Recent studies have already discussed how the position of rectal tumors with respect to the APR has important treatment implications,
since extra- and intra-peritoneal cancers present different systemic spread [1,
4,
12,
13]: as a consequence,
an accurate preoperative identification of the APR may address extra-peritoneal tumors to neoadjuvant therapy [13],
since choosing the optimal treatment for each patient is crucial in relation to the risk of under- or overtreatment.
Rigid rectoscopy may provide a measurement of the distance from the inferior edge of tumors to the anal verge,
and several trials on pre-operative radiation have considered patients using different cut-off measures,
varying in a range of 12-16 cm from the anal verge [14-16],
with a consistent risk of overtreatment.
Paparo et al.
recently found that a cut-off measure <10 cm is sufficient to reach a 100% sensitivity in detecting extra-peritoneal cancers,
preserving an acceptable specificity [4],
results confirmed by the Dutch TME trial,
which showed no beneficial effect of radiation therapy for cancers above 10 cm from the anal verge [14].
If the greatest limitation of endoscopy is represented by the lack of an optimal cut-off measure to distinguish extra- from intra-peritoneal tumors,
due to the great variability of APR location,
MRI recently showed its ability to overcome this limitation,
providing a clear preoperative visualization of APR in addition to the loco-regional staging [4,
5].
Other imaging modalities showed great potentiality in the evaluation of local and systemic involvement of rectal cancer,
such as FDG-PET and FDG-PET/CT,
especially with respect to clinical management and neoadjuvant treatment planning of locally advanced rectal cancer [17,
18]; however,
also CT and CT-WE may provide important informations about the loco-regional features of rectal cancers,
while they are commonly used as the initial staging modality because of their wide availability and their ability to evaluate distant metastasis [19]: in this scenario,
we decided to verify if the APR could be detected on CT and CT-WE as well as on MR images,
allowing an accurate discrimination between intra- and extra-peritoneal rectal cancers.
As seen on MR images,
the APR is a thin hypointense/hyperdense linear structure running from the tip of the seminal vesicles in men or from the utero-cervical angle in women to its attachment on the anterior rectal wall [4,
5].
If motion artifacts are the main causes that may prevent the visualization of the APR on MR images,
beam hardening artifacts due to the presence of metallic hip prosthesis represent a major limitation for the detection of the APR and of tumor location on CT images (Figure 3).
In our study,
the diagnostic performances of MRI and CT in distinguishing extra- vs intra-peritoneal tumors were comparable,
but in 2 cases CT did not allow the visualization of the APR,
while it was detected in all MR examinations.
However,
when compared to MR imaging,
the intrinsic inferior spatial and contrast resolution of CT can be partially balanced out by the employment of post-processing techniques,
as usually happens for emergency finalities (e.g.
bowel obstruction) [20]: in particular,
the use of CT curved planar reformations to obtain a straight representation of the rectum helped us in the evaluation of the level at which the different structures involved are localized (Figure 2A).
Nevertheless,
the use of these tools does not always compensate for the intrinsic limitation of CT in local evaluation: in fact,
we were not able to identify the location of the APR in two patients,
one because of the presence of beam hardening artifacts,
the other one because of the poor quality of visualization.
Furthermore,
it is important to notice that despite we expected a better diagnostic performance of CT-WE compared to conventional CT,
because of bowel distention that should allow an easier and more accurate measurement of the distances between the anal verge and the APR and the tumors’ inferior edge respectively,
we did not detect significant differences between the two techniques.
Limitations of our study are mainly represented by the low prevalence of intra-peritoneal cancers,
but this percentage,
despite the low number of patients included in the current study,
is coherent with our previous results and with the common clinical practice [4,
5].
All examinations were performed before surgery and/or neoadjuvant chemoradiation,
so potential tumor’s shrinkage did not influenced our results.
We have already pointed out that a minor methodological limitation is represented by the use of the anal verge as landmark for the lower part of the external sphincter on MR and CT images,
possibly leading to clinical and anatomical discrepancies; however the excellent correlation between MRI and rigid rectoscopy [4] and between CT and MRI measurements pushed us to consider the anal verge as a reliable marker at cross-sectional imaging.
The results of our study further confirm the great clinical value of MRI in determining the extra- vs intra-peritoneal location of rectal cancers by the direct visualization of APR in a preoperative setting; CT and CT-WE showed as well to have the possibility to play a potential supporting role in the evaluation of rectal cancer,
but it appears unlikely that CT could currently replace MRI in local staging.
Furthermore,
MRI and CT measurements of the distance between the tumor’s inferior edge and the anal verge demonstrated a strong correlation.