Neoadjuvant FOLFOX 4 versus FOLFOX 4 with Cetuximab versus immediate surgery for high-risk stage II and III colon cancers: a multicentre randomised controlled phase II trial – the PRODIGE 22 - ECKINOXE trial

Thus based on the RAS status of the primary tumor the design will be: Arm A vs. Arm B vs. Arm C for RAS WT patients and Arm A vs. Arm C for those RAS mutated (Fig. 1)

Arm A: simplified FOLFOX-4 alone every 2 weeks for 4 cycles, followed by colectomy, followed by simplified FOLFOX-4 alone every 2 weeks for 8 cycles.

Arm B: Cetuximab plus simplified FOLFOX-4 every 2 weeks for 4 cycles, followed by colectomy, followed by Cetuximab plus simplified FOLFOX-4 every 2 weeks for 8 cycles.

Arm C (control arm): Colectomy followed by simplified FOLFOX-4 alone every 2 weeks for 12 cycles. Adjuvant chemotherapy will be given to all patients with stage III CC and according to the local practices for stage II cancer.

For inclusion in the study, all of the following inclusion criteria must be fulfilled: (i) pathologically confirmed colon adenocarcinoma (≥15 cm from the anal verge); (ii) assessment of RAS status of the primary colon cancer on biopsies (WT or mutated); (iii) CC classified by abdominal CT scan: high risk T3 (disruption of muscle wall and extension into pericolic fat with more than 5 mm protrusion into adjacent mesenteric fat) - T4 (penetration within adjacent organs) and/ or N2 (more than 3 clustered lymph nodes above 1 cm in shortest diameter); (iv) non metastatic CC (lung, liver, peritoneal); (v) non complicated primary tumor (obstruction, perforation, bleeding). Patients with obstructive CC and treated by defunctioning stoma can be included; (vi) absence of synchronous colorectal cancer ; (vii) age ≥ 18 years et ≤ 75 years; (viii) ECOG performance status 0–1 ; (ix) no prior chemotherapy or abdominal or pelvic irradiation; (x) life expectancy of ≥ 5 years; (xi) no history of colorectal cancer; (xii) patients with childbearing potential should use effective contraception during the study and the following 6 months; (xiii) laboratory data including : white blood cell count ≥ 3 × 10⁹/L with neutrophils ≥ 1.5 × 10⁹/L, platelet count ≥ 100 × 10⁹/L, hemoglobin ≥ 9 g/dL (5,6 mmol/l), total bilirubin ≤ 1.5 × ULN (upper limit of normal), ASAT and ALAT ≤2.5 × ULN, Alkaline phosphatase ≤1.5 × ULN, serum creatinine ≤ 1.5 × ULN; (xiv) signed written informed consent obtained prior to any study specific screening procedures.

Patients are not eligible for this study if any of the following exclusion criteria apply: (i) contra-indication to iodinated contrast medium injection including allergy to iodinated contrast medium and renal insufficiency proscribing iodinated injection; (ii) rectal cancer located within 15 cm from the anal verge by endoscopy or under the peritoneal reflection at surgery or having received radiation therapy prior to surgery; (iii) complicated primary CC (obstruction, bleeding, perforation); (iv) synchronous colorectal cancer; (v) metastatic spread at baseline assessment (lung, liver, peritoneal); (vi) history or current evidence on physical examination of central nervous system disease or; (vii) peripheral neuropathy ≥ grade 1 Common Toxicity Criteria for Adverse Events (CTCAE) v.3.0; (viii) known hypersensitivity reaction to any of the components of study treatments; (ix) presence of inflammatory bowel disease; (x) HNPCC syndrome or polyposis; (xi) major surgical procedure, open biopsy or significant traumatic injury within 28 days prior to study treatment start. Incompletely healed wounds or anticipation of the need for major surgical procedure during the course of the study; (xii) clinically relevant coronary artery disease or history of myocardial infarction in the last 12 months, or high risk of uncontrolled arrhythmia; (xiiii) pregnancy (absence to be confirmed by ß-hCG test) or breast-feeding period; (xiv) previous malignancy in the last 5 years; (xv) medical, geographical, sociological, psychological or legal conditions that would not permit the patient to complete the study or sign informed consent; (xvi) any significant disease which, in the investigator’s opinion, would exclude the patient from the study.

The primary objective is to assess the histopathological response to neoadjuvant chemotherapy measured by the Tumor Regression Grade (TRG), as simplified to three categories by Ryan [19]. The TRG-Ryan is based on the relationship between fibrosis and the amount of residual tumor cells is defined as follows (Fig. 2). The histopathological finding of fibrous stroma containing few viable tumor cells can be naturally present in almost 10 % of CCs not treated with chemotherapy prior to surgery [15]. Therefore, this parameter represents an objective measure of evaluation in the control (no pre-operative chemotherapy) and study (neoadjuvant chemotherapy) arms. This limits the risk of wrongly concluding therapy efficacy, thereby meeting the methodological standards of a Phase II randomized study. The histopathological response to neoadjuvant chemotherapy will be centrally assessed by 2 of the 3 members of our Central Review Committee of Pathologists who will be blinded to the patients’ treatment. The objective of the blinded comparison of TRG by two independent pathologists is to assess: (i) the inter-observer reproducibility of this grading system in CC after neoadjuvant chemotherapy; (ii) the specificity of histological "regression" criteria by comparing, without knowing the treatment arms, the histological features of tumors operated on immediately versus that of tumors treated by neoadjuvant chemotherapy.

Secondary endpoints are the following: (i) postoperative morbidity (occurring within 60 postoperative days) graded 0 to 5 according to the Clavien-Dindo classification [20]. If grade 3 or 4 complications occur at a rate ≥ 10 % in either neoadjuvant chemotherapy arm, the arm will be declared hazardous (see section statistical analysis); (ii) the safety of neoadjuvant therapy will be assessed by the systematic collection of toxicity data related to chemotherapy, the total number of cycles administered pre-operatively (documenting reasons for early termination), and the systematic collection of GI symptoms. In addition, complications related to the in situ tumor that could delay the date of surgery or require emergency colonic surgery or stenting, will be documented. These parameters will be evaluated after administration of the first cycle of neoadjuvant chemotherapy; (iii) Regression-free survival at 3-years defined as the time interval between the date of randomization and the date of first recurrence (local, regional or metastatic) or date of death (all causes) if no recurrence; (iv) Disease-free survival at 3 years defined as the time interval between the date of randomization and the date of first recurrence (local, regional or metastatic), or second cancer, or the date of death without recurrence (all causes) and date of second cancer; (iv) Quality of life (evaluated using the EORTC quality of life (QoL) QLQ-C30 and QLQ-CR29); Quality of life will be assessed at the inclusion, after neoadjuvant chemotherapy, at 1 month after surgery and every 6 months thereafter. The QLQ-C30 consists of 30 items with both multi-items scales and single item measures. The internal validation of QLQ-C30 allowed to identify 15 scales and to generate 15 scores: 5 scores of functional scale (physical functioning, role functioning, cognitive functioning, emotional functioning, social functioning), a global health status scare, a financial difficulties scale and 8 symptoms scale (fatigue, nausea/ vomiting, pain, dyspnoea, insomnia, appetite loss, constipation, diarrhoea). These scores vary from 0 (worse) to 100 (better) for functional and global health scale and from 0 (better) to 100 (worse) for symptoms scales. Targeted QoL dimensions for this trial are Global Health, physical and emotional functioning, fatigue and pain of QLQ-30 as well as dimensions of CR29. A 5 points difference in QoL is considered as minimal clinically important difference (MCID). Time to QoL deterioration will be defined as time interval between randomization and first decrease in QoL score greater or equal to 5 points. Patients without QoL decrease greater or equal to 5 points will be censored at the last follow-up. (v) Quality and completeness of the surgical excision evaluated on pathological criteria (number of lymph nodes examined, the completeness of removed mesocolon, surgical margins); (vi) radiological staging and radiological response. The initial radiological staging based on abdominopelvic CT scan will be compared with the histopathological pTNM staging in the control group. CT scan images will be interpreted by the individual study centers, and reviewed by 2 of the 3 members of our Central Review Committee of Radiology. This dual independent reading will assess inter-observer reproducibility of the radiological response to treatment with a test of concordance. The radiological response will be defined by RECIST (v1.1) [21] on CT scan and evaluated in the neoadjuvant chemotherapy arms. These criteria will be applied to the primary colon tumors (largest diameter) and lymph nodes (smallest diameter). The sum of the diameters of the target lesions will be determined on the pretreatment CT scan (initial staging) and compared with the sum of these measures on the reevaluation CT scan (performed within 3 weeks after the completion of neoadjuvant chemotherapy). Interpretation of the reevaluation CT scan will be centralized and carried out independently by 2 of the 3 members of our Central Review Committee of Radiologists blinded to the pretreatment CT scan staging. This dual independent reading will assess inter-observer reproducibility of RECIST criteria applied to CC, and determine if there is a difference in radiological response between the two chemotherapy regimens under investigation in the study; (vii) Correlation between histopathological and radiological response. The radiological response of the primary tumor as defined by RECIST v1.1 criteria (taken for target lesion, only the primary tumor) will be compared with the histopathological response as defined by TRG-Ryan. Evaluation of this correlation will be centralized and carried out by the Central Review Committee of Radiology and Pathology; (viii) The 5-point regression grading system [22] adapted to CC by our team in a pilot study [15], will also be evaluated in this trial. This assessment will be centralized and carried out independently and blindly (without knowing the treatment arms) by 2 of the 3 members of our Central Review Committee of Pathology. This assessment will determine the inter-observer reproducibility of this grading system in CC after chemotherapy and to assess the specificity of histological "regression" criteria using this grading system. Comparison of the reproducibility (test of concordance) of the two systems of grading regression (TRG-Ryan and TRG-Mandard) will identify the most appropriate system for assessing response to chemotherapy in CC.

The screening visit must be performed within 3 weeks before the randomisation. During this visit, inclusion and exclusion criteria are checked and validated. The complete pre-therapeutic work-up includes a physical examination, standard laboratory tests with CEA level, complete colonoscopy with biopsies and a CT scan of the thorax, abdomen and pelvis. Clinical tumoural staging (cTNM) will be based on the data obtained from CT scan. The determination of the RAS status (KRAS and NRAS mutations located within the exons 2, 3 and 4) of the primary tumor will be performed from biopsies obtained during the diagnostic colonoscopy, by each centers at the “plateformes de génétique somatique des tumours” labelled by the National Institut of Cancer (INCa). Once baseline assessments are completed, check and validation of inclusion and exclusion criteria for the study will be performed at the oncological multidisciplinary meeting of each center.

The treatment protocols are standardized with regard to radiological and pathological parameters, based on the recommendations of I’INCa, the FFCD, the French Society of Pathology, the French Society of Digestive Endoscopy and the French Society of Digestive Surgery. Treatment application depends on the RAS status of the primary tumor (Fig. 1) and must start within 14 days after the randomisation.

Neoadjuvant chemotherapy requires an implantable subcutaneous venous access and will be administrated for 4 cycles. The following assessments will be performed prior to each cycle (every two weeks) before chemotherapy administration: documentation of concomitant medication and concurrent procedures; physical examination; vital signs and ECOG performance status; blood sampling for hematology and clinical biochemistry and recording of adverse events since last cycle (including primary tumor related symptoms or complications). The end of treatment visit must be performed 15 days after the 4^th cycle administration.

For all patients, follow-up assessment will be performed until recurrence and/or death. For the study, the recurrence free survival will be assessed at 3 years. For patients included in the present trial, the follow-up will be systematic and performed thereafter this period according to the good clinical practices. The follow-up schedule and investigations proposed for the present study are those recommended by the FFCD [8].

Every 6 months, the following investigations will be performed: tumor assessment, CEA measurement, survival status, additional cancer therapy. Moreover, the following assessments will be performed during the follow-up: colonoscopy 3 years after surgery then every 3 to 6 years, assessment of neurological toxicity, and assessment of electrolytes (magnesium, calcium, potassium) at the second follow up visit (12 months after surgery) if still abnormally decreased at end of treatment assessment.

Patients RAS Wild Type will be randomized between arm A, B and C whereas RAS mutated patients will be only randomized between arm A and arm C. According to Simon (Optimax) 2 steps design, it will be required to include 33 patients/ arm (with unilateral alpha type one error of 5 % and a power of 95 %) to test following hypotheses: (i) H₀: an histological response (TRG1-Ryan) rate of 10 % is uninteresting; (ii) H₁: an histological response (TRG1-Ryan) higher than 10 % is required to rate pursue investigations in phase III trial. A response rate of 35 % is expected.

At the first step (interim analysis), after inclusion of the first 13 patients in each arm, if the number of histological response (TRG1-Ryan) is ≤ 1, inclusion in this arm will be stopped and if the number of histological response (TRG1-Ryan) is ≥ 2, inclusion of the next 20 patients will be done. Calculated power at the interim analysis is 97 %.

At the 2^nd step (final analysis), after inclusion of 33 patients in each arm, if the number of histological response (TRG1-Ryan) is ≤ 6, this arm will be declared uninteresting for further investigations in phase III trial and if the number of histological response (TRG1-Ryan) is ≥ 7, this arm will be declared interesting for further investigations in phase III trial. At the second step calculated power will be 96.3 % and alpha type one error will be 2 %. In control arm no interim analysis is plan for efficacy. Then in this arm, at the second step, calculated power will be 97.2 % and alpha type one error will be 4.2 %. If ≥ 10 % of complications ≥ grade III (Clavien and Dindo classification) within 60 postoperative days are reported, this arm will be declared unsafe.

Interim and final results of this phase II study will be reviewed by an independent data monitoring committee (IDMC). Trial will be pursued in phase III with 2 or 3 arms according to IDMC recommendations. Control arm (no neoadjuvant treatment) will be kept as control arm in phase III whatever the results of phase II study. Required number of patients is 33 × 3 = 99 patients for RAS WT patients and 33 × 2 = 66 for those with RAS mutated. Taking into account prevalence of RAS WT patients of 50 %, it will be necessary to screen 192 patients to obtain 99 patients RAS WT and 66 patients mutated for RAS with a minimal probability of 90 %.

Fifty seven French centres will participate in the study: Amiens University Hospital, Caluire et Cuire Hospital in Lyon, University Hospital of Clermont-Ferrand, General Hospital Center of Orléans, European Georges Pompidou University Hospital in Paris, Claude Huriez University Hospital in Lille, North University hospital of Bordeaux, University Hospital of Nîmes, Purpan University Hospital in Toulouse, Paoli Calmettes Cancer Institute in Marseille, Henri Mondor University Hospital in Créteil, North University Hospital in Marseille, Saint-Louis University Hospital in Paris, Pitié Salpêtrière University Hospital in Paris, Ambroise Paré University Hospital in Boulogne-Billancourt, St André University Hospital in Bordeaux, General Hospital Center in Beauvais, Croix-Rousse University Hospital in Lyon, Beaujon University hospital in clichy, Pessac University Hospital of Bordeaux, the University Hospital in Strasbourg, Pontchaillou University Hospital in Rennes, General hospital in Bézier, Robert Debré University Hospital in Reims, Avicenne University Hospital in Bobigny, Cochin University Hospital in Paris, St Anne Clinic in Strasbourg, University Hospital of Nantes, Jean Mermoz Private Hospital in Lyon, University Hospital in Grenoble, La Timone University Hospital in Marseille, Saint Antoine University Hospital in Paris, General Hospital of Pau, Paul Strauss Cancer Institute in Strasbourg, University Hospital of Dijon, Saint Joseph Private Hospital in Paris, University Hospital of Rouen, Rangueil University Hospital in Toulouse, Sainte Barbe Clinic in Strasbourg, Sainte Catherine Institut in Avignon, The Asclepios House Clinic in Avignon, General Hospital of Bayonne, Eugene Marquis Cancer Institut of Rennes, Intercommunal Hospital Center of Creteil, General Hospital of Mont de Marsan, Bichat University Hospital in Paris, Grand Sud Polyclinic of Nîmes, Kenval Polyclinic of Nîmes, Valdegour Clinic Oncogard Center of Nîmes, General Hospital of Boulogne Sur Mer, University Hospital of Poitiers, University Hospital of Brest, Saint Joseph-Saint Luc Hospital in Lyon, Synergia Polyclinic in Carpentras, General Hospital of Avignon, University Hospital of Saint Etienne.

This study protocol was approved by the Institutional Review Board, the Ile de France IX ethic committee on November 2011 and the AFSSAPS (Agence Française de Sécurité Sanitaire des Produits de Santé) on September 2011 under the registration number A110936-41. The institutional promoter is the Assistance Publique Hôpitaux de Paris, France. The trial has been registered on ClinicalTrial.gov website under the identification number NCT01675999. This study received a grant from the French National Cancer Institute (INCA) in 2010. The study complies with the Declaration of Helsinki rules and the principles of the Good Clinical Practices guidelines. Informed consent will be obtained from each patient in a written form prior to randomisation. Patient safety and all potential threats to the patients will be monitored every 6 months by an independent data safety monitoring board (DSMB) and, additionally, at the discretion of the DSMB or Promoter. The DSMB also will evaluate the primary endpoint data. Qualified personnel at the sponsor site will also meet every three months to review safety data, including adverse events and serious adverse events. Any information deemed to potentially affect the safety of the trial will be brought to the attention of the DSMB. An Independent committee (IDMC) including at least 1 clinicien, 1 methodologist and pharmacologist will be created before trial initiation. This IDMC should statute on efficacy and safety at the interim analysis regarding decision criteria of Simon design, post operative morbidity.