Treatment patterns and survival analysis in patients with unresectable stage III EGFR-mutated non-small cell lung cancer

Purpose: To investigate the treatment patterns and survival outcomes in patients with unresectable Stage III EGFR-mutated non-small cell lung cancer (NSCLC). Materials and methods: A retrospective analysis was conducted on patients with unresectable Stage III EGFR-mutated NSCLC spanning from 2012 to 2022. Treatment patterns were outlined, and survival comparisons between different treatment groups were performed using Kaplan-Meier methods. Results: A total of 88 patients were included: 62.5% received TKI alone, 26.1% received TKI+chemotherapy, 4.5% received radiotherapy, 4.5% participated in clinical trials, and 2.4% received TKI+antiangiogenic drugs. Prior to propensity score matching, TKI+chemotherapy and TKI alone groups demonstrated similar progression-free survival (hazard ratio [HR] = 1.56, 95% confidence interval [CI]: 0.87-2.80; P = 0.134), overall survival (HR = 1.12, 95% CI: 0.59-2.13; P = 0.733), and locoregional-free survival (HR = 1.46; 95% CI: 0.75-2.81; P = 0.267). However, TKI+chemotherapy showed reduced distant metastasis-free survival compared to TKI alone (HR = 2.39, 95% CI: 1.11-5.18; P = 0.022). After propensity score matching, no significant differences were observed in progression-free survival (P = 0.435), overall survival (P = 0.205), locoregional-free survival (P = 0.706), and distant metastasis-free survival (P = 0.171) between the TKI+chemotherapy and TKI alone groups. Conclusions: The addition of chemotherapy to TKI did not enhance survival outcomes compared to TKI monotherapy in patients with unresectable Stage III EGFR-mutated NSCLC.

Despite its status as the standard treatment, several studies have reported that CCRT may lead to poorer survival outcomes in EGFR-mutated patients [16,17].Conversely, in clinical practice, EGFR-tyrosine kinase inhibitor (TKI) is often recommended for stage III cases based on survival data from trials designed for stage IV EGFR-mutated NSCLC [18][19][20][21][22][23][24].To date, the optimal treatment strategy for unresectable stage III EGFRmutated NSCLC patients remains unclear.Therefore, this study aims to investigate the treatment patterns and survival rates in this specific patient population.AGING

Patient characteristics
The selection process of patients is visually represented in Figure 1.Out of the 5,477 patients tested for EGFR status, 88 were deemed eligible for the study.A comprehensive summary of these patients' characteristics can be found in Table 1.Notably, the median followup time for these patients was 19 months, with an interquartile range of 11-29 months.

Treatment patterns
The various treatment approaches administered to the patients are detailed in Figure 2. Overall, 62.5% of the patients received TKI alone, while 26.1% received TKI+chemotherapy.Radiotherapy was administered to 4.5% of the patients, 4.5% participated in clinical trials, and 2.4% received a regimen consisting of TKI+antiangiogenic drugs.
Due to the small numbers of patients in the radiotherapy, antiangiogenic therapy, and clinical trials subgroups, the survival analysis primarily compared the outcomes between the TKI alone group and the TKI+chemotherapy group.The specific characteristics of patients in these two subgroups before and after PSM are detailed in Table 2.

PFS
In the unmatched cohort, the median PFS for the TKI alone group was 17 months, compared to 11 months for the TKI plus chemotherapy group (Figure 3A).The 1-year PFS rates were 62.8% and 41.6%, respectively, for the two groups.Similarly, the 2-year PFS rates were 41.3% and 31.2%,respectively.The univariable regression analysis revealed no statistically significant difference in PFS between the two groups (HR = 1.56, 95% CI: 0.87-2.80;P = 0.134; Table 3).Furthermore, multivariable Cox regression analysis confirmed that TKI+chemotherapy was not an independent prognostic factor for PFS (HR = 1.47, 95% CI: 0.73-2.97;P = 0.276; Table 3).In the PSM cohort, the median PFS for the TKI alone group and the TKI+chemotherapy group was 18 months and 12 months, respectively (Figure 3B).The corresponding 1-year PFS rates were 64.3% and 45.9%, while the 2-year PFS rates were 35.7% and 34.4%.

OS
In the unmatched cohort, the median OS was 29 months for the TKI alone group and 28 months for the TKI+chemotherapy group (Figure 4A).The 1-year OS rates were 77.7% and 82.0%, respectively, for the two groups, while the 2-year OS rates were 63.7% and 63.2%.According to the univariable regression analysis, no significant difference in OS was observed between the two groups (HR = 1.12, 95% CI: 0.59-2.13;P = 0.733; Table 4).Furthermore, the multivariable Cox regression analysis also supported this finding by showing that TKI+chemotherapy was not an independent prognostic factor for OS (HR = 1.57, 95% CI: 0.69-3.54;P = 0.280; Table 4).
In the PSM cohort, the median OS was 74 months for the TKI alone group and 28 months for the TKI+chemotherapy group (Figure 4B).The 1-year OS was 82.9% and 80.0%, respectively, for these two groups, while the 2-year OS was 63.5% and 60.0%, respectively.
In the PSM cohort, the median LRFS for the TKI alone group was 75 months, compared to 49 months for the TKI+chemotherapy group (Figure 5B).The respective 1-year LRFS rates were 74.7% and 49.0%, while the 2year LRFS rates were 37.4% and 36.8%.
In the PSM cohort, similar trends were observed with median DMFS not being reached for the TKI alone group and being 25 months for the TKI+chemotherapy group (Figure 6B).For the TKI alone and TKI+chemotherapy groups, the respective 1-year DMFS rates were 82.2% and 61.9%, while the respective 2-year DMFS rates were 64.7% and 55.7%.

DISCUSSION
The findings of this study revealed two important insights.Firstly, it highlighted the prevailing trend in clinical practice where the majority of patients with unresectable stage III EGFR-mutated NSCLC were primarily treated with TKI monotherapy, as opposed to chemoradiotherapy.Secondly, our data suggested that the addition of chemotherapy to TKI therapy as an initial treatment strategy did not provide survival benefits over TKI alone.
There could be several explanations for this observation.One possible explanation was the evidence from various clinical trials that have demonstrated the efficacy of TKI as a standard treatment for stage IV EGFR-mutated NSCLC patients [18][19][20][21][22][23][24].These trials had also included a subset of stage III patients, who achieved superior PFS with TKI monotherapy compared to CCRT [16,17].Another explanation could be the lack of difference in OS (HR = 0.71, 95% CI: 0.34-1.47)and cancer-specific survival (HR = 0.65, 95% CI: 0.31-1.35)between TKI monotherapy and CCRT, as reported in a previous    study [25].Additionally, the absence of radiation oncologists in defining treatment strategies could also contribute to the low utilization of radiotherapy [26].
While TKI monotherapy has shown better PFS than CCRT, it is crucial to recognize that it is not a curative treatment.TKI alone did not improve OS for stage III patients compared to stage IV patients [27].Moreover, patients who experienced recurrences after CCRT had the option of salvage TKI therapy, which had shown significant OS improvement [28][29][30].Therefore, recommending EGFR-TKI monotherapy as the initial treatment for all stage III patients may not be appropriate.
The FLAURA2 study had emphasized the improved PFS in stage IV patients receiving TKI plus chemotherapy [31,32].This raises the intriguing possibility of whether similar advantages could be observed in patients with stage III disease.Currently, Table 7. Adverse events between TKI alone and TKI+chemotherapy groups.however, there is a paucity of evidence addressing this question directly.In our study, we observed that the median PFS was 17.0 months in the TKI alone group and 11.0 months in the TKI+chemotherapy group, with no statistically significant difference between the two groups (HR = 1.56, 95% CI: 0.87-2.80;P = 0.134).This finding was consistent with several other studies that had reached similar conclusions [25,29,33,34].
While these findings contribute valuable insights, it is crucial to interpret them with caution.A majority of participants in our study (89.7%) received firstgeneration TKIs, which had not demonstrated improved PFS when combined with chemotherapy in stage III patients [35].It is noteworthy that third-generation TKIs have shown greater efficacy compared to their firstgeneration counterparts [18][19][20].Therefore, there is a compelling rationale to investigate the therapeutic potential of combining third-generation TKIs with chemotherapy in stage III EGFR-mutated NSCLC patients.
It is important to acknowledge the limitations of our study.The retrospective design inherently results in potential confounders, such as age, sex, and T stage, being imbalanced between the treatment groups.
Although we attempted to account for these factors through multivariable analysis and PSM, the possibility of unmeasured confounders remains.Moreover, the relatively small sample size of the TKI+chemotherapy group may have limited our ability to detect subtle differences in survival outcomes between the two treatment strategies.
In conclusion, our findings suggested that the addition of chemotherapy to TKI therapy did not improve survival outcomes in unresectable stage III EGFRmutated NSCLC patients compared to TKI alone.AGING However, given the limitations of our study, including its small sample size, these results should be interpreted cautiously.Larger prospective studies and clinical trials are essential to validate these findings and provide more definitive guidance on the optimal treatment approach for these patients.

Patient selection
We conducted a comprehensive search to identify NSCLC patients at Guangxi Medical University Cancer Hospital between September 2012 and January 2022.
The inclusion criteria for patient selection were as follows: (1) histologically or cytologically confirmed NSCLC, (2) EGFR testing performed and mutation confirmed, (3) definitive EGFR subtypes identified for the EGFR mutation, (4) staging according to the 8th edition of the American Joint Committee on Cancer (AJCC) criteria confirming stage III disease, and (5) presence of unresectable disease.Patients meeting any of the following criteria were excluded from the study: (1) those who did not receive any treatment, (2) patients who underwent surgery as their initial treatment modality, and (3) individuals with incomplete or missing clinical information.

Endpoint
The primary endpoint of the study was progression-free survival (PFS).The secondary endpoints included overall survival (OS), locoregional-free survival (LRFS), and distant metastasis-free survival (DMFS).

Treatment-related toxicities
Treatment-related toxicities were evaluated and categorized using the Common Toxicity Criteria for Adverse Events version 4.0 (CTCAE v4.0).

Statistical analysis
The continuous factor of age was transformed to categorical factor according to the median value.Categorical variables were assessed using the χ 2 test or Fisher's exact test.To compare PFS, OS, LRFS, and DMFS across different treatment modalities, the Kaplan-Meier method coupled with log-rank test statistics was utilized.Univariable regression analysis was performed to pinpoint potential prognostic indicators.Furthermore, multivariable Cox regression analysis was conducted to identify independent prognostic factors after adjusting for variables.The results were presented in terms of hazard ratios (HRs) along with corresponding 95% confidence intervals (CIs).
To minimize the impact of selection bias when comparing outcomes between different treatment modalities, a matched case-control analysis via propensity score matching (PSM) was executed.During the calculation of propensity scores, one-to-one matching without replacement was achieved within a logistic regression model.The nearest-neighbor matching algorithm based on the propensity score was applied for factors with a caliper of 0.8 on the logistic regression model.
The statistical analyses for this study were carried out using SPSS Statistics Version 26.0 software (IBM Co., Armonk, NY, USA) and R software (version 4.2.1).
A two-tailed P-value of less than 0.05 was deemed statistically significant.

Figure 1 .
Figure 1.Flowchart of patient selection process.

Figure 3 .
Figure 3. Progression-free survival between TKI alone and TKI+chemotherapy groups.(A) The unmatched cohort.(B) The propensity score matching cohort.

Figure 4 .
Figure 4. Overall survival between TKI alone and TKI+chemotherapy groups.(A) The unmatched cohort.(B) The propensity score matching cohort.

Figure 5 .
Figure 5. Locoregional-free survival between TKI alone and TKI+chemotherapy groups.(A) The unmatched cohort.(B) The propensity score matching cohort.

Figure 6 .
Figure 6.Distant metastasis-free survival between TKI alone and TKI+chemotherapy groups.(A) The unmatched cohort.(B) The propensity score matching cohort.

Table 4 . Univariable and multivariable Cox regressions of overall survival.
ECOG, Eastern Cooperative Oncology Group; AJCC, American Joint Committee on Cancer; TKI, tyrosine kinase inhibitor; EGFR, epidermal growth factor receptor; HR, hazard ratio; CI, confidence interval.