Evaluation of Second-Line Anti-VEGF after First-Line Anti-EGFR Based Therapy in RAS Wild-Type Metastatic Colorectal Cancer: The Multicenter “SLAVE” Study

Background: The optimal anti-angiogenic strategy as second-line treatment in RAS wild-type metastatic colorectal cancer (mCRC) treated with anti-EGFR (Epidermal Growth Factor Receptor) based first-line treatment is still debated. Methods: This multicenter, real-world, retrospective study is aimed at evaluating the effectiveness of second-line Bevacizumab- and Aflibercept-based treatments after an anti-EGFR based first-line regimen. Clinical outcomes measured were: objective response rate (ORR), progression free survival (PFS), overall survival (OS) and adverse events (AEs) profiles. Results: From February 2011 to October 2019, 277 consecutive mCRC patients received Bevacizumab-based (228, 82.3%) or Aflibercept-based (49, 17.7%) regimen. No significant difference was found regarding ORR. The median follow-up was 27.7 months (95%CI: 24.7–34.4). Aflibercept-treated group had a significantly shorter PFS compared to Bevacizumab-treated group (5.6 vs. 7.1 months, respectively) (HR = 1.34 (95%CI: 0.95–1.89); p = 0.0932). The median OS of the Bevacizumab-treated group and Aflibercept-treated group was 16.2 (95%CI: 15.3–18.1) and 12.7 (95%CI: 8.8–17.5) months, respectively (HR= 1.31 (95%CI: 0.89–1.93) p = 0.16). After adjusting for the key covariates (age, gender, performance status, number of metastatic sites and primary tumor side) Bevacizumab-based regimens revealed to be significantly related with a prolonged PFS (HR = 1.44 (95%CI: 1.02–2.03); p = 0.0399) compared to Aflibercept-based regimens, but not with a prolonged OS (HR = 1.47 (95%CI: 0.99–2.17); p = 0.0503). The incidence of G3/G4 VEGF inhibitors class-specific AEs was 7.5% and 26.5% in the Bevacizumab-treated group and the Aflibercept-treated group, respectively (p = 0.0001). Conclusion: Our analysis seems to reveal that Bevacizumab-based regimens have a slightly better PFS and class-specific AEs profile compared to Aflibercept-based regimen as second-line treatment of RAS wild-type mCRC patients previously treated with anti-EGFR based treatments. These results have to be taken with caution and no conclusive considerations are allowed.


Introduction
With the exception of intensive first-line regimens [1,2], it is now been years that the treatment algorithm of metastatic colorectal cancer (mCRC) patients includes a backbone of fluoropyrimidine-based chemotherapy combined with either oxaliplatin or irinotecan for the first-line approach, followed by the alternative regimen for the second-line treatment. EGFR (Epidermal Growth Factor Receptor) antibodies (Panitumumab and Cetuximab) or anti-angiogenic agents (Bevacizumab, Aflibercept, and Ramucirumab) (Vascular endothelial growth factor [VEGF] pathway inhibitors) are added to these backbones across treatment lines, according to the RAS genotype [3]. However, the optimal use and sequencing of these agents has yet to be determined [4].
RAS wild-type mCRC patients represent about 40-50% of the overall mCRC population [5] and a common first-line treatment strategy for these patients includes the combination of chemotherapy with anti-EGFR agents [6][7][8][9]. A growing amount of evidences, derived from both retrospective and phase I-II prospective studies, highlights the possibility to obtain clinical benefit from continuing EGFR inhibitors after first-line disease progression in a subset of molecularly selected mCRC patients [10]. However, to date, according to ESMO guidelines [11], the recommended second-line options after an anti-EGFR based first-line treatment include both Bevacizumab-based and Aflibercept-based regimens. The efficacy of Bevacizumab in the second-line setting was assessed in two phase III studies (E3200 and ML18147), which respectively analyzed the effect of adding Bevacizumab to FOLFOX in anti-angiogenesis naïve patients previously treated with FOLFIRI [12], and the efficacy of maintaining Bevacizumab across multiple lines of treatment [13]. On the other hand, the efficacy of Aflibercept was assessed in a phase 3 trial (VELOUR), which analyzed the effect of adding Aflibercept to FOLFIRI as a second-line treatment in mCRC patients progressed to an oxaliplatin-containing regimen, including patients who had previously received Bevacizumab [14]. Therefore, the use of Aflibercept in clinical practice is limited to patients previously treated with oxaliplatin and in combination with an irinotecan-containing regimen. To date, no head to head clinical trial compared Bevacizumab and Aflibercept as second-line treatment in RAS wild-type mCRC patients.
The present study is aimed at evaluating the effectiveness of second-line Bevacizumab-based and Aflibercept-based treatments after a first-line anti-EGFR based regimen in RAS wild-type mCRC patients in a multicenter real-world cohort.

Patient Eligibility
This retrospective analysis evaluated consecutive RAS wild-type mCRC patients, treated with either Bevacizumab-based or Aflibercept-based systemic therapy, at medical oncology department of 13 Italian and one Spanish institutions (Table S1), from February 2011 to October 2019.
Eligibility criteria were: age ≥ 18 years; histologically confirmed diagnosis of CRC; measurable metastatic disease; confirmed KRAS (exons 2, 3, 4) and NRAS (exons 2, 3, 4) wild-type genotype; having received an anti-EGFR-based (Panitumumab or Cetuximab) first-line treatment (fluoropyrimidines and/or oxaliplatin and/or irinotecan) and an anti-VEGF based (Bevacizumab or Aflibercept) second-line treatment (fluoropyrimidines and/or oxaliplatin and/or irinotecan) at disease progression. All patients alive at the time of data collection provided informed consent to participate to this retrospective observational non-interventional study. The procedures followed were in accordance with the precepts of good clinical practice and the Declaration of Helsinki. The study was approved by the respective local ethical committees on human experimentation of each institution, after previous approval by the coordinating center (University of L'Aquila, Internal Review Board protocol number 55741, approved on 11 October 2019). The datasets used during the present study are available from the corresponding author upon reasonable request.

Study Design
This is a retrospective, multicenter, observational study, aimed at evaluating the effectiveness of second-line treatments according to the anti-angiogenic regimen received (Bevacizumab-based and Aflibercept-based regimens) in consecutive patients.
The measured clinical outcomes were objective response rate (ORR), progression free survival (PFS), overall survival (OS) and cumulative toxicity. Patients were assessed with radiologic imaging according to the local clinical practice of the participating centers; disease responses were evaluated with the RECIST criteria (version 1.1) [15]. ORR was defined as the portion of patients experiencing an objective response (complete response or partial response) as best response, according to RECIST criteria (version 1.1) [15]. PFS was defined as the length of time from the beginning of second-line treatment to disease progression or death resulting from any cause or to the last contact [16]; OS as the length of time between the beginning of second-line treatment to death resulting from any cause or to the last contact [16]. For PFS as well as for OS, patients without events were considered as censored at the time of the last follow-up. The data cut-off period was January 2020.
Cumulative toxicity, defined as the maximum grade of toxicity experienced was registered according to National Cancer Institute Common Terminology Criteria (NCI-CTC) for Adverse Events (AEs) (version 4 up to January 2018, version 5 from January 2018) and grouped according to severity (grade [G] 1-2 and 3-4). Toxicities were summarized and compared among subgroups according to three key subgroups: VEGF inhibitors class-specific AEs (hypertension, arteriovenous thromboembolic events, fistulae, gastrointestinal perforation, proteinuria, bleeding), hematologic AEs (leukopenia, neutropenia, anemia, thrombocytopenia), and non-hematologic AEs (nausea, vomiting, diarrhea, asthenia, anorexia, mucositis, hand-foot syndrome). Only AEs which occurred in more than 5% of patients were included in the safety analysis.

Molecular Profile Assessment
All the molecular analyses were performed according to the local clinical practice of the participating centers. KRAS, NRAS and BRAF mutational status was assessed with Sanger sequencing, real-time PCR techniques and next-generation sequencing (NGS) (such as: OncoGenBasic-S1 kit, Seqplexing (Valencia, Spain); Pyromark Q96 ID System, Qiagen (Hilden, Germany); EasyPGX and Myriapod Colon Status, Diatech Pharmacogenetics (Jesi, Italy)). MSI (microsatellite instability) status and/or MMR (mismatch repair) proteins expression were assessed with molecular sequencing (Sanger, Real-Time PCR and NGS) and Immunohistochemistry (IHC) (such as: Applied Biosystem 3500 DX genetic analyzer, Thermo Fisher Scientific (Waltham, MA, USA); Ultraview Universal Detection Kit and Ventana platform, Roche Tissue Diagnostics and Ventana Medical Systems (Tucson, AZ, USA)).

Statistical Analysis
Baseline patients' characteristics were reported with descriptive statistics and compared among subgroups with the Chi-square test. Chi-square test was also used to compare ORR and the incidence of AEs across subgroups. Logistic regression was used for the multivariate analysis of ORR. Median PFS and median OS were evaluated using the Kaplan-Meier method. Median period of follow-up was calculated according to the reverse Kaplan-Meier method. Cox proportional hazards regression was used for the univariate and multivariate analysis of PFS and OS. The alpha level for all analyses was set to p < 0.05. Hazard Ratios (HRs) with 95% confidence intervals (CIs) were calculated using the logistic regression model. All statistical analyses were performed using MedCalc Statistical Software version 18.11.3 (MedCalc Software bvba, Ostend, Belgium; http://www.medcalc.org; 2019).

Patients Characteristics
A total of 277 consecutive RAS wild-type mCRC patients were treated with Bevacizumab-based (228, 82.3%) or Aflibercept-based (49, 17.7%) second-line regimens. The median age was 64.5 years (range: 29-84). Patients features (overall and according to subgroups) are summarized in Table 1. A significantly higher rate of primary tumor resection was reported for the Bevacizumab-treated group (78.9%), compared to the Aflibercept-treated group (49%) (p < 0.0001). According to the clinical indication of Aflibercept, also the previously received first-line regimens (p < 0.0001) and second-line chemotherapy backbone (p = 0.0148) were significantly different.

Clinical Outcomes Analysis
The activity profile for the overall population and according to subgroups is summarized in Table 2. In the overall population the ORR was 25.8%. No significant ORR difference was found between patients who received Bevacizumab-based and Aflibercept-based regimens.

Toxicity Analysis
The toxicity profile for the overall study population and according to subgroups is summarized in

Toxicity Analysis
The toxicity profile for the overall study population and according to subgroups is summarized in Table 5. The incidence of G1/G2 VEGF inhibitors class-specific AEs was 23.7% and 32.7% in the Bevacizumab-treated group and in the Aflibercept-treated group, respectively (p = 0.1908). The incidence of G3/G4 VEGF inhibitors class-specific AEs was 7.5% and 26.5% in the Bevacizumab-treated group and in the Aflibercept-treated group, respectively (p = 0.0001) (Figure 2). The incidence of G1/G2 non hematologic AEs was 36.4% and 59.2% in the Bevacizumab-treated group and in the Aflibercept-treated group, respectively (p = 0.0033), while the incidence of G3/G4 non hematologic AEs was 4.4% and 10.2%, respectively (p = 0.1032). The incidence of G1/G2 hematologic AEs was 24.6% and 22.4% in the Bevacizumab-treated group and in the Aflibercept-treated group, respectively (p = 0.7545), while the incidence of G3/G4 hematologic AEs was 3.1% and 18.4%, respectively (p < 0.0001). Table 5. Adverse events in overall, Bevacizumab-based and Aflibercept-based population.

Maintenance Regimens and Post-Progression Treatments
A total of 67 patients (29.4%) and nine patients (18.4%) underwent a maintenance therapy after an induction phase, in the Bevacizumab-treated group and in the Aflibercept-treated group, respectively (p = 0.2236). A total of 136 patients (69.4%) and 24 patients (60%) were treated with a third-line systemic therapy, among those who discontinued second-line treatment in the Bevacizumab-treated group and Aflibercept-treated group, respectively (p = 0.5930). Table 6 summarized maintenance treatments characteristics, causes of second-line discontinuation and third-line treatments.

Maintenance Regimens and Post-Progression Treatments
A total of 67 patients (29.4%) and nine patients (18.4%) underwent a maintenance therapy after an induction phase, in the Bevacizumab-treated group and in the Aflibercept-treated group, respectively (p = 0.2236). A total of 136 patients (69.4%) and 24 patients (60%) were treated with a third-line systemic therapy, among those who discontinued second-line treatment in the Bevacizumab-treated group and Aflibercept-treated group, respectively (p = 0.5930). Table 6 summarized maintenance treatments characteristics, causes of second-line discontinuation and thirdline treatments.  Table 6. Second-and third-line treatment characteristics in overall, Bevacizumab-based and Aflibercept-based population.

Discussion
This observational retrospective study intends to provide further data outside the clinical trial framework. To the best of our knowledge this is the first study aimed at comparing the effectiveness of Bevacizumab-based and Aflibercept-based second-line regimens in RAS wild-type mCRC patients. Moreover, the phase III E3200, ML18147 and VELOUR studies enrolled patients who had not previously received EGFR inhibitors [12][13][14], therefore, little is known about the clinical outcomes with Bevacizumab and Aflibercept in this setting.
Findings from preclinical studies showed that acquired resistance to EGFR inhibitors derives from the emergence of novel mutations in the RAS protein family and that KRAS mutant isoforms could be a VEGF expression inducer, which in turn is targetable by anti-angiogenic treatments [24][25][26][27]. Data from the first-line setting further suggest that an EGFR-based first-line therapy might create a favorable precondition for second-line treatments with VEGF-targeted antibodies [28], particularly in left-sided colon cancer [29]. Regarding the sequential use of Bevacizumab or Aflibercept after an anti-EGFR therapy, three retrospective studies [30][31][32], two of which were conducted only among Asian population, showed that the clinical outcomes of mCRC patients treated with a second-line anti-angiogenic therapy seemed to be comparable with those reported in the phase III studies [12][13][14].
Despite the unbalanced grouping of the study population according to the received regimens (82.3% Bevacizumab-based vs. 17.7% Aflibercept-based), most of the patients characteristics were balanced between the subgroups, such as elderly patients, number of metastatic sites and primary tumor location (see Table 1). On the other hand, there was a statistically significant difference of primary tumor resection rate between Bevacizumab-treated group and Aflibercept-treated group (78.9% vs. 49%, p < 0.0001), and this might have affected the clinical outcomes [33]. The clinical indication of Aflibercept (limited to patients previously treated with oxaliplatin and in combination with an irinotecan-containing regimen) explain instead the significant differences according to the previously received first-line regimen and to the second-line chemotherapy backbone. The prevalence of left-sided tumors (74.4%) and the probable attitude not to treat with first-line EGFR-inhibitors BRAF mutant patients [34], are aligned to the BRAF mutational status (almost 90% of patients were BRAF wild-type), identifying a study population with good prognosis overall [35].
Even though studies results comparisons are not methodologically correct, some speculations are allowed. The median PFS of the Bevacizumab-treated group (7.1 months) was comparable to the PFS reported in the E3200 and ML18147 trials (7.3 and 5.7 months, respectively) [12,13], whereas the median PFS of the Aflibercept-treated group (5.6 months) was slightly worse than the PFS reported for the experimental arm of the VELOUR study (6.9 months) [14]. The median OS of the Bevacizumab-treated group (16.2 months) was slightly better than the OS reported in the experimental arms of E3200 and ML18147 trials (12.9 and 11.2 months, respectively) [12,13], while the median OS of the Aflibercept-treatment group (12.7 months) was comparable to the OS of the experimental arm of the VELOUR study (13.5 months) [14]. Additionally, the ORR of Bevacizumab-treated (25.7%) and Aflibercept-treated (26.1%) groups resulted to be higher compared to the experimental arms of the E3200 (23%), the ML18147 (5%) and the VELOUR (19.9%) trials. Surely, in addition to some study populations' differences, the genotype selection of our cohort (only RAS wild-type patients were eligible) might also partially explain these discrepancies. Interestingly, genotype based post-hoc analyses reported an OS of 15.4 months for KRAS wild-type patients of the experimental arm of the ML18147 [36], and an OS of 16.0 months for RAS wild-type patients of the experimental arm of the VELOUR trial [37]. Moreover, we have to take into account that most of our patients received active third-line regimens, such as Regorafenib and Trifluridine-tipiracil, which might have affected the OS.
Intriguingly, the multivariate analysis revealed that the Aflibercept-treated group had a statistically significant shorter PFS compared to the Bevacizumab-treated group (HR = 1.44 (95%CI: 1.02-2.03); p = 0.0399), whereas a not significant trend towards a shorter OS was reported (HR = 1.47 (95%CI: 0.99-2.17); p = 0.0503). Concerning safety data, we found a significant higher incidence of G3/G4 VEGF inhibitors class-specific AEs among Aflibercept-treated patients, compared to the Bevacizumab-treated patients (26.5% vs. 7.5%, p = 0.0001). This aspect might be also related to the different pharmacodynamic mechanisms of action of Bevacizumab (a monoclonal antibody which targets VEGF-A) and Aflibercept (a fusion protein which targets both VEGF-A, VEGF-B and placental growth factor (PIGF)) [38]. Furthermore, a statistically significant difference in the incidence of G1/G2 non hematologic AEs (36.4% vs. 59.2%, p = 0.0033) and G3/G4 hematologic AEs (3.1% vs. 18.4%) to the detriment of the Aflibercept-treated patients, was found. The latter aspect could be related to the different chemotherapy backbone (FOLFIRI in 32% of Bevacizumab-treated group and 100% in the Aflibercept-treated group, p = 0.0148).
Our results suggest a slightly better clinical performance for second-line Bevacizumab-based regimens compared to Aflibercept-based regimens. In our opinion, the different safety profile might had affected the effectiveness of Aflibercept-based regimens compared to Bevacizumab-based regimens, leading to a higher discontinuation rate (17.5% vs. 9.2%, respectively) and a worse PFS.
According to the RAISE trial results [39], it would have been interesting to take into consideration Ramucirumab-based second-line regimens, however, Ramucirumab is not reimbursed in Italy as second-line treatment in mCRC patients.
Results from important prospective phase II-III studies, comparing different sequencing strategies of available biological agents for RAS wild-type patients, are awaited. The STRATEGIC-S1 trial (NCT01910610) [40] is an international, open-label, randomized, multicenter phase III trial designed to compare two standard treatment strategies in unresectable RAS wild-type mCRC patients: an oxaliplatin-based second-line regimen with Bevacizumab after fist line FOLFIRI-Cetuximab vs. an irinotecan-based second-line regimen with Bevacizumab after a first-line OPTIMOX Bevacizumab, followed by an anti-EGFR based third-line treatment. The DISTINCTIVE trial (NCT04252456) [41] is a prospective phase II trial, designed to evaluate the efficacy of FOLFIRI-Aflibercept as second-line treatment of RAS wild-type mCRC patients after an oxaliplatin/fluoropyrimidines-based first-line regimen combined with either Panitumumab or Cetuximab.
There are some obvious limitations in this study, including its retrospective design, which expose to selection bias, therefore the results must be taken with caution. Further analysis with a larger sample size and a prospective translational design are certainly needed to better define and personalize the anti-angiogenic strategy as a second-line treatment in RAS wild-type mCRC patients.

Conclusions
Our analysis seems to reveal that Bevacizumab-based regimens have a slightly better efficacy and safety profile compared to Aflibercept-based regimens as second-line treatment of RAS wild-type mCRC patients who received first-line anti-EGFR based treatments. These results have to be taken with caution and no conclusive consideration are allowed.
Supplementary Materials: The following are available online at http://www.mdpi.com/2072-6694/12/5/1259/s1, Table S1: List of participating centers. Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Conflicts of Interest: Alessio Cortellini received grants as a speaker by MSD and Astra-Zeneca, grant consultancies by BMS, Roche, Novartis and Astellas. Daniele Rossini received personal fees from Takeda. Ingrid Garajova received grants as a speaker by Amgen and Takeda. The other authors declare no conflict of interest.