Lesion Detection

In patients with known malignant tumor, PET/CT may be beneficial in the detection of metastases, particularly those distant from the primary tumor. In a study of colon cancer patients by Cohade and colleagues, 30% more lesions were detected by PET/CT than by CT alone. In this same study, the accuracy of tumor staging increased from 78% to 89% with PET/CT. Similar results were reported in a review by Subhas and colleagues on the use of PET in patients with cervical and endometrial cancer. These authors showed greater than 95% sensitivity and 76% specificity in detecting disease recurrence. Schiepers and colleagues reported sensitivities of 82% to 99% and specificities of approximately 99% for initial staging and posttherapy evaluation of non-Hodgkin lymphoma. In addition to detecting distant metastases in patients with newly diagnosed malignant tumors, PET and PET/CT are useful for detecting metastases in patients with rising tumor markers despite negative anatomic imaging (Figs. 4-1 and 4-2). Unfortunately, very small (<6 mm) lesions such as peritoneal implants often are not evident on PET despite increased metabolism.

Figure 4-1 A, Coronal MIP reconstruction of PET data from a patient with rising CEA after surgery for colon cancer and negative contrast-enhanced CT. B, Coronal PET/CT image in the same patient as in A.

Figure 4-2 Coronal PET image in a patient with gastrohepatic lymph node metastasis after surgery for colon cancer with rising CEA. No enlarged lymph nodes were visible on contrastenhanced CT.

When considering the use of PET to detect malignant lesions in the abdomen and pelvis, the avidity of 18F-FDG for various malignancies should be considered. The most common abdominal and pelvic tumors known to be strongly positive on 18F-FDG PET in the majority of cases are listed in Box 4-2.

Abdominal malignancies commonly regarded as showing no or little hypermetabolic activity on 18F-FDG PET include mucinous adenocarcinoma, well-differentiated hepatocellular carcinoma, and well-differentiated prostate cancer (Fig. 4-3). Testicular cancer shows variable 18F-FDG uptake. While high-grade lymphomas usually are hypermetabolic, low-grade lymphoma may be difficult to detect with PET/CT. Up to half of all renal cell carcinomas are not 18F-FDG avid (Fig. 4-4). 18F-FDG PET for renal cell carcinoma presents additional challenges even when the tumor is hypermetabolic. Renal cell carcinoma may be isointense to surrounding renal parenchyma on PET images, and renal excretion of 18F-FDG can obscure primary renal lesions adjacent to the renal collecting system. Despite these limitations, PET may be occasionally useful for finding distant metastatic deposits from renal cell carcinoma. In particular, 18F-FDG PET seems to be very sensitive for early detection of lytic bone metastases when compared with plain film, CT or bone scan. For tumors with variable avidity for 18F-FDG, if PET/CT shows hypermetabolic activity in lesions with anatomic correlates, then the study may be a reliable indicator of extent of disease.

Figure 4-3 Axial PET image at the level of the pancreatic head in a patient with a mucinous adenocarcinoma of the pancreas. Mucin-producing tumors can show variable uptake on PET, limiting sensitivity of this modality for these lesions.

Figure 4-4 Coronal fused PET/CT image in a patient with large right renal cell carcinoma that is not particularly FDG avid.

The utility of 18F-FDG PET is not limited to the detection of malignant neoplasms. The highly sensitive nature of PET imaging may be exploited for the evaluation of patients with fever of unknown origin, vasculitis, arthroplasty infections, and inflammatory disorders such as sarcoidosis (Fig. 4-5). Because PET and PET/CT are emerging technologies, not all potential applications are reimbursed by insurers. As with any imaging modality, incidental findings are common on PET and PET/CT scans (Figs. 4-6 and 4-7). Normal physiologic FDG uptake (e.g., bowel, brown fat) is also common and must be distinguished from pathologic FDG activity.

Figure 4-5 Coronal PET image in a patient with sarcoidosis (same patient as in Fig. 4-3).

Figure 4-6 A, Coronal PET image of a patient with metastatic colon cancer being evaluated for possible hepatic resection of solitary hepatic lesion. The patient also had early appendicitis accounting for increased metabolic activity near the cecum. B, Axial CT image of the same patient as in A shows thickened appendix with fat stranding corresponding to increased activity on PET scan and appendicitis at surgery.

Figure 4-7 A, Coronal PET image shows a hypometabolic pelvic mass (arrow) with concomitant urinary stasis due to unexpected intrauterine pregnancy. B, Sagittal reformation of the same patient as in A showing unexpected intrauterine pregnancy accounting for findings on PET portion of exam.