*Improving clinical management of colon cancer through CONNECTION, a nation-wide Colon Cancer Registry and Stratification effort.*

Colorectal cancer is the second most common cancer in the Netherlands with an
incidence around 16,000 patients at 2016. Patients with high-risk stage II
colon cancer (defined as T4N0) may be offered adjuvant systemic chemotherapy
and patients with stage III colon cancer routinely undergo curative surgery
combined with adjuvant chemotherapy when fit for treatment. Despite this
intensive treatment, 20% of the patients with stage II and 30-35% of the
patients with stage III colon cancer will still develop metastatic disease. In
addition, on estimate, 50% of the patients with high-risk stage II and stage
III colon cancer would probably never develop metastases after surgery and are
therefore over-treated with adjuvant chemotherapy. Furthermore, none of risk
factors defining high risk stage II disease seem to have predictive value.
Thus, accurate prognostication and selecting patients for adjuvant chemotherapy
is still an unmet need.
The consensus molecular subtypes (CMS) in colorectal cancer, consisting of four
molecular subtypes, is currently considered the most robust classification of
colorectal cancers with clear distinctive biological features.
It has been suggested that the four subtypes may have prognostic and/or
predictive value. Guinney et al. showed a clear relapse free survival and
overall survival advantage for CMS1-3 compared to CMS4 based on a patient
cohort with stage I to IV colon and rectum carcinoma with divergent therapy
strategies.

Although the available evidence does not clearly delineate the predictive and
prognostic value of CMS, the biological features of the subtypes and the scarce
literature available provide some handles. First, CMS1 tumours, which are
generally MSI and hypermutated, show high response to immunotherapy. Second,
CMS4 tumours, characterized by a mesenchymal phenotype, seem to have a worse
prognosis when compared to CMS1-3. This may be due to biological features
resulting in more metastatic potential and/or a poor response to chemotherapy.
Roepman et al showed a similar overall survival for patients treated with and
without adjuvant chemotherapy for patients with colorectal cancer classified as
a mesenchymal phenotype [4]. Last, CMS2 and CMS3 remain subtypes for which the
response to chemotherapy is unknown. Recent data however indicate that the
response to oxaliplatin is only observed in a fraction of colon cancers, which
all belong to the CMS2 subtype. Combined, these observations provide further
support for the idea that subtypes might be useful to predict response to
chemotherapy.

In de MOSAIC study 2246 patients with stadium II or III colon carcinomas were
randomised between adjuvant treatment with 5FU/LV and with 5-FU/LV+ oxaliplatin
(FOLFOX-4). The DFS at five and six years postoperatively were both
significantly better in the oxaliplatin group compared to the 5FU/LV group
((DFS at 5 years HR 0,80; 95%CI: 0,68-0,93; p=0,003) (OS at 6 years HR 0,84;
95%CI: 0,71-1,00; p=0,046)). In other randomised adjuvant studies, comparable
results were found with 5FU/LV or capecitabine in combination with oxaliplatin
for stadium II and III colon carcinomas.
The IDEA Collaborative group (International Duration Evaluation of Adjuvant
chemotherapy) performed a prospective, pre-planned pooled analysis of 6
concurrently conducted randomised phase III trials to evaluate whether 3 months
(3m) adjuvant treatment with FOLFOX/CAPOX is non-inferior 6-months (6m) for
DFS. DFS was defined as the time from enrolment to relapse, second CRC, and
death (all causes). Non inferiority was to be declared if the 2-sided 95%
confidence interval (CI) for DFS hazard ratio (HR 3m v 6m) was below 1.12. The
analysis included 12834 patients with stage III colon cancer from 12 countries.
A total of 40% of the patients received CAPOX. Overall, the 3-year DFS rate was
74.6% (3m) and 75.5% (6m), with an estimated hazard ratio (HR) of 1.07 (95%CI,
1.00-1.15). The 3-months versus 6-months DFS HRs were 1.16 (95%CI, 1.06-1.26)
and 0.95 (95% CI, 0.85-1.06) for FOLFOX and CAPOX treated patients,
respectively. The investigators therefore concluded that while non-inferiority
was not established for the overall cohort, non-inferiority of 3-months versus
6-months oxaliplatin-based adjuvant therapy was supported for CAPOX (figure 3).
Importantly, grade 3 or higher neurotoxicity was observed less frequently in
the 3-months versus 6-months arm (3% vs 16% FOLFOX, 3% vs 9% CAPOX, p<0.0001)
which showed a benefit for a 3-months regimen. In the Netherlands,
approximately 70% of the patients with colon cancer who are offered adjuvant
chemotherapy receive CAPOX.

Neoadjuvant therapy has been implemented in the treatment of various GI
malignancies including oesophageal, gastric and rectal cancers to 1) to reduce
the chance of surgical irradicality, 2) treat microscopic disease that is not
addressed by surgery and 3) because patients might better tolerate full
intensity chemotherapy when administered prior to surgery rather than
postoperatively. Furthermore, neoadjuvant chemotherapy offers the opportunity
to evaluate tumour response prior to surgery.
At present, there is very limited data on neoadjuvant chemotherapy in
resectable colon cancer patients. Important arguments for the absence of
studies on neoadjuvant chemotherapy in resectable colon cancer include amongst
others that 1) resectability is seldom an issue, 2) the effect to chemotherapy
is uncertain in some patients as discussed above and 3) tumour progression
during neoadjuvant chemotherapy might cause bowel obstruction. The FOxTROT
Collaborative Group (2012) was the first to set up a trial with neoadjuvant
chemotherapy in patients with locally advanced resectable colon cancer and
concluded that preoperative chemotherapy was feasible with acceptable toxicity
and perioperative morbidity. The results demonstrated that the risk for bowel
obstruction was very low with only 1 patient (1%) developing recipient bowel
obstruction in the preoperative chemotherapy group.
After this successful pilot study, the FOXTROT study group analyzed the effect
of neoadjuvant chemotherapy in more than 1000 patients with colon carcinoma.
These results have not yet been published, but they have been presented at the
annual ASCO meeting. In this study, patients with a cT3-4N0-2M0 tumor were
randomized 2:1 for 3 courses of neoadjuvant and 9 courses of adjuvant FOLFOX
or 12 courses of adjuvant FOLFOX. Administering neoadjuvant chemotherapy was
safe in accordance with previously presented results with fewer major surgical
complications and significant down-staging and reduced risk of incomplete
resection. Furthermore the risk of a recurrence after 2 years was smaller in
the experimental group (13.6% vs 17.2% HR 0.75 (0.55-1.04), p = 0.08).
Thus, neoadjuvant chemotherapy in colon cancer is feasible and suitable to
evaluate tumour response. This led us to propose a study to investigate the
role of subtypes on therapy response in a novel neoadjuvant setting that allows
to determine therapy efficacy of individual colon cancer subtypes.

Primary objective:
Evaluation of the pathological tumour response to neoadjuvant systemic
chemotherapy per CMS subtype in patients with microsatellite stable high risk
stage II and stage III colon cancer.
Secondary Objectives:
• Evaluation of the radiological tumour response to neoadjuvant systemic
chemotherapy per CMS subtype.
• Assessment of prognostic value of pathological and radiological response to
neoadjuvant systemic therapy for recurrence free survival (RFS) at two and
three years and overall survival (OS) at 5 and 10 years.
• Assessment of differences in CMS classification before and after neoadjuvant
systemic therapy.
• Assessment of pathological response by Ki-67, Caspase-3 and
cystatic-cytotoxic effects on HE-stained tissue slides.
• An indirect comparison will be made for 2 and 3 y RFS and (5 and 10Y) OS
between the proposed schedule and the standard of care schedule with 4 adjuvant
cycles of CAPOX using a matching control cohort including patients who
retrospectively would have been eligible for participation in CONNECTION-IIfrom
the Dutch colorectal cancer cohort PLCRC METC-12-510/D. Evaluation of
diagnostic accuracy of ctDNA measurements for monitoring treatment response to
neoadjuvant treatment and detection of residual disease.
• Results from our study will be compared with the results from the FOXTROT
trial

Prospective multicentre intervention study

2 courses neo-adjuvant en 2 courses adjuvant capecitabine and oxaliplatin
(CAPOX)

As a result of this study, we can possibly prevent a large number of patients
from being over-treated or treated insufficiently in the future, and we can
look for better and more effective treatment options for patients with poor
survival outcome (currently 30% of treated patients). To analyse the predictive
value of the CMS subtypes, the response to chemotherapy must be quantified and
the most reliable outcome measure is the pathological tumour response. This
question can therefore only be answered if we study this in a neo-adjuvant
setting.
We expect that patients will have comparable survival benefits compared to
adjuvant chemotherapy, and potentially a subgroup of patients will have an
increased survival benefit from participating in this study because of (early)
chemotherapy. Most patients in this study would receive the same chemotherapy
as standard of care adjuvant treatment instead of neoadjuvant. The regimen used
and it*s possible side effects are therefore well known. The side effects to be
expected in the neoadjuvant setting are assumed not to differ from those in the
adjuvant setting. Plus, patients will probably be in a better physical
condition when they undergo the preoperative courses of chemotherapy because
they did not undergo surgery yet.
Nevertheless, there are several risks associated with participating in this
study and patients should be well-informed on these possible risks. One of the
burdens are the side effects of the chemotherapy.
The study population will be treated with 2 courses of neo-adjuvant and 2
courses of adjuvant chemotherapy. According to current guidelines in daily
clinical practice, patients with an indication for adjuvant chemotherapy are
being offered the same chemotherapy but then 4 courses post-operatively, based
on the results of the IDEA study..
This IDEA study, of which the results have been published recently, was a
prospective, pre-planned pooled analysis of 6 concurrently conducted randomised
phase III trials (n=12,834 patients) to evaluate the non-inferiority of 3
months* adjuvant treatment with FOLFOX/CAPOX compared with 6 months. Based on
this study the number of courses of adjuvant chemotherapy in the guideline was
adjusted from 8 (6 months) to 4 (3 months). The well known most common side
effects for the proposed chemotherapy schedule are peripheral neuropathy,
haematological and gastrointestinal side effects, and hand-foot syndrome. By
adjusting the number of courses from 8 to 4 on the basis of the IDEA study
mentioned above and the current guideline, the load for patients is already
considerably less than before. Results from the IDEA study showed that adverse
events occurred significantly less often in the 3-months versus 6-months arm
[11].
Another potential risk in this study design is progression from disease
during the waiting period until surgery. This risk has been investigated in the
FOXTROT study. They found that in both the adjuvant chemotherapy group and the
neo-adjuvant chemotherapy group one patient needed an acute operation due to
bowel obstruction. In addition, one patient in the neo-adjuvant chemotherapy
group was found to have metastases. However, it is unlikely that this has been
a consequence of postponing surgery due to neo-adjuvant treatment [8]. This
conclusion is supported by the fact that in the adjuvant chemotherapy group two
patients were found to have peritoneal metastases during surgery. We therefore
expect that the chance of progression during the waiting period until the
operation is small. However, to monitor this properly, we perform an extra CT
scan after 2 courses of neo-adjuvant chemotherapy, in order to adjust the
treatment in the event of disease progression. In the FOXTROT study, no
differences were seen between the two groups in the duration of admission and
side effects of the chemotherapy. Fewer major surgical complications occurred
in the preoperative group as well as a reduced risk of imcomplete resection.
Due to erroneously pre-operative staging on CT-scans, there might be patients
with actual pT1-3N0 tumours who will receive chemotherapy in the present study,
who would otherwise not qualify for adjuvant chemotherapy according to current
standard of care. This will be true for approximately 26% of patients at most,
but will most likely be less based on the results of two studies previously
performed[12, 13]. However, of these patients, about 10% of patients that are
normally not treated with adjuvant chemotherapy do develop recurrent disease
for which this chemotherapy could be effective. Moreover, the chemotherapy
concerns a maximum of 4 courses in total with limited toxicity.