Women with ovarian cancer often present at advanced stage of disease. The outcome depends mainly on the stage of disease at first diagnosis, but also on the quality of treatment. For individualised tumour treatment, detailed assessment of tumour extension using modern imaging is crucial. Ultrasound remains the initial and most important imaging method for ovarian cancer detection. Although increasing evidence shows that ultrasound is an accurate technique to stage and follow up ovarian cancer, it requires an experienced examiner capable of examining both the pelvis and the abdomen. Computed tomography is the most commonly used imaging modality for preoperative staging and follow up. Magnetic resonance imaging remains a second-line imaging method for solving problems, mainly in the pelvis. Positron emission tomography combined with computed tomography is the optimal imaging technique for suspected recurrence, particularly in women with rising CA 125 levels, but negative results of conventional imaging methods.
Introduction
Ovarian cancer is the most aggressive gynaecologic malignancy, accounting for about one-half of all deaths related to gynaecological cancer, with a 5-year survival rate of around 40% . Despite advances in surgery, chemotherapy, and intensive ongoing research, survival has not significantly increased. The most important factor for survival is the disease stage at diagnosis. About 70% of women present when the cancer is at an advanced stage (i.e. it has metastasised to the upper abdomen or beyond the abdominal cavity) . One of the reasons for late detection of ovarian cancer was thought to be its asymptomatic nature until later stages, and its location deep in the pelvis. It is now recognised that most women diagnosed with ovarian cancer actually have symptoms, but they can easily be confused with those of the gastrointestinal tract (e.g. meteorism, changes in bowel habits, unexplained weight loss, and abdominal swelling) . Another important factor influencing the prognosis of women with ovarian cancer is the referral to a gynaecologic oncology centre for further diagnosis and staging, debulking surgery, and interdisciplinary tumour board evaluation . Although such centralised care is recommended in many countries, a large proportion of women with ovarian cancer remain treated by general surgeons and clinicians .
The goal of preoperative (clinical) staging of ovarian cancer is (1) the confirmation of a malignant adnexal mass and exclusion of a primary tumour in the gastrointestinal tract or pancreas, whose metastatic spread might mimic primary ovarian cancer; (2) assessment of tumour burden and mapping of the distribution of metastases; and (3) diagnosis of possible complications (e.g. bowel obstruction, hydronephrosis, or venous thrombosis) . Ovarian cancers spread mainly by local extension, by intra-abdominal dissemination, and by lymphatic dissemination, and rarely through the blood stream . The International Federation of Gynecology and Obstetrics (FIGO) Committee on Gynecologic Oncology is responsible for the staging system that is used internationally today . It is also useful, however, to be aware of the equivalents TNM (primary Tumour, regional lymph Nodes and distant Metastases) staging system developed within the International Union Against Cancer and the American Joint Commitee on Cancer . The two staging classification systems are presented in Table 1 . At present, surgical staging remains the gold standard for staging of ovarian cancer . Operative findings, before tumour debulking, determine the stage that may be modified by histopathologic as well as clinical or radiological findings . Histologic confirmation of the disease should be made on the basis of biopsies of all suspicious sites relevant for staging, such as omentum, mesentery, liver, diaphragm, pelvic, and paraaortic lymph nodes. Imaging studies and serum tumour markers may be helpful in diagnosis and follow up of the tumours. Serum CA-125 results can give some information on the specific nature of an adnexal mass (e.g. the median CA125 value for advanced ovarian cancer has been reported to be more than 400 U/mL compared with 99 U/mL for metastatic cancer) . Measurement of carcinoembryonic antigen should be considered to rule out a primary tumour other than ovarian cancer . If the CA 125 kU/L/carcinoembryonic antigen (ng/mL) ratio is 25 or less, a primary gastric, colon, or breast carcinoma should be excluded using imaging, endoscopy (gastroscopy, colonoscopy), and biopsy .
| TNM | FIGO stage | Clinical staging (Imaging findings) | Additional cytological/histological findings necessary to stage disease |
|---|---|---|---|
| T1 | I | Tumour limited to the ovary/ovaries | |
| T1a | IA | Tumour limited to the ovary, intact capsule and no tumour on the external surface. | No malignant cells in the ascites or peritoneal washings. |
| T1b | IB | Tumour limited to both ovaries, intact capsule and no tumour on the external surface. | No malignant cells in the ascites or peritoneal washings. |
| T1c | IC | Tumour on surface or ruptured capsule | Ascites or peritoneal washings positive for malignant cells. |
| T2 | II | Tumour growth involving one or both ovaries with pelvic extension | |
| T2a | IIA | Extension, metastases, or both to the uterus, Fallopian tubes, or both | No malignant cells in the ascites or peritoneal washings. |
| T2b | IIB | Extension to other pelvic tissues | No malignant cells in the ascites or peritoneal washings. |
| T2c | IIC | Tumour either IIA or IIB | Ascites or peritoneal washings positive for malignant cells. |
| T3 and/or N1 | III | Tumour involving one or both ovaries, with peritoneal metastases outside the pelvis, including liver surface implants, regional lymph node metastases, or both. | |
| T3a | IIIA | Tumour grossly limited to the true pelvis, ascites | Microscopic peritoneal metastases beyond the pelvis. |
| T3b | IIIB | Peritoneal implants of abdominal surfaces 2 cm or less. | Macroscopic peritoneal metastases ≤ 2 cm beyond the pelvis. |
| T3c and/or N1 | IIIC | Peritoneal implants of abdominal surfaces >2 cm, retroperitoneal or inguinal lymph node infiltration, or both. | Macroscopic peritoneal metastases >2 cm beyond the pelvis, regional lymph node infiltration, or both. |
| M1 | IV | Distant metastasis/metastases, including parenchymal liver metastases, cytological confirmation of malignant cells in pleural effusion, or both. | Distant metastasis/metastases, pleural effusion with positive cytology, or both. |
a Supplementary use of image-guided aspiration or biopsy provides the additional information necessary for preoperative staging and increases its accuracy. FIGO, International Federation of Gynecology and Obstetrics; TNM, The TNM classification of malignant tumours.
The precise assessment of tumour extent is the basis for the evaluation of feasibility of surgery. The ultimate goal of debulking surgery is complete macroscopic tumour resection (i.e. optimal cytoreduction) . Postoperative residual tumour is the strongest independent prognostic factor after tumour stage . Imaging aims to identify women unfit for surgery by depicting disease extent beyond the reach of surgery ( Table 2 ) . For those selected women, delayed primary surgery (i.e. interval debulking surgery) after three courses of neoadjuvant chemotherapy is an option . Before starting neoadjuvant chemotherapy, an image–guided tru-cut biopsy of the primary tumour or one of the metastases should prove the presence of an ovarian carcinoma . Many scoring systems based on CA 125 levels, ultrasound, computed tomography, magnetic resonance imaging (MRI), positron emission tomography (PET) combined with computed tomography, laparoscopy, performance status, the American Society of Anesthesiologists physical status classification system (ASA score), age, coincidental morbidity, and FIGO stage have been explored to define preoperatively if the woman is suitable for undergoing optimal cytoreduction . These studies, however, have at least two limitations. First, with the purpose of minimising the risk of misclassifying woman who potentially could have achieved complete cytoreduction as unsuitable for primary surgery, the studies demonstrated high rates (40–70%) of unnecessary laparotomy procedures . Second, when promising computed tomography-based scoring systems underwent external validation, low positive predictive values of computed tomography predictors of suboptimal cytoreduction were reported, suggesting that most women predicted to have suboptimal cytoreduction would in fact be optimally cytoreduced . This implies that computed tomography predictors should be used with caution. An ideal imaging method that allows a precise selection of women to an appropriate treatment does not exist. To choose the optimal treatment, one needs to consider patient characteristics, tumour biology, and also infrastructure, surgical skills, and experience . A standard staging imaging modality should always provide a structured report containing a detailed evaluation of all potentially non-resectable tumour sites as described in Table 2 . Such a description is essential when deciding on treatment.
| Criteria for patient selection to primary chemotherapy and interval debulking surgery in FIGO stages IIIC and IV ovarian carcinoma | |
|---|---|
| Tumour growth pattern determined using imaging or during surgery | Non-resectable metastases in the liver parenchyma. |
| Tumour involving duodenum, large parts of the pancreas (not only tail), or both. | |
| Tumour infiltrating large vessels of celiac trunc and hepatoduodenal ligamentum, and metastases behind the porta hepatis. | |
| Deep tumoral involvement of the superior mesenteric artery and mesenterial root of the small bowel. | |
| Diffuse and confluent carcinomatosis of the stomach, small bowel, or both, involving such large parts that resection would lead to a total gastrectomy or a short bowel syndrome. | |
| Not completely resectable extra-abdominal metastases. | |
| Patients’ characteristics | Impaired performance status and co-morbidity not allowing maximal surgical effort to achieve complete resection. |
| Patient’s non-acceptance of potential supportive treatment, such as blood transfusions or temporary stoma. | |
a Inoperability because of tumour spread does not depend on size of metastasis but on location and extent of infiltration . Pleural fluid with cytological proof of tumour but without detectable macroscopic tumour deposits on imaging is not regarded as a marker of primary inoperability. FIGO, International Federation of Gynecology and Obstetrics.
Initial response rates to the standard primary treatment (i.e. surgical cytoreduction and adjuvant platinum-based combined chemotherapy) are about 80%. On completion of treatment, the treatment effect (i.e. re-staging) is evaluated using the same imaging methods as for initial staging. In about 20–30% of women with early stage disease and 50–75% of those with advanced disease who obtain a complete response after first-line chemotherapy, disease will ultimately recur . Recurrent disease usually involves the pelvis and the abdomen . After first-line treatment, the women are usually monitored using a combination of clinical assessment and measurement of serum CA 125 levels supplemented by imaging if clinically indicated . A large Medical Research Council OV05/EORTC 55955 collaborative trial on follow up (the Medical Research Council OV05/European Organisation for Research and Treatment of Cancer 55955) , showed that early diagnosis of recurrence shortens the disease-free period and treatment-free period without any identifiable effect on overall survival. This study was criticised because the only treatment modality of a recurrent tumour was chemotherapy. Data from retrospective studies, however, have shown that surgery for platinum-sensitive recurrent ovarian cancer might be beneficial if it results in complete resection of the disease . Randomised trials (DESKTOP III and GOG 213) evaluating the effect of surgery in recurrent tumours are ongoing. Thus, the results of MRC OV05/EORTC 55955 trial must be interpreted with caution, and the monitoring of patients should not be abandoned. The clinical evaluation is not only useful to detect tumour recurrence, but also to evaluate the complications of disease progression or treatment. CA 125 levels can be used to predict tumour recurrence in women who are clinically tumour free. Rising CA 125 levels may precede clinical recurrence, with a median lead time of 3–5 months . CA 125 monitoring is recommended in the European Society of Gynaecological Oncology statement for patients after complete response to primary treatment for ovarian cancer who (1) have been, or are being treated, as a part of a clinical trial; (2) are considered for (future) studies on second-line treatment; (3) will not have routine follow up (every 3 months), including regular imaging; and (4) are eligible for secondary surgery at recurrence . Imaging is usually used to investigate suspected relapse with rising CA 125 or clinically suspicious symptoms. Women should be educated about the signs and symptoms suggestive of recurrence (e.g. pelvic pain, bloating, early satiety, obstruction, weight loss, and fatigue). Women treated with fertility-sparing surgery should always be monitored with ultrasound examinations .
On the basis of the above information, the following factors are crucial for patient management: early detection of a suspicious mass and referral to a gynaecologic oncology centre for further diagnosis; staging and debulking surgery; interdisciplinary tumour board evaluation; and follow up . The evolution in imaging has substantially changed the management of women with ovarian cancer. Imaging provides clinicians with the information necessary for the provision of optimal individualised treatment of women with newly diagnosed or recurrent ovarian cancer.
Imaging techniques
In this chapter, we focus on the estimation of ovarian cancer extent using ultrasound, computed tomography, MRI, and PET with computed tomography. We discuss their indications, advantages and limitations as presented in Tables 3 and 4 . Conventional imaging with ultrasonography, computed tomography and MRI may provide false–negative results because of their inability to identify disease when normal anatomic landmarks have been lost because of surgery or radiation. They may yield also false–positive results because of their inability to distinguish between viable tumour masses and masses of necrotic or scar tissue . The combined functional and anatomical methods, such as PET with computed tomography and perfusion sequences or diffusion weighted MRI (DWI) are therefore being used as additional imaging methods for initial staging and for the detection of recurrent disease .
| Indication | Imaging techniques | |||
|---|---|---|---|---|
| Ultrasound | Computed tomography | Magnetic resonance imaging | Positron emission tomography with computed tomography | |
| Detection and characterisation of adnexal mass | First-line imaging | – | Second-line imaging (indeterminated mass on ultrasound) c | – |
| Staging | First-line imaging a | First-line imaging | Second-line imaging (for the pelvis or if CT is contraindicated | Second-line imaging d (distant metastases) |
| Follow up | First-line imaging a | First-line imaging | Second-line imaging (pelvis) | Second-line imaging (rising CA 125 but negative standard imaging or if surgical resection is planned) |
| Image-guided biopsy | First-line technique | Second-line technique b | – | – |
a Ultrasound is a method of choice for staging and follow up if an experienced examiner is available and acoustic conditions are satisfactory. It is an ideal technique after fertility-sparing surgery;
b Ultrasound-guided biopsy is better tolerated by the patient, computed tomography guidance is reserved for less accesible tumour sites in the abdomen;
c Magnetic resonance imaging increases the specificity to detect malignancy . It is not known, however, if magnetic resonance imaging adds anything if the ultrasound examination has been carried out by and experienced ultrasound examiner (see chapter on adnexal masses in this issue of the Best Practice and Research Clinical Obstetrics and Gynaecology );
d Second-line imaging (positron emission tomography with computed tomography) may improve staging accuracy in suspected stage IV disease, in case of indeterminate lymph node appearance or when there is contraindication to contrast-enhanced computed tomography).
| Ultrasound | Computed tomography | Magnetic resonance imaging | Positron emission tomography with computed tomography | |
|---|---|---|---|---|
| Costs (approximate) | 1 x | 2 x | 4 x | 6 x |
| Availability | Universal | Most hospitals | Specialised centres | Specialised centres, such as university hospitals |
| Examination duration (minutes) | 15–20 | 1 | 30–45 | 30 |
| Dynamic examination a | Yes | No | No | No |
| Preparation before imaging | None | 4 h fasting | Antiperistaltic agents | 4 h fasting and 1 h physical rest before examination |
| Contrast agent | None | Iodine-based | Gadolinium-based b | FDG-radiotracer and iodine-based |
| Radiation exposure | None | 10–20 mSv | None | 20–30 mSv |
| Limitation for application | None | Contraindication for iodine based contrast agent: renal insufficiency, hyperthyreoidism, iodine allergy | Claustrophobia, metal components, cochlear implants, cardiac pacemaker | Contraindication for iodine-based contrast agent: renal insufficiency, hyperthyreoidism, iodine allergy |
| Dependence of expertise | Yes | Yes | Yes | Yes |
a Ultrasound can provide information on site-specific tenderness and information how pelvic structures move in relation to each other (sliding effect);
b In people with renal insufficiency gadolinium-based contrast media must be used with caution. FDG, 18 F-fluoro-deoxyglucose.
Imaging techniques
In this chapter, we focus on the estimation of ovarian cancer extent using ultrasound, computed tomography, MRI, and PET with computed tomography. We discuss their indications, advantages and limitations as presented in Tables 3 and 4 . Conventional imaging with ultrasonography, computed tomography and MRI may provide false–negative results because of their inability to identify disease when normal anatomic landmarks have been lost because of surgery or radiation. They may yield also false–positive results because of their inability to distinguish between viable tumour masses and masses of necrotic or scar tissue . The combined functional and anatomical methods, such as PET with computed tomography and perfusion sequences or diffusion weighted MRI (DWI) are therefore being used as additional imaging methods for initial staging and for the detection of recurrent disease .
| Indication | Imaging techniques | |||
|---|---|---|---|---|
| Ultrasound | Computed tomography | Magnetic resonance imaging | Positron emission tomography with computed tomography | |
| Detection and characterisation of adnexal mass | First-line imaging | – | Second-line imaging (indeterminated mass on ultrasound) c | – |
| Staging | First-line imaging a | First-line imaging | Second-line imaging (for the pelvis or if CT is contraindicated | Second-line imaging d (distant metastases) |
| Follow up | First-line imaging a | First-line imaging | Second-line imaging (pelvis) | Second-line imaging (rising CA 125 but negative standard imaging or if surgical resection is planned) |
| Image-guided biopsy | First-line technique | Second-line technique b | – | – |
a Ultrasound is a method of choice for staging and follow up if an experienced examiner is available and acoustic conditions are satisfactory. It is an ideal technique after fertility-sparing surgery;
b Ultrasound-guided biopsy is better tolerated by the patient, computed tomography guidance is reserved for less accesible tumour sites in the abdomen;
c Magnetic resonance imaging increases the specificity to detect malignancy . It is not known, however, if magnetic resonance imaging adds anything if the ultrasound examination has been carried out by and experienced ultrasound examiner (see chapter on adnexal masses in this issue of the Best Practice and Research Clinical Obstetrics and Gynaecology );
d Second-line imaging (positron emission tomography with computed tomography) may improve staging accuracy in suspected stage IV disease, in case of indeterminate lymph node appearance or when there is contraindication to contrast-enhanced computed tomography).
| Ultrasound | Computed tomography | Magnetic resonance imaging | Positron emission tomography with computed tomography | |
|---|---|---|---|---|
| Costs (approximate) | 1 x | 2 x | 4 x | 6 x |
| Availability | Universal | Most hospitals | Specialised centres | Specialised centres, such as university hospitals |
| Examination duration (minutes) | 15–20 | 1 | 30–45 | 30 |
| Dynamic examination a | Yes | No | No | No |
| Preparation before imaging | None | 4 h fasting | Antiperistaltic agents | 4 h fasting and 1 h physical rest before examination |
| Contrast agent | None | Iodine-based | Gadolinium-based b | FDG-radiotracer and iodine-based |
| Radiation exposure | None | 10–20 mSv | None | 20–30 mSv |
| Limitation for application | None | Contraindication for iodine based contrast agent: renal insufficiency, hyperthyreoidism, iodine allergy | Claustrophobia, metal components, cochlear implants, cardiac pacemaker | Contraindication for iodine-based contrast agent: renal insufficiency, hyperthyreoidism, iodine allergy |
| Dependence of expertise | Yes | Yes | Yes | Yes |
a Ultrasound can provide information on site-specific tenderness and information how pelvic structures move in relation to each other (sliding effect);
b In people with renal insufficiency gadolinium-based contrast media must be used with caution. FDG, 18 F-fluoro-deoxyglucose.
Ultrasound
Ultrasound is the primary imaging technique in ovarian cancer. Over the past decade, massive technical improvements in ultrasound have been made. If carried out by an experienced sonographer, ultrasound has an invaluable role in the primary diagnosis of ovarian cancer and in the assessment of tumour extent in the pelvis and abdominal cavity; in the evaluation of the effectiveness of treatment using RECIST criteria (Response Evaluation American Joint Committee on Cancer Criteria in Solid Tumours); and in follow up . Ultrasound is also a standard procedure for monitoring women treated with fertility-sparing surgery . Furthermore, it is an ideal technique to guide tru-cut biopsy for the collection of material for histology .
Advantages
Ultrasound is a commonly available, non-invasive, and inexpensive imaging method that can be carried out without any risk or discomfort to the patient. Ultrasound also has the advantage of being a dynamic and interactive examination that can provide information on how pelvic and abdominal structures move in relation to each other . Doppler ultrasonography can be added to gray-scale imaging to assess the presence, architecture, and density of blood vessels. Doppler sometimes helps to discriminate between benign and malignant ovarian tumours, and to evaluate changes in tumour viability after treatment . The use of ultrasound contrast media, especially for the differential diagnosis of hepatic lesions, and for ovarian lesions difficult to classify as benign or malignant, may also increase the sensitivity and specificity of this imaging modality . If the insertion of a transvaginal probe is not technically possible, the transrectal approach allows assessment of the pelvis with equal image quality and without the need for special preparation (e.g. clysma or fasting) ( Fig. 1 ). The combination of transvaginal and transabdominal ultrasound allows the complete assessment of the of the pelvis and abdomen for staging of ovarian cancer ( Fig. 2 ) .
Limitations
Operator skill in interpreting transvaginal and transabdominal ultrasound images, equipment, and patient body habitus, are thought to account for variability in diagnostic accuracy when performing a gynaecologic ultrasound examination for ovarian cancer detection and staging . Systematic documentation using videoclips, three-dimensional volumes, or static images in standard sections is mandatory for subsequent additional analysis or future evaluation of treatment effect. The image quality of the retroperitoneum is often limited in obese people, and in people who have undergone multiple laparotomies, causing adhesion of the intestinal loops. The evaluation of the liver in the presence of a large amount of ascites is difficult. Ultrasound has inherent limitations for the examination of the thorax and skeleton.
Technique and imaging investigations
A high-end ultrasound machine with sensitive Doppler, and equipped with endocavitary (microconvex) and transabdominal convex and linear array probes, is needed. A detailed review on how to scan gynaecological cancers for staging (e.g. methodology, terminology, clinical implementation) has been published . Staging of ovarian cancer by ultrasound includes a transvaginal scan of the whole pelvis and a transabdominal scan from the inguinal region to the diaphragm. Pleural effusion might be detected at transabdominal scan. If clinically indicated, a linear probe will enable evaluation of supraclavicular lymphadenopathy ( Fig. 3 ).
Transvaginal sonography is the most widely available technique to evaluate the gynaecological internal organs, but also adjacent pelvic structures (e.g. bladder, rectum, sigmoid colon, pelvic lymph nodes, and peritoneum) ( Fig. 4 ) . A high-resolution vaginal probe enables detailed assessment of all layers of the rectal and sigmoid colon, the bladder wall, and precise evaluation of the depth of tumour invasion ( Fig. 5 ). The high tissue resolution also helps to identify infiltrated pelvic lymph nodes based on their rounded shape and increased size, and also on the sonomorphological and vascular changes in the node . The structured report of transvaginal scan includes the description of (1) ovarian tumours using IOTA terminology (the International Ovarian Tumour Analysis), with a description of the capsule and external surface of the tumour ; (2) the uterus (including information about cervical or endometrial involvement, myometrial pathology, ventral, lateral, and dorsal parametria involvement); (3) bladder, ureteric, and bowel invasion, including the assessment of the surface of the ileal bowel loops; (4) pelvic peritoneal involvement (i.e. carcinomatosis), including the type of peritoneum affected, the extent of peritoneal involvement (focal or sheet-like), and the size of implants. Parietal carcinomatosis lines the pelvic wall, visceral carcinomatosis covers the visceral organs, and mesenterial carcinomatosis originates from the visceral peritoneum, which comes into contact with itself forming mesenterial peritoneum ; (5) infiltration of visceral lymph nodes located around the pelvic organs and of retroperitoneal lymph nodes around the external and internal iliac vessels up to inter-iliac bifurcation; (6) pelvic side wall invasion; and (7) evidence of bowel obstruction, hydroureter, venous obstruction, or thrombosis.
