Local consolidative therapy in oligometastatic non-small-cell lung cancer after effective systemic treatment: who will benefit?

Current randomized evidence regarding the application of LCT in oligometastatic NSCLC

Several randomized prospective studies have suggested that LCT for residual lesions in oligometastatic NSCLC after effective systemic chemotherapy prolongs progression-free survival (PFS) and OS and may even be curative for some patients with acceptable toxicity. Forty-nine patients with limited metastases who did not develop disease progression after first-line drug therapy were randomly assigned to either maintenance systemic therapy or LCT (radiotherapy or surgery) for all active residual disease in a multicenter phase II study, as reported by Gomez et al. (NCT01725165). The LCT arm demonstrated significant benefits in PFS and OS (median PFS, 14.2 months vs. 4.4 months; P = 0.022 and median OS, 41.2 months vs. 17.0 months; P = 0.017). The incidence of grade ≥ 3 toxicity was 20% in the LCT arm compared to 8% in the non-LCT arm^5,6. Another randomized phase II trial by Iyengar et al. (NCT02045446) included 29 patients with oligometastatic NSCLC using radiotherapy as the sole LCT method. LCT administered after induction chemotherapy prolonged the median PFS (9.7 months vs. 3.5 months; P = 0.01). The incidence of grade ≥ 3 toxicity was 29% in the LCT arm compared to 20% in the non-LCT arm⁷. The majority of participants in these two trials received induction chemotherapy. Both trials were recommended for premature cessation by the Ethics Committee based on the efficacy results of the interim analysis.

A randomized phase II trial (NCT03595644) assessed the efficacy and safety of radiotherapy as LCT after induced first-line EGFR-tyrosine kinase inhibitors (TKIs), as reported by Peng et al. Sixty-one patients with oligometastatic NSCLC harboring EGFR-sensitive mutations who had received first-generation EGFR-TKIs were included and randomly assigned to receive radiotherapy for primary and/or metastatic sites. The results showed that patients receiving radiotherapy had a significantly prolonged median PFS (17.6 months vs. 9.0 months; P = 0.016) and OS (33.6 months vs. 23.2 months; P = 0.026). No grade ≥ 3 toxicity was observed in either group⁸. Recent data from the unicentric phase II NORTHSTAR trial (NCT03410043), which added LCT (radiotherapy or surgery) for patients with EGFR-mutant advanced NSCLC who did not progress within 6–12 weeks of osimertinib (a third-generation EGFR-TKI), reported that osimertinib plus LCT was well tolerated without a significant increase in grade ≥ 3 adverse events compared to osimertinib alone (29% vs. 16%; P = 0.09), with 2.3% of patients experiencing grade 3 pneumonitis⁹. Survival data is anticipated.

In contrast, findings for immunotherapy appeared different. The phase II/III NRG-LU002 trial (NCT03137771) reported negative results, including 215 patients with oligometastatic NSCLC receiving first-line chemotherapy or immunotherapy with or without subsequent LCT (radiotherapy and/or surgery), with 90.7% of patients receiving immunotherapy. The 1- and 2-year PFS and OS rates were not statistically different between the LCT and non-LCT arms (PFS: HR = 0.93; P = 0.664; OS: HR = 1.05, P = 0.821). Patients receiving LCT with maintenance systemic therapy experienced a higher incidence of grade ≥ 3 pneumonitis (10% vs. 1%). Results from the NRG-LU002 trial do not support the routine use of LCT in oligometastatic NSCLC after immunotherapy¹⁰. Reasons for the negative outcome of the NRG-LU002 trial remain unclear because further details have not yet been disclosed. The NRG-LU002 trial mainly included patients without disease progression after 4 cycles of immunotherapy, of whom the proportion with stable disease (SD) was 57.8%. It appears that this population did not truly identify patients who benefit from aggressive LCT. Specifically, it remains unclear if patients with the best overall SD response will benefit less from local treatment compared to patients with a partial response to immunotherapy.

A summary of the randomized clinical trials evaluating the application of LCT in patients with oligometastatic NSCLC is shown in Table 1.

Unsolved issues in clinical practice

The results from different randomized clinical trials appear controversial, leading to unresolved issues in clinical practice (i.e., how are patients selected who might benefit from LCT after effective systemic treatment?).

In these trials, the same imaging criteria are often used to select NSCLC patients with oligometastasis along with a uniform time frame for systemic treatment. However, it is crucial to recognize the heterogeneity of tumors, which presents a challenge: the enrollment population may not be sufficiently precise if based solely on the same clinical characteristics. One key to treatment selection in patients with oligometastatic NSCLC is identifying patients likely to benefit from aggressive LCT. Precision biomarkers are needed to aid in selecting these beneficiary populations.

Optimal exposure duration of prior systemic therapy

In theory, patients typically exhibit the lowest tumor burden at the time of maximum response to systemic therapy when LCT targeting residual sites is most likely to yield substantial efficacy with minimal side effects. However, defining the time of maximum response poses a challenge. There is a lack of research on the optimal duration of induced systemic therapy before local treatment. However, most studies have mandated that induced systemic therapy last at least 3 months. Some phase III clinical trials reported that > 80% of responders experienced the first response to EGFR-TKIs or chemotherapy within 12 weeks, which may partly explain the cut-off duration. It was also noted that the most significant tumor shrinkage usually occurs within 40 days after the initiation of EGFR-TKI treatment and further shrinkage becomes difficult after 90 days¹¹. Greater than 80% of first responses to immunotherapy occur within 12 weeks after 4 cycles of administration¹². However, exploratory analysis data indicate that in 75% of responders receiving chemoimmunotherapy tumors continue to regress to maximum tumor remission after achieving initial remission with a median time from the first response to maximum tumor response reported to be 109 days (ranging from 26–438 days)¹³. Persistent tumor regression represents one of the differences in response characteristics between immunotherapy and other anti-tumor therapies, such as conventional chemotherapy and targeted therapy. Unlike conventional chemotherapy and targeted therapy, which directly target tumor cells, immunotherapy induces indirect antitumor effects by restoring systemic antitumor immunity. The immune cells activated by immunotherapy provide a more comprehensive and lasting attack on tumor cells. Consequently, an exposure duration of 4 cycles before LCT may not be sufficient to achieve optimal efficacy during the immunotherapy period, which might also be a protentional reason for negative outcomes in the NRG-LU002 trial.

Elusive inclusion of oligometastatic disease by imaging

The inclusion criteria for oligometastasis varied among different trials, although all utilized the number of metastases for patient enrollment. The trial by Gomez et al. included patients with < 3 metastases (excluding the primary tumor) after frontline systemic therapy. Each lesion was counted separately but regional lymph node metastases were counted collectively as one lesion. Central nervous system lesions were counted towards the total number of metastases even if local treatment had been administered before randomization. The majority of patients in this trial had 0–1 metastases (65.3%), followed by 2–3 metastases (34.7%)⁵. The trial by Iyengar et al. allowed patients with up to 6 sites of extracranial disease (including the primary tumor, with ≤ 3 sites in the liver or lung) prior to the initiation of LCT. Up to 2 contiguous vertebral metastases were considered a single site of disease. The median number of disease sites prior to induction chemotherapy was 2–3, with a range of 2–6⁷. The trial by Peng et al. admitted patients with ≤ 5 discrete distant metastases with most exhibiting 1 metastatic organ (51.5%)⁸. The NRG-LU002 trial enrolled patients with ≤ 3 extracranial metastases after induction systemic therapy, with the majority of patients presenting with 1 lesion (59%), followed by 2 lesions (26%), 3 lesions (14%), and 4–5 lesions (1%)¹⁰.

The absence of biomarkers renders imaging the most relevant diagnostic method for defining oligometastatic disease. Despite some consensus on an image-based definition of oligometastatic disease, clinical outcomes after LCT for this condition vary widely¹⁴. There is still a lack of biological or pathologic evidence suggesting that a clinical condition exists which separates oligometastatic NSCLC with a defined number of detectable metastases from those with widespread metastases. Furthermore, it remains unclear whether patients with synchronous oligometastasis exhibit different biological characteristics compared to patients with metachronous oligometastasis or whether induced oligometastasis differs from repeated oligometastasis. From the perspective of precision medicine, incorporating biomarkers that reflect the underlying biological characteristics of the tumor and immune contexture could enhance these definitions.

Potential factors on selecting benefit population

Novel treatment paradigm in oligometastatic NSCLC based on biomarkers

ctDNA-guided maintenance treatment or observation after LCT

It has been reported that a decreased ctDNA burden post-LCT at an early follow-up time compared to pre-LCT levels reflect the reduction in tumor burden achieved by LCT. Additionally, the first detection of increasing ctDNA occurred at a median of 5.0–6.7 months before radiographic progression in NSCLC¹⁹. These findings support the application of ctDNA monitoring for the dynamic assessment of disease changes in advanced NSCLC after LCT, further guiding subsequent treatment. An exploratory proof-of-concept study based on a prospective trial proposed a new treatment paradigm of ctDNA-guided maintenance treatment or observation for oligometastatic NSCLC after LCT. This study included patients with oligometastatic NSCLC harboring driver mutations after TKI therapy and LCT. Treatment discontinuation was performed in those with radiologically undetectable disease, undetectable ctDNA, and normal serum carcinoembryonic antigen (CEA) levels following LCT. Surveillance, including radiologic examination and ctDNA and CEA analyses, was conducted every 2–3 months. Participants received retreatment with prior TKIs for progressive disease or ctDNA or CEA positivity, whichever occurred first. Treatment was discontinued again during surveillance once a complete response was achieved and molecular indicators were negative. The median PFS was 18.4 months and the median treatment break duration was 9.1 months with a median duration of preceding TKI therapy of 12.0 months in the cohort of 60 patients enrolled. The TKI retreatment response rate was 96% and no grade ≥ 3 adverse events were observed²⁰. These results suggest that this adaptive de-escalation maintenance TKI strategy based on ctDNA status is feasible for oligometastatic NSCLC after LCT. A novel treatment paradigm based on biomarkers in oligometastatic NSCLC is shown in Figure 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Novel treatment paradigm in oligometastatic NSCLC based on biomarkers. Patients with metastatic non-small cell lung cancer (NSCLC) who met the criteria of radiologic oligometastasis after an optimal duration of induced systemic therapy receive peripheral blood molecular testing, such as circulating tumor DNA (ctDNA) and carcinoembryonic antigen (CEA). Those with negative plasma molecular testing could be defined as molecular oligometastasis and receive local consolidative therapy (LCT) for all residual diseases. After LCT, regular molecular and imaging monitoring was recommended. Treatment discontinuation could be performed in those with undetectable radiologic and peripheral blood molecular disease following LCT.

Although the above results are encouraging, the limitations of ctDNA-guided treatment paradigm must be noted. Currently, ctDNA detection via liquid biopsy technology still has false negatives and the risk of ctDNA-guided “drug holidays” cannot be overlooked, especially for patients with brain metastases. In addition, because of tumor heterogeneity ctDNA analysis may not fully reflect all mutations in the tumor and the value of detecting non-driver genes mutations in ctDNA analysis remains controversial, which may lead to limited effectiveness in individual patients. The applicability of this treatment paradigm to oligometastatic patients with non-EGFR mutations requires further exploration. These limitations still need to be gradually overcome in future studies to further improve their clinical application value.

Future directions

The definition, assessment, and overall management of treatment strategies in oligometastatic NSCLC remain controversial and challenging. We conclude that three aspects should be considered when evaluating the benefits of LCT for oligometastatic NSCLC: 1) systemic treatment strategies and durations before LCT; 2) radiologic criteria for oligometastatic disease; and 3) pre-treatment ctDNA status before LCT. Future research is warranted to further clarify the criteria for defining oligometastatic disease and to determine optimal treatment strategies to improve patient survival and quality of life. Additionally, the rational use of biomarkers for personalized treatment and the collaboration of multidisciplinary teams will become central to future research efforts.

Author contributions

Conceived and designed the analysis: Jiayi Deng, Qing Zhou.

Collected the data: Jiayi Deng, Mingyi Yang.

Wrote the paper: Jiayi Deng, Mingyi Yang.