Lung cancer is the leading cause of cancer related mortality worldwide, with most patients having advanced disease at the time of diagnosis (1). However, much progress has been made recently in treating non-small cell lung cancer (NSCLC), the most common type of lung cancer. Tyrosine kinase inhibitors (TKIs), anti-angiogenesis agents, and immune checkpoint inhibitors have dramatically changed the landscape of NSCLC treatment (2–4).. In addition, combination therapy with different pharmaceuticals has proven highly effective due to the ability to affect multiple pathways involved in the progression of the disease (5).

Drug resistance has been the most important factor limiting the success, in terms of overall survival, of systemic anticancer therapy for advanced lung cancer. Once a cancer has developed resistance to a given chemotherapeutic agent, the usual strategy is to initiate a different therapy using non-cross resistant drugs (6). Since the Response Evaluation Criteria In Solid Tumors (RECIST) was introduced in 2000 (Version 1.0) and updated in 2009 (Version 1.1), to assess tumor response of cancer therapeutics by RECIST and to stop current anti-cancer treatment if evaluated as disease progression has been the standard of care of lung cancer management (7, 8). In a systematic review, Davies et al. described that median overall survival ranged from 4.6 months to 12.8 months from the time of second-line treatment initiation in advanced NSCLC (9). Approximately 30% of patients received third-line treatment and only 2.5 to 17.7% patients received fourth-line therapy (9). Currently, there is an unmet need to prolong the duration of each line of effective therapy.

In oncology, treatment beyond progression (TBP) is an expression that indicates the continuation of ongoing therapy after disease progression has resumed (10). TBP is therefore defined as treatment continuing in spite of disease progression, according to the Response Evaluation Criteria In Solid Tumors (RECIST). TBP is carried on while the patient tolerates the current therapy well, is clinically stable, without main organ dysfunction, and provides updated consent (11).

Randomized studies in which patients either continue or discontinue an anti-cancer agent after disease progression are essential to conclude whether TBP is effective. However, there might have substantial difference of characteristics between patients who continue a treatment and those who discontinue treatment (6). Generally, patients who are doing well are left on their medicine while those who are struggling are moved onto a new therapy. As a result, randomized controlled trials which investigate lung cancer treatment are scant. Nevertheless, there are retrospective observation studies. These studies are not designed to compare TBP or treatment discontinuation, which may result in selection bias, and have produced inconsistent results because of small sample sizes and part of subgroup analysis. However, they are still of great reference value clinically. Therefore, we conducted the present systematic review and meta-analysis to provide evidence for the effects of TBP on lung cancer survival.

This systematic review and meta-analysis have been performed following the PRISMA checklist. The study protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO) database (CRD42021285147).

To the best of our knowledge, this is the first systematic review and meta-analysis to focus on whether or not the TBP treatment strategy provided survival benefit for NSCLC patients. Our findings suggest that TBP may improve ppPFS, ppOS and OS from initiation of drugs.

In recent years, immunotherapy, mainly consisting of checkpoint inhibitors including anti-programmed death 1 and anti–programmed death-ligand 1, has dramatically changed cancer treatment paradigms. Immune checkpoint inhibitors stimulate the immune system to attack tumors instead of targeting tumor cells directly, exhibiting different patterns of response to immunotherapy (45). These include alterations in tumor biology reflecting anticancer efficacy following initial radiographic PD (46). Because uncertainty about whether immunotherapy was discontinued and late benefit from treatment continuation among some patients, most clinical trials of immunotherapies permit treatment beyond RECIST-defined PD as long as performance status remains acceptable, the patient provides consent, no serious toxic effects, and no impending end organ damage is observed (47, 48).

Immunotherapy TBP may be a rational treatment choice for the following reasons. First, about 0.6 to 5.8% patients with NSCLC may initially experience increased size of tumor lesions during immunotherapy treatment, followed by a delayed partial response (49). This phenomenon is called “pseudo-progression,” and possibly results from infiltration and recruitment of lymphocytes in the tumor (50). Second, the interaction between the tumor and the immune system may be a long term process which could possibly result in undulant clinical effects, such as undulating tumor growth and shrinkage (48). Third, radiotherapy and chemotherapy may have synergistic effects when combined with immunotherapy via the release of tumor antigen, causing a proinflammatory environment and resulting in activation and clonal expansion of T cells (51, 52). Forth, lesion-level heterogeneity at the time of RECIST-defined PD was common in immunotherapy-treated patients and they these patients may demonstrate ongoing disease control in a subset of tumor sites (53).

In our meta-analysis, TBP significantly prolonged ppOS without statistically significant benefit for ppPFS and OS from initiation of drugs. Enomoto et al. demonstrated no significant difference in ppOS between TBP and the other treatment group. The definition of TBP (nivolumab ≥ 2 weeks after first PD using RECIST v1.1) may explain the less favorable results obtained with nivolumab beyond progression in this study compared with other studies, such as Ricciuti et al. (nivolumab ≥ 6 weeks after first PD using RECIST v1.1) (33, 41). Metro et al. showed no significant difference in ppPFS after comparing pembrolizumab beyond progression and salvage chemotherapy (31). Despite the study’s small sample size, pembrolizumab TBP could be beneficial in select patients. Among the nine patients in the TBP group with the addition of local ablative radiotherapy and PD in no more than two organ sites, the ppPFS rates at 6 and 12 months were high at 88.9% and 71.1%, respectively. Based on previous studies, TBP with immunotherapy may be beneficial in specific circumstances, such as oligo-progression, PD without new lesion, in patients with good performance status, and with add on treatment (41, 43, 53).

For many years, first and second generation EGFR TKIs represented milestones of first line treatment in NSCLC patients with EGFR mutations (54). Osimertinib, a third generation TKI, could overcome treatment resistance acquired after use of first line TKIs, such as T790M. In first line settings, T790M showed better efficacy compared to first and second generation TKIs (55, 56). However, most patients receiving EGFR TKIs eventually developed TKIs resistant progressions. Moreover, EGFR-mutant lung cancer patients showed poor responses to immunotherapy treatment (57). To achieve better survival among EGFR mutant lung cancer patients, we need to prolong treatment duration with EGFR TKIs. In clinical practice, continuing EGFR TKIs still benefited some patients with EGFR mutation who developing acquired resistance to Erlotinib or gefitinib, suggesting that part of tumor cells remained sensitive to EGFR-TKIs (58).

Oxnard et al. explained this phenomenon in vitro using EGFR-mutant NSCLC cell lines; they pointed out that resistant tumors are likely a mixed components of EGFR-TKIs-sensitive and resistant cells (59). In our subgroup analysis, TBP with EGFR TKIs significantly prolonged ppPFS and OS from initiation of drugs but had no significant benefit in ppOS. Mok et al. and Ding et al. showed that gefitinib plus chemotherapy was not beneficial for patients with acquired resistance to first line gefitinib, however, patients with T790M negative tumors may be the select patients who can benefit from continuation of gefitinib beyond progression (26, 44). In previous studies, patients with gradual progression rather than dramatic progression, oligo-progressive disease, and added on therapy using local ablative treatments may also be among the select patients who benefit from TBP (32, 36, 60, 61).

Since the 2007 discovery of ALK rearrangement in NSCLC, tremendous strides have been made in the treatment of ALK positive NSCLC, best exemplified by the approval of six ALK TKIs (62). Our data show that TBP with ALK TKIs may further prolong ppOS and OS from initiation of drugs. Results from Chiari et al. revealed negative results of TBP; unsurprisingly, shifting to second generation ALK TKIs produced better ppPFS than TBP with first generation TKI, because second generation ALK TKIs may overcome some mechanisms of resistance to first generation ALK TKIs (16, 63). Therefore, a reasonable treatment strategy could be to maximize treatment duration of TBP with each line of ALK TKIs, then shifting to the next line ALK TKIs which impact resistance pathways produced by the previous line ALK TKIs.

Anti-angiogenesis agents, such as bevacizumab, which is a recombinant, humanized monoclonal antibody that targets vascular endothelial growth factor (VEGF), have been approved for treatment of non-squamous NSCLC in combination with chemotherapy, target therapies, and immunotherapy (64). Targeted action against angiogenesis can cause normalization or regression of existing tumor vasculature and the inhibition of new and recurrent tumor vessel growth (65). Furthermore, due to the multiple effects of VEGF on the tumor immune microenvironment, targeting VEGF with anti-angiogenesis agents enhances the anti-cancer immune response (66). Given the mechanism of action of anti-angiogenesis agents, there is a rationale for TBP with added-on TKIs or chemotherapy, on purpose to maintain an angiogenesis blockade (67). Our data supports that TBP with anti-angiogenesis agents prolongs ppPFS and ppOS. However, Takeda et al. reported no significant survival benefit of TBP with bevacizumab; subgroup analysis of their data revealed that patients whose disease progressed starting at least six months after the initiation of first-line chemotherapy and those who achieved a complete or partial response to first-line treatment gained more advantage from bevacizumab continuation (24). In another study that observed negative results of TBP, Higashiguchi et al. nonetheless claimed that they could not deny the possibility of the benefits of TBP with bevacizumab, because it was associated with a better response rate and the OS of the TBP group with bevacizumab looked slightly better than that of the non-TBP group (20).

The results of this study have some limitations. First, as in any meta-analysis, analysis of results of the study limited to the data reported by the authors. Precisely, some authors do not present the HRs, and we could only calculate HRs and the associated statistics based on the information given in the study report. Second, most of the studies were observational, not randomized controlled trials. Observational studies are likely to have greater potential biases than randomized studies because randomized studies adjust known and unknown confounders to balance across different groups. Therefore, we should always interpret results cautiously when observational studies are included in reviews and meta-analyses. Third, this meta-analysis did not include data on individual patients. As a result, it was not possible to adjust patient variables such Eastern Cooperative Oncology Group performance status, age, and race. Fourth, the between-study heterogeneity became lower but was still high after subgroup analysis. Factors that could potentially explain the heterogeneity may include the definition of TBP, which differed in each study; moreover, more than half of the studies did not provide one. Fifth, potential bias of recruitment may contribute to meaningful OS differences. Patients who were selected to TBP groups may have, to a varying degree, better condition such as performance status than those who were not.

This study provides further evidence in support of TBP for NSCLC, however, these results require cautious interpretation. Currently, clinicians and patients are left with uncertainty about how best to deal with disease progression. Treatment decisions will continue to depend on many points, including the availability of other therapeutic agents, clinician’s instincts, and the patient’s evaluation of benefits and risks. Large, randomized, prospective controlled trials to investigate the efficacy of TBP in lung cancer treatment and the biomarkers to predict the populations who may benefit from TBP are warranted.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

W-KK and Y-JL designed the study. W-KK and Y-JL designed the statistical plan. W-KK and Y-JL performed the key analyses. W-KK and Y-JL generated and collected the data. C-FW assisted in data interpretation. W-KK wrote the manuscript. C-FW and Y-JL revised the manuscript. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.