IMforte and DeLLphi-304: Small But Significant Improvements in the Treatment of ES-SCLC

Lung cancer is the second most common cancer in men and women in the United States. While overall rates of lung cancer diagnoses are declining due to decreases in cigarette smoking, it remains the leading cause of death from cancer. In 2020, it had more deaths than breast, prostate, and colon cancer combined.¹

Small-cell lung cancer (SCLC) comprises about 15% of all lung cancer diagnoses in the U.S.² While advancements in the treatment of lung cancer have extended overall survival (OS) in these patients, those benefits have mainly been limited to non-small cell lung cancer (NSCLC). The 2-year OS in women with NSCLC has risen from 35% in 2001-2002 to 52% in 2017-2018. In men, that increase went from 26% to 41%.¹ In SCLC, that increase in women went from 16% to 18%, and the increase in men went from 12% to 15%. Unlike NSCLC, SCLC has a much more aggressive biology, rapidly proliferates, and lacks targetable oncogenic drivers. Most SCLC is driven by loss-of-function mutations in the tumor suppressors (TP53 and RB1), which are not targetable by biomarker-guided strategies.

Extensive stage SCLC (ES-SCLC) is the most aggressive form of SCLC. It is SCLC which has spread extensively throughout the body and is considered incurable. The development of immunotherapy has made incremental improvements on OS in ES-SCLC, but it remains a disease characterized by rapid relapse, limited durability of response, and few post-progression treatment options. In 2019, the Food and Drug Administration (FDA) approved pembrolizumab and atezolizumab for ES-SCLC, and durvalumab was approved in 2020.³ While pembrolizumab’s approval was later withdrawn, atezolizumab and durvalumab remain approved for ES-SCLC as standards of therapy in the first-line setting. Both agents are typically given in combination with platinum-based chemotherapy for four cycles, followed by immunotherapy maintenance.

Long-term follow-up data are available for both agents. Atezolizumab's phase III extension study of the IMpower133 trial, IMbrella A, found that atezolizumab-based chemotherapy followed by atezolizumab maintenance had a 3-year OS of 16% and a 5-year OS of 12%.⁴ Similarly, in the 3-year follow-up for the CASPIAN trial, durvalumab was found to have a 3-year survival rate of 17.6% versus 5.8% in the placebo arm.⁵ Both agents have clearly extended survival in ES-SCLC, but over four in five patients will still pass away within three years as immunotherapy does not induce long-term responses in most ES-SCLC patients.

Outcomes after progression on first-line therapy are not favorable in ES-SCLC. A recent analysis of the Flatiron Health database was performed in the post-immunotherapy era, covering ES-SCLC patients from January 2020 to December 2023. Within this patient population, 82% of these patients received chemoimmunotherapy, and 56% of these patients received maintenance immunotherapy. Additionally, 40% of patients received second-line treatment, and 13% received third-line treatment, highlighting the high attrition rates after first-line therapy.⁶ Second-line therapy typically involves lurbinectedin or topotecan. Topotecan has demonstrated response rates of 24.3% and a median survival of 25 weeks in the second-line setting.⁷ Lurbinectedin, which received accelerated approval for ES-SCLC in 2020, is an alkylating agent that binds to guanine residues in the minor groove of DNA. It forms adducts and can inhibit transcription, induce double-strand breaks resulting in apoptosis, and modulate the tumor microenvironment.⁸ Lurbinectedin demonstrated response rates of 35.2% and a median OS of 9.3 months in a phase II basket trial.⁹

Prior to the checkpoint inhibitor era, maintenance chemotherapy was once proposed as a strategy to extend survival. This idea was studied in the E7593 trial during the pre-immunotherapy era of ES-SCLC. Four cycles of topotecan were compared against placebo in patients who responded to four cycles of cisplatin and etoposide. OS was not significantly improved, after resulting in a median OS of 8.9 months for observation versus 9.3 months for topotecan. Topotecan also increased the rates of most adverse events, particularly hematological adverse events.¹⁰ As such, the standard chemotherapy regimen remained platinum-based therapy with no maintenance at the time.

However, lurbinectedin’s tumor microenvironment modulation properties have been theorized to lead to synergistic effects and improved efficacy when combined with immunotherapy, thus prompting the examination of lurbinectedin alongside checkpoint immunotherapy in the maintenance setting.¹¹

IMforte

IMforte, which enrolled patients from November 2021 to January 2024, was a randomized, open-label, global phase III trial that compared lurbinectedin plus atezolizumab maintenance versus atezolizumab maintenance after first-line induction therapy with atezolizumab, carboplatin, and etoposide in patients with ES-SCLC. Patients were included if they were Eastern Cooperative Oncology Group (ECOG) 0 or 1 and had an ongoing response or stable disease after four cycles of atezolizumab, carboplatin, and etoposide. Lurbinectedin was given at a dose of 3.2 mg/m2 IV every 3 weeks, and atezolizumab was given at a dose of 1,200 mg IV every 3 weeks. The two primary endpoints were independent review facility (IRF) assessed progression free survival (PFS) and OS.

The trial was designed differently than the unsuccessful E7593 trial. First, primary granulocyte colony-stimulating factor (G-CSF) prophylaxis was required for patients receiving lurbinectedin, which would reduce neutropenic events and maintain dose intensity. Second, patients with central nervous system (CNS) metastases or consolidative thoracic radiation were excluded. And finally, lurbinectedin was given indefinitely, as opposed to a limited duration.

Of the 660 patients who were enrolled in IMforte during induction chemoimmunotherapy, only 73% (483 patients) were able to reach maintenance therapy. These 483 patients were randomized into either lurbinectedin plus atezolizumab maintenance (242 patients) or atezolizumab maintenance (241 patients). Baseline characteristics were generally well-balanced except for age, with 49% of patients on lurbinectedin plus atezolizumab below age 65 compared to 37% for patients on atezolizumab. Within the patient population, 79.5% (384/483) of the patients were from Europe or the Middle East, 12.6% (61/483) were from Asia, and 6.4% (31/483) were from North America. Additionally, 14.7% (71/483) of patients received prophylactic cranial irradiation (PCI), 33.3% (161/483) were current smokers, 194/483 (40.2%) had liver metastases, and 86.7% (419/483) had a complete or partial response to induction therapy.

With median follow-up of 15 months, the final IRF-assessed PFS showed a clinically and statistically significant improvement in the lurbinectedin plus atezolizumab arm versus the atezolizumab arm (median 5.4 versus 2.1 months; HR 0.54, 95% CI 0.43-0.67, p<0.0001). Interim OS was also significantly improved in the lurbinectedin plus atezolizumab arm versus the atezolizumab arm (median 13.2 versus 10.6 months, HR 0.73, 95% CI 0.57-0.95, p=0.017).

In the subgroup analysis, lurbinectedin plus atezolizumab showed a PFS benefit in all subgroups. For OS, lurbinectedin plus atezolizumab also showed a benefit in most subgroups, but not in patients with prior PCI, elevated lactate dehydrogenase, or patients from North America. Further investigation would be needed to delineate a true distinction between these subgroups.

Lurbinectedin maintenance increased rates of most adverse events. All-cause treatment-related adverse events were 83% for lurbinectedin plus atezolizumab maintenance versus 40% for atezolizumab. This included rates of infection (38% vs 26%) and neutropenia (11% vs 2%), despite primary G-CSF prophylaxis. Rates of grade 3-4 adverse events were also higher in the lurbinectedin arm (38% vs 22%), as were rates of corticosteroid use (17% vs 8%). This difference in grade 3-4 toxicities was mainly driven by anemia (8% vs 1%), neutropenia (7% vs 0%), and thrombocytopenia (5% vs <1%). There were 12 occurrences of fatal adverse events for lurbinectedin plus atezolizumab compared to 6 for atezolizumab. The study determined that two events were treatment-related in the lurbinectedin plus atezolizumab arm, and one event was treatment-related in the atezolizumab arm.¹²

The First Bispecific for ES-SCLC

Tarlatamab-dlle (Imdelltra™) is a first-in-class bispecific delta-like ligand 3 (DLL3)-directed CD3 T-cell engager indicated for ES-SCLC after disease progression on or after platinum-based chemotherapy.¹³ DLL3 is a protein typically internalized by normal cells but is expressed on the surface of SCLC cells.14 Based on prior phase I and II studies, tarlatamab-dlle was approved using a step-up dosing scheme of 1 mg IV on cycle 1 day 1, 10 mg IV on cycle 1 day 8, and 10 mg IV on cycle 1 day 15 of a 28-day cycle.^13,15,16 Dosing thereafter is 10 mg IV every 2 weeks until disease progression or toxicity. For the first two doses, it is recommended that patients are monitored for 22-24 hours in an appropriate healthcare setting. The dosing schema and monitoring parameters were approved to reduce the incidence and increase identification of cytokine release syndrome (CRS).

Patients on tarlatamab-dlle should be monitored for CRS (black box warning [BBW]), immune effector cell-associated neurotoxicity syndrome (ICANS) (BBW), cytopenias, infection, hepatotoxicity, hypersensitivity, dysgeusia, musculoskeletal pain, and nausea.¹³

Tarlatamab-dlle gained accelerated approval from the FDA in 2024 following the phase II DeLLphi-301 trial showing a 40% objective response rate (ORR), median PFS of 4.9 months and median OS of 14.3 months.¹⁷

DeLLphi-304

DeLLphi-304 was a randomized, multinational, phase III, open-label trial designed to compare tarlatamab-dlle with chemotherapy in the second-line setting for patients with ES-SCLC whose disease had progressed during or after platinum-based chemotherapy.  Included patients needed to have an ECOG score of 0 or 1, and patients with brain metastases were permitted as long as they were asymptomatic and clinically stable. Patients were randomized 1:1 to either tarlatamab-dlle or chemotherapy and then stratified based on prior PD-L1 inhibitor exposure, chemotherapy-free interval, presence of brain metastases, and intended chemotherapy. Options in the chemotherapy arm were topotecan, lurbinectedin, and amrubicin. The tarlatamab-dlle arm received dosing per the package insert with slightly altered monitoring. Patients were monitored in the hospital for 48 hours at the start of the trial, but the protocol was modified to allow 6-8 hours of post-infusion monitoring in the outpatient setting for some patients. Treatment in all arms continued until investigator-determined disease progression, unacceptable toxicity, withdrawal of consent, or death. Patients were allowed to continue trial treatment after progression if the investigator determined a clinical benefit. The primary end point was OS with secondary endpoints of PFS, patient reported outcomes, ORR, disease control, duration of response (DOR), and safety tested hierarchically.

Between May 31, 2023 and July 30, 2024, 509 patients were enrolled in DeLLphi-304. Overall, 254 patients were randomized to tarlatamab-dlle, and 355 patients were randomized to chemotherapy. In the chemotherapy arm, 73% received topotecan, 18% received lurbinectedin, and 9% received amrubicin. Baseline characteristics were well matched between the two groups and with 69% being male, 97% being Asian or White, 67% with ECOG 1, 71% with prior PD-L1 inhibitor exposure, and 45% with brain metastases. The breakdown for chemotherapy-free interval was as follows: 44% <90 days, 32% between 90 to <180 days, and 24% ≥180 days.

At data cutoff, the median duration of follow-up for OS was 11.2 months in the tarlatamab-dlle arm versus 11.7 months in the chemotherapy arm. The median OS was significantly longer in the tarlatamab-dlle arm (13.6 months versus 8.3 months; HR 0.60, 95% CI, 0.47-0.77, p <0.001). The proportional-hazards assumption was found to not be valid for median PFS, so an estimated restricted mean PFS time at 12 months was evaluated and found to be 5.3 months in the tarlatamab-dlle arm versus 4.3 months in the chemotherapy arm (HR 0.71, 95% CI 0.59-0.86, p = 0.002). Additional secondary endpoints tested in the tarlatamab-dlle versus chemotherapy arms included ORR (35% versus 20%), median time to response (1.5 months versus 1.4 months), and median DOR (6.9 months versus 5.5 months). Patient-reported symptom scores for dyspnea and cough were significantly better in the tarlatamab-dlle arm.

Treatment-related adverse effects (TRAEs) differed between the two arms. Patients on chemotherapy had higher rates of anemia (64% vs 31%), neutropenia (31% vs 11%), thrombocytopenia (25% vs 6%), and febrile neutropenia (11% vs 2%). While tarlatamab-dlle had lower rates of hematological toxicities, it had higher rates of CRS (56% vs 1%), pyrexia (27% vs 11%), and neurologic adverse events (56% vs 35%). Overall, 27% of tarlatamab-dlle patients and 62% of chemotherapy patients experienced grade 3 or higher TRAEs, driven by differences in anemia (4% vs 29%), neutropenia (6% vs 23%), and leukopenia (2% vs 14%).

Patients in the tarlatamab-dlle arm were monitored for CRS after completion of the infusion. CRS was experienced by 56% of patients with a majority being grade 1 (42%) or grade 2 (13%), with only three patients (1%) experiencing grade 3 CRS. Treatment for CRS included glucocorticoids (16% of patients), tocilizumab (4%), supplemental oxygen (8%), and vasopressor support (<1%). The most common neurologic adverse event was dysgeusia (24%). ICANS occurred in 6% of patients in the tarlatamab-dlle arm with one grade 5 event and the remainder being grade 1 or 2.¹⁷

Discussion

Despite the positive results of both the IMforte and DeLLphi-304, these trials also demonstrate the challenges in treating ES-SCLC.

While IMforte showed a significant improvement in OS, it excluded patients with brain metastases. It is imperative to note that 10-25% of SCLC patients are diagnosed with brain metastases at diagnosis, and an additional 40-50% will develop brain metastases in the course of their disease.³ These patients were excluded due to the failure of an earlier lurbinectedin trial which included patients with brain metastases.¹⁸ The addition of mandatory G-CSF prophylaxis helped maintain dose intensity and reduced the risk of febrile neutropenia, but likely contributed to the increased rate of adverse events. Dose reductions may be a reasonable alternative to G-CSF prophylaxis in the palliative setting, where there is a balance between dose intensity and quality-of-life.¹⁹ Quality-of-life data from IMforte has not yet been published.

Maintenance chemotherapy may also not have a long-term benefit. The trial demonstrated that lurbinectedin is a reasonably well tolerated maintenance chemotherapy option, and that it extended OS. However, the PFS curves cross around 18 months.¹² Lurbinectedin, like many other agents in ES-SCLC, did not appear to produce durable long-term results, although further follow-up is needed.

DeLLphi-304 provided the first phase III data on BiTE therapy in ES-SCLC. As opposed to IMforte, DeLLphi-304 included patients with brain metastases, which made the results more generalizable to the ES-SCLC population seen in practice. It also provided a framework for outpatient monitoring of CRS and ICANS. The results of the trial and the protocol modification to allow six-to-eight-hour monitoring may increase accessibility and decrease healthcare costs through the outpatient treatment setting.

One critique of the DeLLphi-304 protocol is that the chemotherapy arm did not have an option for platinum-based treatment if the patient had an extended chemotherapy-free treatment interval. Current guideline recommendations allow for re-treatment with platinum doublet therapy if the patient has had prolonged disease-free time after primary treatment. If the investigators included platinum-based treatment, it would have helped delineate which line of therapy to pursue at the time of recurrence or relapse.

Finally, it is important to note that majority of patients did not respond to tarlatamab-dlle. While the trial did show significantly longer OS and PFS versus chemotherapy, 55% of the patients in the tarlatamab-dlle arm had either progressive disease or stable disease while on trial. Of the 35% who responded, the median DOR was 6.9 months.¹⁷ These numbers are better than chemotherapy in the second-line setting but also suggest that more work needs to be done to find more long-term responses.

Further improvements could be made by combining concepts from these two trials. A recent phase Ib study of 88 patients added maintenance tarlatamab to maintenance durvalumab or atezolizumab. The median PFS was 5.6 months, and the median OS was 25.3 months. The PFS rate at 12 months was 34%, and OS rate at 12 months was 82%. These results suggest that tarlatamab could be added to maintenance immunotherapy and possibly increase the number of durable responses.²⁰ The phase III DeLLphi-305 trial is still ongoing but may provide confirmation of this combination.

Despite most patients not reaching the 2-year survival mark after diagnosis, the IMforte and DeLLphi-304 trials demonstrate promising advancements in the ES-SCLC space. While long term efficacy and safety data are still pending, these trials provide promising options for ES-SCLC patients.