NOS2 polymorphisms in prediction of benefit from first-line chemotherapy in metastatic colorectal cancer patients

Background Macrophages play a crucial role in the interaction between tumor and immune system, and iNOS is known as a surrogate marker of M1 macrophages activation. The goal of the study was to investigate the role of iNOS polymorphisms as prognostic marker in mCRC patients. Materials and methods Functional significant polymorphisms in the promoter of INOS gene were analyzed by PCR-based and direct DNA sequencing in 4 cohorts of patients receiving bevacizumab based first-line chemotherapy: two evaluation cohorts (TRIBE ARM A and ARM B) and two validation cohorts (FIRE 3 arm A and MOMA). The relation of the SNPs with PFS and OS was evaluated through Kaplan-Meier method and log-rank test. Subgroup analyses according to RAS status were preplanned. Results In the exploratory cohort 1 (TRIBE A), patients with CCTTT any>13repeats (N = 57) showed improved median PFS compared with patients carrying the ≤13/≤13 repeats variant (N = 152) (HR, 0.64; 95%CI 0.44–0.92, p = 0.010). Similar results were shown adopting the >26repeats/≤26 repeats (HR, 0.56; 95%CI 0.36–0.87, p = 0.005). In RAS mutant, patient with any>13 repeats (N = 24) had improved PFS results compared with those carrying the ≤13/≤13 repeats variant (N = 81) (HR, 0.51; 95%CI 0.30–0.87, p = 30.009). Similar results were found adopting the >26 repeats/≤26 repeats cut off: (HR, 0.52; 95%CI 0.27–0.98, p = 0.035). These data were partially confirmed in the exploratory cohort 2 (TRIBE B): a better median PFS was observed in patients with >26 repeats vs ≤26 repeats (N = 205) patients. However, these data were not confirmed in the two validation cohorts. Conclusion We failed to replicate the exploratory findings in both validation sets. The CCTTT polymorphic region of the INOS gene does not predict outcome in mCRC receiving bevacizumab based first line chemotherapy. Further investigations are needed to reveal mechanisms between tumor, immune system and chemotherapy response.


Materials and methods
Functional significant polymorphisms in the promoter of INOS gene were analyzed by PCRbased and direct DNA sequencing in 4 cohorts of patients receiving bevacizumab based first-line chemotherapy: two evaluation cohorts (TRIBE ARM A and ARM B) and two validation cohorts (FIRE 3 arm A and MOMA). The relation of the SNPs with PFS and OS was evaluated through Kaplan-Meier method and log-rank test. Subgroup analyses according to RAS status were preplanned.

Introduction
Significant progresses have been made over the last years in the treatment of mCRC. Currently first-line therapy of patients affected by mCRC is based on the use of a combination of cytotoxics with monoclonal antibodies either anti-VEFG (bevacizumab) or anti-EGFR (cetuximab, panitumumab). Among cytotoxics, fluoropirimidine-based treatments represent the standard of care as monotherapy, doublet combinations (FOLFOX, XELOX or FOLFIRI) or triplet combination (FOLFOXIRI) [1][2][3][4][5][6].
In the last years, immunotherapy showed promising results in solid tumors especially in lung cancer and in melanoma, demonstrating an overall survival (OS) and progression-free survival (PFS) benefit in several trials [7,8]. The role of immunotherapy in mCRC is currently under investigation. The anti-PD1 antibodies pembrolizumab and nivolumab and the combination of nivolumab plus the anti-CTLA4 ipilimumab showed promising results in ongoing clinical trials in a subgroup of patients with microsatellite instability in advanced lines of treatment [9,10]. Up today, many efforts are ongoing, firstly, in order to identify a role for immunotherapy in earlier lines of treatment and its possible applications in combination with chemotherapy; secondly to unveil, biomarkers leading to identify a wider range of patients possible benefitting from this strategy [11][12][13][14].
The interaction of tumor cells with stroma is closely connected with immune modulation and represents an appealing research field. In this complex mechanism macrophages play a crucial role in the balance of pro and anti tumorigenic stimuli. In particular, macrophages derived from monocyte precursors undergo specific differentiation depending on the local tissue environment [15].
The classically activated M1 macrophages are characterized by the production of high levels of pro-inflammatory cytokines, high production of reactive nitrogen and oxygen intermediates, and the promotion of Th1 responses, through the upregulation of inducible nitric oxide synthase (NOS2 or iNOS) resulting in tumoricidal effect [15][16][17][18].
INOS is a NADPH-dependent enzyme catalyzing the production of nitric oxide (NO) from L-arginine [19] and is known as a surrogate marker of M1 macrophages activation [20,21]. Its relation with colorectal cancer has been investigated and high levels of NOS2 expression where related to colorectal cancer progression and development and with poor prognosis [22]. A possible explanation for such phenomenon is based on the relationship between high macrophage expression and increase of VEGF production and stimulation of angiogenesis and tumor growth [23,24]. Several polymorphisms have been identified in this gene and many studies identified a highly polymorphic pentanucleotide (CTTTT)n repeat, at position-2.6 kb in the promoter of iNOS [25] as possible a surrogate of NOS2 expression [26,27].
Based on the above reported considerations, we hypothesized that iNOS polymorphisms might allow clarifying mechanisms related to the interaction between immune system and CRC.and might have a prognostic role in mCRC patients receiving bevacizumab-based first line treatment in three large and modern phase II/III clinical trials.

Patients' population
This study investigated 4 independents cohorts of mCRC patients receiving bevacizumab based first-line treatment: The above-mentioned studies were carried out between 2007 and 2015 in Italy, Germany and Austria. Main inclusion criteria included: first occurrence of metastatic disease in patients with histological diagnosis of adenocarcinoma of colon or rectum. Further details on inclusion and exclusion criteria have previously been described [28][29][30].
In TRIBE ARM A FOLFIRI consisted of a 180 mg/m2 intravenous infusion of irinotecan followed by a 200 mg/m2 intravenous infusion of leucovorin, a 400 mg/m2 intravenous bolus of fluorouracil, and a 2400 mg/m2 continuous infusion of fluorouracil for 46 h. In TRIBE ARM B and MOMA FOLFOXIRI consisted of a 165 mg/m2 intravenous infusion of irinotecan, followed by an 85 mg/m2 intravenous infusion of oxaliplatin given concurrently with 200 mg/m2 leucovorin for 120 min, followed by a 3200 mg/m2 continuous infusion of fluorouracil for 48 h. Treatment was administered every 2 weeks for up to 12 cycles followed by maintenance treatment with fluorouracil, leucovorin until disease progression, intolerable toxicities, or patient withdrawal. Tumour assessment by CT scan was done every 8 weeks until disease progression.
This study was approved by The Health Sciences Review Board (HSIRB) -The Office for the Protection of Research Subjects 3720 South Flower Street, Third Floor Los Angeles, CA 90089-0706. All patients signed an informed consent.
• functional or predicted functional relevance of gene transcription or protein expression. Functional significance was predicted based on information provided by the National Institute of Environmental Health Science SNP Function Prediction, Queen's University F-SNP, and the location of the SNP in the protein-coding region of the gene (http://snpinfo.niehs. nih.gov/snpinfo/snptag.htm).
Genomic DNA was extracted from peripheral whole blood in TRIBE and MOMA trials and from formalin-fixed paraffin-embedded (FFPE) tissues in FIRE-3 using the QIAmp Kit (Qiagen, Valencia, CA, USA) according to the manufacturer's protocol (www.qiagen.com).
PCR-based direct DNA sequence analysis using ABI 3100A Capillary Genetic Analyzer and Sequencing Scanner v1.0 (Applied Biosystems, Waltham, MA, USA) was performed for genotyping the SNPs. The extracted DNA was amplified using the primer sets shown in Table 1, and analyzed by PCR-based direct DNA sequencing. For quality control purposes, a random selection of 10% of the samples were re-examined for each polymorphism, and the genotype concordance rate was 100%. The investigator analyzed the sequencing data using the ABI Sequencing Scanner v1.0 (Applied Biosystems, Life Technologies, Grand Island, NY, USA) and was blinded to the clinical data set.
For the primary end point analysis, NOS2 CCTTT repeats cut-offs were defined based on the maximum chi-square method for PFS and categorized as 13/ 13 repeats for patients with 13 repeat in each allele and any >13 repeats for patients with at least on allele with >13 repeats.
Exploratory analyses were conducted adopting the sum of repeats in each allele and the predefined cut off was 26 repeat/ >26 repeats.
Hardy-Weinberg Equilibrium was assessed by means of Chi-square test.

Statistical analysis
The primary aim of the study was the identification of a prognostic/predictive role for the selected SNPs (NOS2 rs27779248 and CCTTT repeat) in mCRC patients treated with first-line chemotherapy with bevacizumab. Primary endpoint was PFS. The secondary endpoints were OS and tumor response. Table 1 shows SNPs characteristics and adopted primers. The distribution of patient baseline clinicopathologic characteristics were compared using the Chi-square test between the four cohorts.
The impact of NOS2 SNPS in relation with FOLFIRI plus bevacizumab treatment was evaluated in TRIBE arm A and in FIRE-3 as exploratory and validation cohorts respectively.
The impact of NOS2 SNPS in relation with the addiction of oxaliplatin was evaluated in TRIBE arm B and in MOMA as exploratory and validation cohorts respectively.
The relation of the NOS2 SNPs with PFS and OS was evaluated through Kaplan-Meier method and log-rank test in the overall population. OS was defined as the time from randomization to death due to any cause, PFS was defined as the time from randomization to first documented disease progression or death due to any cause. Patients were censored at the time of last follow-up if no event observed. Patients were defined as responders when achieving complete or partial response and non-responders when stable or progressive disease occurred as defined by RECIST 1.0 criteria. The relation of the NOS2 SNPs with PFS and OS was evaluated through Kaplan-Meier method and log-rank test in the overall population. Multivariable Cox regression model was used to assess the independent effect of two SNPs on PFS and OS when adjusting for the baseline demographic, clinical, and tumor characteristics that were associated with outcomes. Correlation of SNPs with response was estimated by the multivariable logistic regression, adjusting for the same baseline characteristics. Subgroup analyses according to RAS status were preplanned.
With 209 patients (163 PFS events) in the exploratory cohort 1, we would have 80% power to detect a minimum hazard ratio (HR) of 0.59 to 0.65 on PFS for a SNP with a minor allele frequency from 0.1 to 0.5 using a two-sided 0.05 level log-rank test. When applied the same model and test, we would have at least 81%, 89%, and 74% power to detect the same HR in the exploratory cohort 2 (225 patients, 161 PFS events), validation cohort 1 (280 patients, 239 PFS events), and validation cohort 2 (178 patients, 125 PFS events), respectively.
All statistical analyses were performed by SAS 9.4 (SAS Institute, Cary, NC, USA). All tests were two sided at a significant level of 0.05.

Results
Baseline clinicopathologic characteristics in the four cohorts are summarized in Table 2. Differences in terms of age, primary tumor site, primary tumor resection, performance status, KRAS and RAS status were identified across patients enrolled in the trials ( Table 2).

TRIBE ARM A-Exploratory cohort 1
Associations of clinicopathologic characteristics of exploratory cohort 1 with outcome are presented in S1 Table. Median follow up was 50.6 months; median PFS and OS were respectively 9.5 months and 25.8 months. Resection of the primary tumors, ECOG performance status, and BRAF status were significantly associated with PFS and OS. Furthermore age, primary tumor site, time to metastases, and previous adjuvant chemotherapy significantly correlated with OS (S1 Table).
Genotyping results. Associations between NOS2 SNPs and outcomes in exploratory cohort 1 are summarized in Table 3.

TRIBE ARM B-Exploratory cohort 2
Patients characteristics and association with outcome for exploratory cohort 2 are presented in S2 Table. The median follow up was 46.4 months; median PFS and OS were 11.7 and 30.3 months, respectively. Time to metastases and resection of the primary tumors were significantly associated with PFS and OS. Primary tumor site, number of metastases, ECOG performance status, BRAF status were also significantly associated with OS (S2 Table). Genotyping results. Associations between CCTTT repeats and outcomes in exploratory cohort 2 are summarized in Table 3.
Subgroup analyses: among RAS mutant patients no significant outcome differences in PFS were observed when adopting the any > 13 vs 13/ 13 repeats cut off (HR for univariate,  (Table 4).

FIRE -3-Validation cohort 1
Clinicopathologic characteristics and associations with outcome for validation cohort 1 are presented in S3 Table. The median follow up was 40.8 months; median PFS and OS were 10.3 and 24.7 months, respectively. Primary tumor site, ECOG performance status, and BRAF status were significantly associated with PFS and OS. Additionally, number of metastatic sites, liver limited disease, and time to metastasis were also significantly associated with OS (S3 Table). Genotyping results. Associations between CCTTT repeats and outcomes in validation cohort 1 are summarized in Table 3.

MOMA-Validation cohort 2
Associations between clinical features and outcomes in validation cohort 2 are summarized in S4 Table. The median follow up was 25.3 months, median PFS and OS were respectively 9.2 and 25.3 months. Resection of the primary tumors and ECOG performance status were significantly associated with PFS and OS; liver limited disease was also associated with OS (S4 Table).
Genotyping results. Associations between CCTTT repeats and outcomes in exploratory cohort 2 are summarized in Table 3.

Discussion
The identification of prognostic markers in mCRC patients, and in particular in the first-line treatment setting represents a crucial issue in modern clinical oncology. Prognostic factors help identifying patient subgroups with peculiar tumor phenotypes requiring personalized NOS2 polymorphisms in metastatic colorectal cancer patients treatment strategies. Thanks to the recent progresses and emerging data on immunotherapeutic treatment strategies for mCRC, the interaction of immune system with colorectal cancer is becoming an appealing research field and immune-related markers are emerging as possible new prognostic or predictive factors in this setting.
For the first time a polymorphic region in the promoter of NOS2 gene has been analyzed as surrogate marker of M1 macrophages activity and possible prognostic marker in patients receiving a first-line treatment with bevacizumab enrolled in modern, randomized phase III and II clinical trials [32][33][34]. We specifically selected patients receiving anti-angiogenetic drugs in modern clinical trials due to the strong association of macrophage activity and angiogenesis regulation.
The exploratory cohorts involved patients treated in the phase III TRIBE trial receiving first line treatment with FOLFIRI + bevacizumab (exploratory cohort 1) or FOLFOXIRI + bevacizumab ((exploratory cohort 2). Based on the already observed correlation of high number of NOS2 repeats with increased NOS2 expression, we hypothesized a better outcome in patients NOS2 polymorphisms in metastatic colorectal cancer patients with higher number of NOS2 repeats. The role of NOS2 repeats was firstly identified in the cohort 1 but only partially confirmed in the cohort 2.
To challenge our data, we identified two validation cohorts: one treated with FOLFIRI + bevacizumab in the FIRE-3 study and the other treated with FOLFOXIRI + bevacizumab in the MOMA study. No significant clinical differences were observed among the additional 2 cohorts of patients, thus we failed to replicate the exploratory results in both validation sets. Based on the power analyses calculation, no major concerns on sample size can be raised.
In the present study we moved from a strong biological rationale [35,36] and we selected a relatively innovative marker, since only a few studies focused on NOS2 polymorphisms and their connection with M1 macrophage tumoral infarction [37]. Moreover nitric oxide is tightly involved in angiogenesis regulation and tumor growth [38].
As additional strengthen points we have to consider that more than 900 patients were included in the present study; study population presented comparable clinical features and was enrolled in practice changing clinical trials of first line chemotherapy for mCRC. The adoption of a rigorous study design with 2 validation cohorts increases the relevance of our findings.
As limitation points, we have to mention the lack of confirmatory preclinical data obtained from our samples, supporting the correlation between number of CCTTT repeats and iNOS expression. Unfortunately, adequate tumoral tissue was not available to solve this issue.
As possible explanation of our negative result, we have to consider that a single SNP might not catch the complexity of interactions between tumor cells and microenvironment that are regulated both by immune related factors and angiogenesis related mediators. Moreover we have to acknowledge that other sources of NO production might increase the complexity of the scenario.
In particular NO is produced not only by NOS2 but also by the endothelial NOS which plays a central role in maintaining endothelial cell functional integrity. [19]. Previous studies correlated VEGF inhibition mediated by bevacizumab with a decrease in eNOS expression and thus in NO production [21].
However, despite the negative results, our findings underline the importance of investigating mechanisms regulating the interactions of tumor development, chemotherapy response and immune system in order to identify new immunotherapeutic strategies effective in patients receiving chemotherapy.
Supporting information S1