Primary rectal tumours on CT may appear as: intraluminal focal mass; plaque thickening of the viscus involving one or several quarters of its circumference, associated with an intraluminal component.
It should be considered that the normal thickness of the rectal wall is less than or equal to 3 mm . The finding of focal or diffuse thickening should definitely indicate a neoplastic disease when the wall is over 5-6 mm thick [19,20]. However most lesions, on examination, are over 2 cm .
The CT density of the neoplastic tissue is about 40-60 HU, usually hypodense as compared to the normal wall and to date there have been no definite findings to relate the densitometric values of the lesions with the various histologic types. It has been observed that the finding of calcifications within the lesion suggest the diagnosis of adenomucinous carcinoma [18,19].
In our personal experience we have noted that mucinous carcinoma, which is relatively infrequent, on CT has an inhomogenous appearance with typical areas of hypodensity related to the high extracellular contents of mucin, alternating with solid tissue characterised by contrast enhancement. Potentially malignant villous adenoma when bulky is shown as an intraluminal mass with homogenous water-like density interfaced with a branching appearance of the remaining portions.
Because CT does not recognise the 5 layers of the rectal wall (contrary to transrectal sonography and MR imaging), it is not able to differentiate the neoplasm limited to the mucosa (T1) to that involving the muscolaris (T2). It has now been definitely established that the more specific role of CT is to differentiate the advanced stages of the disease (T3-T4): the extent of involvement of pararectal fat and of adjacent organs, adenopathy and metastasis are recognised. It is in fact well known that the diagnostic accuracy of CT is proportional to the stage.
The infiltration of the perirectal space, namely the
advance from stage T2 to stage T3, can be detected on CT (77% sensitivity, 64% specificity, 73% accuracy) based on the identification of irregular external margin at the level of neoplasm with an 'indented' profile and hyperdense strands in the perirectal fat originating from the neoplasm (Figs. 4 and 5).
This is undoubtedly one of the more delicate phases of staging: extremely thin strands of neoplastic tis sue beyond the wall must be recognised or the extent of disease should be suspected even in the presence of minimal densitometric alterations of the perirectal fat [21-27]. The ability to visualise the infiltration of contiguous adipose planes is an important parameter for diagnosis as well as for therapy, as pre- or intraoperative radiotherapy may be hypothesised . However, it should be kept in mind that also peritu-moral lymphangitis, vascular ectasia adjacent to the wall or perirectal inflammation may be responsible for CT hyperdensity of the perirectal fat or of strands mimicking a stage T3 disease [15,19, 20, 28] (Fig. 5).
In the assessment of tumour infiltration in the perirectal fat, it is important to distinguish the involvement of the perirectal fascia. This condition represents the first sign of the advance from stage T3 to stage T4. The fat tissue adjacent to the external rectal surface is covered with the perirectal fascia to form the adipose rectal capsule. The perirectal fascia is normally recognised on CT. In the presence of inflammatory or neoplastic processes it thickens at times asymmetrically and its identification is even easier [18, 22] (Figs. 4 and 5).
CT plays a major role in the definition of stage T4 (100% sensitivity, 92% specificity, 93% accuracy), namely in the identification of signs of infiltration of the anatomical structures and of adjacent pelvic organs (perirectal fascia, seminal vesicles, uterus/vagina, prostate, pelvic muscles, bone segments, etc.) [17,20,21,24]. Two major CT signs indicate a direct involvement of pelvic organs: (1) the loss of adipose cleavage planes between the neoplasm and the adjacent organ. However it should be underlined that the obliteration of the adipose plane may be due also to lymphatic or vascular problems, or cachexia, with no real infiltration . In some cases, the excessive gas distension may be per se the cause of the loss of cleavage with adjacent structures [19, 27]. (2) The finding of direct infiltration by the tumour or the observation of a 'bridge' to the tumour and the adjacent organ with densitometric features similar to those of the rectal tumour.
It is questionable whether the simple thickening of the perirectal fascia should be considered a sign of disease spread [22, 24].
From reports in the literature concerning series of rectal tumour staging, the most frequently involved organs are bladder and prostate followed by the seminal vesicles, ureters, vagina, uterus, ovaries and small intestine [18,19].
The involvement of the bladder is readily established based also on the presence of air within it due to the formation of fistulous tracts. The finding of hydronephrosis is suggestive of ureteral infiltration [18,19].
It should be stressed that sometimes it is difficult
to distinguish a primary rectal tumour from one originating in the prostate, uterus or ovary, secondarily involving the rectum [19, 28].
On CT, definite involvement of muscular structures (levator ani, internal obturator, coccygeal, piri-form and greatest gluteal muscles) is defined based on the detection of enlargement of the involved muscle. It should be noted that because of the normal lack of adipose planes between the levator ani, the most caudal portion of the rectum and the anal sphincter, in this site the assessment of the extramural infiltration is frequently impossible .
Direct infiltration of adjacent bone structures (the sacrum and coccyx) can be diagnosed based on the finding of neoplastic tissue adjacent to gross areas of bone destruction. In other cases, the finding is not so clear and only the use of suitable windows for the study of the bone can evidence a minimum infiltration with areas of osteolysis limited to the cortex .
Was this article helpful?