Correlation between treatments and outcomes of patients with EGFR-mutated non-small-cell lung cancer that transitioned into small-cell lung cancer: an international retrospective study

Abstract

Background

Histological transformation to small-cell lung cancer (SCLC) is an under-recognized but clinically relevant mechanism of resistance in epidermal growth factor receptor (EGFR)-mutant non-small-cell lung cancer (NSCLC). While platinum-based chemotherapy (CT) has emerged as standard treatment after small-cell transformation, optimal treatment strategies are not defined, particularly with regard to the role of immune checkpoint inhibitors (ICIs) or continuation of EGFR-tyrosine kinase inhibitors (TKIs).

Materials and methods

We conducted an international, retrospective, observational study on patients with EGFR-mutated NSCLC that transformed to SCLC, focusing on treatment and outcomes.

Results

Twenty-five patients across 11 centers were included. Twenty-four changed systemic treatment following SCLC transformation: CT (12 patients), CT plus EGFR-TKI (5 patients), CT plus ICI (6 patients), or CT plus ICI plus EGFR-TKI (1 patient). Median follow-up was 9 months (range 1-45 months). All patients experienced relapse with a median progression-free survival of 2 months [95% confidence interval (CI) 2-3 months] and 23 patients died, with a median overall survival (OS) of 9 months (95% CI 7-16 months). Median OS was comparable between patients treated with CT alone [9.5 months; 95% CI 5 months-not reached (NR)], CT plus EGFR-TKI (8 months; 95% CI 8 months-NR), and CT plus ICI (10 months; 95% CI 7 months-NR). Three patients survived >24 months from SCLC transformation; these patients were treated with CT alone (one patient), CT plus EGFR-TKI (one patient), and CT plus ICI and EGFR-TKI (one patient). OS was longer if SCLC transformation occurred >12 months from initial NSCLC diagnosis (19 patients) versus those with transformation within 12 months (6 patients): 31 versus 8 months (P < 0.001).

Conclusions

The prognosis of patients with transformed SCLC remains poor, with only a minority achieving meaningful survival. Survival was similar in patients treated with CT alone or with the addition of ICI or EGFR-TKI. A time interval >12 months between NSCLC diagnosis and SCLC transformation was associated with longer survival and may reflect a different biology than early transformation.

Highlights

• •Transformed SCLC is an extremely rare condition.
• •The prognosis of patients with transformed SCLC remains poor, with only a minority achieving long OS.
• •The addition of EGFR-TKI or ICI to CT alone could not improve survival.
• •An interval >12 months between NSCLC diagnosis and SCLC transformation seems to be a positive prognostic factor.

Introduction

Since the discovery of activating, sensitizing mutations in epidermal growth factor receptor (EGFR), EGFR-tyrosine kinase inhibitors (TKIs) have become the standard first-line systemic treatment for patients with advanced EGFR-mutated non-small-cell lung cancer (NSCLC). Unfortunately, the majority of these tumors develop secondary resistance, mediated by various mechanisms.¹ One under-recognized mechanism is the histological transformation from NSCLC into small-cell lung cancer (SCLC), a process known as SCLC transformation.² This accounts for 5%-15% of EGFR-TKI resistance.^3,4

When transforming to SCLC, tumors undergo significant molecular and biologic change and exhibit many classical characteristics of SCLC, such as a high nuclear-to-cytoplasmic ratio, expression of neuroendocrine markers, and inactivation of the RB1 and p53 tumor suppressors.5, 6, 7 Despite these common characteristics, transformed SCLC should be considered a distinct entity.8, 9, 10 Transformed SCLC is more aggressive and has a worse prognosis, and there are currently no unified treatment guidelines. Unfortunately, due to the rarity of this pathology, there is a scarcity of data on this patient population, especially concerning the efficacy of chemotherapy (CT) with immune checkpoint inhibitors (ICIs) or the value of continuing EGFR-TKIs.

To increase this knowledge, we conducted an international, retrospective, observational study on patients with EGFR-mutated NSCLC that transformed into SCLC, focusing on the correlation between treatment and outcomes.

Materials and methods

We conducted a multicenter, international, retrospective, observational study on patients diagnosed with EGFR-mutated NSCLC that transformed into SCLC from 2013 to 2023. The study was approved by the Scientific Ethics Committee of the coordinating institute (CE Humanitas Prot. GAV 997/22 on 23 December 2022) and of all the other involved centers.

Patient records were reviewed to extract data from all the patients. Mandatory inclusion criteria were a previous diagnosis of EGFR-mutated NSCLC, biopsy-confirmed transformation to SCLC (synaptophysin and/or chromogranin stains were conducted for SCLC diagnosis), and available information on SCLC treatment and outcome, including time to progression after initial SCLC treatment, and date of death or last follow-up.

Patients with mixed tumors were excluded from this analysis.

Descriptive analyses were carried out for all selected patients. Progression-free survival (PFS) and overall survival (OS) functions were estimated using the Kaplan–Meier method. PFS was calculated from the date of biopsy documenting SCLC transformation to the date of tumor progression or death, whichever occurred first. OS was calculated from the date of biopsy documenting SCLC transformation to the date of death or last follow-up. The log-rank test was used to compare OS between patients who received different systemic treatment approaches. In the comparison of OS between patients who received CT plus ICI versus others and those who received CT plus EGFR-TKI versus others, one patient treated with CT plus EGFR-TKI plus ICI was excluded.

Patient characteristics

Patient characteristics are reported in Table 1. Twenty-seven patients with EGFR-mutated NSCLC that transformed into SCLC diagnosed between 2013 and 2023 were identified across 11 centers and 25 out of 27 patients were included in this analysis (Figure 1). Two patients were excluded due to unconfirmed SCLC transformation upon pathological revision (mixed tumor: one patient; large-cell neuroendocrine tumor: one patient). Data on smoking status were available for 24 patients: 9 patients were current/former smokers (cigarette smokers) and 15 patients were never smokers.

Table 1.

Patient characteristics (total: 25 patients)

Characteristics	Number of patients
Sex (female/male)	19/6
NSCLC
Median age at diagnosis, years (range)	61 (40-78)
Histology
Adenocarcinoma	24
Not otherwise specified carcinoma	1
EGFR mutation
Exon 19 deletion	16
L858R	6
Other	3
First-line systemic treatment (patients)	24
First-generation EGFR-TKI	6
Afatinib	2
Osimertinib	16
Transformed SCLC
Median age at diagnosis, years (range)	62 (44-80)
Disease extension
Limited disease	1
Extensive disease	24
Molecular profiling (patients)	14
EGFR mutation	14
First-line systemic treatment (patients)	24
Chemotherapy alone	12
Platinum plus etoposide	11
Cyclophosphamide doxorubicin vincristine	1
Chemotherapy plus ICI	6
Platinum, etoposide, plus atezolizumab	5
Platinum, etoposide, plus durvalumab	1
Chemotherapy plus EGFR-TKI	5
Platinum, etoposide, plus osimertinib	4
Platinum, etoposide, plus erlotinib	1
Chemotherapy plus ICI plus EGFR-TKI	1
Platinum, etoposide, afatinib, and atezolizumab	1

EGFR-TKI, epidermal growth factor receptor-tyrosine kinase inhibitor; ICI, immune checkpoint inhibitor; NSCLC, non-small-cell lung cancer; SCLC, small-cell lung cancer.

Figure 1.

Median overall survival (A) and median progression-free survival (B) in the overall population.

mOS, median overall survival; mPFS, median progression free survival.

The histology of NSCLC was adenocarcinoma in 24 patients (carcinoma not otherwise specified, 1 patient). The median age of patients at the time of primary NSCLC diagnosis was 61 years (range 40-78 years), and the median age at the time of SCLC transformation was 62 years (range 44-80 years). Among the 25 patients, 19 were female and 6 were male. Twenty-four patients received systemic treatment for NSCLC, all with an EGFR-TKI (exclusive radiotherapy: one patient).

Following the detection of transformed SCLC, two patients firstly exhibited single-lesion histologically confirmed progression of adenocarcinoma, and then had a diffuse progression of histologically proven small-cell carcinoma.

NSCLC molecular features

All primary tumors harbored an EGFR mutation: exon 19 deletion (16 patients), L858R mutation (6 patients), exon 15 deletion (2 patients), and exon 18 deletion (1 patient). Coexisting T90M mutation was detected in two cases.

A standard-of-care gene panel sequencing [next-generation sequencing (NGS)] was carried out in 10 samples and many co-mutations/amplifications were detected in 8 patients.

Single co-mutation/amplification:

• -p53 mutations (2 patients)
• -CTNNB1 (1 patient)
• -PIK3CA (1 patient)
• -EGFR amplification (1 patient)

Multiple coexisting co-mutations/amplifications (single-patient profile):

• -BRAF SNV, BRCA1 R332W
• -PIK3CA E970K, BRCA2 copy number loss, MYC amplification, FGFR1 amplification, CCNE1 amplification
• -ALK P835L mutation, ABL mutations, BRCA2 mutation, NF1 mutation, MSH6 mutation

SCLC molecular features

The EGFR sequence was analyzed in 14 SCLC specimens, all exhibiting the same mutation present in the corresponding primary NSCLC.

A standard-of-care gene panel sequencing (NGS) was carried out in 10 patients, showing RB1, TP53, and PIK3CA mutations in 3, 2, and 2 patients, respectively.

One patient also had an AKT1-E17K mutation combined with EGFR and MET amplification, and another patient showed an EP300 mutation.

Systemic treatment for transformed SCLC

Systemic treatments after SCLC transformation are summarized in Table 1. Twenty-four out of 25 patients underwent a change in systemic treatment after SCLC transformation. One patient died 1 month after the SCLC transformation without receiving systemic treatment due to rapid clinical deterioration. Twelve patients were treated with CT alone, five patients with CT plus continuation of an EGFR-TKI, six patients with CT plus ICI, and one patient with CT plus ICI plus an EGFR-TKI.

Patient outcomes

The median follow-up was 9 months (1-45 months). All patients relapsed after SCLC therapy with a median PFS (mPFS) of 2 months [95% confidence interval (CI) 2-3 months] (Figure 1). The mPFS by specific systemic treatment was 2 months [0 months-not reached (NR)], 5 months (3 months-NR), and 2 months (2 months-NR) for patients who received CT alone, CT plus ICI, and CT plus EGFR-TKI, respectively.

Twenty-three patients died due to tumor progression, with a median OS (mOS) of 9 months (95% CI 7-16 months) (Figure 1). The mOS by specific systemic treatment was 9.5 months (5 months-NR), 10 months (7 months-NR), and 8 months (8 months-NR) for patients who received CT alone, CT plus ICI, and CT plus EGFR-TKI, respectively. The mOS of the 5 patients treated with CT plus EGFR-TKI was similar to the OS of the 18 other patients (8 versus 9.5 months, P = 0.86). The patient treated with CT plus EGFR-TKI plus ICI was excluded from this comparison (Figure 2). The mOS of the 6 patients treated with CT plus ICI was similar to the mOS of the 17 other patients (10 versus 9 months, P = 0.89). The patient treated with CT plus EGFR-TKI plus ICI was excluded from this comparison (Figure 2). Three patients achieved an OS >24 months: one patient received CT alone, one patient received CT plus EGFR-TKI, and one patient received CT plus EGFR-TKI plus ICI.

Figure 2.

Median overall survival by treatment. Chemotherapy vs other (A); chemotherapy + Immune checkpoint inhibitors vs other (B); chemotherapy + tirosin kinase inhibitors vs other (C); time to transition < 12 months vs > 12 months (D).

mOS, median overall survival; Cht, chemotherapy; TTT, time to transition; ICI, immune checkpoint inhibitor; TKI, tirosin kinase inhibitor.

The time interval between diagnosis of NSCLC and SCLC transformation was highly variable, ranging from 2 to 87 months (median 24 months). The mOS for the 19 patients with an interval >12 months was longer than the mOS for the 6 patients with an interval ≤12 months (mOS 31 versus 8 months; P < 0.001) (Figure 2). The OS of the two patients who initially exhibited oligoprogression of adenocarcinoma was 30+ and 18+ months.

An overview of the tumors’ molecular features, treatments, and patients’ outcomes is presented in Figure 3.

Figure 3.

Overview of tumor molecular features, treatments, and patients’ outcomes. Afa, afatinib; atezo, atezolizumab; Ampl, amplification; beva, bevacizumab; C/P, carboplatin–pemetrexed; CAV, cyclophosphamide plus adriamycin plus vincristine; CNV, copy number value; CP, platinum–pemetrexed; durva, durvalumab; EP, platinum–etoposide; osi, osimertinib; P/E, platinum–etoposide; tax, paclitaxel; TMZ, temozolomide; TXT, docetaxel; VNB, vinorelbine. ^aPIK3CA E970K, copy number loss, BRCA2 copy number loss, MYC amp, FGFR1 amp, CCNE1 amp. ^bALK P835L, mutations in ABL, BRCA2, NFT, MSH6. ^cAKT1 E17K, EGFR CNV 11, MET CNV 8. Figure created with BioRender.com.

Discussion

SCLC transformation is a poorly understood mechanism of resistance to targeted therapy, best described in EGFR-mutant NSCLC treated with EGFR-TKI therapy. The optimal management strategy is unclear, and clinical practice varies. De novo SCLC is treated with a combination of platinum plus etoposide CT given with a programmed death-ligand 1 (PD-L1) inhibitor, based on the survival benefit conferred by immunotherapy seen in the phase III IMpower 133 and CASPIAN trials with atezolizumab and durvalumab, respectively.^11,12 In contrast, responses to programmed cell death protein 1 and PD-L1 inhibitors in non-transformed EGFR-mutant NSCLC are rare.¹³ While EGFR-mutant NSCLC transformed to SCLC has many similarities morphologically and genomically to de novo SCLC, it is unclear whether the same benefit from immunotherapy exists in this population, which largely comprised patients with no smoking history. It is also unclear whether targeting EGFR after transformation is of value. While the transformed EGFR cells have largely lost EGFR protein expression,² resistance is often polyclonal, and there may be benefit from continuing to suppress any sensitive populations that coexist with those that have transformed to SCLC. While prospective trials are needed, they are at the moment lacking.

This retrospective analysis identified 25 patients with EGFR-mutant NSCLC transformed to SCLC from 11 centers over a 10-year period, confirming the extreme rarity of this condition.

Molecular features of transformed SCLC were consistent with the available data already reported, with TP53, RB1, and PIK3CA mutations identified in a minority of patients.¹⁴

It is noteworthy that 9 out of 25 patients (36%) were current or former smokers at the time of EGFR-positive NSCLC diagnosis. This proportion of patients with a smoking history is higher than expected in the population of EGFR-positive NSCLC. This observation aligns with the findings of Saalfeld et al. and Wang et al.,^15,16 who reported 55% and 34% of smokers, respectively, supporting further investigation. Following the detection of transformed SCLC, two patients firstly exhibited single-lesion histologically confirmed progression of adenocarcinoma, and then had a diffuse progression of histologically proven small-cell carcinoma. This observation supports the utility of carrying out a further rebiopsy at the time of transformed SCLC progression, particularly in cases of oligoprogression. Of interest, the OS of these two patients was 18+ and 30+ months.

The analyzed patient population received different SCLC systemic treatments, primarily CT alone, CT plus EGFR-TKI, and CT plus ICI. There was no clear difference in OS between these strategies and it remains unclear whether ICI or EGFR-TKI provide value beyond platinum-based CT. What is clear is that the prognosis for patients was very poor, with an mPFS of 2 months and an mOS of only 9 months. There were some patients who achieved longer survival. Of major interest, the time interval between NSCLC diagnosis and SCLC transformation appeared to carry prognostic value, with better outcomes for patients who experienced transformation after 12 months from the initial diagnosis. It is possible that late transformation stems from a different biology than early transformation.

A comparison between our results and those reported from other authors is limited by the scarcity of existing data in the literature.

The largest series on transformed SCLC has been published by Saalfeld et al. including 47 patients; 17 were treated with CT alone, 20 were treated with CT plus ICI, and 10 received CT plus EGFR-TKI.¹⁵ Also, in this series, PFS and OS were similar across the various treatment strategies.

The prognosis of this patient population reported in five retrospective series with over 10 patients appears similar to that found in our analysis, with an mOS ranging from 10 to 14 months.^3,4,16,17 Numerous case reports and small case series do not dramatically differ from these results.18, 19, 20, 21

Conclusion

EGFR-mutated NSCLC that transforms into SCLC carries a poor prognosis, with only a minority showing a long OS. An interval of >12 months between NSCLC diagnosis and SCLC transformation seems to be a favorable prognostic factor. The addition of ICI or EGFR-TKI to CT does not appear to significantly improve patient survival.

Our findings should be interpreted with caution, given the retrospective design of the study and the limited sample size of the patient cohort analyzed. Prospective studies in this setting, with planned NGS analysis, tissue rebiopsy, liquid biopsy, and delta-like ligand 3 staining, remain a relevant and pressing unmet need.