Date of Award

5-2022

Document Type

Dissertation

Degree Name

Master of Medicine (MMed)

First Supervisor/Advisor

ROSE NDUMIA

Second Supervisor/Advisor

SHAHIN SAYED

Third Supervisor/Advisor

STANLEY LUCHTERS

Department

Imaging and Diagnostic Radiology (East Africa)

Abstract

With emerging evidence showing that the incidence of breast cancer in Kenya is increasing, there is renewed emphasis on screening, early detection, and management of breast cancer. Early breast cancer (EBC) has a better prognosis and is potentially curable. Axillary nodal metastasis is an important prognostic indicator in breast cancer. Axillary ultrasound is the best modality for assessing, and managing, the affected nodes. It is a readily available and versatile tool with several sonographic nodal characteristics suggestive of metastatic involvement having been described in previous studies. Axillary ultrasound (AUS) performance in predicting nodal metastasis has yielded variable results which are mostly dependent on whether the authors included patients with larger tumors or additionally performed fine needle aspiration for cytology (FNA) or core needle biopsy (CNB) for nodes with suspicious ultrasound features. Sentinel lymph node biopsy (SLNB) is performed for node-negative patients while axillary lymph node dissection (ALND) is performed for those with confirmed nodal metastases following either of FNA, CNB or SLNB. Based on the varying study findings and the lack of a similar study in Kenya, this study is to evaluate the diagnostic accuracy of axillary ultrasound (AUS) in predicting axillary node metastatic involvement among patients being managed for EBC in Aga Khan University Hospital, Nairobi.

Primary objective: To assess the diagnostic accuracy of axillary ultrasound in predicting metastatic axillary nodal disease among patients managed for early primary breast cancer at Aga Khan University Hospital, Nairobi.

Methods: Our retrospective cross-sectional diagnostic accuracy study included cases diagnosed with early primary breast cancer with available AUS and histology results, used as a reference standard, of lymph nodes evaluated following either SLNB or ALND over a five-year period, from January 2016 to December 2020. We excluded male patients, those with intervening neoadjuvant chemotherapy and those with breast cancer recurrence. A total of 108 women diagnosed with EBC fulfilled the criteria.

Results: AUS was found to have sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy of 59.0%, 88.4%, and 74.2%, 79.2% and 77.8% respectively. ROC curves were plotted for nodal cortical thickness and short axis diameters iv revealing optimal cortical thickness cutoff of 2.7 mm with improved sensitivity of 79.7% albeit with lower specificity of 66.7%.

Conclusion: AUS performed similarly in a low resource setting where most ultrasounds are performed by sonographers with variable years of experience as in the western world in preoperative evaluation for nodal metastases in EBC.

Included in

Radiology Commons

Share

COinS