Efficacy and safety of metastasis directed stereotactic radiotherapy in NSCLC patients progressing under targeted- or immunotherapy.

Background: Metastasis directed treatment (MDT) is increasingly performed with the attempt to improve outcome in non-small cell lung cancer (NSCLC) patients receiving targeted- or immunotherapy (TT/IT). This study aimed to assess the safety and efficacy of metastasis directed stereotactic radiotherapy (SRT) concurrent to TT/IT in NSCLC patients. Methods: A retrospective multicenter cohort of stage IV NSCLC patients treated with TT/IT and concurrent (≤30 days) MDT was established. 56% and 44% of patients were treated for oligoprogressive disease (OPD) or polyprogressive disease (PPD) under TT/IT, polyprogressive respectively. Survival was analyzed using Kaplan-Meier and log rank testing. Toxicity was scored using CTCAE v4.03 criteria. Predictive factors for overall survival (OS), progression free survival (PFS) and time to therapy switch (TTS) were analyzed with uni- and multivariate analysis. Results: MDT of 192 lesions in 108 patients was performed between 07/2009 and 05/2018. Concurrent TT/IT consisted of EGFR/ALK-inhibitors (60%), immune checkpoint inhibitors (31%), VEGF-antibodies (8%) and PARP-inhibitors (1%). 2y-OS was 51% for OPD and 25% for PPD. After 1 year, 58% of OPD and 39% of PPD patients remained on the same TT/IT. Second progression after MDT was oligometastatic (≤5 lesions) in 59% of patients. Severe acute and late toxicity was observed in 5.5% and 1.9% of patients. In multivariate analysis, OS was influenced by the clinical metastatic status (p=0.002, HR 2.03, 95% CI 1.30-3.17). PFS was better in patients receiving their first line of systemic treatment (p=0.033, HR 1.7, 95% CI 1.05-2.77) and with only one metastases-affected organ (p=0.023, HR 2.04, 95% CI 1.10-3.79). was 6 months longer in patients with one metastases-affected organ (p=0.031, HR 2.53, 95% CI 1.09-5.89). Death was never therapy-related. Conclusions: Metastases-directed SRT in NSCLC patients can be safely performed concurrent to TT/IT with a low risk of severe toxicity. To find the ideal sequence of the available multidisciplinary treatment options for NSCLC and determine what patients will benefit most, a further evaluated in a broader context within prospective clinical trials is needed continuation of TT/IT beyond progression combined with MDT for progressive lesions appears promising but requires prospective evaluation. Trial registration: retrospectively registered


Introduction
The development of targeted-, and immunotherapy (TT/IT) has improved the prognosis of stage IV non-small cell lung cancer (NSCLC) significantly. However, due to the development of resistance, most patients will eventually develop progressive disease. Since stereotactic radiotherapy (SRT) has emerged as a safe and locally effective modality for treatment of oligometastases, its use as part of the multimodality treatment in this situation is increasing [1], as radical local radiotherapy of persistent or progressive (oligo-) metastases could improve outcome [2][3][4][5][6].
Three randomized trials have shown that addition of a localized metastases-directed treatment (MDT) to systemic therapy improved progression free survival (PFS) as well as overall survival in NSCLC patients with oligometastatic disease [7][8][9] . However, the definition of oligometastatic disease varies in literature, while research on the identification of a biological oligometastatic state is ongoing [10]. This becomes obvious in a systematic review on oligometastatic NSCLC patients, where a variation of 5-year OS between 8.3% and 86% was observed [11]. The subcategorization of (oligo-)metastatic patients remains a challenge, but is important as the value of local treatments varies between these patients [3,12].
Besides oligometastatic disease, SRT might also play a role as "salvage" treatment in the management of patients who have (oligo-)progressive metastatic disease while under TT/IT [12]. Once a stable oncological status is achieved under systematic therapy and no severe side effects develop, a patient preferably continues this drug for as long as possible, assuming that the further lines of systematic treatment are characterized by a worse therapeutic ratio.
Unfortunately, intrinsic or acquired resistance to systematic drugs develops in nearly all patients [13]. Here, MDT could possibly prevent or delay the switch of systematic therapy by radical local treatment of all progressive metastatic sites. This study aimed to evaluate efficacy and safety of metastasis directed SRT (MDT) in NSCLC patients who are progressive under immuno-or targeted therapy.

Materials and Methods
A retrospective international multicenter registry study was established by the German Society for Radiation Oncology (DEGRO) working group for radiosurgery and stereotactic radiotherapy to collect data on stage IV patients receiving SRT with concurrent targeted-or immunotherapy (TOaSTT study antibodies, tyrosine kinase inhibitors and/or immune-checkpoint inhibitors. Cranial SRT was defined as delivery of up to 5 fractions, or one fraction with a minimum of 16Gy. Stereotactic body radiotherapy (SBRT) was defined as delivery of ≤10 fractions with a minimum total dose of 50 Gy (2 Gy equivalent, α/β of 10 Gy; the α/β-ratio of 10 represents the intrinsic radiosensitivity of NSCLC metastases [14]).
To further evaluate clinical scenarios in which MDT is currently often performed, the group of oligoprogressive disease patients (OPD, ≤5 metastases) and polyprogressive disease patients Descriptive statistical analysis was performed with SPSS v25.0 statistic software package (IBM Corp., Armonk, NY, USA). Kaplan-Meier survival curves with log-rank analysis for comparison of subgroups was used to evaluate OS, PFS, and TTS. The Fisher`s exact and Chi-square test was used to compare differences between two independent groups. Univariate and backward multivariate Cox regression analysis was performed to identify independent variables for OS, PFS and TTS. A p-value of less than 0.05 was deemed statistical significant.

Patient characteristics
Data of 108 patients from 16 participating centers was included. Baseline patient, tumor and treatment characteristics are summarized in Table 1. Median patient age was 63 (range 33-80) years, 90% of patients had an ECOG performance score of ≤1, 41% had minimal comorbidities (age-adjusted Charlson Comorbidity Index ≤3). The most frequent histological NSCLC subtype was adenocarcinoma (80%). Multiorgan metastatic disease was present in 71% of patients, with a median of 2 (range 1-7) involved organs per patient. Reported reasons for MDT in PPD patients were palliation of symptoms (26%), prevention of future complications (79%), attempt to extend treatment with current systemic therapy (11%) and to induce a possible immunomodulation effect (9%).

Targeted therapy / Immunotherapy
Overall, 56% of patients received first line systemic therapy, while 44% were under secondline therapy or more (Table 1). Systemic therapy concurrent to MDT consisted of EGFR/ALKinhibitors (60%), immune checkpoint inhibitors (31%), VEGF-antibodies (8%) and PARPinhibitors (1%). In 10% of patients, these were combined with chemotherapy. Sixty-seven percent of patient started TT/IT before MDT, with a median of 269 days (range 1-180 days), 8% started IT/TT at the same time of SRT and 25% started IT/TT a median of 14 days (range 1-30) after SRT. Overall, in 28% of patients their systemic therapy was paused during MDT with a median of 10 (range 2-42) days. Targeted therapy was paused in 35% of the patients, for a median of 3 days before and 3 days after MDT (range 1-21 days). Immune checkpoint inhibitors were paused in 15% of patients, for a median of 7 days before and after MDT (range 1-19 days). Bevacizumab was paused in 20% of patients for a median of 14 days before up to 14 days after MDT (range 7-21 days).

Effectivity and factors influencing survival
Median follow-up was 18.7 (range 1-102) months. Median OS was 18.1 months, 2 year OS was 39% ( Figure 1). Cause of death was tumor-related in 85% and never therapy-related. There was no significant difference in OS between patients with-or without brain metastases (p=0.181). There was also no significant difference in OS between patients receiving IT, TT, AAT or PARPi (p=0.765). In univariate analysis, metastatic status (OPD vs. PPD) and number of affected organs were significant predictors of OS (Table 2). In multivariate analysis, metastatic status remained the only independent factor predicting OS (p=0.002, HR 2.03 (95% CI; 1.3-3.17). Local control after SRT was 84% after 2 years, there was no difference between OPD and PPD. Overall, median PFS was 8.7 months. In OPD, this was median 10.4 months and 7.4 months for PPD patients, or 25% and 8% after 2 years, respectively ( Figure 1). In the univariate analysis, metastatic status, previous lines of systemic therapy and affected organs were significant predictors of PFS. In multivariate analysis, the number of previous lines of systemic therapy (P=0.033, HR 1.7 (95%CI 1.05-2.77)) and number of affected organs (p=0.023, HR 2.04 (95% CI 1.10-3.79) remained independent factors predicting PFS ( Table 2).

Patterns of disease progression
In case of progression, only one organ was affected (range 1-4) in 69% of OPD patients and  (Table 3). No grade 5 toxicity occurred.

Discussion
This analysis of stage IV NSCLC patients receiving MDT for progressive or persistent metastases under targeted-, or immunotherapy showed survival rates in the OPD group that appear promising compared to literature on patients receiving TT/IT alone [15][16][17]. The concept of targeting OPD while continuing systemic therapy beyond progression is increasingly performed [18,19]. This is based on the observation that 1) further lines of systematic treatment are characterized by a worse therapeutic effect, 2) MDT obtains good results with limited toxicity in the primary oligometastatic situation [7][8][9], and 3) whole genomic sequencing studies showing that through parallel evolution, subpopulations of metastatic clones are capable to metastasize themselves [20]. This may indicate that MDT of these lesions could possibly improve prognosis. Literature on the concept of MDT for OPD is still limited and includes, next to small retrospective studies [2][3][4][5][6], a phase II study, which showed that MDT of ≤ 6 progressive metastases in platinum-refractory NSCLC patients who received erlotinib resulted in a better PFS and OS than would be expected in patients receiving TT alone [21]. In all available studies, metastatic patients in varying phases of their treatment were included, which increases the difficulty to interpret their results and transfer the data to clinical practice. Furthermore, the increasing use of immunotherapy in this population has so far not been taken into account. Our study therefore analyzes real-life data of the efficacy and safety of MDT performed in metastatic NSCLC patients progressive under targeted-, as well as immunotherapy.
MDT resulted in a good OS in OPD patients, and allowed continuation of systemic therapy within the first year for many patients. The effect of MDT in the patient group treated with palliative intent was less pronounced in terms of OS and PFS. However, these patients were most frequently treated with SRT for brain metastases. Some TT/IT are characterized by good penetration of the blood-brain barrier; however, upfront MDT of cerebral metastases might improve patient outcome and quality of life [22][23][24][25].
When a new progression occurred after MDT, this was often again OPD and a repeatirradiation was performed in most of these patients. The effectiveness of repeat-irradiations has been previously published for prostate cancer recurrences and treatment of brain metastases, generally resulting in an excellent local control with limited toxicity [22]. In our multivariate analysis, it was shown that the number of lines of systemic therapy influenced PFS. This reflects that the reducing efficacy of subsequent lines of therapy drives distant progression and appears to be more important than metastatic burden at time of MDT.
Repeat-MDT instead of therapy-switch may therefore play an increasingly important role.
However, experiences of repeat-irradiation remain limited and further studies need to investigate carefully the concept of repeat local treatment.
Although the concept of MDT for OPD under TT/IT is practiced with increasing frequency, several uncertainties remain. First of all, with the currently limited knowledge on the molecular background of resistant clones in NSCLC, it is not known whether preferably all metastases or just selected progressive metastases should be targeted, and what the threshold of the number and volume of targeted metastases should be [26]. Current decisionmaking is based on imaging and clinical criteria [11,27], but studies on the biological status of metastatic NSCLC is urgently needed, including integration of liquid biopsy for staging and response assessment. Secondly, the timing of MDT currently remains a clinical decision, usually based on CT or PET-images. However, progression under TT/IT can be very slow, or after immunotherapy, a pseudoprogression could be observed. Thirdly, there might be different strategies in combining systemic therapies with MDT, as the effects of concurrent treatment may be diverse. For example, combining immunotherapy with MDT could possibly strengthen the antitumor immune response [28], whereas abscopal effects of radiotherapy are exceedingly unlikely in patients not receiving immunotherapy [26].
A limitation of this study is its retrospective nature which, however, is a way of a meaningful evaluation of a quickly changing clinical field and associated limitations in standardization of reporting of factors, such as toxicity and local control. Furthermore, as all patients received a combined modality treatment, our study does not allow to evaluate the influence of the continued TT/IT alone on outcome. It may be possible, that patients with less metastatic sites have had a better prognosis irrespective of MDT. Especially under immunotherapy, residual sites potentially could remain stable for a longer time. However, for patients receiving targeted therapy a further progress of residual sites can be expected after several months. A first study comparing targeted therapy as monotherapy to a combined modality therapy with SRT indicated a benefit of combined therapy compared to targeted drugs alone [29] and will be further investigated in the randomized HALT (NCT03256981) and STOP-NSCLC NCT02756793 trials. Another limitation is the combined analysis of patients treated with IT/TT in one study population. This was done because the intention of MDT was similar irrespective of the systemic therapy, namely ablation of (oligo-) progressive metastases while the otherwise effective systemic therapy is continued.
In conclusion, metastases-directed SRT in NSCLC patients can be safely performed concurrent to TT/IT with a low risk of severe toxicity. To find the ideal sequence of the available multidisciplinary treatment options for NSCLC and determine what patients will benefit most, a further evaluated in a broader context within prospective clinical trials is needed continuation of TT/IT beyond progression combined with MDT for progressive lesions appears promising but requires prospective evaluation.

Ethics approval and consent to participate
This study was approved by the ethics committees at all participating sites (BASEC-Nr. 2016-01807). A informed consent of all participants was obtained.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.