO.2020.38.15_suppl.9516

6. Retevmo. Prescribing information. Eli Lilly and Company; 2020. Accessed March 9, 2021.

7. FDA approves selpercatinib for lung and thyroid cancers with RET gene mutations or fusions. FDA. Updated May 11, 2020. Accessed March 5, 2021.

8. Gainor JF, Curigliano G, Kim DW, et al. Registrational dataset from the phase I/II ARROW trial of pralsetinib (BLU-667) in patients (pts) with advanced RET fusion+ non-small cell lung cancer (NSCLC). J Clin Oncol. 2020;38(suppl 15):9515. doi:10.1200/JCO.2020.38.15_suppl.9515

9. FDA approves pralsetinib for lung cancer with RET gene fusions. FDA. Updated September 8, 2020. Accessed March 5, 2021.

10. Gavreto. Prescribing information. Blueprint Medicines Corporation; 2020. Accessed March 5, 2021.