Abstract
Small cell lung cancer (SCLC) is an aggressive cancer, with most cases diagnosed as extensive-stage (ES-SCLC). Platinum and etoposide chemotherapy is the mainstay of first-line treatment, achieving high initial response rates. However, treatment resistance develops quickly, leading to poor overall survival and limited efficacy of subsequent therapies, especially for platinum-resistant disease. The addition of immune checkpoint inhibitors (ICIs) to first-line chemotherapy for ES-SCLC has resulted in modest improvements in survival. For limited-stage SCLC (LS-SCLC) treated with radical chemo-radiotherapy, the ICI durvalumab is now approved as a consolidation therapy to reduce relapse risk. Further trials are investigating ICIs concurrently with chemo-radiotherapy and/or as consolidation or maintenance therapy. For relapsed SCLC, treatment options include chemotherapies such as topotecan or lurbinectedin and carboplatin/etoposide rechallenge. The delta-like ligand 3-targeting bispecific T-cell engager (BiTE), tarlatamab, has been approved by the FDA for ES-SCLC with disease progression on or after platinum-based chemotherapy and is being evaluated in earlier lines of treatment. Other BiTEs are also in early-phase development, with promising early activity. Several antibody-drug conjugates, including sacituzumab govitecan, are being tested in clinical trials and have demonstrated encouraging efficacy. Novel targeted therapies aimed at overcoming resistance to chemotherapy and immunotherapy are also in preclinical development. Despite these advancements, progress remains hindered by the absence of validated biomarkers for predicting treatment outcomes. The identification of SCLC transcriptional subtypes with distinct therapeutic vulnerabilities offers hope for better treatment stratification. The SCLC-I transcriptional subtype, tumour mutational burden and tumour immune cell signatures are promising biomarkers for longer-term survival benefit from ICIs. Circulating tumour DNA and circulating tumour cells have demonstrated potential for prognostication, molecular subtyping and tumour monitoring. Further research remains essential to support treatment stratification, prolong treatment responses, overcome resistance and ultimately improve outcomes for this devastating disease.
Keywords
Introduction
Small cell lung cancer (SCLC) is a high-grade neuroendocrine carcinoma and represents around 15% of all lung cancers. 1 It is characterised by a rapid proliferation and early development of metastases. Around one-third of patients are diagnosed with limited-stage SCLC (LS-SCLC), where disease has not advanced beyond one hemithorax or regional lymph nodes, 2 and this can be treated with curative intent. The remaining two-thirds present with extensive-stage SCLC (ES-SCLC), where disease has spread beyond the hemithorax or regional lymph nodes. Here, palliative systemic therapy ± consolidation radiotherapy is the only treatment option. SCLC has a poor prognosis, with a median overall survival (mOS) of <2.5 years for LS-SCLC and <1 year for ES-SCLC. 1
Although SCLC is initially highly sensitive to treatment, resistance develops quickly in most cases, leading to short durations of response, high relapse rates and poor outcomes with subsequent treatments. 3 SCLC has a high frequency of somatic mutations and correspondingly high tumour mutational burden (TMB), along with marked inter- and intra-tumoural heterogeneity that increases after chemotherapy.4,5 These features, along with phenotypic plasticity, contribute to its aggressive nature and development of treatment resistance. 6 The majority of SCLC tumours carry inactivating mutations of tumour suppressor genes, TP53 and RB1. 4 MYC-family oncogene amplification is observed in approximately 25% of cases. 5 Other less frequently altered genes are linked to cell cycle control and apoptosis (e.g. TP73, RBL1, RBL2), NOTCH signalling (e.g. NOTCH1-4), epigenetic modification (e.g. CREBBP, EP300, KMT2A-D), cytoskeletal and adhesion regulation (e.g. ALMS1, ASPM, COL22A1) and kinase signalling pathways (e.g. PIK3CA, PTEN). 4 SCLC tumours generally exhibit a neuroendocrine phenotype, with a smaller subset displaying non-neuroendocrine features. 7 While mutational profiling has so far not identified broad subtypes, transcriptional profiling has identified distinct subtypes based on differential expression of neurogenic transcription factors. The neuroendocrine subtypes include ASCL1 (SCLC-A), NEUROD1 (SCLC-N) and ATOH1, while POU2F3 (SCLC-P) represents the non-neuroendocrine phenotype.8–12 A recently described ‘inflamed’ subtype (SCLC-I) has been characterised by low expression of ASCL1, NEUROD1 and POU2F3, along with elevated levels of immune checkpoint and other immune-related genes. 13 Delta-like ligand 3 (DLL3) is an inhibitory NOTCH ligand abnormally expressed on the surface of neuroendocrine SCLC cells, particularly in the ASCL1-driven subtype, but is largely absent in normal tissues. 14 This tumour-specific expression makes DLL3 a promising biomarker and therapeutic target in SCLC.
Despite these growing insights into the heterogeneity of SCLC, there are currently no validated biomarkers to guide treatment selection or precision medicine treatments. This highlights a critical need for more effective treatments, improved biomarker-directed approaches and personalised therapies to enhance clinical outcomes and OS. This narrative review focuses on published phase II and III clinical trials in SCLC between 2018 and 2024, along with earlier pivotal studies and includes key phase I and ongoing trials to highlight emerging therapeutic approaches. It summarises current evidence on standard-of-care treatments, novel therapies in clinical trials, emerging therapeutic targets and prognostic and predictive biomarkers for therapy selection.
Current treatment approaches for SCLC
First-line treatment for LS-SCLC
Treatment of SCLC is stratified according to LS-SCLC and ES-SCLC. For very early-stage LS-SCLC (T1-2N0M0), surgical resection can be considered, followed by adjuvant platinum/etoposide chemotherapy. A retrospective study demonstrated that for resected pT1-2N0M0 tumours, adjuvant chemotherapy ± radiotherapy produced a better mOS compared to no adjuvant therapy (66.0 vs 42.0 months, p < 0.01). 15 In a retrospective study of 507 stage I/II patients, those undergoing lobectomy and adjuvant chemotherapy (including radiotherapy if node-positive) had a longer mOS compared to those receiving concurrent chemo-radiotherapy alone (48.6 vs 28.7 months, p < 0.0001). 16 However, the decision to proceed with surgery rather than chemo-radiotherapy must be weighed against the risk of progression during the wait, which may depend on the resources and capacity of treatment centres.
For the majority of LS-SCLC cases where surgery is not feasible or recommended, chemotherapy and thoracic radiotherapy are given with curative intent. 3 Concurrent chemo-radiotherapy typically consists of cisplatin/etoposide chemotherapy, and thoracic radiotherapy given concurrently with cycle 2 in twice-daily doses. According to the CONVERT trial, this regimen has a median progression-free survival (mPFS) of 15.4 months and mOS of 30.0 months. 17 For patients unable to receive concurrent chemo-radiotherapy, sequential chemo-radiotherapy is an alternative option, but is associated with poorer survival. A phase III trial in Japan reported a reduced mOS of 19.7 months in LS-SCLC patients receiving sequential chemo-radiotherapy, compared to 27.2 months with concurrent chemo-radiotherapy. 18 Carboplatin can be considered as an alternative to cisplatin if there are contraindications or concerns about toxicity. A retrospective study of LS-SCLC patients treated with concurrent or sequential chemo-radiotherapy showed that cisplatin was associated with a slightly better, but not statistically significant, improvement in OS and locoregional control compared to carboplatin. 19 The phase III CONVERT trial showed that mOS was not significantly different between LS-SCLC patients receiving once-daily versus twice-daily radiotherapy from cycle 2 for LS-SCLC (30 vs 25 months, p = 0.14). 17 However, as this trial was powered to show superiority rather than equivalence of once-daily radiotherapy, twice-daily radiotherapy remains the standard of care.
Despite radical chemo-radiotherapy, LS-SCLC relapse rates are high, with ~70% relapsing within 5 years. 20 This highlights the need for additional treatments to reduce recurrence risk and improve survival. Treatment with the programmed death-ligand 1 (PD-L1) immune checkpoint inhibitor (ICI), durvalumab, has been recently approved following chemo-radiotherapy. The phase III ADRIATIC trial demonstrated that patients with LS-SCLC who had no evidence of disease progression following concurrent chemo-radiotherapy, receiving consolidation durvalumab for up to 24 months, had a longer mPFS (16.6 vs 9.2 months, p = 0.02) and mOS (55.9 vs 33.4 months, p = 0.01) compared to those receiving placebo, 21 resulting in FDA approval of durvalumab in December 2024. Results are awaited for a third cohort receiving consolidation with both durvalumab and the cytotoxic T-lymphocyte-associated protein (CTLA-4) inhibitor, tremelimumab. Limitations of the ADRIATIC trial include the lack of biomarker assessment (e.g. PD-L1) to guide future treatment stratification. Prophylactic intracranial radiation (PCI) was also administered at investigator discretion (54% in both arms) although OS did not significantly differ between PCI and no-PCI subgroups. 21 Grade 3–4 toxicity was comparable in the durvalumab and placebo groups, although long-term toxicity data are awaited. In contrast, results from the NRG-LU005 trial showed that patients with LS-SCLC receiving concurrent chemo-radiotherapy with atezolizumab followed by consolidation atezolizumab for up to 1 year showed no significant difference in mPFS (11.5 vs 12.0 months, hazard ratio (HR) = 1.00, 95% confidence interval (CI), 0.80–1.25) or mOS (39.5 vs 33.1 months; HR = 1.11, 95% CI, 0.85–1.45) compared to chemo-radiotherapy alone. 22 However, unlike the ADRIATIC study, NRG-LU005 enrolled patients after only one cycle of chemotherapy and did not require confirmation of disease control following chemo-radiotherapy, suggesting that upfront concurrent ICI may not be the optimal approach for LS-SCLC. Ongoing trials are evaluating other ICIs for LS-SCLC, both upfront in combination with chemo-radiotherapy and as consolidation or maintenance therapy (Table 1). The phase III DeLLphi-306 trial is investigating the DLL3-targeting BiTE tarlatamab compared to placebo after concurrent chemo-radiotherapy (NCT06117774).
Clinical trials of ICIs in combination with chemo-radiotherapy for LS-SCLC.
cCRT, concurrent platinum and etoposide-based chemo-radiotherapy; ICI, immune checkpoint inhibitor; LS-SCLC, limited-stage small cell lung cancer; ORR, objective response rate; OS, overall survival; PCI, prophylactic cranial irradiation; PFS, progression-free survival; sCRT, sequential platinum and etoposide-based chemo-radiotherapy.
First-line treatment for ES-SCLC
Platinum/etoposide chemotherapy is the cornerstone of first-line treatment for ES-SCLC. A meta-analysis of 36 clinical trials comparing different chemotherapy regimens found a significantly better survival from regimens containing cisplatin and/or etoposide. 23 Carboplatin is generally preferred to cisplatin due to similar response rates but more favourable toxicity profile. 24 Platinum/etoposide chemotherapy is typically given for four to six cycles followed by monitoring, as studies have not shown significant improvement in OS with maintenance chemotherapy. 25 Irinotecan and gemcitabine have demonstrated similar efficacy and outcomes when combined with carboplatin instead of etoposide.26,27
Recent clinical trials have explored the integration of ICIs with platinum/etoposide chemotherapy (Table 2). The PD-L1 inhibitors, atezolizumab or durvalumab, are recommended in combination with platinum/etoposide for first-line treatment based on the IMPower133 and CASPIAN trials, respectively.28,29 In the phase III IMPower133 trial, ES-SCLC patients receiving first-line atezolizumab with carboplatin/etoposide, followed by maintenance atezolizumab, had a longer mOS compared to patients receiving placebo with carboplatin/etoposide (12.3 vs 10.3 months, p = 0.0154). 30 Similarly, in the CASPIAN phase III trial, ES-SCLC patients receiving durvalumab with platinum/etoposide, followed by maintenance durvalumab, had a longer mOS compared to platinum/etoposide alone (13.0 vs 10.3 months, p = 0.0047). 29 A limitation of IMpower133 is it did not include biomarker analyses such as PD-L1 or TMB. While CASPIAN included biomarker analysis, this was exploratory and constrained by incomplete data (~54% patient samples evaluable for PD-L1 and ~33% for tissue TMB), imbalance in sample availability across treatment arms and no blood-based TMB assessment.29,31 In addition, neither trial incorporated consolidation thoracic radiotherapy, a recognised component of ES-SCLC management with chemotherapy alone. Furthermore, the absolute survival benefits with ICI were modest. In IMPower133, 18-month OS was 34.0% in the atezolizumab/carboplatin/etoposide group compared to 21.0% in the carboplatin/etoposide group, indicating only 13.0% experienced prolonged OS benefit. 28 Similarly, in CASPIAN, 3-year OS was 17.6% in the durvalumab/carboplatin/etoposide group, compared to 5.8% in the carboplatin/etoposide group, showing 11.8% experienced a longer-term OS benefit. 32 These results raise concerns over long-term benefit and cost-effectiveness in unselected patients and highlight the need for predictive biomarkers to select patients most likely to benefit from ICIs.
Key clinical trials of immune checkpoint inhibitors combined with platinum/etoposide chemotherapy for first-line treatment of ES-SCLC.
For up to 2 years following 3–4 cycles of first-line platinum/etoposide without progression.
CI, confidence interval; ES-SCLC, extensive-stage small cell lung cancer; HR, hazard ratio; ORR, objective response rate; OS, overall survival; PFS, progression-free survival.
Programmed cell death protein 1 (PD-1) inhibitors have also been investigated in combination with chemotherapy. In the phase III ASTRUM-005 trial, ES-SCLC patients receiving serplulimab with carboplatin/etoposide followed by maintenance serplulimab had a longer mOS compared to placebo with carboplatin/etoposide (15.4 vs 10.9 months, p < 0.001). 33 As a result, serplulimab is the first PD-1 inhibitor approved in combination with chemotherapy for ES-SCLC.
Three phase III trials conducted in ES-SCLC patients in China have also demonstrated improved outcomes with PD-L1/PD-1 inhibitors.34–36 In the phase III CAPSTONE-1 trial, patients receiving adebrelimab (anti-PD-L1) with carboplatin/etoposide followed by maintenance adebrelimab had a longer mOS compared to those receiving placebo with carboplatin/etoposide (15.3 vs 12.8 months, p = 0.0017). 34 In the phase III RATIONALE-312 trial, patients receiving tislelizumab (anti-PD-1) with platinum/etoposide, followed by maintenance tislelizumab, had a longer mOS compared to placebo with platinum/etoposide (15.5 vs 13.5 months, p = 0.004). 35 Finally, in the phase III EXTENTORCH trial, ES-SCLC patients receiving toripalimab (anti-PD-1) with platinum/etoposide, followed by maintenance toripalimab for up to 2 years, had a longer mOS compared to placebo with platinum/etoposide (14.6 vs 13.3 months, p = 0.03). 36
Clinical trials of other ICIs have been conducted with mostly negative results. The PD-1 inhibitor, pembrolizumab, combined with platinum/etoposide numerically improved mOS compared to platinum/etoposide alone in the phase III KEYNOTE-604 trial, but this did not reach statistical significance (10.8 vs 9.7 months, p = 0.0164).37,38 Ipilimumab was investigated in a phase II trial with platinum/etoposide first-line but did not improve mOS compared to platinum/etoposide alone. 39 The TIGIT inhibitor, tiragolumab, was investigated in the SKYSCRAPER-02 phase II trial in combination with carboplatin/etoposide but did not produce an improvement in mPFS or mOS. 40 Additionally, ICIs have been investigated as a maintenance-only therapy. The CheckMate-451 phase III trial showed that in ES-SCLC patients receiving up to four cycles of platinum/etoposide chemotherapy, maintenance ipilimumab (anti-CTLA-4) plus nivolumab (anti-PD-1) or nivolumab alone, for up to 2 years did not improve mOS compared to placebo. 41 Combination ICI with chemotherapy has so far not demonstrated improved outcomes compared to single-agent ICI; the addition of the CTLA-4 inhibitor, tremelimumab, to durvalumab/platinum/etoposide in the CASPIAN trial did not improve mOS. 29
The phase III IMforte trial investigated lurbinectedin added to maintenance atezolizumab in patients with ES-SCLC who had no progression after four cycles of carboplatin/etoposide. 42 Patients were randomised to receive maintenance atezolizumab with or without lurbinectedin until progression. mPFS was longer with lurbinectedin plus atezolizumab maintenance compared to atezolizumab alone (5.4 vs 2.1 months, p < 0.0001), as was mOS (13.2 vs 10.6 months, p = 0.017), both measured from the start of maintenance. Grade 3–4 adverse events were more frequent in the combination arm (38% vs 22%), mainly due to haematological toxicity. Patients who had received consolidative thoracic radiotherapy were excluded. These findings support lurbinectedin plus atezolizumab as a promising maintenance strategy in ES-SCLC, although tolerability in more comorbid populations remains to be assessed, as does the role of consolidative thoracic radiotherapy.
The modest survival benefits with the addition of anti-PD-L1/PD-1 ICIs to chemotherapy, and the lack of benefit from other classes of ICI, highlights the need for better strategies to overcome immune evasion and resistance to ICIs. Biomarker-guided stratification is also essential to identify patients most likely to benefit from immunotherapy. Several mechanisms of immune evasion in SCLC may be therapeutically targetable. 43 Although SCLC has a high TMB, its immunogenicity is low due to defects in antigen presentation, 44 driven in part by defective major histocompatibility complex (MHC) class I expression that limits CD8+ T cell recognition, 45 and reduced MHC class II expression, which impairs CD4+ T cell activation. 46 SCLC tumours are also typically ‘immune cold’, with low levels of tumour-infiltrating lymphocytes (TILs), particularly CD8+ T cells. 47 Vascular endothelial growth factor (VEGF) secreted by tumour and stromal cells also disrupts immune cell adhesion and migration, supporting rationale for combining VEGF inhibitors with ICIs. 48 SCLC cells also produce low levels of chemokines required for TIL recruitment. 49 Innate immune responses mediated by natural killer (NK) cells are impaired due to reduced expression and epigenetic silencing of NK-activating ligands such as NKG2DL. 50 In addition, the tumour microenvironment is enriched with immunosuppressive cells, including regulatory T cells, tumour-associated macrophages, myeloid-derived suppressor cells and cancer-associated fibroblasts, all of which contribute to suppression of anti-tumour immune responses. 51 Therapies that restore antigen presentation, promote immune cell infiltration and/or reverse immunosuppression could enhance the effectiveness of existing ICIs and guide development of novel immunotherapies in SCLC.
Role of consolidation thoracic radiotherapy in ES-SCLC
Consolidation thoracic radiotherapy is usually given following platinum-based chemotherapy in ES-SCLC patients who achieve a response. In a prospective study of 210 patients treated first-line with three cycles of cisplatin/etoposide, patients with at least a partial response at local sites and complete response at distal sites were randomised to thoracic radiotherapy with carboplatin/etoposide or four further cycles of cisplatin/etoposide. 52 Patients with a complete response at distant sites also received PCI. Patients receiving thoracic radiotherapy had a longer mOS than those who did not (17 vs 11 months, p = 0.041). In the phase III CREST trial of ES-SCLC patients who responded to first-line platinum/etoposide and underwent PCI, 6-month PFS was 24% with thoracic radiotherapy versus 7% without (p = 0.001). Although 1-year OS was not significantly different, 2-year OS was greater with thoracic radiotherapy (13% vs 3%, p = 0.004). 53 Patients with residual intrathoracic disease after chemotherapy experienced a better mPFS and mOS with thoracic radiotherapy but there was no significant difference for patients without residual disease. 54 It remains uncertain whether patients receiving first-line PD-(L)1 inhibitor alongside chemotherapy, followed by maintenance, benefit from consolidation thoracic radiotherapy. A meta-analysis of 15 retrospective studies showed that ES-SCLC patients receiving first-line chemo-immunotherapy followed by consolidation thoracic radiotherapy had improved survival rates compared to those who did not receive thoracic radiotherapy. 55 This is currently being investigated in prospective clinical studies. 56
Role of PCI in the first-line treatment of SCLC
Due to the high risk of relapse with brain metastases, PCI is routinely offered to LS-SCLC patients, and ES-SCLC patients <70 years, who achieve a response to first-line treatment.3,57 However, the reduction in risk of brain metastases should be balanced against the risk of chronic neurological toxicity. 58 A meta-analysis of seven clinical trials involving patients with LS-SCLC or ES-SCLC who achieved complete response after first-line treatment showed that PCI reduced the 3-year cumulative incidence of brain metastasis by 25.3% and increased 3-year survival rates by 5.4%. 59 A recent meta-analysis of 223 studies of patients with LS-SCLC or ES-SCLC demonstrated that PCI was associated with a longer OS compared to no PCI (HR = 0.59, p < 0.001). 60 The incidence of brain metastases was lower in patients who received PCI based on 63 of the studies (relative risk, 0.45, p < 0.001). However, in a post hoc analysis of nine studies where patients underwent MRI to exclude brain metastases at the end of first-line therapy, mOS was not significantly different in patients who received PCI versus no PCI (HR = 0.74, p = 0.08), suggesting that PCI might not be superior to surveillance for improving OS. 60
In a phase III trial of patients with ES-SCLC who responded to chemotherapy, patients who received PCI had a lower risk of developing symptomatic brain metastases (HR = 0.27, p < 0.001) and a lower 1-year cumulative risk of brain metastases (14.6% vs 40.4%). PCI was also associated with a longer mOS (6.7 vs 5.4 months, p = 0.003). 53 However, this study did not include baseline brain imaging prior to enrolment to confirm the absence of brain metastases. In a phase III trial, patients with ES-SCLC who responded to platinum-based chemotherapy and had no metastases on brain MRI prior to enrolment were randomised to PCI or observation with MRI scans for up to 24 months. 61 mOS was not significantly different in the PCI and observation groups (11.6 vs 13.7 months, p = 0.094).
Further prospective studies are needed to determine whether PCI is superior to surveillance for improving OS and to evaluate the impact of neurotoxicity, particularly with the addition of PD-(L)-1 inhibitors to standard first-line treatments. PRIMALung (EORTC-1901) is a prospective randomised trial comparing 3-monthly MRI surveillance with PCI to assess for non-inferiority in terms of OS for patients with LS-SCLC and ES-SCLC who have responded to first-line treatment. 62
Treatment of transformed SCLC
Histological transformation of non-SCLC (NSCLC) to SCLC is an increasingly recognised mechanism of acquired resistance to targeted therapies, most frequently reported in EGFR-mutant adenocarcinoma. Transformation has been observed in 3%–14% of patients with EGFR-mutant NSCLC treated with EGFR tyrosine kinase inhibitors (TKIs).63,64 Although the precise molecular mechanisms of transformation remain unclear, lineage plasticity – characterised by a phenotypic switch from an epithelial to a neuroendocrine state – is considered to be the main driver. 65 Tumour heterogeneity may also contribute, as some tumours display mixed NSCLC and SCLC histology, suggesting the expansion of a pre-existing SCLC subclone under selective pressure of TKI therapy. 66 Most transformed tumours retain their original EGFR-activating mutations, and inactivating mutations of TP53 and RB1, typical of de novo SCLC, are frequently present.67,68 While transformation is most commonly observed in EGFR-mutant cases, it has also been reported in NSCLC with ALK or ROS1 fusions receiving TKI therapy, and following ICI therapy. 69
There are currently no formal clinical guidelines tailored to transformed SCLC and treatment decisions are typically extrapolated from the management of de novo SCLC. In a retrospective multicentre study, platinum/etoposide chemotherapy achieved an objective response rate (ORR) of 54% and mPFS of 3.4 months in 53 patients. 70 Taxane-based therapy produced an ORR of 50% after a median of two prior treatment lines post-transformation. In 17 patients who received single-agent PD-1 inhibitors or combination ipilimumab/nivolumab therapy, there was no response to treatment, indicating the lack of effectiveness of ICI in this group of patients. Based on these findings, both platinum/etoposide and taxanes are considered potentially effective treatments for patients with transformed SCLC.
Given the poorer prognosis of transformed SCLC compared to de novo SCLC, and its resistance to ICI, tailored therapeutic strategies are being explored. A phase II trial (NCT04538378) is evaluating durvalumab plus olaparib in patients with EGFR-mutant NSCLC that has undergone SCLC transformation. A phase I study (NCT03567642) is investigating whether combining osimertinib with platinum/etoposide can prevent transformation of EGFR/TP53/RB1-mutant NSCLC. 71 Further studies in larger cohorts are needed to clarify the mechanisms of transformation and treatment resistance, identify high-risk patients, detect transformation earlier and guide development of dedicated trials for this subgroup.
Second-line and subsequent therapies for SCLC
SCLC that progresses during first-line platinum-based chemotherapy or <90 days post-treatment is considered platinum-resistant, whereas SCLC that relapses ⩾90 days after the end of first-line platinum-based chemotherapy is considered platinum-sensitive. 72 A meta-analysis showed that patients with platinum-sensitive SCLC had a higher ORR to second-line treatments compared to platinum-resistant SCLC (27.7% vs 14.8%, p = 0.0001). 73 Therefore, for patients with platinum-resistant SCLC, enrolment in clinical trials or best supportive care (BSC) has often been recommended. 1
Second- and further-line chemotherapies that have demonstrated modest improvements in OS include topotecan, lurbinectedin, cyclophosphamide/doxorubicin/vincristine (CAV) and liposomal irinotecan. 3 Table 3 summarises the results from key clinical trials evaluating chemotherapy agents in relapsed SCLC. In a randomised trial of SCLC patients who relapsed >60 days after completing first-line chemotherapy, intravenous (IV) topotecan produced a mOS similar to CAV (25.0 vs 25.7 weeks, p = 0.795). 74 However, a greater proportion of patients receiving IV topotecan experienced symptom improvement compared to CAV, particularly in the following areas: dyspnoea (27.9% vs 6.6%, p = 0.002), anorexia (32.1% vs 15.8%, p = 0.042), hoarseness (32.5% vs 13.2%, p = 0.043), fatigue (22.9% vs 9.2%, p = 0.032) and interference with daily activities (26.9% vs 11.1%, p = 0.023). 74 Oral topotecan was investigated in a phase III trial of patients with relapsed SCLC and demonstrated a longer mOS combined with BSC compared to BSC alone (25.9 vs 13.9 weeks). 75 Patients receiving topotecan also had slower deterioration in quality of life and greater symptom control. Oral and IV topotecan have demonstrated similar efficacy and mOS as second-line treatments in a phase III trial of platinum-sensitive SCLC patients. 76 Liposomal irinotecan was investigated in the phase III RESILIENT trial in patients with either platinum-resistant or platinum-sensitive SCLC who had progressed after first-line platinum-based chemotherapy. Patients receiving liposomal irinotecan or IV topotecan had a similar mOS (7.9 vs 8.3 months, p = 0.3094), although ORR was higher for liposomal irinotecan (44.1% vs 21.6%, p < 0.0001) and grade ⩾3 toxicity lower (42.0% vs 83.4%). 77
Key clinical trials of chemotherapy and immune checkpoint inhibitor treatments for relapsed SCLC.
BSC, best supportive care; CAV, cyclophosphamide/doxorubicin/vincristine; HR, hazard ratio; IV, intravenous; ORR, objective response rate; OS, overall survival; PFS, progression-free survival; PO, oral; SCLC, small cell lung cancer; TTP, time to treatment progression.
Rechallenge with carboplatin/etoposide is a favourable option for patients with platinum-sensitive SCLC. In a multicentre retrospective study involving 112 patients, ORR was 45% to second-line platinum/etoposide in patients with a treatment-free interval of >90 days following first-line platinum-based chemotherapy, and mPFS and mOS were 5.5 and 7.9 months, respectively. 78 Subgroup analysis comparing three relapse-free intervals (90–119, 120–149 and >150 days) indicated no significant difference in mPFS. 78 A recent retrospective multicentre study of 93 patients with platinum-sensitive disease who received rechallenge carboplatin/etoposide following first-line chemo-immunotherapy showed similar results; ORR was 59.1%, with mPFS of 5 months and mOS of 7 months. 79 In a phase III trial in France, second-line rechallenge of platinum-sensitive SCLC with six cycles of carboplatin/etoposide produced a longer mPFS compared to six cycles of oral topotecan (4.7 vs 2.7 months, p = 0.0041). 80 Triple combination chemotherapy with cisplatin/etoposide/irinotecan was compared with IV topotecan in the phase III JCOG0605 trial of patients with platinum-sensitive relapsed SCLC. 81 mOS was longer with combination chemotherapy (18.2 vs 12.5 months, p = 0.0079) but >85% of patients had grade 3–4 toxicities, meaning this regimen is not favoured in clinical practice.
Lurbinectedin, a transcription inhibitor, is a newer second-line treatment option for relapsed SCLC. In a phase II trial, it had an ORR of 35.2% in patients with platinum-resistant or platinum-sensitive SCLC and no brain metastases following previous chemotherapy treatment. 82 When indirectly compared with topotecan using data from the ATLANTIS trial in a matched patient population with a chemotherapy-free interval ⩾30 days and no brain metastases, lurbinectedin demonstrated a higher ORR (41.0% vs 25.5%, p = 0.0382), longer mOS (10.2 vs 7.6 months) and lower toxicity rate. 83 In ATLANTIS, patients with relapsed SCLC following platinum-based chemotherapy had a similar mOS when receiving lurbinectedin/doxorubicin or physician’s choice of IV topotecan or CAV (8.6 vs 7.6 months, p = 0.70). 84 However, the lurbinectedin/doxorubicin group experienced less toxicity. As lurbinectedin alone was not directly compared with lurbinectedin/doxorubicin, the relative benefit of adding doxorubicin remains unclear. In a phase II trial, lurbinectedin/irinotecan was evaluated in patients with ES-SCLC with disease progression on or after platinum-based chemotherapy. 85 Among 74 patients with a chemotherapy-free interval >30 days, the ORR was 51.4%, with a mPFS of 5.6 months and mOS of 12.7 months, highlighting it as a promising treatment option in this setting. The phase III LAGOON trial (NCT05153239) is comparing lurbinectedin/irinotecan with lurbinectedin or topotecan alone. 86
Paclitaxel, irinotecan and temozolomide have demonstrated modest efficacy as single-agent treatments in second-line and beyond for SCLC in studies involving small patient cohorts. In a phase II study involving 24 patients with platinum-resistant ES-SCLC, paclitaxel produced a 29% ORR and mOS of 100 days. 87 In a retrospective study of 185 ES-SCLC patients where 93% received paclitaxel as third of fourth lines therapy, ORR was 17% with a mOS of 3.3 months. 88 In a phase II trial of temozolomide for patients with relapsed SCLC after one or two treatment lines, ORR was 23% in 48 platinum-sensitive patients and 22% in 16 platinum-resistant patients. 89 In a phase II study, irinotecan was administered to 30 patients who had previously received ⩾1 platinum-based chemotherapy. The ORR was 41.3%, mPFS 4.1 months and mOS 10.4 months. 90
PD-1 and PD-L1 ICIs have been investigated in second- and third-line settings for SCLC, with predominantly negative results (Table 3). CheckMate-331 was a phase III trial showing that nivolumab produced a similar mOS to topotecan or amrubicin chemotherapy in both platinum-sensitive and platinum-resistant relapsed SCLC (7.5 vs 8.4 months, p = 0.11). 91 The phase I/II CheckMate-032 trial of third-line nivolumab following previous platinum-based therapy demonstrated an ORR of 11.9% and 12-month OS of 28.3%. 92 Similarly, a phase I study of pembrolizumab after ⩾2 lines of therapy demonstrated an ORR of 19.3%. 93 However, neither of these studies met mOS endpoints and are not recommended in the second or third-line settings. Atezolizumab showed a lack of efficacy second-line following previous platinum/etoposide chemotherapy in the phase II IFCT-1603 trial, with an ORR of 2.3%, although mOS was 9.5 months compared to 8.7 months for topotecan or re-challenge chemotherapy (p = 0.60). 94 The benefit of rechallenging patients with PD-(L)1 inhibitors after first-line chemo-immunotherapy remains unclear, as second-line trials of PD-(L)1 inhibitors were conducted before this became standard first-line treatment.
Table 4 provides a summary of current FDA-approved systemic therapies for SCLC and associated survival outcomes reported in clinical trials.
FDA-approved systemic therapies for the treatment of SCLC.
cCRT, concurrent platinum and etoposide-based chemo-radiotherapy; ES-SCLC, extensive-stage small cell lung cancer; IV, intravenous; LS-SCLC, limited-stage small cell lung cancer; mOS, median overall survival; mPFS, median progression-free survival; ORR, objective response rate; PO, oral.
Novel targeted therapies for SCLC
Despite initial sensitivity to platinum-based therapy, SCLC rapidly acquires resistance, resulting in short-lived responses and poor long-term outcomes. This has driven the development of novel therapies aimed at overcoming resistance and improving the durability of response. While many agents are currently under investigation in relapsed or refractory SCLC, increasing efforts are focused on incorporating them into first-line treatment. Table 5 summarises key clinical trials of novel systemic therapies and treatment combinations for SCLC.
Key clinical trials of novel systemic therapies and treatment combinations for SCLC.
Following four cycles of platinum/etoposide without progression.
Following 2–4 cycles of carboplatin/etoposide + atezolizumab without progression.
Following ⩽3 cycles of platinum/etoposide (ES-SCLC or LS-SCLC) ± concurrent or sequential chemo-radiotherapy (LS-SCLC), with no progression.
Followed by matching maintenance therapy with benmelstobart ± anlotinib.
Following 4 cycles of carboplatin/etoposide/atezolizumab without progression.
BD, twice a day; CAV, cyclophosphamide/doxorubicin/vincristine; DCR, disease control rate; ES-SCLC, extensive-stage small cell lung cancer; HR, hazard ratio; I-DXd, Ifinatamab Deruxtecan; IV, intravenous; LS-SCLC, limited-stage small cell lung cancer; ORR, objective response rate; OS, overall survival; PARP, Poly (ADP-ribose) polymerase; PFS, progression-free survival, TDS; three times a day.
Bispecific T-cell engagers
BiTEs are antibodies that bind both an antigen on tumour cells and CD3 on T cells, engaging T cells to target cancer cells and induce T cell-mediated lysis. Tarlatamab is a BiTE which works by targeting DLL3 on SCLC cells and CD3 on T cells. 95 In the phase II DeLLphi-301 trial, tarlatamab was administered to 220 patients with both platinum-sensitive and platinum-resistant SCLC who had relapsed after one platinum-based treatment and ⩾1 other line of therapy. 95 The ORR was 40% in patients receiving 10 mg and 32% in those receiving 100 mg and duration of response was ⩾6 months in 59% of cases. mPFS was 4.9 months in the 10 mg group and 3.9 months in the 100 mg group, and mOS was 14.3 months in the 10 mg group and not reached in the 100 mg group. Cytokine release syndrome was the most common toxicity, occurring in 51% of the 10 mg group and 61% of the 100 mg group, primarily during the first cycle of treatment. However, only 3% of patients discontinued due to drug toxicity. Given the good ORR and favourable toxicity profile, tarlatamab received FDA approval in May 2024 for the treatment of ES-SCLC following progression on or after platinum-based chemotherapy. Tarlatamab is being further evaluated in the phase III DeLLphi-304 trial as a second-line treatment for SCLC in comparison with standard of care treatment (lurbinectedin, topotecan or amrubicin; NCT05740566). It is also under investigation in the first-line maintenance setting in the DeLLphi-305 trial, in combination with durvalumab versus durvalumab alone after induction with carboplatin/etoposide and durvalumab (NCT06211036).
Beyond tarlatamab, further BiTEs and T-cell therapies targeting DLL3 are in development. 14 The BiTE, BI 764532, which is also a DLL3/CD3 T cell engager, is being investigated in the phase II DAREON-5 trial in neuroendocrine tumours as a second-line and beyond treatment, with promising efficacy observed in a cohort of 36 SCLC patients.96,97 Further phase I studies are exploring BI 764532 in the first-line setting in combination with platinum/etoposide chemotherapy with or without anti-PD-L1 therapy (NCT06132113, NCT06077500), as well as in relapsed/refractory SCLC in combination with single-agent chemotherapy (NCT05990738). HPN328 is a novel DLL3/CD3 T-cell engager featuring an additional albumin-binding domain designed to prolong its half-life. In a phase I/II trial involving patients with relapsed or refractory ES-SCLC, HPN328 demonstrated an ORR of 50% in an initial cohort of 25 SCLC patients. 98
The high response rates reported with BiTE therapies to date support their earlier integration into the treatment pathway – not only for relapsed/refractory SCLC following platinum-based chemotherapy, but potentially in the first-line setting. Further clinical data are needed to fully assess the durability of response, impact on OS and safety, particularly in combination with chemotherapy and/or ICIs.
Antibody-drug conjugates
Antibody-drug conjugates (ADCs) are currently being investigated in multiple clinical trials and work by delivering potent cytotoxic agents via antibodies that target cell-surface proteins overexpressed in SCLC. Rovalpituzumab tesirine (Rova-T) was one of the first ADCs investigated in SCLC, targeting DLL3 and linked to the DNA-damaging agent pyrrolobenzodiazepine. The phase II TRINITY study assessed Rova-T as a third-line or beyond treatment in 339 patients with confirmed DLL3 expression in SCLC tumours. 99 Rova-T produced an ORR of 12.4% and mOS of 5.6 months. Grade 3–4 toxicities affected 54% of patients and grade 5 toxicities occurred in 10%, with 3% being drug-related. In the phase III TAHOE study, Rova-T was compared to IV topotecan as a second-line therapy in DLL3-high ES-SCLC. 100 Patients receiving Rova-T had a worse mOS compared to topotecan (6.3 vs 8.6 months). The limited ORR, lack of OS benefit compared to standard treatment and high toxicity of Rova-T mean it was not approved as a treatment for SCLC.
Clinical trials are currently underway of newer ADCs targeting Trop-2 and other markers. 101 Sacituzumab govitecan is an ADC that binds to Trop-2 and delivers the topoisomerase I inhibitor, SN-38, to tumour cells. Trop-2 is a transmembrane protein overexpressed in various cancers, including SCLC. 102 In the phase II TROPiCS-03 basket trial, sacituzumab govitecan was evaluated as a second-line treatment in patients with ES-SCLC who had progressed after one prior line of platinum-based chemotherapy and PD-(L)-1 therapy. 103 A total of 43 patients received sacituzumab govitecan, resulting in an ORR of 41.9%, mPFS of 4.4 months and mOS of 13.6 months. This demonstrates promising efficacy for sacituzumab govitecan in the second-line setting and would need further investigation in a phase III trial.
Ifinatmab deruxtecan (I-DXd) is an ADC that targets B7-H3, a transmembrane protein overexpressed in several tumour types including SCLC, and is attached to the topoisomerase I inhibitor, deruxtecan. I-DXd is under investigation for the treatment of both relapsed/refractory and treatment naïve ES-SCLC. In the phase II IDeate-Lung01 trial (NCT05280470), which enrolled 88 patients with ES-SCLC who had received 1–3 prior lines of therapy including platinum-based chemotherapy, the 12 mg/kg cohort of I-DXd demonstrated an ORR of 52.4%, mPFS of 5.5 months and mOS of 11.8 months. 104 A phase III study, IDeate-Lung02 (NCT06203210), is now comparing I-DXd with standard second-line chemotherapy (topotecan, lurbinectedin or amrubicin). The phase Ib/II IDeate-Lung03 trial (NCT06362252) is evaluating I-DXd in combination with atezolizumab as part of first-line therapy, both in the induction and maintenance settings, either following or concurrently with induction carboplatin/etoposide/atezolizumab.
ABBV-796 is a novel ADC targeting seizure-related homolog 6 (SEZ6), a transmembrane protein overexpressed in SCLC. It is conjugated to a topoisomerase I inhibitor and is currently being evaluated in a phase I clinical trial (NCT05599984) as monotherapy and in combination with budigalimab (anti-PD-1), carboplatin or cisplatin. Preliminary data from the SCLC cohort (n = 15) reported an ORR of 40%, suggesting promising early clinical activity in the relapsed/refractory setting. 105
Overall, sacituzumab govitecan and other ADCs offer a promising treatment option for SCLC, with encouraging ORRs. Their broader use will depend on manageable toxicity and further validation in ongoing trials.
Angiogenesis inhibitors
Angiogenesis inhibitors have been investigated as part of combination treatments for SCLC. Bevacizumab, an anti-VEGF-A monoclonal antibody, was evaluated in several phase II studies in combination with chemotherapy for ES-SCLC but did not demonstrate improvement in PFS or OS. 106 A phase III RCT in Italy of cisplatin/etoposide ± bevacizumab for first-line treatment of ES-SCLC did not show a significant improvement in OS. 107 Other large molecule anti-angiogenic agents, such as ziv-aflibercept and endostar, have been investigated, as well as small molecule TKIs with anti-VEGF receptor (VEGFR) activity, such as sunitinib, vandetanib, sorafenib, cediranib and nintedanib. Their use in combination with first-line chemotherapy or as single agents in second-line and beyond failed to demonstrate a significant improvement in OS. 106 In the ALTER1202 phase II trial in China, anlotinib, a multi-kinase inhibitor with anti-VEGFR activity, demonstrated an improvement in mOS compared to placebo in patients with relapsed platinum-resistant or platinum-sensitive SCLC after ⩾2 lines of chemotherapy (7.3 vs 4.9 months, p = 0.0029). 108 The ORR was only 4.9%, although disease control rate was 66.7%. Anlotinib is now approved as a third-line or beyond treatment for SCLC in China.
Angiogenesis inhibitors have also been investigated in combination with chemo-immunotherapy. The phase III ETER701 trial in China investigated anlotinib and benmelstobart (anti-PD-L1) with carboplatin/etoposide for first-line treatment of ES-SCLC versus anlotinib and placebo with carboplatin/etoposide versus double placebo with carboplatin/etoposide. 109 The addition of benmelstobart and anlotinib to carboplatin/etoposide was associated with a longer mOS (19.3 vs 11.9 months, p = 0.0002). However, there was no significant difference in mOS with the addition of anlotinib only to carboplatin/etoposide (13.3 vs 11.9 months, p = 0.1723). This suggests that the addition of an ICI may enhance the benefit from angiogenesis inhibitor therapy. A key limitation was the absence of a treatment arm with a PD-L1 inhibitor alone combined with carboplatin/etoposide, which would be representative of the current standard of care. Nonetheless, this trial suggests a potential OS benefit with the addition of an angiogenesis inhibitor to first-line chemo-immunotherapy, and benmelstobart and anlotinib combined with platinum/etoposide has been approved in China for first-line treatment of ES-SCLC. Additional evidence supporting the use of angiogenesis inhibitors in combination with chemo-immunotherapy first-line is provided by a phase Ib/II trial, which evaluated surufatinib, a multi-kinase inhibitor with anti-VEGFR activity, and PD-1 inhibitor, toripalimab, in combination with cisplatin/etoposide for first-line treatment of ES-SCLC. 110 This trial, involving 35 patients, reported a mPFS of 6.9 months and mOS of 21.1 months. The combination of angiogenesis inhibitors with ICI is also being investigated in the second-line setting. In a phase II trial, patients with SCLC who received ⩾1 prior line of platinum-based chemotherapy were treated with the PD-1 inhibitor, sintilimab, combined with anlotinib. 111 Among 37 patients, ORR was 56.8%, mPFS 6.1 months and mOS 12.7 months. In the PASSION phase II trial, ES-SCLC patients who received one previous line of chemotherapy were treated with the PD-1 inhibitor, camrelizumab, and VEGFR2 inhibitor, apatinib, demonstrating a mPFS of 3.6 months and mOS 8.4 months. 112 In platinum-sensitive versus platinum-resistant patients, the ORR (37.5% vs 32.3%), mPFS (3.6 vs 2.7 months) and mOS (9.6 vs 8.0 months) were similar. These studies suggest that combining angiogenesis inhibitors with ICIs may be beneficial in both first-line and beyond.
PARP inhibitors
Poly (ADP-ribose) polymerase inhibitors (PARPi) have been investigated as part of combination therapies for SCLC in first-line and maintenance settings. In the phase II ECOG-ACRIN 2511 trial, veliparib with cisplatin/etoposide demonstrated a significant improvement in mPFS for first-line treatment of ES-SCLC compared to placebo with cisplatin/etoposide (6.1 vs 5.5 months, p = 0.01), but did not significantly improve mOS (10.3 vs 8.9 months, p = 0.17) or ORR (71.9% vs 65.6%, p = 0.57). 113 In another phase II trial, veliparib with platinum/etoposide first-line followed by veliparib maintenance did not improve mPFS or mOS compared to placebo with platinum/etoposide. 114 Further studies have investigated PARPi as a maintenance only treatment in the first-line setting. The phase II SWOG S1929 trial compared maintenance talazoparib plus atezolizumab with maintenance atezolizumab after first-line treatment in ES-SCLC patients with SLFN11-positive tumours. 115 The combination led to a modest but statistically significant improvement in mPFS (2.9 vs 2.4 months, HR = 0.66, p = 0.019), although preliminary results indicate no significant difference in mOS. This trial paves the way for more biomarker-directed approaches to trial design based on SCLC subtypes. In the STOMP phase II trial, patients with LS-SCLC or ES-SCLC who responded to first-line treatment were randomised to receive olaparib or placebo maintenance with no significant difference in mPFS or mOS. 116 Similarly, in the ZL-2306-005 phase II trial in China for ES-SCLC patients who responded to first-line platinum-based chemotherapy, niraparib maintenance was associated with a marginally longer mPFS (1.54 vs 1.36 months, p = 0.024) but mOS was not significantly different (9.92 vs 11.43 months, p = 0.905). 117
PARPi are also being investigated in the second-line and beyond. In a phase II trial, veliparib with temozolomide was compared to temozolomide alone in patients with recurrent platinum-sensitive or platinum-resistant SCLC. 118 Although the ORR was higher with veliparib and temozolomide (39% vs 14%, p = 0.016), there was no significant improvement in 4-month PFS or mOS. Similarly, olaparib with temozolomide was investigated in a phase I/II trial of previously treated SCLC with promising results; an ORR of 41.7%, mPFS of 4.2 months and mOS of 8.5 months. 119 In the phase I/II MEDIOLA trial, olaparib with durvalumab in patients with relapsed SCLC after platinum-based chemotherapy had an ORR of 10.5% and mPFS 2.4 months, suggesting limited efficacy of the PARPi and ICI combination, although mOS was 7.6 months. 120
Overall, while PARPi like veliparib and olaparib have shown some promise in improving ORR and PFS in combination with chemotherapy, they have so far not demonstrated significant improvements in OS. Biomarkers such as SLFN11 may play an important role in predicting which patients might benefit from PARPi and more research is needed to establish if there is a role for PARPi in the treatment of SCLC.
ATR inhibitors
The ATR inhibitor, berzosertib, has been investigated in combination with topotecan compared to topotecan alone in a phase II trial for SCLC patients who relapsed after ⩾1 previous chemotherapies. 121 Among 57 patients, the ORR (26% vs 6%, p = 0.15) and mPFS (3.9 vs 3.0 months, p = 0.44) were not significantly different. However, the mOS was significantly longer for berzosertib with topotecan (8.9 vs 5.4 months, p = 0.03), suggesting promise for this treatment combination, and outcomes were similar in both platinum-sensitive and platinum-resistant subgroups.
Emerging therapeutic targets
Recent advances in high-resolution omics technologies, including single-cell and spatial transcriptomics, alongside developments in bioinformatics and artificial intelligence, have significantly improved our understanding of molecular heterogeneity, subtypes and mechanisms driving phenotypic plasticity and therapy resistance in SCLC. 6 These tools, combined with preclinical in vitro and in vivo models, are being used to identify novel targets for systemic therapies. 123
Targeting MYC-family genes is an emerging therapeutic strategy in SCLC. MYC-family gene amplifications promote proliferation and transdifferentiation from a neuroendocrine to non-neuroendocrine phenotype via NOTCH signalling, a key mechanism of chemoresistance. 6 Aurora kinases stabilise MYC proteins, and their inhibition leads to MYC degradation and tumour apoptosis in preclinical models. 124 Similarly, USP7 inhibition has been shown to reverse platinum/etoposide resistance in MYCN-overexpressing mouse models. These findings highlight Aurora kinases and USP7 as potential therapeutic targets in MYC-family-driven SCLC. 125
Epigenetic deregulation is a hallmark of SCLC and a promising therapeutic avenue. EZH2 silences tumour suppressors and downregulates MHC class I via H3K27me3, promoting immune evasion. 126 It also represses SLFN11, a key determinant of sensitivity to DNA-damaging agents. In mouse models of SCLC, EZH2 inhibition restored SLFN11 expression and antigen presentation, re-sensitising tumours to topoisomerase inhibitors and preventing acquired resistance. 126 LSD1 is another emerging epigenetic target. By repressing NOTCH signalling and inflammatory gene expression, it maintains the neuroendocrine phenotype. Inhibition of LSD1 disrupts this repression, shifting tumours towards an inflamed phenotype and enhancing response to anti-PD-1 therapy in preclinical models. 127
CDK7 inhibition is another potential strategy to enhance ICI efficacy. CDK7 regulates cell-cycle progression, and its selective inhibition induces replication stress and genome instability, triggering immune signalling and T cell-mediated tumour surveillance. 128 In preclinical SCLC models, CDK7 inhibition synergised with anti-PD-1 therapy to improve tumour control and survival. 128
Novel immunotherapeutic approaches are also under investigation. CAR-T cells targeting DLL3 (e.g. AMG 119) have entered early-phase trials, with encouraging early signs of activity and tolerability. 129
Extrachromosomal DNA (ecDNA), found in ~20% of SCLC tumours, is associated with increased tumour aggressiveness, enhanced heterogeneity through non-chromosomal inheritance and poor prognosis. 130 In mouse models of SCLC, ecDNA carrying MYC-family gene amplifications was associated with acquired and cross-resistance to DNA-damaging chemotherapies. 131 ecDNA-positive tumours often exhibit replication stress and transcription-replication conflicts, displaying selective vulnerability to CHK1 inhibition in preclinical studies. 132 ecDNA therefore holds potential both as a biomarker of therapy resistance and a therapeutic target in SCLC. 133
These emerging therapeutic strategies mark a shift towards mechanism-based, biomarker-guided approaches to overcome chemotherapy and immunotherapy resistance, underscoring the need for biomarkers to inform therapy selection and personalise treatment in SCLC.
Current status of predictive and prognostic biomarkers for therapy selection in SCLC
Despite its heterogeneity, SCLC treatment currently lacks any form of stratification based on molecular subtypes or biomarkers. Whole genome sequencing studies have demonstrated a high rate of heterogeneous somatic genetic mutations in SCLC tumours.4,5 However, mutation profiling has so far failed to identify distinct subtypes or actionable targets and has limited utility for predicting therapy response. 123 There is a crucial need for biomarkers that can predict responses to current and emerging therapies, enable personalised treatment approaches and improve survival outcomes. Figure 1 shows candidate biomarkers of therapy response recently investigated in clinical studies.
Tissue-based and blood-based candidate biomarkers for therapy selection in small cell lung cancer investigated in clinical studies.
Transcriptional subtypes
Preclinical studies suggest that the transcriptional subtypes of SCLC may possess distinct vulnerabilities to systemic therapies, highlighting their potential as biomarkers for guiding therapy selection. 13 In the IMpower133 trial, patients with the SCLC-I subtype had the greatest improvement in mOS with the addition of atezolizumab to chemotherapy (18 vs 10 months), suggesting it has predictive value for survival benefit from PD-L1 inhibitors. 13 Similarly, in the CASPIAN trial, patients with the SCLC-I subtype receiving durvalumab ± tremelimumab/carboplatin/etoposide had the longest mOS of 24.0 months, compared to 12.1 months for SCLC-N, 11.5 months for SCLC-A and 7.0 months for SCLC-P. 134 In the IMpower133 study, high tumour DLL3 expression was found in SCLC-A and SCLC-N subtypes, suggesting potential vulnerability to DLL3-targeting drugs. 13 The SCLC-A subtype could be further stratified by high and low SLFN11 expression, a potential biomarker for response to DNA-damaging drugs. 13 In vitro assays showed sensitivity of SCLC-A to BCL2 inhibitors, SCLC-N to Aurora kinase inhibitors, SCLC-I to BTK inhibitors and SCLC-P to cisplatin, PARPi and antimetabolites. 13 Prospective clinical studies incorporating transcriptional subtyping are necessary to determine its clinical utility for predicting systemic therapy responses.
Biomarkers for immunotherapy response
PD-L1 expression in tumour tissues is widely used as a predictive biomarker for ICIs in NSCLC and other cancers. 135 In the IMpower133 and CASPIAN trials, PD-L1 expression in tumours was predominantly found in immune cells, rather than tumour cells, and was generally low and variable.28,31 In these trials, PD-L1 expression did not predict ORR, PFS or OS with the addition of a PD-L1 inhibitor to chemotherapy. Similarly, blood-based TMB was not predictive of OS benefit with the addition of anti-PD-L1 therapy. However, in CASPIAN, PD-L1 predicted improvement in OS with the addition of both PD-L1 and CTLA-4 inhibitors to carboplatin/etoposide. 31 In the CheckMate-032 trial, ORR was higher in patients with high versus low TMB: 46.2% versus 22.2% with ipilimumab/nivolumab, and 21.3% versus 4.8% with nivolumab. 136 1-year OS was also longer in patients with high versus low TMB for both ipilimumab/nivolumab (62.4% vs 19.6%) and nivolumab (35.2% vs 22.1%) groups. This suggests TMB can predict survival benefit from ICIs.
Tumour immune cell signatures may also aid ICI response prediction. Pre-treatment SCLC samples from IMpower133 showed that the SCLC-I subtype could be differentiated into neuroendocrine and non-neuroendocrine subsets based on immune gene signatures. 9 The SCLC-I neuroendocrine subset, characterised by low tumour-associated macrophage signals but high T-effector signals, showed longer OS with atezolizumab plus carboplatin/etoposide versus carboplatin/etoposide alone (HR = 0.26, 95% CI, 0.12–0.57). In non-neuroendocrine SCLC-I, characterised by high tumour-associated macrophage and high T-effector signals, there was little OS difference between groups (HR = 0.85, 95% CI, 0.53–1.37). This suggests specific SCLC-I subtypes may predict improved survival with PD-L1 inhibitor immunotherapy.
Other immune-related biomarkers have also been explored. In a retrospective study, 135 pre-treated ES-SCLC tumours were classified as inflamed or non-inflamed based on PD-1 expression and CD8+ TIL density. 137 Neoantigen load was assessed using whole-exome sequencing. mPFS was higher in inflamed tumours treated with chemo-immunotherapy compared to chemotherapy alone (10.8 vs 5.1 months, p = 0.002). 12-month PFS was longer in patients with a high frameshift neoantigen load (16.1% vs 0%). This suggests PD-1, CD8+ T cell infiltration in tumours and neoantigen load may predict benefit from ICI. An analysis of 182 pre-treatment tumour samples from the CASPIAN trial showed that high versus low CD8A expression/CD8 cell density, CD4 expression, antigen-presenting/processing machinery, MHC I and II gene signatures and higher MHC I expression were predictive of OS benefit from addition of tremelimumab to chemo-immunotherapy. 134
Serum autoimmune antibodies and cytokines have demonstrated potential as predictors of response to ipilimumab. Serum anti-voltage-gated-calcium-channel and anti-voltage-gated-potassium channel antibodies and antibodies against intracellular neuronal antigens were compared in 47 patients who received ipilimumab with carboplatin/etoposide and 38 patients who received carboplatin/etoposide in a phase II trial. 138 Patients with a positive autoimmune profile at baseline had a better mPFS with the addition of ipilimumab (8.8 vs 7.3 months, p = 0.036), although mOS was not significantly different. Serum cytokines were also globally increased with the addition of ipilimumab to chemotherapy. 139 Higher baseline IL-2 predicted sensitivity to ipilimumab, whereas high IL-6 and TNF-α predicted resistance. An increase in IL-4 in patients treated with chemo-immunotherapy was associated with improved OS. These findings are limited by small sample sizes, independent and unmatched cohorts, and a focus on CTLA-4 inhibition, which is not currently used in standard treatment for SCLC.
Biomarkers for response to targeted therapies
DLL3 has gained attention as a predictive biomarker due to the development of DLL3-targeting therapies. Although widely expressed in SCLC, its expression is heterogeneous. 140 In the phase II Rova-T study, ORR was similar in patients with high DLL3 (⩾75% of tumour cells) and DLL3-positive tumours pre-treatment (14.3% vs 13.2%). 99 In the DeLLphi-301 study, responses to tarlatamab were seen in both DLL3 positive and negative patients. 95 This suggests DLL3 expression levels might not predict response to DLL3 targeted therapies. Further studies in larger cohorts with improved DLL3 stratification are needed to confirm these findings.
SLFN11 has emerged as a potential predictive biomarker for DNA-damaging therapies in several cancers and pre-clinical SCLC models. 141 In a phase II trial, SLFN11-positive pre-treated SCLC tumours treated with veliparib/temozolomide showed a longer mPFS (5.7 vs 3.6 months, p = 0.009) and mOS (12.2 vs 7.5 months, p = 0.014) compared to SLFN11-negative tumours. 118 No differences in mPFS or mOS were observed with temozolomide alone regardless of SLFN11 expression. This suggests SLFN11 might predict outcomes with veliparib. Pre-clinical studies using SCLC cell lines and xenograft models have demonstrated that low levels of SLFN11 are linked to resistance to lurbinectedin, while the absence of SLFN11 correlates with increased sensitivity to ATR inhibitors. 142 This suggests SLFN11 could have potential as a predictive biomarker for responses to lurbinectedin and ATR inhibitors. In a phase II trial, 38% of relapsed SCLC patients with methylated O-6-methylguanine-DNA methyltransferase (MGMT) DNA-repair gene responded to temozolomide, compared to 7% with unmethylated MGMT, 89 suggesting MGMT methylation may have utility as a biomarker for temozolomide response, consistent with studies in glioblastoma. 143
Liquid biopsies
Due to the limited availability of tumour tissue in SCLC, there has been increased interest in blood-based liquid biopsies as a minimally invasive route to molecularly characterise tumours. 144 Circulating tumour cells (CTCs) and circulating tumour DNA (ctDNA) have been detected in the blood of 70%–95% and >90% of SCLC patients, respectively,145,146 suggesting strong potential as biomarkers for prediction and monitoring of treatment response.
Higher CTC numbers are associated with poorer PFS and OS in patients undergoing platinum-based chemotherapy, although most studies did not show a significant relationship between change in CTC count and tumour response.147–151 Higher CTC levels have been linked to worse PFS and OS in LS-SCLC patients receiving chemo-radiotherapy. 152 Single-cell sequencing of CTCs can identify somatic mutations and/or copy number alterations (CNAs), with CNA scores predictive of PFS and OS for patients on first-line chemotherapy. 153 Detecting SLFN11 expression from CTCs has demonstrated potential for prediction and monitoring of response to DNA-damaging therapies. 154 PD-L1 expression on CTCs is also being explored as a biomarker for ICI response. 155
Cell-free DNA (cfDNA) in SCLC patients can be analysed by profiling of mutation, methylation, fragmentation and/or structural changes. 133 High ctDNA levels are associated with worse PFS and OS.156–158 ctDNA mutation profiling has identified actionable mutations and chemotherapy resistance mechanisms,159–161,146,162 although these have not yet demonstrated utility for matching to targeted treatments. Reduction in blood-based TMB is associated with a better prognosis 163 and detectable ctDNA after curative-intent treatment predicts worse PFS and OS. 164 Longitudinal studies of cfDNA mutations have shown that variant allele frequency and blood-based TMB reduce with chemotherapy treatment 165 and changes in ctDNA levels correlate with response to chemotherapy.158,166 Longitudinal analysis of genome-wide CNAs showed their detectability is associated with response to chemotherapy and relapse/progression. 167 Most of these studies are limited by small sample sizes and the lack of utility of mutational profiling for molecular subtyping and treatment stratification in SCLC.
There is growing interest in epigenomic profiling of cfDNA for subtyping, monitoring, prognosis and predicting therapy response. 133 Methylation profiling of cfDNA has shown utility for identifying transcriptional subtypes of SCLC,10,168 as has combined mutational and nucleosome cfDNA profiling. 169 This approach enables minimally invasive identification of SCLC transcriptional subtypes, which may prove valuable in the future for treatment stratification and longitudinal monitoring for subtype switching. Combined methylation and nucleosomal profiling of cfDNA has also shown potential to detect transformation of EGFR-mutant NSCLC to SCLC, which could enable earlier identification of transformation and inform switching of treatment. 170 Therefore, cfDNA holds significant potential as a tool for minimally invasive subtyping and monitoring of tumours to aid therapy selection.
Conclusion
Advances in drug development and growing insight into SCLC subtypes are beginning to reshape the treatment landscape towards more effective and personalised approaches. Recent progress includes the addition of ICIs to chemotherapy in ES-SCLC and their use as adjuvant therapy following radical chemo-radiotherapy in LS-SCLC. ICIs are also being evaluated in combination with chemo-radiotherapy or as consolidation in LS-SCLC. However, outcomes remain poor due to the rapid development of resistance and high relapse rates, and only a minority of patients experience long-term survival benefit from ICIs.
Beyond first-line treatment, chemotherapy options such as lurbinectedin, topotecan and platinum/etoposide rechallenge offer modest survival benefits. The DLL3-targeting BiTE, tarlatamab, has demonstrated significant improvement in survival outcomes and received FDA approval for relapsed SCLC, and is now being investigated in earlier treatment lines. Other BiTEs are also in development, with promising early results. Several ADCs, including sacituzumab govitecan and I-DXd, are under clinical investigation and have shown encouraging efficacy in previously treated SCLC. PARPi and angiogenesis inhibitors have so far demonstrated limited efficacy as monotherapies in SCLC, and may offer greater benefit when used in combination treatments, such as with ICIs.
Molecular profiling and computational biology are enhancing our understanding of SCLC subtypes, revealing mechanisms of resistance to therapy and helping to identify new therapeutic targets. Ultimately, mechanism-driven combination strategies may be required to overcome resistance and achieve durable responses. Introducing novel therapies earlier in the treatment pathway to delay or prevent resistance and reduce relapse risk is likely to become an emerging focus. Although no predictive biomarkers or molecular subtypes are currently used in clinical practice, candidates under investigation include transcriptional subtypes, SLFN11, TMB, immune signatures and liquid biopsies. There remains an urgent need for more effective therapies, improved molecular subtyping and reliable biomarkers to optimise therapy selection and enable personalisation of SCLC treatment.
