Nitroglycerin Plus Intracranial Radiotherapy for Brain Metastases in NSCLC Patients

Non-small cell Lung cancer (NSCLC) is the most frequent type of lung cancer worldwide. Brain metastases (BM) are the most frequent neurological complications related to NSCLC and it is estimated that 20% to 40% of these patients will present them at some point during the progression of their disease, leading to a poor prognosis.

As only a selected group of NSCLC patients with single metastases or small lesions are candidates for surgical resection or SBRT, the standard treatment for multiple BM is whole-brain radiation therapy (WBRT). Although 80% of the patients who receive WRBT can initially respond to treatment, 50% of these patients will have disease progression. This further has been related to genetic and environmental factors that may lead to radio-resistance, which is the capacity of a cell to stand the effects of radiant energy. To face this problem, other options such as chemotherapy or radio-sensitizers have shown slight benefit in terms of progression-free survival (PFS) but not in overall survival (OS).

Different studies have suggested tumor microenvironment (TME) has an important role in treatment response of BM-derived from melanoma and NSCLC. The demand for O2 by tumors during cancer is an unsteady phenomenon, caused by continuous metabolic profile changes, immune response activity, and TME interactions. Hypoxia occurs in most tumors and plays an important role in tumor progression. In NSCLC exists a clear relationship between hypoxia and radioresistance. Hence, there is interest into modulate TME hypoxia to improve treatment response. Hypoxia regulates the expression of genes that encode growth factors such as endothelin-1 (ET-1), growth factor-derived platelets-B (PDGF) and vascular endothelial growth factor (VEGF), as well as genes that regulate the production of gas molecules such as nitric oxide (NO) and carbon monoxide (CO). The hypoxia-inducible factor (HIF) is a transcription factor that regulates the cellular response to hypoxia, functioning as a regulator of oxygen homeostasis. Several studies have shown that over-expression of HIF-1α is capable of inducing resistance to chemotherapy, radiotherapy and decreasing overall survival in NSCLC primary tumors.

The administration of nitric oxide donors, such as nitroglycerin, reduces the tumor resistance related to hypoxia by inducing the direct proteolysis of HIF-1α. In tumor cells increases oxygen pressure, increased blood flow, activation of p53 and apoptosis have shown a synergistic effect with ionizing irradiation in an experimental model. According to several randomized phase II studies, the addition of transdermal nitroglycerin to vinorelbine and cisplatin treatment can significantly improve the OS and time to progression in patients with locally advanced non-small cell lung cancer. However, in phase III the addition of nitroglycerin to carboplatin-based did not show benefit in PFS, OS and health-related quality of life (PMID: 26347110). Therefore, the aim of this study was to assess, if the addition of transdermal NTG to standard WRBT treatment among stage IV NSCLC patients with BM, could have a significant impact in PFS and OS as primary end-point and disease control rate (DCR) and overall response rate (ORR) as secondary endpoints.

METHODS Experimental design We conducted a phase II clinical trial with a parallel design study among patients treated at the Instituto Nacional de Cancerología from January 2014 to May 2017. The Median follows up was 18 months. Patients with histologically confirmed NSCLC and documented BM defined as the presence of one or more intra-axial enhancing lesions on gadolinium-enhanced brain magnetic resonance imaging (MRI) were included. Candidate patients who underwent radiosurgery or surgical resection were excluded.

Sample size. The sample size calculation was estimated for a two-sample proportion difference in ORR for NSCLC patients who received WBRT plus TN. Although in one of our previous studies we found a difference of 40% in ORR among locally advanced NSCLC patients 30, we prefer to be more conservative in our estimations taking into account our different study populations. Thus, our sample estimation assumptions were performed for a difference in ORR (delta) of 30% between population (p1= 25% vs p2= 65%) with a study power was set at 0.80 and the two-sided type I error (alpha) was set as 0.05. Therefore, the estimated sample number is 108 (54 per group). We also add a 5% of extra patients, to prevent the possible loss of patients due to the progression of the disease.

Endpoints. The primary outcome was overall response rate (ORR); Secondary endpoints were disease control rate (DCR), intracranial progression-free survival (ICPFS) and overall survival (OS). The radiographic response was assessed by two independent blinded radiologists according to the Response Evaluation Criteria in Solid Tumors (RECIST) guideline version 1.1 by comparing the pre-and post-treatment images. Any in-field tumor progression or the appearance of new malignant lesions denoted progressive disease. ORR was defined as the sum of complete and partial response, whereas DCR was defined as the sum of ORR and stable disease. For PFS, time to event was defined and calculated from the date of randomization until radiographic disease progression, treatment discontinuation due to either unacceptable toxicity death by any cause. For OS, time-to-event was defined as time from randomization until death by any cause or loss to follow-up. Observations for patients who did not experience the event were censored at patient-specific last follow-up. Finally, treatment toxicity was evaluated with CTCAE criteria.

Intervention assignment and informed consent Patients were randomized 1:1 and allocated to either the control arm who received whole-brain radiotherapy (WBRT) (30 Gy in 10 fractions, in 10 days of treatment) or the experimental arm who received WBRT and the addition 36 mg of transdermal nitroglycerin (TN) with release of 10 mg in 24 hours, for 24 hours with a 12-hour rest interval (to avoid saturation of receptors). All patients signed the informed consent letter, prior to any procedure. The project was approved by the local scientific and bioethics committee (number). Further details can be found in clinicaltrials.gov (NCT¿?). Patients received chemotherapy (CT) platinum-based or TKI according to mutation status.

Response assessment BM was documented among all patients at NSCLC diagnosis and staging by either computed tomography (CT) of the head gadolinium-enhanced brain magnetic resonance imaging (MRI) at baseline. Then, all patients underwent an MRI prior to treatment and at the end of treatment (median 15 days). The imaging studies were carried out in a Signa HDxt 1.5 T scanner (GE Healthcare). They include conventional sequences in multiple planes, T1, T2, T2 Flair, Echo gradient, Diffusion, volumetric acquisition in axial plane T1 Spoiled Gradient recalled (SPGR), without contrast and after this (Gadolinium). Cerebral perfusion was evaluated by echo-planar sequence with the following parameters: multiphase (25 - 45 phases) with an acquisition time of 1:10, with a PSI of 5 and a flow rate of 5; with a volume of 15ml of gadolinium (+) 20ml of saline at a speed of 5ml/second.

The radiographic response of intracranial tumors was assessed by two blinded and independent blinded radio-oncologist (MY \& FM) according to the Response Evaluation Criteria in Solid Tumors (RECIST) guideline version 1.1 by comparing the pre-and post-treatment intracranial images [15]. Any in-field tumor progression or the appearance of new malignant lesions denoted progressive disease. Objective response was defined as the sum of complete and partial response.

DNA Extraction Biopsies were taken using CT-guided tru-cut or bronchoscopy and were analyzed by the pathology department for their histologic diagnosis and neoplastic cellularity quantification (>50%); they were later embedded in paraffin until processed for DNA extraction. Genomic DNA was extracted from the areas of paraffin slides using a standard procedure and a QIAamp DNA FFPE tissue kit (TMQIAGEN), following manufacturer's instructions.

Determination of EGFR and KRAS mutational status EGFR exon 19, exon 20 and exon 21 gene mutations were detected using the Therascreen RGQ PCR kit (TMQIAGEN, Scorpions ARMS method), which combines both the ARMS and Scorpions technologies for detecting the mutations by real-time polymerase chain reactions (PCR). Real-time PCR was performed using a Rotor-Gene Q 5plex HRM (TMQIAGEN), following manufacturer's instructions.

Statistical analysis Continuous variables were summarized as arithmetic means or medians, with standard deviation or interquartile range for descriptive purposes according to normal data distribution assessed by means of the Shapiro-Swilk test. Meanwhile, categorical variables were summarized as frequencies and percentages. For dichotomous outcomes (eg. objective response rate (ORR), disease control rate (DCR), the percentage (incidence rate) and 95% CI are presented. Inferential comparisons were made using the T-test or two-way ANOVA for continuous variables and with either Mann-Whitney U test or Kruskall-Wallis test, conforming to the data distribution and a number of groups. The χ2 test or Fisher exact test was used for assessing the statistical significance of categorical variables. To address the effect of treatment of secondary outcomes (eg. OS, ICPFS) we performed univariate survival analysis. Time-to-event was estimated using the Kaplan-Meier method and comparisons among the subgroups were analyzed using the log-rank test. For survival curve analysis, all the variables were dichotomized according to their median. After we performed a forward stepwise Multivariable Cox regression model and hazard ratios (HR) were calculated along with their corresponding 95% CIs as a measure of association. Statistically and clinically significant and borderline significant variables (p <.10) were included for the adjustment in the multivariate Cox regression model. Significantly Kaplan-Meir curves were plotted. Statistical significance was determined as P <0.05 using a 2-tailed test. Stata software version 14 was used for all statistical analyses.