Imaging techniques:
The primary modality for evaluation of adnexal masses is ultrasound.
According to IOTA simple rules,
if at least one of the M features is present (an irregular solid tumor,
presence of ascites,
papillary projections,
irregular multilocular solid tumors larger than 10 cm and strong blood flow) and no B feature is present,
this suggests a high likelihood for malignancy (4,
5).
If at least one B feature is present (a unilocular cyst with solid component less than 7 mm,
acoustic shadowing,
smooth multilocular tumor less than 10 cm and absence of blood flow) and no M feature is present,
this suggests a high likelihood of benignity(6).
If there are both M and B features or if there are no clear M or B features,
then the test is inconclusive and a second stage test is recommended.
Contrast enhanced CT thorax abdomen and pelvis is the standard of care for ovarian cancer staging (7,
8). CT imaging provides clinically relevant information including size and location of peritoneal deposits and enlarged nodes.
Ascites in advanced ovarian cancer makes identification of lesions easier as peritoneal surfaces and bowel loops are more clearly delineated against the fluid.
It is however limited in identification of small peritoneal metastases.
CT is also very helpful in identification of disease sites that indicate stage IV disease.
Multi-parametric MRI can be used in lieu of CT and performs better in detection of subtle peritoneal disease.
PET/CT is mainly a problem-solving tool and currently has limited utility in initial ovarian cancer staging.
Stage I:
Tumor is limited to the ovaries or fallopian tubes,
in the case of fallopian tube cancer.
Stage IA tumor is unilateral with intact ovarian capsule and disease on the ovarian surface.
Imaging findings include a unilateral well-circumscribed solid or mixed solid and cystic mass,
septa > 3 mm thick.
Solid enhancing tissue and papillary projection into the cysts are hallmarks of malignancy.
The solid component and thick septa demonstrate vascularity on color Doppler on ultrasound and enhancement with CT or MRI (6,
9).
Presence of tumor in both ovaries with intact capsule and no disease on ovarian surface corresponds to stage IB.
Stage IC is diagnosed if there is ascites positive for malignancy on cytology or if there is rupture of the capsule and surgical spill during surgery (10).
Stage II
Stage II corresponds to disease in the true pelvis.
Implants on the uterus and tubes corresponds to stage IIA,
while implants on the bladder and sigmoid colon correspond to IIB.
Stage III
Retroperitoneal nodal disease and peritoneal metastases denote stage III.
Stage IIIA is nodal metastases and microscopic peritoneal seeding.
The commonest site of nodal metastases is to the para-aortic and paracaval nodes at the level of the renal hila,
following the anatomic lymphatic drainage of the ovaries (11).
Stage IIIB relates to macroscopic peritoneal metastases less than 2 cm.
Macroscopic metastasis larger than 2 cm correspond to stage IIIC.
Important review areas while searching for peritoneal metastases include the right paracolic gutter,
liver capsule and diaphragm.
Review of coronal images is also helpful especially for diaphragmatic disease.
Peritoneal disease is more conspicuous in the presence of ascites.
Smooth elliptical outline of serosal deposits on the surface of the liver and spleen is useful in differentiating these from visceral metastases (11).
Stage IV
Disease outside the abdomen and pelvic cavity and hematogenous metastases are the hallmark of stage IV disease.
Cytology proven malignant pleural effusion is stage IV.
Nonregional lymph nodes are also considered stage IV disease.
Transdiaphragmatic seeding (rather than hematogenous spread) is thought to be responsible for cardiophrenic nodal metastasis.
Up to 21% of women have stage IV disease at presentation (12).
Comparison of imaging modalities
Contrast enhanced CT has been compared with whole body MRI and diffusion-weighted images in a number of studies with regards to diagnostic performance.
Some of the data focusing on important sites of disease in the peritoneal cavity and retroperitoneum is presented (13,
14,
15) in the table 2.
It is evident that diffusion-weighted imaging and whole-body MRI performed better than CT and may in future be the imaging modality of choice.
Pitfalls in ovarian cancer staging
Radiologists need to be aware of the pitfalls as these could potentially result in inaccurate staging.
Most of the pitfalls are due to inherent inability of the imaging modality to either identify disease,
or due to physiological appearances which mimic disease.
The main pitfall on imaging is the detection of the extent of peritoneal disease.
Small peritoneal deposits less than 5 mm are not easily identified with CT.
These may still be obscured by bowel even on MRI which has better tissue contrast.
Even PET/CT is limited in identifying small peritoneal deposits and detecting low volume peritoneal involvement.
Conversely,
disease under the surface of the diaphragm may be overlooked if only the axial images are reviewed.
Generally small volume peritoneal deposits on the surface of bowel or liver may be difficult to appreciate on CT and PET/CT.
Diffusion weighted imaging and post contrast T1 fat-saturated imaging may be the most accurate imaging currently available,
and the evidence is building to support this view.
Use of size criteria in determining nodal disease is inaccurate unless the node measures over 10 mm in short axis,
when it is highly likey to be involved.
Using lower cut-offs increases sensitivity bbut with loss of specificity.
False-positive nodes can also be due to inflammatory nodal enlargement.
PET/CT usually detects involved nodes greater than 8 – 10 mm in short axis.
However metastatic nodes with necrosis,
cystic or mucinous degeneration may not be FDG avid on PET imaging.
Diffusion-weighted imaging (DWI) is very sensitive but nonspecific as normal nodes may also restrict (16).
DWI has poor anatomic localization and is more susceptible to image degradation by artifacts.