Differential prognosis of metastatic colorectal cancer patients post-progression to first-line triplet chemotherapy plus bevacizumab, FIr-B/FOx, according to second-line treatment and KRAS genotype

Clinical outcome post-progression to first-line triplet chemotherapy (CT) plus bevacizumab (FIr-B/FOx) was evaluated in metastatic colorectal cancer (MCRC) patients (pts). Second-line treatment was selected according to fitness, KRAS genotype, previous efficacy and safety. Efficacy was evaluated and compared according to treatment or KRAS genotype, using log-rank analysis. Among 54 pts, median overall survival (OS) post-progression was 12 months, significantly better in 40 (74.1%) treated compared to 14 (25.9%) who died without further treatment. Second-line surgical treatment, 4 pts (7.4%), medical treatment, 36 pts (66.7%): triplet CT plus targeted agent, 10 (18.5%); triplet regimens, 19 (35.2%); doublet/monotherapy, 7 (13%). At follow-up of 14 months, objective response rate (ORR) was 38%, metastasectomies 12.5%, progression-free survival (PFS) 10 months, OS 14 months. According to treatment, ORR, metastasectomies, PFS and OS were significantly favourable in triplet CT plus targeted agent compared to triplet, respectively: 80%, 40%, 13 months, not reached; 28%, 6%, 8 months, 11 months. PFS and OS were significantly worse in c.35 G>A mutant compared to wild-type and/or other mutant patients. Prognosis after progression to first-line FIr-B/FOx may be significantly favourable in MCRC pts re-challenged with intensive regimens, and unfavourable in c.35 G>A KRAS mutant patients.


Introduction
Clinical management of MCRC is faced with different options and lines of treatment strategies according to the fitness of the patients, extension of metastatic disease and KRAS genotype (1)(2)(3)(4)(5). First line triplet regimens significantly increased PFS up to 7.2-10.6 months and OS up to 19.9-26.1 months over doublet regimens, also integrated with secondary resection of liver metastases in liver-limited (L-L) disease (2,4,6). After progression to first line treatment strategy, 50-80% MCRC pts receive a second line treatment (4,(7)(8)(9)(10)(11). Randomized clinical trials and pooled analysis demonstrated that exposure of patients to all three most active chemotherapeutic drugs is associated with the longest OS and similar efficacy (7), regardless of the sequence of administration. OS after progression does not correlate with any second line treatment (8).
Panitumumab confirmed the significantly positive predictive effect of KRAS wild-type status, with ORR of 17%, median PFS 12.3 weeks, median OS 8.1 weeks, compared to mutant genotype (22,23). In KRAS wild-type pts, the addition of panitumumab to FOLFIRI significantly increased ORR of 35% and PFS 5.9 months, with a trend toward increased OS (24).
More intensive first line medical treatment consisting of triplet chemotherapy plus targeted agent can increase activity, thus increasing resection rate of liver metastases and clinical outcome of MCRC pts (1,2,6,25,26). We recently proposed a phase II study of BEV addition to triplet chemotherapy, according to FIr-B/FOx schedule (1) reaching ORR of 82%, 54% liver metastasectomies in L-L disease, median PFS 12 months, median OS 28 months (1,3). KRAS wild-type pts with L-L disease may achieve significantly greater benefit from integration with liver metastasectomies compared to other/multiple metastatic (O/MM) pts, with respect to KRAS mutant pts (3,5).
The present study evaluated clinical outcome of the fit MCRC pts after progression to FIr-B/FOx and, retrospectively, the prognostic relevance of second line treatments and KRAS genotype.

Materials and methods
Patient eligibility. Sixty-seven fit MCRC pts were enrolled in previously reported phase II study (1)   Darmstadt, Germany), 400 mg/m 2 initial dose, then 250 mg/ m 2 /week; cycles every 4 weeks. Triplet FIr/FOx regimen, doublets and mono regimens were administered according to previously reported schedules (27,28).
Study design. Pts were assessed at the time of progression to first line treatment and every 2-3 cycles of second line treatment. A multidisciplinary team, consisting of a medical oncologist, liver surgeon, radiologist, evaluated resectability, according to previously reported resectability categories (3). Clinical criteria of activity and efficacy were: ORR, resection rate of metastases, PFS, OS. ORR was evaluated according to RECIST criteria (29); pathologic complete response was defined as absence of residual cancer cells in surgically resected specimens. Clinical evaluation of response was made by CT-scan; PET was added based on investigator assessment. Liver metastasectomies were defined as: R0, if radical surgery; R1, if radiofrequency was added. Surgery was recommended >4 weeks after BEV discontinuation. PFS and OS were evaluated using the Kaplan-Meier method (30). PFS was defined, as the length of time from the beginning of treatment and disease progression or death (resulting from any cause) or to the last contact; OS as length of time between beginning of treatment and death or to last contact. Prognostic relevance of second line treatments and of KRAS genotype was retrospectively assessed, using log-rank test to compare PFS and OS (31).
Mutational analysis. KRAS and BRAF genetic analyses were performed on paraffin-embedded tissue blocks from primary tumor and/or metastases, through selection of tumor cells, and DNA extraction, as previously described (5). Genotype status was assessed for KRAS codon 12-13 and BRAF c.1799 T>A (V600E) mutations by SNaPshot ® multiplex screening for KRAS mutations and KRAS/BRAF mutations in 36 and 32 samples, respectively (32,33); direct sequencing was performed to detect KRAS mutations in 26 samples. SNaPshot multiplex assay was performed as reported (32,33). Briefly, KRAS exon 2 and BRAF exon 15 were simultaneously PCR-amplified using specific primers and analyzed using the ABI PRISM SNaPshot Multiplex kit (Applied Biosystems, Foster City, CA, USA) with five primers including at their 5'-end an additional tail allowing their simultaneous detection. Sense primers allowing the extension at nucleotides KRAS c.34G, c.35G, c.37G, c.38G and BRAF c.1799T were used and multiplex SNaPshot reaction was performed as reported (32). KRAS exon 2 sequence was performed from PCR-amplified tumor DNA using the Big Dye V3.1 Terminator kit (Applied Biosystems), electrophoresis in ABI PRISM 3130xl Genetic Analyzer (Applied Biosystems), and analysis using the GeneMapper Analysis Software version 4.0 (Applied Biosystems).

Discussion
Among fit MCRC pts treated with first line FIr-B/FOx regimen, adding BEV to triplet chemotherapy, 74.1% underwent a second line treatment, in the range of reported 50-80% (7-11); 25.9% died without receiving further antitumoral treatment. Median OS post-progression to FIr-B/FOx was 12 months, including untreated pts and significantly Overall       (22% of L-L disease), all previously challenged with firstline FIr-B/FOX regimen and secondary surgery. Doublet FOLFOX4 schedule, or OXP associated to CPT-11 reported significantly increased ORR of 22 and 28%, and PFS 6.2 and 5.3 months, compared to CPT-11 alone, respectively. Median OS was 13 months, significantly increased only with OXP/ CPT-11 regimen (14,15). Randomized studies of cetuximab plus CPT-11 in EGFR-overexpressing pts, previously treated with CPT-11 or with 5-FU/OXP, respectively showed significantly improved ORR of 16.4 and 22.9% and PFS 4 months (12,20). Triplet FOLFOX4-BEV association, after progression to 5-FU/CPT-11, demonstrated significantly increased ORR 22.7%, median PFS 7.3 months, and median OS 12.9 months (16). Recently, FOLFIRI-aflibercept, after progression to OXP-containing chemotherapy, gained significantly increased median OS 13.5 months (34). A randomized trial reported that BEV associated with 5-FU-based chemotherapy, after first line BEV-containing regimen, significantly improved clinical outcome (35). In KRAS wild-type pts, triplet panitumumab/ FOLFIRI regimen reported significantly increased ORR of 35% and median PFS 5.9 months (23,24). Thus, OS after progression does not correlate with any second line treatment (8) in clinical trials and few secondary resections of metastases were reported after second line treatment (7). Retrospective analysis of 32 pts (24%) achieving OR and progressed >3 months, who were re-challenged with FOLFOXIRI, reported significantly longer PFS (8.2 months) and OS (19.3 months), with respect to doublet regimens (10,11). In our present analysis, second line triplet regimens, proposed to 19 pts (47.5%) achieved ORR 28%, secondary metastasectomies 6%, median PFS 8 months, median OS 11 months. Re-challenge of triplet chemotherapy associated to targeted agent, proposed to 10 pts (25%), with previous OR, long PFS (≥10 months), off-treatment interval ≥3 months and no previous limiting toxicities, achieved ORR 80%, that correlated with 40% secondary surgical resections, median PFS 13 months, and 2-year OS 80% (median OS not reached at median follow-up 31.5 months). PFS and OS were significantly favourable in pts treated with triplet chemotherapy plus targeted agent compared to triplet regimens. Present data confirm that re-challenge of intensive medical treatment is feasible in a selected subgroup of MCRC pts, with high activity, efficacy and effectiveness of secondary metastasectomies. Prospective studies will address if medical and surgical re-challenge can be the standard multidisciplinary second line strategy.

KRAS KRAS treated wild-type mutant ------------------------------------------------
Direct comparison of PFS and OS in KRAS wild-type compared to mutant pts failed to significantly differentiate prognosis in second line, as it was previously reported in first line treated MCRC pts (5,36,37). In KRAS mutant pts harbouring the prevalent c.35 G>A transversion, median PFS and OS were significantly worse compared to KRAS wild-type pts and/or other than c.35 G>A KRAS mutant pts, due to increased aggressiveness and resistance to medical treatment (38). Present data confirm our recent findings of significantly  worse prognosis of c.35 G>A KRAS mutant pts treated with first line FIr-B/FOx (39), even in a small cohort of MCRC patients. Here we report for the first time the c.35 G>A KRAS mutant genotype as prognostic biomarker of unfavourable clinical outcome, significantly related to worse efficacy (PFS) of second line treatments. Further prospective studies will confirm prognostic and predictive value of c.35 G>A KRAS mutation in MCRC patients.
In conclusion, clinical outcome of MCRC progressing to first line FIr-B/FOx regimen may be significantly favourable in pts re-challenging triplet chemotherapy associated with targeted agent compared to other second line treatments and significantly worse in c. 35 G>A mutant compared to wild-type and other mutant KRAS patients.