DWI MR imaging for dedicated staging of patients with peritoneal seeding.

1.1 Background
Peritoneal seeding is a well-known mechanism of spread in advanced
gastrointestinal and ovarian cancer. Peritoneal carcinomatosis (PC) has
significant implications for not only treatment options but also prognosis; it
is the second-most frequent cause of death in colorectal cancer patients after
metastatic disease to the liver [ref]. The last few decades showed a revolution
in the treatment of PC. Presently the prognosis of PC patients has dramatically
improved and where once only palliative treatments and comfort measures were
contemplated, nowadays selected patients benefit from a radical locoregional
approach aiming at long-term disease control. That radical locoregional
approach consists of cytoreductive surgery and hyperthermic intraperitoneal
chemotherapy (CRS-HIPEC). When a complete surgical cytoreduction can be
achieved 5-years survival rates of up to 50% are reported after CRS-HIPEC.
Despite this potential survival gain, CRS-HIPEC can be a morbid procedure. The
morbidity rate of this procedure is about 49% with a severe complication rate
of 21% [4].The mortality rate has been reported to be ranging from 0.9% to 5.8%
[4]. One out of ten patients needs a reoperation after CRS & HIPEC. The high
complication rate also implies that CRS-HIPEC is not only an extensive but also
a costly surgical procedure. Obviously, it is important for patients with a
limited life expectancy to avoid pointless aggressive surgical procedures.
Hence, recognizing patients with a maximum risk-to-benefit ratio for the
procedure is imperative. For ovarian carcinoma the most important factor for
survival is the cytoreduction rate as described in a meta analysis by Bristow
et al. Each 10% increase in maximal cytoreduction was associated with a 1.9
month increase in median survival time [5].
To select patients who could benefit from CRS the PCI is used in
gastrointestinal cancer. The PCI combines the location and the volume of
peritoneal tumor found at laparotomy at 13 abdominal and pelvic regions. The
PCI is widely validated and is a quantitative prognostic indicator for
long-term outcome. In order to prevent unsuccessful laparotomies surgeons and
gynaecologists would like to know whether a complete cytoreduction is feasible.
Currently a diagnostic laparoscopy is often used to assess the abdomen and
establish the PCI. However, with this invasive procedure it is not always
feasible to inspect all relevant areas in the abdomen due to formation of
adhesions and or tumor. So, if preoperative imaging could accurately determine
the PCI it would be a valuable, noninvasive, selection tool to select those
patients who will benefit from CRS-HIPEC. For colorectal cancer patients a
better selection will limit the number of ineffective procedures; for patients
with advanced stage ovarian carcinoma a better diagnostic tool will enable
clinicians to decide whether to perform primary cytoreductive surgery or start
with neo-adjuvant chemotherapy followed by surgical cytoreduction after 3
cylcles of chemotherapy.

1.2 Imaging of Peritoneal metastasis
Imaging has an important role in the assessment of PC, from the initial
diagnosis to the evaluation of disease volume and distribution that may help
select those patients who will benefit from CRS-HIPEC. However due to its lack
of specificity and the small size of lesions the diagnosis and staging of PC
presents a serious challenge for the radiologist. Therefore, the current role
of imaging in the patient selection process is aimed at ruling out
extraperitoneal disease involvement, assessing peritoneal disease volume and
distribution as a guide for surgical planning and evaluating possible signs
that may preclude the achievement of a complete cytoreduction.

1.2.1 Computed Tomography
CT is routinely used for evaluating patients with PC. A meta-analysis of
current literature shows a pooled sensitivity and specificity for the overall
detection of PC with CT of 73% and 90%, respectively [1]. However the accuracy
of CT for PC strongly depends on tumor size, site, morphology and the presence
of ascites. For example, it has been shown that CT scan accuracy for the
detection of peritoneal lesions varies with their location within the abdomen,
being greatest in the gutters, over the free surface of spleen and liver, and
less favorable in the pelvis and midabdomen. Also tumor nodule size has a major
impact on sensitivity as well, ranging from 25% for lesions smaller than 0.5 cm
to 90% for nodules more than 5 cm in size[6].
In addition de Bree et al. reported a wide interobserver variability among
radiologists in the interpretation of CT scans of patients with peritoneal
carcinomatosis of colorectal origin [7]. The role of CT is very limited in the
prediction of the PCI with a steady underestimation of disease. Low et al.
found that in a patient group with a surgical median PCI score of 33, the
median PCI score of CT was 15 [3]. Obviously CT cannot be used as a reliable
selection tool.

1.2.2 Diffusion-weighted MR imaging
Magnetic resonance imaging (MRI) uses the effect of a strong magnetic field on
tissue protons spin motion, resulting in a superior soft tissue contrast
compared with CT. DWI is a non-invasive functional imaging technique that
measures the extracellular movement of water protons. In tissues with a normal
cellularity (most healthy soft tissues) or low cellularity (fluids), there is
ample extracellular space and water protons can diffuse relatively freely. This
movement of water protons causes a signal loss on diffusion-weighted images. In
tissues with increased cellularity (tumour), the extracellular space is limited
and the movement of water protons is restricted. As a result the signal on DWI
remains high. Because DWI suppresses the signal in all normal tissues, the high
signal of malignant tissues stands out which makes DWI a highly promising tool
for detection of malignant tumours. DWI is a well-established technique for
brain imaging, in particular for the detection of brain ischemia. Currently,
the potential value of DWI for extracranial imaging is widely being
investigated with a specific focus on cancer imaging. The value of DWI for
malignant tumour detection has been demonstrated in several cancer types
including prostate cancer, liver tumours, head & neck tumours and
gynaecological malignancies.
In a comparative study reported by Low [2] using CT and MRI, the latter showed
a significantly improved sensitivity for depicting tumor involving the
peritoneum and the intestinal tract. The same author reports on the ability of
MRI to depict subtle peritoneal implants, an important feature that constitutes
a weakness of CT scanning. Moreover, in patients with moderate to high-volume
ascites, it allows for a good evaluation of the parietal or visceral peritoneum
covered by fluid, which is not possible with a CT scan [8]. In a small study
patients were categorized as small volume tumor (PCI 0-9), moderate volume (PCI
10-20), and large volume (PCI > 20) according to their surgical PCI [3]. MRI
could correctly categorize 91% of the patients (in contrast to CT, which could
correctly catogorize 50%).
This means that DWI MRI could play an important role in the diagnosis and
therapeutic management of patients with colorectal and gynecological
malignancy, however more prospective research is needed.

Primary Objective:
Prediction of PCI with DWI MRI, using the surgically determined PCI as the
reference standard.

Secondary Objective(s):
- MR Protocol optimisation
- Comparison of the diagnostic performance of DWI-MRI in assessing the PCI with
the results of CT (=common practice) in colorectal cancer.

This is a prospective observational cohort study and will be conducted in The
Netherlands Cancer Institute. In six months 20 patients scheduled for
cytoreductive surgery for either colorectal carcinoma with peritoneal
carcinomatosa or advanced stage ovarian carcinoma will be included. Patients
with colorectal carcinoma will subsequently be treated with HIPEC. In addition
to the standard diagnostic work-up (CT), patients will receive an additional
MRI scan 0-14 days prior to their scheduled surgery.

The goal of this study is to evaluate the diagnostic performance of DWI MRI and
CT for predicting the PCI.

For the study population, this study has no direct benefits. However, the
results of this study will help improve the performance of MRI for staging and
response evaluation in patients with peritoneal cancer, which can have a
substantial impact on future treatment planning and prognosis as described
above.
The burden for patients exists of:
• The total time of MR imaging will be 35 minutes. MRI does not have any
radiation exposure.
• Gadolinium and Buscopan will be administrated intravenously. A potential side
effect that can occur as a result of the administration is an allergic
reaction. Although it is known that the side effects of Gadolinium /Buscopan
are limited and occur in only a very limited number of patients, the
possibility of serious or life-threatening anaphylactic or anaphylactoid
reactions, including cardiovascular, respiratory and/or cutaneous
manifestations, should always be considered (as with other contrast media).
Furthermore, caution should be exercised in patients with renal insufficiency
due to the possibility of further deterioration in renal function (there have
been reports of nephrogenic systemic fibrosis associated with the use of some
gadolinium-containing contrast agents in patients with acute or chronic severe
renal impairment). Therefore, only patients with a GFR (glomerular filtration
rate) of >30 will be included, as is the recommendation for the clinical use of
Gadolinium.