PDF
Citation
Print
Introduction
Lung cancer is responsible for the highest cancer incidence and mortality globally, with 2 million new cases and 1.8 million deaths annually, and it has been identified as the most prevalent type of fatal cancer in both sexes in domestic data [1,2]. Over 30% of patients who are initially diagnosed with lung cancer have locally advanced non–small cell lung cancer (LA-NSCLC). For these patients, particularly those with unresectable tumors, the historical standard of care (SoC) was definitive concurrent chemoradiotherapy (CCRT) alone, and, despite progress in chemotherapy and radiotherapy (RT), the median overall survival (OS) rate for patients with LA-NSCLC has remained disappointingly low at 9-34 months [3,4]. However, the PACIFIC trial, a phase III randomized controlled trial (RCT), demonstrated a dramatic improvement in progression-free survival (PFS) and OS for patients with stage III non–small cell lung cancer (NSCLC) who received durvalumab following definitive CCRT, which established consolidative durvalumab as the new SoC [5]. The recently released 5-year survival outcomes further confirmed the robust success of durvalumab, with a median OS of 47.5 months compared to 29.1 with a placebo [6].
However, there are some discrepancies between RCTs and real-world clinical settings. The PACIFIC trial included patients with unresectable stage III NSCLC and a European Cooperative Oncology Group performance status (ECOG PS) of 0-1 while excluding those who experienced grade ≥ 2 radiation pneumonitis (RP) after CCRT and had a known history of tuberculosis. Furthermore, protocol regulations included constraints such as mean lung dose (MLD) < 20 Gy, a volume of lung receiving at least 20 Gy (V20) < 35%, and mean esophagus dose < 34 Gy. The stringent criteria in clinical trials and the variations in patient characteristics and healthcare systems across countries highlight the need for validation to ensure the real-world safety and efficacy of the results obtained from RCTs [7].
Durvalumab received the Food and Drug Administration approval in the United States in February 2018, followed by its approval in South Korea in April 2020. Hence, while a considerable number of patients are currently undergoing immunotherapy in real-world settings, there is a scarcity of publications on the assessment of long-term oncologic outcomes in the domestic context. Therefore, we aimed to present the real-world clinical outcomes of consolidative durvalumab at a single center and explore the role of RT in the era of immunotherapy.
Materials and Methods
1. Patients
We retrospectively reviewed the medical records of patients who underwent definitive CCRT with unresectable LA-NSCLC at Asan Medical Center from May 2018 to May 2021. The key inclusion criteria were: patients (1) aged ≥ 18 years, (2) with unresectable LA-NSCLC who require definitive CCRT based on National Comprehensive Cancer Network guidelines, (3) who had received at least two cycles of chemotherapy concurrently with RT, and (4) who had never received RT to the thorax or underwent pulmonary surgery [8]. A total of 171 patients were included for analysis, with 89 receiving consolidative durvalumab after CCRT and 82 receiving CCRT alone (S1 Fig.). The patient’s clinical stage was evaluated based on the 8th edition of the American Joint Committee on Cancer staging system.
2. Treatments
Four-dimensional simulated computed tomography (CT) scans of 2.5-mm slices were performed for all patients in the supine position while free-breathing using a respiratory management system (RPM, Varian Medical Systems, Palo Alto, CA) to create ten respiratory phases. Gross tumor volume was delineated with reference to chest CT and positron emission tomography (PET)–CT images, and lymph nodes (LN) suspected of malignancy were included if they had an axial diameter of 1 cm or larger on CT or demonstrated high fluorodeoxyglucose uptake on PET-CT. A clinical target volume was generated, taking into account microscopic tumor extension, and elective nodal irradiation was not performed. The internal target volume (ITV) was established considering the movement of the tumor during each phase of respiration, and the planning target volume (PTV) was set with a margin of 5-7 mm from the ITV. All patients were treated with an intensity-modulated radiation therapy plan, ensuring that 95% of the PTV received 95% of the prescribed radiation dose. The organs at risk and the constraints applied to these were consistent with those described in previous reports [9]. The median total dose was 60 Gy (range, 50 to 73 Gy), and the fraction size ranged from 1.8 to 2.2 Gy. Patients received concurrent chemotherapy agents of either a paclitaxel-based or etoposide-based regimen or cisplatin in combination with pemetrexed.
Consolidated durvalumab was initiated after the completion of CCRT, with a median time of 34 days (range, 0 to 172 days). Durvalumab was administered at a dose of 10 mg/kg every 2 weeks, and treatment was discontinued if disease progression or unacceptable complications occurred. The median number of cycles administered was 17 (range, 2 to 26).
3. Follow-up and outcomes
Regular follow-up after CCRT included history taking, physical examinations, chest X-rays, and blood tests. Chest CT scans were performed every 6 months for 2 years, followed by annual CT scans for the subsequent 5 years. Freedom from locoregional failure (FFLRF) was defined as the period from the end of CCRT to the occurrence of locoregional failure (LRF), and in-field and out-field recurrence were identified based on the PTV margin. Distant metastasis-free survival (DMFS) was defined as the period from the end of CCRT to the occurrence of distant metastasis or death. In this study, recurrence or progression in the lung, mediastinum, hilum, and/or supraclavicular region at ipsilateral side was classified as LRF, while the progression at the contralateral lung or any pleural metastasis was classified as distant metastasis. PFS and OS were defined as the durations from the end of CCRT to the occurrence of any progression or death and to death from any cause, respectively. RP grade ≥ 2 was assessed using criteria established by the Radiation Therapy Oncology Group and the European Organization for Research and Treatment of Cancer.
4. Statistical analysis
Statistical analyses were performed using R ver. 4.2.2 (R Development Core Team, Vienna, Austria). To compare the distribution of characteristics between the durvalumab and CCRT-alone groups, an independent-sample T-test was performed for continuous variables, and a chi-square test was performed for noncontinuous variables. Survival outcomes were calculated using the Kaplan-Meier method, and to identify clinical factors related to each oncologic outcome, univariate and multivariate analyses were performed using Cox regression analysis. A receiver operating characteristic curve was generated to identify an optimal cutoff value for each dosimetry parameter in predicting RP. A p-value < 0.05 indicated statistical significance for all analyses.
Results
1. Patients characteristics
A total of 171 patients were included in the final analysis, and their characteristics are presented in Table 1. The median age of the entire study population was 65 years (range, 36 to 85 years), and 154 patients (90.1%) had an ECOG PS of 0-1. There were 81 patients (47.4%) with squamous cell carcinoma and 82 (48.0%) with adenocarcinoma. In total, 85 patients (49.7%) had cT3-4 disease, 137 (80.1%) had cN2-3 disease, and 154 (90.1%) were diagnosed with stage III disease. In the programmed cell death ligand 1 (PD-L1) (SP263) Assay (Roche Diagnostics, F. Hoffmann-La Roche Ltd., Basel, Switzerland), expression of ≥ 1% was defined as PD-L1–positive, and out of the 158 patients, 102 (59.6%) were found to be PD-L1–positive.
Table 1.
Characteristics of patients according to the use of durvalumab (n=171)
Values are presented as median (range), number (%), or mean±SD. Because of rounding off, not all of the percentages total 100. CCRT, concurrent chemoradiotherapy; DLCO, diffusing capacity of the lung for CO; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; FEV1, forced expiratory volume in one second; FVC, forced vital capacity; PD-L1, programmed cell death ligand 1; PFT, pulmonary function test; RT, radiotherapy; SD, standard deviation.
2. A pattern of failure with the use of durvalumab
During a median follow-up period of 25.9 months, 108 patients (63.2%) experienced progression or death; the pattern of failure is shown in Table 2. The number of LRF was significantly higher in the CCRT-alone group compared to the durvalumab group, with 48 and 33 cases, respectively (p=0.008). Furthermore, incidence of in-field recurrences was higher within the CCRT-alone group (p=0.027). However, there was no significant difference in distant failures between the two groups (p=0.644). The most common sites of first failure in both groups were lung, regional LN, brain, and bone, in that order.
Table 2.
Recurrence pattern of 171 non–small cell lung cancer patients treated with concurrent chemoradiotherapy
| Outcome | CCRT alone (n=82) | CCRT+Durvalumab (n=89) | p-value |
|---|---|---|---|
| Any recurrence | 55 (67.1) | 47 (52.8) | 0.115 |
| Any LRF | 48 (58.5) | 33 (37.1) | 0.008 |
| Isolated LRF | 18 (22.0) | 11 (12.4) | 0.143 |
| Any DM | 37 (45.1) | 36 (40.4) | 0.644 |
| Isolated DM | 7 (8.5) | 14 (15.7) | 0.231 |
| Both LRF and DM | 30 (36.6) | 22 (24.7) | 0.129 |
| First site of failure | |||
| Lung and/or regional lymph node | 44 (53.7) | 28 (31.5) | 0.005 |
| Lung only | 20 (24.4) | 11 (12.4) | 0.066 |
| Regional lymph node only | 8 (9.8) | 6 (6.7) | 0.661 |
| Both lung and regional lymph node | 16 (19.5) | 11 (12.4) | 0.284 |
| Brain | 4 (4.9) | 5 (5.6) | > 0.99 |
| Bone | 1 (1.2) | 5 (5.6) | 0.252 |
| Others (liver, adrenal gland, distant LN) | 5 (6.1) | 7 (7.8) | > 0.99 |
| Both lung and distant organ | 1 (1.2) | 2 (2.2) | > 0.99 |
| Locoregional failure pattern | |||
| In-field failure | 22 (26.8) | 11 (12.4) | 0.027 |
| Out-of-field failure | 11 (13.4) | 10 (11.2) | |
| Botha) | 15 (18.3) | 12 (13.5) | |
| No. of organ with distant metastasis | |||
| Single organ | 15 (18.3) | 17 (19.1) | 0.807 |
| Multiple organ | 20 (24.4) | 18 (20.2) | |
| Site of distant metastasis | |||
| Pleura | 13 (15.9) | 7 (7.9) | 0.189 |
| Contralateral lung | 21 (25.6) | 12 (13.5) | 0.042 |
| Bone | 13 (15.9) | 17 (19.1) | 0.855 |
| Brain | 13 (15.9) | 14 (15.7) | 0.999 |
| Distant LN | 8 (9.8) | 10 (11.2) | 0.924 |
| Adrenal gland | 5 (6.1) | 3 (3.4) | 0.582 |
| Liver | 2 (2.4) | 3 (3.4) | 0.900 |
| Others | 6 (7.3) | 0 | 0.029 |
The median FFLRF in the durvalumab group was not reached, compared to 19.8 months in the CCRT-alone group, and the 3-year rate was 57.3% versus 38.8% (p=0.008), with an absolute difference of 18.5% (Fig. 1A). Although there was a tendency toward prolonged median DMFS with the addition of durvalumab (CCRT alone vs. CCRT+durvalumab, 24.3 months vs. 37.8 months, p=0.270), but with no significant difference (Fig. 1B).
Fig. 1.
Kaplan-Meier curve for the overall population (n=171). (A) Freedom from locoregional failure (FFLRF), and 1-, 2-, 3-, and 4-year FFLRF rates. (B) Distant metastasis-free survival (DMFS), and 1-, 2-, 3-, and 4-year DMFS rates. (C) Progression-free survival (PFS) and 1-, 2-, 3-, and 4-year PFS rates. (D) Overall survival (OS), and 1-, 2-, 3-, and 4-year OS rates. CCRT, concurrent chemoradiotherapy.
3. Progression-free survival and overall survival
The median PFS was 13.7 months for the CCRT-alone group compared to 20.9 months for the durvalumab group, with corresponding 3-year PFS rates of 28.3% and 46.3%, respectively (p=0.048) (Fig. 1C). The median follow-up period for mortality was 28.2 months, during which 41 patients in the CCRT-alone group and 26 patients in the durvalumab group expired. The median OS with durvalumab was not reached, compared to 35.4 months with CCRT alone (p=0.010) (Fig. 1D). In line with this, the univariate analysis also revealed that the use of durvalumab was associated with improved PFS and OS (Table 3). The multivariate analysis for OS further demonstrated that advanced age, stage IIIC, and receiving CCRT alone were associated with inferior survival (S2 Fig.).
Table 3.
Univariate analysis of clinical factors affecting progression-free survival and overall survival (n=171)
| Variable |
Progression-free survival |
Overall survival |
||
|---|---|---|---|---|
| HR (95% CI) | p-value | HR (95% CI) | p-value | |
| Age, ≥ 65 yr | 1.08 (0.73-1.58) | 0.710 | 1.61 (0.98-2.63) | 0.059 |
| Male sex | 0.61 (0.39-0.96) | 0.031 | 1.49 (0.76-2.92) | 0.246 |
| ECOG PS, 2-3 | 1.02 (0.53-1.96) | 0.950 | 1.64 (0.81-3.32) | 0.166 |
| Underlying lung disease | 1.17 (0.70-1.94) | 0.552 | 1.30 (0.69-2.43) | 0.412 |
| (Ex-)Smoker | 0.80 (0.52-1.21) | 0.283 | 1.73 (0.93-3.23) | 0.086 |
| FVC (%) | 1.01 (1.00-1.03) | 0.181 | 1.00 (0.98-1.02) | 0.772 |
| FEV1 (L) | 0.99 (0.71-1.37) | 0.931 | 0.86 (0.57-1.31) | 0.487 |
| DLCO (%) | 1.00 (0.99-1.01) | 0.477 | 0.99 (0.97-1.00) | 0.083 |
| Clinical stage, III | 1.50 (1.17-1.93) | 0.002 | 1.56 (1.13-2.15) | 0.006 |
| Pathology, SqCC | 1.28 (0.92-1.76) | 0.141 | 0.86 (0.55-1.33) | 0.487 |
| EGFR, mutation | 1.30 (0.74-2.28) | 0.362 | 0.48 (0.18-1.26) | 0.137 |
| PD-L1 positive (> 1%) | 0.76 (0.48-1.20) | 0.242 | 0.71 (0.39-1.28) | 0.259 |
| Total RT dose, ≥ 66 Gy | 0.90 (0.61-1.31) | 0.582 | 1.04 (0.65-1.69) | 0.860 |
| Chemotherapy, TP | 0.68 (0.45-1.03) | 0.070 | 0.57 (0.35-0.95) | 0.032 |
| Use of durvalumab | 0.68 (0.47-1.00) | 0.050 | 0.53 (0.32-0.87) | 0.011 |
| Interval between CCRT and durvalumab, ≤ 42 daysa) | 0.62 (0.29-1.30) | 0.203 | 0.62 (0.23-1.71) | 0.359 |
CI, confidence interval; DLCO, diffusing capacity of the lung for CO; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; FEV1, forced expiratory volume in one second; FVC, forced vital capacity; HR, hazard ratio; PD-L1, programmed cell death ligand 1; RT, radiotherapy; SqCC, squamous cell carcinoma; TP, paclitaxel/cisplatin.
4. Oncologic outcomes by PD-L1 expression
The OS for each group, categorized by PD-L1 expression status and the use of durvalumab, is presented in S3 Fig. In the PD-L1–positive population with PD-L1 expression over 1%, the use of durvalumab resulted in improvements in FFLRF, DMFS, PFS, and OS (3-year rate, CCRT alone vs. CCRT+durvalumab; FFLRF 34.4% vs. 61.3%, DMFS 31.7% vs. 56.3%, PFS 24.7% vs. 50.9%, OS 46.1% vs. 73.8%). However, in the PD-L1–negative population, no significant differences were observed between the two groups (Table 4). Furthermore, this trend was also observed in the analysis of patients, whether stratified by PD-L1 expression levels below or above the 10% cutoff (S4 Fig.).
Table 4.
Survival outcomes in patients with negative or positive PD-L1 expression (n=138)
Among the 102 patients who were PD-L1–positive, the durvalumab group exhibited a higher proportion of younger age and advanced stage. The CCRT-alone group showed a higher percentage of patients receiving induction chemotherapy, with the groups otherwise maintaining a well-balanced distribution (S5 Table). Within this population, a more pronounced survival benefit with durvalumab was observed across all oncologic outcomes compared to the overall population analysis, with an absolute difference of 27.7% in the 3-year OS rates between the two groups (Fig. 2).
Fig. 2.
Kaplan-Meier curve for the patients with programmed cell death ligand 1-positive tumors (n=102). (A) Freedom from locoregional failure (FFLRF), and 1-, 2-, 3-, and 4-year FFLRF rates. (B) Distant metastasis-free survival (DMFS), and 1-, 2-, 3-, and 4-year DMFS rates. (C) Progression-free survival (PFS) and 1-, 2-, 3-, and 4-year PFS rates. (D) Overall survival (OS), and 1-, 2-, 3-, and 4-year OS rates. CCRT, concurrent chemoradiotherapy.
5. RP of grade 2 or higher
The occurrence of RP grade ≥ 2 was noted at a median of 2.8 months (range, 0.3 to 25.2 months) following the completion of CCRT. There were 34 cases (41.5%) in the CCRT-alone group and 40 cases (44.9%) in the durvalumab group, with no significant difference observed between the two groups (p=0.761) (Fig. 3). With each threshold value used to predict RP, it was consistently observed that patients with higher values of PTV, MLD, a volume of lung receiving at least 5 Gy (V5), and V20 were associated with a significantly higher incidence of pneumonitis for each parameter (S6 and S7 Figs.). No significant differences in dosimetry values were found between the durvalumab and CCRT-alone groups (S8 Table).
6. Evaluating eligibility criteria for the PACIFIC study
Real-world adherence to the eligibility criteria of the PACIFIC trial was assessed, identifying 100 patients (58.5%) classified as ineligible, with the reasons documented in S9 Table. In the eligible group, there were 31 patients in the CCRT-alone group and 40 patients in the durvalumab group. The 3-year freedom from local progression (FFLP) for CCRT-alone and durvalumab group were 42.1% and 61.8%, respectively (p=0.220), while the 3-year OS were 60.6% and 73.2%, respectively (p=0.017). In the non-eligible group, which included 51 patients in the CCRT-alone group and 49 patients in the durvalumab group, the 3-year FFLP were 36.4% and 56.7%, respectively (p=0.310), while the 3-year OS were 40.7% and 66.1%, respectively (p=0.013). Furthermore, there was no significant difference in the occurrence of grade ≥ 2 RP based on the use of durvalumab in both eligible and non-eligible patients.
Discussion
In our study, the addition of consolidative durvalumab following CCRT for patients with unresectable LA-NSCLC was found to improve PFS and OS, and its safe application was confirmed through real-world data (RWD) analysis.
A remarkable finding is that the application of durvalumab led to a dramatic improvement in locoregional control (LRC), and it is expected to have contributed to the enhancement of survival outcomes, even though we had a considerable proportion of high-risk patients with stage IIIB (48.0%) and IIIC (14.6%) disease. It is particularly noteworthy that the improvement in LRC was even more pronounced in PD-L1–positive patients, which can be explained by preclinical results elucidating the synergistic effect between RT and immunotherapy [10-12]. Durvalumab functions by blocking the immunosuppressive interaction between PD1 and CD80, leading to the increased activation of T-cells and subsequent detection and elimination of tumor cells. RT can play a role in counteracting tumor immune escape by enhancing MHC-I expression, releasing Toll-like receptor 4, generating tumorspecific cytotoxic T cells, and upregulating PD-L1 expression [13,14]. These synergistic effects are anticipated to be further augmented through the administration of immune checkpoint inhibitors (ICIs) following the completion of CCRT. The optimal timing for implementing ICI remains a subject of debate, with ongoing studies investigating the most optimal treatment sequence [15,16]. Our findings are supported by other clinical data as well. Taugner et al. [17] reported a significant improvement in 1-year FFLRF with respective rates of 78.9% compared to 44.5% for the durvalumab group versus CCRT alone (p=0.002). Similarly, Abe et al. [18] observed a 1-year local control of 86% compared to 62% with the addition of durvalumab (p=0.005). In the PACIFIC trial, despite not specifically suggesting LRF rate as an outcome, the pattern of failure analysis reveals that the durvalumab group exhibited a higher incidence of new lung lesions compared to the placebo group (17.3% vs. 11.8%) [19]. In the Korean extension of PACIFIC (PACIFIC-KR), LRF emerged as the most common type of failure, accounting for 56.2% of cases, and lung and LN involvement were the most prevalent, representing 60.3% and 35.6%, respectively [20].
In our study, we were also able to observe enhancements in both PFS and OS with the addition of durvalumab. Despite the criticism in many studies regarding the disparity in patient inclusion between real-world situations and RCTs, we could confirm, as seen in S10 Table, the improvement of clinical outcomes in various forms of RWD [20-28]. Similarly, despite including a higher proportion of elderly patients and approximately 10% of patients with an ECOG PS of 2, which categorized approximately 58.5% of the patients as ineligible according to clinical trial criteria, our study still demonstrated superior survival compared to RCTs. Furthermore, there was a survival improvement with the use of durvalumab even among those belonging to the ineligible group. Hence, we propose a cautious expansion of durvalumab application in the real-world setting, leveraging these RWD.
Age, clinical stage, and the use of consolidative durvalumab were identified as clinical factors associated with OS, and these findings align with results reported in previous studies. However, in our study, there was a trend towards better survival among patients with epidermal growth factor receptor (EGFR) mutations, which could be attributed to the fact that 40.9% of these individuals received adjuvant tyrosine kinase inhibitor (TKI) treatment. The 3-year OS for patients with EGFR wild type, mutation without TKI, and mutation with TKI were 62.5%, 61.5%, and 88.9%, respectively, with the PACIFIC trial reporting a TKI therapy usage rate of only around 10%.
While we observed comparable incidences of RP grade ≥ 2 across cohorts, suggesting the safe use of durvalumab, our study, in line with previous research, emphasized that the primary determinant for RP occurrence was not the inclusion of durvalumab, but rather the influence of dosimetry factors. This underscores the necessity for increased vigilance during the radiation therapy planning process [29].
Our study had limitations. First, as a retrospective study, it may have had inherent selection bias. Additionally, while our study had a longer follow-up period compared to other investigator-initiated studies, a more mature follow-up is still necessary. On the other hand, this study’s strength lay in its comprehensive exploration of oncologic outcomes, including OS, setting it apart from the majority of non-early access program studies that primarily focused on RP and offered limited information. Additionally, with a substantial patient population, our study, as the largest single-center analysis to date, enhanced the understanding of treatment efficacy by conducting a detailed evaluation of patterns of failure.
In conclusion, this real-world study demonstrated that consolidative durvalumab following CCRT improved LRC, PFS, and OS in patients with unresectable LA-NSCLC. The benefits were more pronounced in patients with PD-L1–positive tumors. These findings support the use of durvalumab as an SoC in this patient population, validating the results of previous clinical trials. Further research and longer follow-up periods are warranted to confirm these results and assess long-term survival outcomes.
XML Download