Bevacizumab and Combination Chemotherapy in rectal cancer Until Surgery (BACCHUS): a phase II, multicentre, open-label, randomised study of neoadjuvant chemotherapy alone in patients with high-risk cancer of the rectum

Study design

The BACCHUS trial is an investigator initiated, multicentre, open-label, prospective,
randomized phase II study. All participants have to provide written informed consent,
signed and personally dated, before inclusion in the trial. The trial EudraCT number
is 2010-022754-17, and is registered on ClinicalTrials.gov (BACCHUS: NCT01650428).

Trial organisation

The sponsor is University College London. Central coordination is financed by Cancer
Research UK (CR UK) and carried out by the CR UK University College London Cancer
Trials Centre (UCL CTC). An independent data monitoring committee (IDMC) will monitor
the conduct and safety of the trial. Participating sites are required to report all
serious adverse events (SAE) as defined by the protocol to UCL CTC in line with applicable
regulations.

Ethics and informed consent

The final protocol was approved by Riverside l Research Ethics Committee (ref: 12/LO/1158).
Appropriate approval from respective local ethics committee is required to join this
trial. This study has been approved by the ethics committee of the following Hospitals
or Universities: Barnet and Chase Farm Hospital, Blackpool Teaching Hospitals, East
and North Herts Hospitals, NHS Greater Glasgow and Clyde Hospitals, Heatherwood and
Wexham Park Hospitals, Hillingdon Hospitals, Imperial College Healthcare NHS Trust,
North Middlesex University Hospital, The Royal Marsden Hospital, University College
Hospital, London. This study is conducted in accordance with the most recent version
of the Declaration of Helsinki and according to GCP. Written informed consent, signed
and personally dated, is obtained from each patient before inclusion in the trial.

Population

Patients with histologically confirmed adenocarcinoma of the rectum require specific
tumour and patient criteria for inclusion. A staging MRI is mandated. The lists of
inclusion and exclusion criteria are presented in Tables 1 and 2.

Table 1. Patient inclusion criteria

Table 2. Patient exclusion criteria

Trial entry has been restricted to patients in whom MRI suggest primary tumour or
lymph nodes do not extend to ?1 mm from, or breach the circumferential resection margin
(CRM)– since even with preoperative chemoradiation up to 30 % of these patients would
have a positive CRM (?1 mm) after TME. Eligibility is also confined to patients with
MRI estimated penetration of the mucularis propria 1 mm and/or patients with cN2
predicted by MRI and extramural vascular invasion (EMVI), but T3 tumours must have
a predicted ?2 mm margin from the mesorectal fascia.

These criteria are likely to form a group of patients making up about 40 % of rectal
cancers overall. Such patients have a 50 % 5 year survival 32] and a local recurrence rate of 6–10 % with surgery alone.

Trial entry has also been restricted to patients younger than 70 years with distal
rectal tumours, 4–12 cm from the anal verge. Accurate clinical staging with MRI is
more diffcult in the low rectum, at lower than 4 cm there is a 5–15 % risk of involved
lateral pelvic lymph nodes, which are not resected at TME. A lack of evidence to suggest
a benefit from oxaliplatin containing adjuvant chemotherapy for stage II colorectal
cancer 34]–36] and insufficient data to support a benefit from adjuvant chemotherapy in Stage III
colorectal cancer in patients over 70 years has informed the decision to exclude patients
over 70 years from this trial 33].

Study objectives and endpoints

The primary objective of the BACCHUS study is to evaluate the efficacy of FOLFOXIRI?+?BVZ
and FOLFOX?+?BVZ in terms of their ability to produce pCR. Secondary objectives include
evaluation of the safety and tolerability of the two regimes and the feasibility of
delivering them, as well assessment of additional measures of efficacy such as progression
free and overall survival.

The primary endpoint is pathological complete response (pCR) at surgery; secondary
endpoints include ORR, CRM negative (R0) resection rate, T and N stage downstaging,
PFS, DFS, OS, local control, 1 year colostomy rate, adverse events, compliance with
chemotherapy treatment, tumour regression grade (TRG), and tumour cell density (TCD).

Survival curves for DFS and OS will be plotted. Cumulative incidence of local recurrence
will be computed accounting for death as competing risk. Differences in survival will
be tested with the log-rank test. Hazard ratios and 95 % confidence intervals (CI)
will be computed using Cox regression. A table will present the completion rate of
the neo-adjuvant treatment, pCR frequency, patients with a R0 resection with 90 and
95 % CI. Frequency and percentages for toxicity will be presented according to the
Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. All proportions
will be presented with 95 % CI.

Randomisation and stratification

Patient Randomisation will be performed centrally at the UCL trials centre. Eligible
patients are randomly assigned to one of the two treatment arms in a 1:1 ratio and
stratified according to treating centre, gender and presence or absence of EMVI. Treatment
process and schedules for the BACCHUS trial are summarised in Figs. 1 and 2.

Fig. 1. Treatment schedule

Fig. 2. The BACCHUS trial. Patients will be randomised to one of two neoadjuvant chemotherapy
regimens

Neoadjuvant Chemotherapy

In both arms, chemotherapy is delivered with bevacizumab. In total, 6 cycles of chemotherapy
are prescribed preoperatively every 2 weeks (bevacizumab omitted during cycle 6).
Adverse events are monitored from informed consent to 3 months after surgery and dose
modification can be made according to specified protocol guidelines.

Assessments/follow up

Response and resectabilty evaluation

Clinical response has not been shown to be a robust surrogate endpoint to predict
outcome. However, patients will undergo response evaluation with MRI of the pelvis
prior to cycle 4 and at the end of all treatment (prior to surgery) according to the
Response Evaluation Criteria in Solid Tumours (RECIST 1.1) and additionally MRI-based
TRG assessment is required 37]. An additional response evaluation according to standard uptake values (SUV) changes
with PET/CT is mandated prior to cycle 4. Patients who do not respond will come off
all trial treatment (allowing the investigator to proceed to whatever treatment is
felt most appropraite ie surgery or SCPRT/CRT followed by surgery).

Tolerability to treatment is evaluated at each visit including physical examination,
vital signs, WHO performance status, clinical laboratory profile, and adverse events,
graded according to NCI-CTCAE v.4.03.

Surgery and histopathology

Surgery should be performed 8–12 weeks after termination of chemotherapy, and a minimum
of 8 weeks after the final dose of bevacizumab. Surgical dissection according to TME
principles should not differ between the two trial groups and can be performed open
or laparoscopically. Surgery may include anterior resection, abdominoperineal resection
or a low Hartmann’s procedure.

Pathological evaluation of resected specimens will be according to guidelines included
in the study protocol. The 5th edition of TNM will be used. In addition, circumferential
resection margin (CRM) will be assessed and a margin of 1 mm or less considered positive.
TRG will be presented as data categorised into five groups- TRG 0, TRG 1, TRG 2, TRG
3 and TRG 4 using the Dworak method. Also, the quality of the resected specimen will
be evaluated with separate scoring for the mesorectum and the anal canal. Formalin
fixed and paraffin embedded (FFPE) tumour tissue obtained at baseline will be evaluated
for KRAS and BRA status and plasma /buffy coat collected at baseline and before the
2
^nd
, 3rd and the 4th cycle (and also if the patient relapses), will be assessed for angiogenic
markers (FFPE and serum) in the BACCHUS trial. Serum obtained at baseline, during
preoperative treatment, postoperatively and at follow up will be evaluated for circulating
tumour DNA.

Adjuvant chemotherapy and follow-up

Patients can be treated with postoperative chemotherapy according to the local protocol
of each participating centre. Patients will be followed up every 6 months for up to
42 months after randomisation, to document progression, recurrence and survival. Postoperative
investigations/surveillance are performed according to local practice.

Statistical considerations and sample size estimation

The primary endpoint for this trial is the pCR of the TME specimen. The proportion
of patients in each arm who achieve a pCR will be presented, along with a 95 % CI.
Within each group the achieved pCR rate will be compared to the historical rate achieved
by radiotherapy alone (5 %). In the United Kingdom patients without a threat to the
circumferential resection margin are likely to be treated with short course preoperative
RT, and not chemoradiation. The study is powered on the assumption that a proportion
of patients will have a pCR. It is well recognised that patients who have a complete
clinical response (cCR) both on imaging and clinical examination will from time to
time refuse surgery. For the purpose of this study, patients who have a sustained
cCR at 12 months will be considered the same as a patient with a complete pathological
response. Patients with a transient clinical response where subsequent relapse is
observed within this 12 month period, will not.

Based on pCR with similar regimens prior to liver resection, and primary tumours responding
better than metastases, we anticipate a pCR rate of 15–20 %. Compared to 5 % pCR rate
historically for radiotherapy alone, a type I error ? =0.05 and a type I I error ?
=0.8, 27 patients for the FOLFOX arm are required. The same number of patients is
required for the FOLFOXIRI arm. Assuming 10 % of patients will be non-evaluable, 30
patients will be recruited to each arm (i.e. a total of 60 patients. NACT will be
considered worth exploring further in a randomised phase III trial if at least 4/27
pCRs are observed. In the instance of more than 27 patients being assessed for pCR,
the first 27 randomised patients per arm will be assessed. The study is not powered
for a direct comparison between the two arms.

Quality assurance/safety

Monitoring will be conducted centrally at UCL CTC and on-site monitoring will be scheduled
if there is any evidence of non-compliance at site. An independent data safety monitoring
committee (IDMC) meeting will be held periodically to review interim analysis, or
as necessary to address any issues.

Translational research

Analyses of both tumour tissue and plasma with tissue microarray, proteomics and genomics
may generate increased knowledge of prognosis and prediction of response to chemotherapy
in the BACCHUS trial. Hence, a schedule for collection of plasma and of fresh tissue
for freezing, at different stages of treatment in each arm, is defined in the study
protocol.

Tumour tissue and blood samples will be stored for future research. At surgical resection
blocks of tumour and normal mucosa will be also be collected. In addition, HE stained
slides from the diagnostic biopsy and resection samples will be collected to undertake
Tumour Cell Density and Tumour regression Grading.

Plasma and Peripheral Blood Leucocytes (PBL) samples will be collected at baseline
and at the following timepoints during treatment :-Baseline (prior to starting treatment
cycle 1), prior to starting treatment cycle 2, prior to starting treatment cycle 3,
prior to starting treatment cycle 4 (preferably at radiological response assessment)
and if patient relapses.

Conventional size–based radiological criteria using RECIST may not be the optimal
method of assessing response to chemotherapy, especially with a regimen integrating
bevacizumab) 37], 38]. Hence, imaging biomarkers will also be explored in terms of mri-based TRG and MRI
diffusion weighted imaging 39]. Exploratory SPECT imaging using Tc99m-maraciclitide as the tracer in a subset of
patients at baseline and post treatment will provide information regarding changes
in angiogenesis with treatment.