FIRE-9 – PORT / AIO-KRK-0418: a prospective, randomized, open, multicenter Phase III trial to investigate the efficacy of adjuvant/additive chemotherapy in patients with definitely-treated metastatic colorectal cancer

Background Additive/adjuvant chemotherapy as concept after local treatment of colorectal metastases has not been proven to be successful by phase III trials. Accordingly, a standard of care to improve relapse rates and long-term survival is not established and adjuvant chemotherapy cannot be recommended as a standard therapy due to limited evidence in literature. The PORT trial aims to generate evidence that post-resection/ablation/radiation chemotherapy improves the survival in patients with metastatic colorectal cancer. Methods Patients to be included into this trial must have synchronous or metachronous metastases of colorectal cancer—either resected (R0 or R1) and/or effectively treated by ablation or radiation within 3–10 weeks before randomization—and have the primary tumor resected, without radiographic evidence of active metastatic disease at study entry. The primary endpoint of the trial is progression-free survival after 24 months, secondary endpoints include overall survival, safety, quality of life, treatments (including efficacy) beyond study participation, translational endpoints, and others. One arm of the study comprising 2/3 of the population will be treated for 6 months with modified FOLFOXIRI or modified FOLFOX6 (investigator´s choice, depending on the performance status of the patients but determined before randomization), while the other arm (1/3 of the population) will be observed and undergo scheduled follow-up computed tomography scans according to the interventional arm. Discussion Optimal oncological management after removal of colorectal metastases is unclear. The PORT trial aims to generate evidence that additive/adjuvant chemotherapy after definitive treatment of colorectal metastases improves progression free and overall survival in patients with colorectal cancer. Trial registration This study is registered with clinicaltrials.gov (NCT05008809) and EudraCT (2020–006,144-18).

Background Application of additive or adjuvant treatment following local treatment of metastases in patients with metastatic colorectal cancer (mCRC) is real practice in oncology, but data from large phase III trials are limited. Colorectal cancer has a 5-year prevalence of 87-107/100.000 persons in Germany [1,2]. Approximately 40-50% of all patients with colorectal cancer develop metastases and benefit from local treatment of metastases, although relapse occurs in 70-80% of these patients [3][4][5][6][7]. A reduction of the relapse rates would accordingly improve the long-term outcome of these patients. Local treatment (surgery, ablation, or radiation) with intent to remove one or several metastases has been integrated into the treatment algorithm of metastatic colorectal cancer [7,8]. The evolution of these strategies and consecutive recommendations in current national and international guidelines focus on therapy of liver metastases, which are present in 75-80% of all patients with colorectal metastases. Nevertheless, other metastatic lesions are also treated accordingly if the course of disease allows for it (i.e., lung, peritoneum, bone, etc.) [9][10][11].
Studies investigating peri-or postoperative therapy in the context of surgery of colorectal metastases have addressed largely homogeneous cohorts of metastatic disease limited to the liver [5,6,12,13]. Unfortunately, these studies have failed to define a standard of care for adjuvant chemotherapy of locally treated colorectal metastases. Several aspects have limited the implementation of systemic therapies in this regard: A pooled analysis of a EORTC phase III trial of adjuvant systemic chemotherapy after surgical resection of colorectal cancer metastases showed only a marginal statistical significance in favor of adjuvant chemotherapy with fluoropyrmidine application (bolus 5-fluorouracil) due to the reduced size of the respective trials (278 patients included, assigned to chemotherapy [n = 138] or surgery alone [n = 140]; hazard ratio = 1.32; 95% CI, 1.00 to 1.76; p = 0.058). Moreover, this regimen is no longer used and has been replaced by more effective and less toxic regimens (oral capecitabine or infusional 5-fluorouracil) [12]. The second major trial in mCRC investigated pre-and postoperative therapy with a combination regimen of 5-FU, folinic acid and oxaliplatin (FOLFOX) vs. surgery alone [5,6]. Importantly, the median number of metastases resected in this trial was "1"-clearly comprising a cohort with favorable risk which due to evolution of therapy is not directly comparable to patients with resected metastases in the context of current treatment algorithms. The study described a trend towards improvement in outcome that did not translate into a clear benefit in survival. Recently, a Japanese phase II or III trial suggested that additive FOLFOX after surgical treatment of colorectal liver metastases improves disease-free survival with an unclear effect on overall survival [14]. However, similar to the EORTC-initiated trial of perioperative chemotherapy [5,6], the majority of patients in the JCOG0603 trial presented with rather low risk characteristics based on number and size of lesions, metachronous/synchronous disease and other characteristics. Therefore, these data cannot be generalized with respect to current treatment interventions for colorectal metastases, and the implications for the proposed trial seem to be limited. Accordingly, standard of care treatment to improve the relapse rates of definitively treated colorectal metastases is not established and the current national and international guidelines for colorectal cancer do not recommend additive/adjuvant chemotherapy due to insufficient evidence on its benefit [7,8]. The FIRE-9 -PORT / AIO-KRK-0418 trial aims to generate evidence that post-resection/ablation/radiation chemotherapy improves the survival in patients with mCRC.

Aim of the study
The main objective of this randomized multicenter trial is to generate evidence that additive/adjuvant therapy after resection, ablation or radiation of metastases with modified FOLFOXIRI or modified FOLFOX6 may improve progression-free survival (PFS) and overall survival (OS) in patients with colorectal cancer. This is of specific importance since improvements in localized but also systemic therapies have resulted in increasing numbers of patients with mCRC undergoing definitive local treatment of metastases [4,[15][16][17][18][19][20]. To support the purely clinical information, a supporting translational study will help to identify subgroups of patients that might benefit from systemic therapy after definitive treatment of colorectal metastases.

Sample size and follow-up
The primary endpoint will be PFS (progression-free survival, defined as progression/relapse or death from any cause) at 24 months after randomization.
Our sample size planning is based on two pooled, early-stopped trials in the adjuvant/additive setting using fluoropyrimidine monotherapies [12] that reported a hazard ratio for PFS in favor of active treatment of 0.76 (the originally reported hazard ratio was reported reversed as 1.32) that translated into a similar effect for the endpoint overall survival. We hypothesize a slightly larger effect for PFS in favor of active therapy due to the 2-3 drug regimens resulting in an estimated hazard ration of 0.70.
For the control arm (surgery alone) with structured follow-up, a progression/relapse/death-free rate of 40% at time point 24 months was observed translating into a 60% progression/relapse/death-rate at this time (according to the reported control arm [12]. With a hazard ratio 0.70 (= λ I /λ C C = 0.0267/0.0382) favoring active treatment, the hypothesized relapse rate at 24 months in the intervention arm is assumed to be 47%. With a power of 80%, a 2-sided alpha of 0.05, a total of 276 events need to be observed in order to detect a difference in progression-free survival of a hazard ratio of 0.70 -favoring active treatment vs. observation (Schoenfeld formula). Assuming an accrual time of 48 months and a follow-up time of 24 months, a drop-out/censoring rate of 30% after 24 months after randomization, a total of 480 patients (320/160 in the respective arms, rounded to receive integers and maintain the allocation ratio) is expected to yield the required number of events if the accrual rate is constant. We account for additional 5% of patients that directly leave the study after randomization and never receive the study medication. In total, 507 patients (480/507≈0.95) are planned to be recruited.

Selection of study population Study population
The study will pragmatically recruit all patients with mCRC that were effectively treated with surgery/ablation/radiation, as the underlying question applies to all types of metastases in the setting of colorectal cancer. Besides this selection, the indicated criteria take into account that patients have to be fit enough to undergo an intervention.

Randomization
Screened and eligible patients will be included in the trial after initiation of the study. Patients will be allocated in a strictly concealed way by 2:1 randomization (intervention versus control) with variable block lengths randomization. Randomization will be stratified by four binary stratification variables, using secuTrial ® (interactive Systems GmbH, Berlin, Germany): • number of treated metastases (> 2 versus 1-2) • pretreatment with systemic therapy for metastatic colorectal cancer yes versus no pretreatment • choice of potential therapy in the trial: (fit for mFOL-FOXIRI versus fit for FOLFOX) • presence of at least one bad prognostic factor (peritoneal metastases resected/ known BRAF mutation/ synchronous metastases defined as evidence of metastases < 12 months vs 12 + months after first diagnosis) versus no bad prognostic factor)

Primary endpoint
Progression-free survival (PFS) is the primary endpoint of the trial. PFS is defined as time from randomization to death or evidence of disease relapse/progression, whatever occurs first within a time frame of 24 months. Blood samples are collected during study and follow-up to create a biobank of patients with and without relapse. Moreover, the relapses will be recorded as part of the study protocol, including the collection of tumor tissue and blood samples, if available, at relapse. The patterns of relapse will be correlated with the initially resected/ ablated metastases clinically and in terms of tumor characteristics (mutations, expressions).

Interventions and comparisons Trial design
The design of the trial is displayed in Fig. 1.

Experimental intervention
One arm of the study comprising 2/3 of the population will be treated (intervention) by a highly active tripletregimen (mFOLFOXIRI) or a standard adjuvant regimen (mFOLFOX6), developed both in palliative and adjuvant treatment settings of gastrointestinal cancer [21][22][23][24]. A guidance is outlined in Table 1, however, at the discretion of the investigator, site specific modifications are permitted, e.g. additional 5-FU bolus, parallel leucovorin and oxaliplatin administration, or variation of the infusion rates. The choice of mFOLFOXIRI/mFOLFOX6 is based on the fact that available data with monotherapy have not provided a standard therapy as they failed to demonstrate superior outcome. Moreover, the success and feasibility of these regimens in adjuvant treatment of colorectal and pancreatic cancer highlights the potential of active treatment of micrometastases to prolong the overall survival. In particular-and certainly similar to pancreatic cancer-the high risk of relapse in the study cohort requires active therapy, aiming to definitely improve the outcome of patients given the risks of this treatment. The postoperative treatment period of six months was chosen based on the recommendation for adjuvant treatment in high-risk patients with UICC III colorectal cancer [25]. As the option of pre-operative therapy is left open to the discretion of the investigators, the study will limit the use of oxaliplatin to a total of 12 cycles FOLFOX(IRI) or 8 cycles of a capecitabine-based 3-weekly regimen), including the pre-study and study treatment. Accordingly, patients with more than 3 months (i.e., 6 cycles of FOLFOX[IRI] or 4 cycles of CAPOX/XELOX) pretreatment will be excluded from study participation.

Control intervention
Since there is no established standard and no evidence of clinical improvement by systemic treatment after treatment of metastases from colorectal cancer, the control arm (1/3 of the population) will offer a structured oncological follow-up with computed tomography (CT) scans of chest and abdomen every 3 months [7].

Re-assessment during active study participation
Radiologic re-assessment with contrast-enhanced CT scans is scheduled 3 months (6 cycles of therapy or 3 months observation) and 6 months (12 cycles of therapy or 6 months of observation) after randomization.

Structured follow-up
Structured follow-up for up to 60 months after randomization should be maintained for both arms of the trial (Fig. 2). It is recommended to perform CT scans and/ or MRI scans every 3 months within the 2 years after randomization. After two years without relapse, intervals are stretched to 6 months in the third and following years after randomization.

Statistical and safety analyses Primary analysis
The null hypothesis to be tested in confirmatory analysis states that the hazard ratio for PFS comparing  intervention versus control equals 1. This hypothesis will be tested by means of Cox-regression adjusting for the below mentioned strata for randomization. The twosided significance level is set to 0.05. The primary analysis will be conducted based on the full analysis set defined as the intention to treat population including all randomized patients who received the treatment at least once.

Secondary analyses
As a sensitivity analysis to the primary efficacy analysis, the same test will be repeated on the per-protocol population including all patients without major protocol violations. As another sensitivity analysis, the primary hypothesis will also be tested with the test for the average hazard ratio1 to potentially account for non-proportional hazards. Event probabilities of PSF will be estimated by Kaplan-Meier-Curves. The above analyses will be repeated for overall survival (OS).

Safety analyses
Absolute and relative frequencies as well as unbiased event-rate estimates (Kaplan-Meier, Empirical cumulative incidence) of AEs, SAEs, event rates of grade 3 and 4 toxicities (NCI-CTCAE) and abnormal biochemical paramters between treatment arms will be reported at different time points together with 95%-confidence intervals. All analyses will be done using validated statistical software.

Discussion
Patients with locally treated metastases of colorectal cancer represent a high-risk cohort for development of recurrent metastases and consecutively death. A definite answer to the question of postoperative systemic chemotherapy is not only scientifically interesting but also an unmet clinical need with increasing prevalence. Therefore, intervention with systemic chemotherapy appears clearly justified, taking into account, that the chosen combination regimen (fluoropyrimidine plus oxaliplatin and optionally also irinotecan) represent standard of care in advanced colorectal cancer. The addition of postoperative therapy is often discussed in tumor boards and chosen for individualized treatment, in particular in previously untreated or shortly treated patients. Participating patients will receive either standard of care on the basis of available evidence, or as intervention, up to 6 months of therapy. Therefore, a potential disadvantage, although possible, seems unlikely. The PORT trial will therefore either provide a basis for routine treatment or alternatively confirm that additive/adjuvant chemotherapy after definitive treatment of colorectal metastases does not provide a clinically relevant benefit and should not be promoted.