Oncology imaging in the abdomen and pelvis: where cancer hides

Steven C Eberhardt,Jennifer A Johnson,Rosaleen B Parsons
DOI: https://doi.org/10.1007/s00261-012-9941-z
Abstract:As the incidence of cancer continues to increase, imaging will play an ever more important role in the detection, diagnosis, staging, surveillance, and therapeutic monitoring of cancer. Diagnostic errors in the initial discovery of cancer or at follow-up assessments can lead to missed opportunities for curative treatments or altering or reinitiating therapies, as well as adversely impact clinical trials. Radiologists must have an understanding of cancer biology, treatments, and imaging appearance of therapeutic effects and be mindful that metastatic disease can involve virtually any organ system. Knowledge of patient history and tumor biology allows for optimizing imaging protocols. The majority of cancer imaging utilizes computed tomography, where contrast enhancement characteristics of lesions can be exploited and detection of subtle lesions can involve manipulation of window width and level settings, multiplanar reconstruction, and maximum intensity projections. For magnetic resonance imaging, diffusion-weighted imaging can render lesions more conspicuous, improve characterization, and help assess therapeutic response. Positron emission tomography with (18)F-labeled fluorodeoxyglucose and sodium fluoride are invaluable in detecting occult existing and new cancerous lesions, characterizing indeterminate lesions, and assessing treatment effects. The most common anatomic "hiding places" for cancer include metastases to solid organs, such as the kidneys and pancreas, gastrointestinal tract, peritoneum and retroperitoneum, neural axis, muscular body wall, and bones. Consistent work habits, employment of appropriate technologies, and particular attention to the above anatomic areas can enhance detection, staging, and reassessments of these complex and often stealthy diseases, ensuring the radiologists' integral role in the cancer care team.
What problem does this paper attempt to address?