Imaging of Lymph Nodes

Contents

Abstract

This chapter provides an overview of lymph node imaging in the female pelvis. We start by outlining the importance and implications of detecting metastatic lymph nodes in gynecological cancer patients. Thereafter, indications are given about the use of different magnetic resonance imaging (MRI) sequences (e.g., T2-weighted and diffusion-weighted MRI), as well as computed tomography (CT), in the context of metastatic lymph node detection. Further, we discuss the potential advantages and disadvantages of intravenous unspecific and tissue-specific contrast agents. The chapter concludes with a description of common imaging findings in benign and malignant lymph nodes, together with corresponding MRI and CT examples.

S. Barbieri • K.H. Härmä • H.C. Thoeny (*) Department of Diagnostic, Interventional and Pediatric Radiology, Inselspital, University of Bern, Bern, Switzerland

e-mail: sebastiano.barbieri@insel.ch; kirsihannele. haermae@insel.ch; harriet.thoeny@insel.ch

1\ Background

The most common gynecological malignancies are endometrial cancer and uterine sarcomas (estimated number of new cases in Europe in 2012: 98,900 women; mortality: 23,700 women), ovarian cancer (incidence: 65,500 women; mortality: 42,700 women), and uterine cervical cancer (incidence: 58,300 women; mortality: 24,400 women) (Ferlay et al. 2013). According to the guidelines published by the

International Federation of Gynecology and Obstetrics (FIGO) cancer of the corpus uteri is up-staged because of metastases to pelvic and/ or para-aortic lymph nodes; similarly, ovarian cancer is up-staged because of positive retroperitoneal lymph nodes (Stage III) and further up-staged if cardiophrenic lymph nodes are considered or proven to be metastatic (Stage IVB) (Berek et al. 2015).

In ovarian cancer, lymph node enlargement above the renal hilum suggests that neo-adju- vant chemotherapy (NACT) might be preferable to primary debulking surgery (PDS) aiming at optimal cytoreduction (Qayyum et al. 2005; Vergote et al. 2010). Cardiophrenic lymph nodes larger than 5 mm are also of interest: upon review of computed tomography (CT) scans of 78 ovarian carcinoma patients, they have been found to be a significant adverse prognostic factor for both progression-free survival and overall survival and to be associated with peritoneal metastases. Thus, the involvement of paracardiac lymph nodes should be regarded as suspicious for stage IV ovarian cancer (Holloway et al. 1997). These results are supported by a more recent retrospective study on 31 patients which suggests that a cut-off value of 7 mm on the short axis can be used to detect metastatic cardiophrenic lymph nodes in advanced ovarian cancer patients with 83% specificity and 63% sensitivity (Raban et al. 2015).

In another study, a significant number of patients (20 out of 30) with advanced epithelial ovarian cancer (EOC) showed supradiaphragmatic lymph node metastases in pretreatment PET-CT findings, suggesting that the route of EOC cells from the peritoneal cavity to the lymphatic system permeates the diaphragm mainly to cardiophrenic lymph nodes and continues to parasternal lymph nodes (Hynninen et al. 2012). The lack of histopathological correlation is a substantial limitation of many studies on cardiophrenic lymph node involvement in EOC patients, due to the fact that it requires additional thoracic surgery and is generally not part of clinical routine. In a recent study, suspicious cardiophrenic

lymph nodes with a short axis greater than 10 mm on preoperative CT were removed in case of optimal intra-abdominal debulking; 90% of patients (27 out of 30) had histologically confirmed metastases (Prader et al. 2016).

In the future, further knowledge about the involvement of small cardiophrenic lymph nodes, together with growing experience and technical advances in transdiaphragmatic cardiophrenic lymph node resection, could have an impact on whether NACT or PDS is chosen to treat advanced ovarian cancer patients.

Lymph node involvement has considerable implications on patients’ treatment management and prognosis. Metastatic lymph nodes have been found to correlate with histological tumor grade and depth of myometrial invasion (Boronow et al. 1984) and to be associated with both progression-free survival and overall survival (Polterauer et al. 2012). Further, pelvic radiotherapy might be unnecessary for endometrial cancer patients with negative lymph nodes after extended pelvic lymph node dissection (Bottke et al. 2007). Recently, risk-scoring systems have been published to predict lymph node involvement and distant metastases in endometrial carcinoma; a cut-off on the corresponding risk score was found to be associated with an acceptable lymphadenectomy rate (Tuomi et al. 2015). Further, lymphatic dissemination and high-risk tumor features as per the risk-scoring systems were found to be predictors of survival independently of tumor stage (Tuomi et al. 2017).

Metastatic pelvic lymph nodes are recognized as an important prognostic factor in uterine cervical cancer as well (Piver and Chung 1975). In a retrospective study on 127 locally advanced cervical cancer patients, the status of lower pelvic lymph nodes was able to predict the pathologically assessed status of upper pelvic lymph nodes and of the parametrium (Ferrandina et al. 2007). In another study on 70 consecutive patients, lymph node metastases were significantly associated with a higher likelihood of uterine body involvement and of higher FIGO

stage. Moreover, the average tumor volume was determined to be larger in node-positive patients (69 cm3) than in node-negative patients (49 cm3) (Narayan et al. 2003).

2\ Indications and Value

of Imaging Techniques

There are no specific indications for the imaging of pelvic lymph nodes by magnetic resonance imaging (MRI) or CT. Lymph node evaluation is always done in conjunction with imaging performed for tumor staging and evaluation of general metastatic spread.

Traditionally, the standard technique used to differentiate normal from metastatic lymph nodes consists in applying a threshold to the short axis diameter of the node: lymph nodes with a short axis of 10 mm or more are considered metastatic. Nonetheless, this method does not allow discriminating between enlarged inflammatory lymph nodes and metastatic ones; further, the employed threshold value is a matter of debate: 10 mm guarantee a relatively high specificity (90% or more), but are also associated with low sensitivity; increasing this threshold value, e.g., to 12 mm, increases sensitivity but decreases specificity (Jager et al. 1996; Yang et al. 2000).

In recent years, diffusion-weighted magnetic resonance imaging (DW-MRI) has emerged as a promising additional technique for detecting metastatic lymph nodes. Due to the impeded diffusivity within metastatic tissue, positive lymph nodes appear as noncontinuous, hyperintense, ovoid or round structures on high b-value images (b in the 800–1000 s/mm2 range) and as hypointense structures on corresponding apparent diffusion coefficient maps (ADC). Several studies on cervical and uterine cancer patients have reported significant differences in ADC values between malignant and benign lymph nodes (Kim et al. 2008; Lin et al. 2008; Koplay et al. 2014). In one study the combined analysis of ADC values and nodal size increased sensitivity to 83%, compared with 25% using morphological MRI alone; at the same time specificities remained high at

99% and 98%, respectively (Lin et al. 2008). The smallest metastatic lymph node detected by combined diffusion-weighted and morphological MRI was 5 mm in short axis diameter. A recent meta-analysis of studies on cervical cancer patients indicates that, using DW-MRI, metastatic lymph nodes might be detected with a pooled sensitivity of 86% (95% confidence interval, CI: 84–89%) and a pooled specificity of 84% (CI: 83–89%) (Shen et al. 2015). Nevertheless, there is a considerable overlap between ADC values measured in positive and negative lymph nodes. Another study, possibly due to the relatively small patient sample, did not determine a significantly lower ADC in metastatic lymph nodes (Nakai et al. 2008). It is worth mentioning that also the authors of the latter study conclude that DW-MRI improved detection of lymph nodes compared with morphological MRI alone. While several threshold values on ADC have been proposed for the detection of metastatic lymph nodes (e.g., ADC value below 0.10 × 10−3 mm2/s in (Lin et al. 2008)), the measured ADC values might depend on the hardware characteristics and field strength of the employed MR scanner which should therefore be carefully calibrated at each center (Donati et al. 2014). It should also be noted that ADC measurements might be underestimated due to the presence of fatty hila, overestimated because of necrosis, or generally affected by partial volume artifacts since lymph nodes are small relative to image resolution. Finally, to detect metastatic lymph nodes, we advise to always correlate DW-MRI with T2-weighted MRI (Thoeny et al. 2014).

Another interesting development is represented by hybrid imaging techniques such as positron emission tomography (PET)-CT and PET-MRI. A meta-analysis of cervical cancer studies indicates that the pooled sensitivity of PET or PET-CT for detecting nodal metastases is 54% (CI: 46–61%), significantly higher than morphological MRI (38%, CI: 32–43%), but similar to CT alone (52%, CI: 42–62%). On the contrary, the pooled specificity of PET or PET-CT is 97% (CI: 96–98%), significantly