SGM-101 tumor-targeted fluorescence endoscopy to enable discrimination of malignant from benign tissue in rectal polyps with suspected T1 adenocarcinoma or high grade dysplasia: a feasibility study

Selection of patients with early rectal cancer, T1 rectal cancer (T1RC) or
high-grade dysplasia (HGD), for local resection is based on endoscopic imaging,
endorectal ultrasound and/or MRI. However, all these imaging modalities have
their limitations in the accurate detection of small areas of cancer in large
rectal polyps [1-4]. Fluorescent guided endoscopy may offer the opportunity to
aid in detecting these small cancerous areas in colorectal polyps.

Detection of T1RC or HGD in large rectal polyps is essential to select patients
for the appropriate (endoscopic) resection technique [5]. Completely benign
polyps, solely containing low-grade dysplasia (LGD), could be resected by
piecemeal endoscopic mucosal resection (pEMR), a fast technique on which almost
all endoscopists are trained. However, for rectal polyps with T1RC or HGD,
curative resection must be achieved by en-bloc resection techniques such as
endoscopic submucosal dissection (ESD) or endoscopic intermuscular dissection
(EID) [5]. These techniques are more complex and expensive and only a small
number of endoscopists are trained on these techniques [2]. pEMR of lesions
with HGD or T1RC must be avoided because they may result in irradical
resections (R1) or in inconclusive histological margin assessment since the
polyp is not resected en-bloc [2]. Often, this leads to unnecessary additional
abdominal surgical resections.
Furthermore reported sensitivities for optical diagnosis of T1RC in a polyp are
low, varying from 20.8% to 77.8% [3, 4, 6]. Because of the limited accuracy of
optical diagnosis, the current Dutch guideline suggest that rectal polyps >3cm
should be resected in an en-bloc manner, although it has been demonstrated that
only 10.2% of rectal polyps >3cm contain a malignant component [6]. Similarly,
in recent studies it was shown that 10-15% of entirely benign polyps were
removed by abdominal surgery rather than endoscopically. This has major
implications for the patient since surgical resection is associated with
considerable morbidity [7]. Therefore, we are in need of a technique that
provide the endoscopist with real-time information about specific molecular
features of the tumor to differentiate between benign polyps with LGD amenable
for pEMR and those that contain HGD or T1RC that require en-bloc resection.
Tumor targeted fluorescence-guided surgery (FGS) has emerged as a technique
with the potential to enable real-time lesion visualization based on specific
molecular features rather than on morphology [8]. Recently it was shown that
carcinoembryonic antigen (CEA) is overexpressed in approximately 75% of
HGD/T1RC [9]. Additionally, expression in LGD was (nearly) absent in 66% of
low-grade dysplastic tissue and in 98% of normal rectum tissue, making it a
suitable marker for distinguishing T1RC and HGD from LGD and normal tissue. CEA
can be targeted by SGM-101, an anti-CEA antibody attached to a fluorophore
which has been studied in several clinical studies for patients with colorectal
cancer undergoing surgery [10-14]. We hypothesize that the use of SGM-101
during endoscopy aids the endoscopist in discriminating benign polyps with
solely LGD from polyps containing HGD/T1RC.

This study has been transitioned to CTIS with ID 2024-510769-41-00 check the CTIS register for the current data.

The main objective is to investigate the feasibility of a tumour-targeted
fluorescent tracer SGM-101, combined with the use of the CE-marked
fluorescence-laparoscope of Quest Medical Imaging, to discriminate between
normal, LGD and malignant tissue (HGD, T1RC) in patients with suspected T1/HGD
rectal cancer.
It consists of two phases for which primary objectives are defined separately:
Phase I: Dose optimization phase to determine the optimal dose of SGM-101
Phase II: feasibility assessment of tumor-targeted fluorescence endoscopy with
the use of SGM-101 in discriminating normal, LGD and malignant tissue (HGD,
T1RC).

This is a single center prospective, non-randomized phase 2 proof of concept
study, on the performance of SGM-101 to discriminate HGD/T1RC from LGD in
patients that will be scheduled for endoscopic local en-bloc resection in the
Leiden University Medical Centre (LUMC).

In total 20 patients will be included. The first 3 patients with large rectal
polyps in the distal and mid-rectum scheduled for a local endoscopic en-bloc
resection, will receive a single dose of 10mg SGM-101 4 days (+-1) prior to
surgery, the optimal dose in our previous study in patients undergoing surgery
for advanced colorectal cancer [12]. During abdominal surgery, visualization of
the tumor can be hampered by overlying colon and other tissue. During
endoscopic intraluminal imaging, the tumor can directly be visualized, and a
smaller dose of SGM-101 could suffice. Therefore, the consecutive 3 patients
will receive 5mg 4 days (+-1) prior to endoscopy. If after analysis of the
first 6 patients normal tissue and LGD are oversaturated as assessed by
intensity and TBR then the next group will consist of 3 patients receiving 2mg.
In the unlikely event that no adequate differentiation can be achieved between
tumor (T1RC/HGD) and dysplasia (LGD) in the first 6 patients due to low maximum
fluorescence intensity and a low TDR, dose-escalation to 15mg SGM-101 will be
considered. The remaining patients will receive the dose of the group that
provided the highest T1RC/HGD to LGD ratio - from now on defined as
tumor-to-dysplasia ratio (TDR). If no difference between these treatment groups
is observed, we will continue with the lowest dose. If uncertainty about the
optimal dose persists after this first interim analysis, a second interim
analysis will be conducted after 13 patients have been included (10 patients in
one dosing group and 3 patients in the other). Based on the findings from this
analysis, the dose for the remaining 7 patients can be adjusted to 2 mg, 5 mg,
10 mg, or 15 mg. Imaging will be performed with the CE-marked Quest Spectrum
system intraluminally and on the back table, respectively with the laparoscope
and open imaging camera. Under propofol sedation a standard transanal minimal
invasive surgery (TAMIS) seals port (Applied medical) will be placed and
fluorescent imaging will be performed using the Quest Spectrum laparoscope.
After careful fluorescent imaging, a local endoscopic en-bloc resection will be
performed according standard care [15]. Postoperative, ex-vivo imaging will be
performed with the Quest open Camera and the Pearl Trilogy fluorescence imager
(LI-COR, Lincoln, Nebraska). The resection specimen is pinned on cork for
standard assessment by the pathologist. Additional formalin-fixed
paraffin-embedded (FFPE) slides will be analyzed with a dedicated 700nm
NIR-fluorescence scanner to allow for macro- and microscopic histological
evaluation and the accompanied difference in SGM-101 fluorescence distribution
between T1RC/HGD, LGD and normal rectum tissue.

Patients will need to make one extra visit to the LUMC for infusion of SGM-101
4 days (+-1) prior to their intervention. This admission comprises of infusion
(45 minutes) and observation (3 hours). In over its >5 years of use, SGM-101
has not caused drug related adverse effects in >250 patients, suggesting that
toxicity associated with its use should be minimal. A standard TAMIS seals port
will be placed and fluorescent imaging will be performed using the Quest
Spectrum laparoscope. The procedure will be extended by approximately 30
minutes. In this study, no action will be taken on fluorescence. The risks for
the patients are therefore deemed negligible.