Treatments after progression to first-line FOLFOXIRI and bevacizumab in metastatic colorectal cancer: a pooled analysis of TRIBE and TRIBE2 studies by GONO

FOLFOXIRI/bevacizumab (bev) is a first-line regimen of proven activity and efficacy in metastatic colorectal cancer. The upfront exposure to three cytotoxics raises concerns about the efficacy of treatments after progression. We performed a pooled analysis of treatments after progression to upfront FOLFOXIRI/bev in patients enrolled in two randomised Phase 3 studies (TRIBE and TRIBE2) that compared FOLFOXIRI/bev to doublets (FOLFOX or FOLFIRI)/bev. Response rate, progression-free survival (2nd PFS) and overall survival (2nd OS) during treatments after progression were assessed. The RECIST response in first line and the oxaliplatin and irinotecan-free interval (OIFI) were investigated as potential predictors of benefit from FOLFOXIRI ± bev reintroduction. Longer 2nd PFS was reported in patients receiving FOLFOXIRI ± bev reintroduction compared to doublets ± bev or other treatments (6.1 versus 4.4 and 3.9 months, respectively, P = 0.013), and seems limited to patients achieving a response during first line (6.9 versus 4.2 and 4.7 months, respectively, P = 0.005) and an OIFI ≥ 4 months (7.2 versus 6.5 and 4.6 months, respectively, P = 0.045). First-line FOLFOXIRI/bev does not impair the administration of effective second-line therapies. First-line response and longer OIFI seem associated with improved response and 2nd PFS from FOLFOXIRI ± bev reintroduction, without impacting 2nd OS.

intensified strategy, concerns still exist about the most appropriate treatment to administer after the failure of FOLFOXIRI/bev. Options include the same triplet regimen and bev, doublet chemotherapy and an anti-angiogenic agent, anti-EGFR (Epidermal Growth Factor Receptor)-based regimens only in the case of RAS wild-type tumours, 1,2 or later lines options (regorafenib or trifluridine/tipiracil) in the case of clearly refractory disease. Therefore, an optimal selection of the patients that could gain benefit from the reintroduction of the triplet, compared to a different strategy, is a challenge in the choice of the treatment after progression. In this setting, a retrospective pooled analysis of the OPTIMOX-1 and OPTIMOX-2 trials that evaluated 5fluorouracil-based maintenance or chemoholidays following 6 months of upfront therapy with FOLFOX, demonstrated that patients with an oxaliplatin-free interval lasting at least 6 months were those who reported better outcomes following the reintroduction of oxaliplatin. 9 Drawing from these considerations, we focused on therapies administered after first-line progression to patients treated with FOLFOXIRI/bev in the TRIBE and TRIBE2 studies and analysed their outcomes. In addition, with the aim of identifying predictors of benefit from the reintroduction of FOLFOXIRI/bev, we explored the association of clinical parameters of benefit from the first-line therapy, including the oxaliplatin and irinotecan-free interval (OIFI), with the efficacy of treatments after progression.

METHODS
Study design TRIBE (NCT00719797) and TRIBE2 (NCT02339116) are two Phase 3 randomised, open-label, multicentre trials that involved 1187 unresectable previously untreated mCRC patients. As previously detailed, 3,8 in the TRIBE study, patients were randomised in a 1:1 ratio to receive up to 12 cycles of FOLFIRI/bev or FOLFOXIRI/bev followed by maintenance with 5-fluorouracil plus bev until disease progression, unacceptable adverse events, or consent withdrawal in both arms. Treatments after progression were left at investigators' choice. 3 In the TRIBE2 study, patients were randomised in a 1:1 ratio to FOLFOX/bev followed by FOLFIRI/ bev after disease progression or FOLFOXIRI/bev followed by the reintroduction of the same agents after disease progression; all treatments were administrated up to eight cycles followed by 5fluorouracil plus bev maintenance until disease progression, unacceptable adverse events or consent withdrawal. 8 The present analysis included all patients enrolled in the TRIBE and TRIBE2 studies that progressed to upfront FOLFOXIRI/bev.

Definition of endpoints
Treatments received after the first evidence of disease progression were collected. Overall response rate (ORR) was defined as the proportion of patients achieving partial or complete response during the treatment according to RECIST criteria version 1.0 and version 1.1 in TRIBE and TRIBE2 trials, respectively; 2nd progression-free survival (2nd PFS) was defined as the time from the beginning of the second-line treatment to the evidence of disease progression or death, whichever occurred first; 2nd overall survival (2nd OS) was defined as the time from the beginning of second-line treatment to death.
We investigated the association of RECIST response during the first-line therapy (CR + PR versus SD + PD) and of the oxaliplatin and irinotecan-free interval (OIFI), defined as the time from the last simultaneous administration of oxaliplatin and irinotecan in first line to disease progression (≥ versus <4 months), with clinical outcome during treatments after progression.

Statistics
The chi-square test and Kruskall-Wallis test were used when appropriate, to compare clinical and biological features, and ORR between groups. Second PFS and 2nd OS were determined according to the Kaplan-Meier estimates method, and survival curves were compared using the log-rank test. Odds ratios (OR) and 95% confidence intervals (CI) were estimated with a logistic regression model, while hazard ratios (HR) and 95% CI were estimated with a Cox proportional hazard model.
All statistical tests were two-sided, and P values ≤0.05 were deemed significant. Statistical analyses were performed using SAS version 9.4 (SAS Institute, Inc., Cary, NC). The data cut-off for the present analysis was July 31, 2014 and July 30, 2019 for TRIBE and TRIBE2, respectively.

RESULTS
Among 1187 patients enrolled in TRIBE and TRIBE2 studies, 586 received FOLFOXIRI/bev as upfront treatment and after a median follow up of 43.2 months, 524 of them experienced disease progression. In total, 419 (80%) patients received treatment after progression. The majority (70%) of those who did not receive any further treatment died within 3 months after progression. FOLFOXIRI ± bev was administered in 176 (42%) cases, while 123 patients (29%) received a doublet (irinotecan-or oxaliplatinbased in 74 (60%) and 49 (40%) cases, respectively) ± bev. The remaining 120 patients (29%) received other treatments, including anti-EGFR-based regimens in 68 cases (57%). Among patients receiving anti-EGFR-based therapies, 23 (34%) had a RAS and BRAF wild-type tumour, while a RAS or BRAF mutation was found in 16 (23%) and 6 (9%) cases, respectively. RAS and BRAF mutational status were not assessed in the remaining 23 (34%) patients ( Fig. 1). Table 1 summarises patients' characteristics according to their treatment after progression. Therapies other than FOLFOXIRI ± bev or doublets ± bev were more frequently chosen for patients who achieved less clinical benefit from first-line FOLFOXIRI/bev (i.e., those who had not achieved RECIST response, and with an OIFI < 4 months).

DISCUSSION
Starting the therapeutic route of mCRC patients with the upfront exposure to the three active cytotoxics raised doubts about the efficacy and feasibility of treatments after progression. Two Phase 3 randomised trials demonstrated a survival advantage with firstline FOLFOXIRI/bev compared with doublets/bev thus showing that the benefit from the intensified upfront regimen was not limited to the first step of patients' treatment but translated into a long-term effect. Moreover, the initial advantage was not subsequently lost, thus suggesting that treatments after progression could be still efficacious. 3  In the present pooled analysis of the TRIBE and TRIBE2 study, we confirm the feasibility of treatments after progression to first-line FOLFOXIRI/bev showing that the 80% of patients progressed to the first-line therapy were then able to receive a subsequent systemic therapy. Though acknowledging all the limitations of cross-trials comparisons including the heterogeneity of different studies' inclusion criteria, outcome results reported by patients receiving doublets + /− bev were similar to those of trials investigating the continuation of bev beyond progression or the efficacy of other anti-angiogenic agents combined with doublets in terms of OS and slightly worse in terms of PFS. [10][11][12][13] Notably, the reintroduction of FOLFOXIRI ± bev was chosen in the 42% of patients receiving a treatment after progression. When interpreting this percentage, it should be mentioned that while the reintroduction strategy was prospectively planned in the TRIBE2 study, this was not recommended or suggested in the TRIBE trial, where treatments after progression were left at investigators' choice. The reintroduction of FOLFOXIRI ± bev provided a significant 2nd PFS benefit compared to both doublets ± bev and other strategies, with consistent activity results. 3,8 Reintroducing oxaliplatin after progression to a first-line oxaliplatin-based therapy is a valuable therapeutic option along the continuum of care of mCRC patients, as demonstrated by several retrospective and prospective analyses. 9,14,15 This strategy is likely more efficacious for those patients who achieved higher benefit from the upfront therapy, with a longer interval between the last administration of oxaliplatin and the evidence of disease progression. The shorter duration of the induction therapy in TRIBE2 (4 months) compared with TRIBE (6 months) led to similar PFS results (median PFS: 12 months in both studies) and may have favourably affected the reintroduction rate, as a result of the lower incidence of serious neurotoxicity. In the targeted agents' era, prolonging this time interval is made possible thanks to the efficacy of maintenance strategies able to delay disease progression. Evidence about the efficacy of reintroducing the triplet is currently much more limited. 16 In order to identify patients more likely to achieve benefit from the reintroduction of the triplet ± bev, we investigated the association of parameters of benefit from first-line FOLFOXIRI/bev with the efficacy of treatments after progression. Patients who had achieved an objective response during the first-line therapy reported a significant advantage from the reintroduction of the triplet instead of other treatments after progression in terms of both objective responses, and 2nd PFS that was not observed among individuals who did not respond to the first-line therapy. Similarly, a trend towards longer 2nd PFS was observed in favour of the triplet ± bev in patients with a ≥ 4 months interval from the last administration of oxaliplatin and irinotecan to the first evidence of disease progression. Consistently with our results, in the pooled analysis of the OPTIMOX-1 and OPTIMOX-2 trials higher ORR and PFS were reported in the subgroup of patients receiving oxaliplatin reintroduction after two disease assessments by means of CT scan. 9 The retrospective nature of our work prevents us from drawing conclusions about the different efficacy of triplet ± bev versus doublets ± bev versus other options as treatments after progression to upfront FOLFOXIRI/bev. In fact, the lack of randomisation does not allow to exclude that clinical outcomes reported with administered therapies might have been affected by unbalances in prognostic factors at the time of their beginning. In particular, alternative treatment strategies were chosen more frequently in Table 2. Overall response rate, 2nd progression-free survival and 2nd overall survival according to second-line treatments.  patients who achieved less benefit from the upfront therapy probably as a result of more aggressive tumour biology or resistance to the upfront approach. To this regard, we were not able to assess the role of anti-EGFR-containing regimens in second line compared with other strategies in RAS and BRAF wild-type patients due to the limited sample size. In the PRODIGE-18 trial, favourable results were reported with the continuation of bev in combination with a switched chemotherapy as second-line therapy of KRAS wild-type patients treated with first-line chemotherapy plus bev compared with second-line chemotherapy plus anti-EGFRs. 17 At the same time, the association of the second-line treatment outcome with the benefit from the upfront regimen may be explained by the better prognosis of those tumours but also by the effect of the intensified upfront therapy in modifying tumour biology. In fact, it is well established that achieving response with the first-line therapy positively affects patients' long-term outcome independently of prognostic features at baseline. Finally, we were not able to assess the safety profile of therapies after progression since adverse events that occurred were not collected in the TRIBE study. TRIBE2 showed no increased toxicity with FOLFOXIRI/bev compared with FOLFIRI/bev after progression with the exception of a higher incidence of neuropathy in a population clinically selected according to general conditions but also to the previous tolerance during the first-line therapy. A clear limitation is the lack of quality of life data from both trials. 3,8 In conclusion, upfront FOLFOXIRI/bev does not impair the efficacy of treatments after progression. The reintroduction of     Fig. 3 2nd PFS in patients achieving (a) or not (b) an objective response to first-line FOLFOXIRI/Bev according to treatment received after progression. Bev bevacizumab, CI confidence interval.
Treatments after progression to first-line FOLFOXIRI and bevacizumab in. . . D Rossini et al.
FOLFOXIRI ± bev may provide an incremental benefit in ORR and 2nd PFS compared to other options with no significant OS advantage, and should be considered especially in fit patients, with good treatment tolerance, no moderate/severe residual neurotoxicity who experienced objective response during the firstline therapy and with a ≥4 months OIFI.