63-year-old male was referred to the endocrine clinic for the management of advanced thyroid cancer. Nine years before presentation, he underwent a thyroidectomy for a 7-cm Hürthle cell carcinoma with metastatic nodes in the lateral neck. A radioactive iodine ablation showed uptake only in the thyroid bed.
- Within 1 year, he had a macroscopic nodal recurrent in the lateral neck. A radioactive iodine whole-body scan did not show uptake in the metastatic tumor. A PET scan with 18F-fluorodeoxyglucose (FDG) showed increased uptake in nodes in the anterior mediastinum and lateral neck consistent with non-iodine avid metastatic thyroid cancer. His thyroglobulin levels have risen from 12 ng/mL in October 2010 to 20 ng/mL in October 2012 to the current levels of 1,607 ng/mL, suggesting progression of the metastatic thyroid cancer. His other significant medical history includes type 2 diabetes treated with metformin, pioglitazone and insulin glargine (Lantus, Sanofi-Aventis), and recent renal stones.
A recombinant human thyroid-stimulating hormone 18F-FDG PET/single-photon emission CT (SPECT) scan was performed. There were multiple hypermetabolic pleural-based masses in the right thoracic cavity. The two largest masses measured 4 cm x 2.4 cm and 3.3 cm x 3.5 cm with maximal standardized uptake value (SUVmax) of 13, consistent with metastatic disease. It was noted that there was intense 18F-FDG accumulation in the intestines (Figure 1, patient 1) compared with a different patient not taking metformin (Figure 1, patient 2). The CT scan showed a staghorn calculus in the right renal collecting system and a small left renal cyst. No renal mass was seen. The fused images of the 18F-FDG SPECT and CT scans clearly demonstrated the FDG accumulation occurred in the intestines. Thyroid cancer is not known to metastasize to the intestines. A literature search was performed.
A prospective study demonstrated that intestinal 18F-FDG uptake was significantly increased in patients with type 2 diabetes who take metformin compared with patients with diabetes who do not take metformin or controls without diabetes.
When seen, the intestinal uptake occurs in all segments of the intestine. Metformin treatment was associated with increased uptake in the small intestine (OR=15.9, P<.0001) and the colon (OR=95.3, P<.0001) compared with patients not taking metformin. The uptake is typically intense and diffuse in the small and large intestine in both the bowel wall and lumen (Figure 2). The pattern can be variable with some patients showing focal or segmental uptake as shown in this patient (Figure 1, patient 1) compared with minimal intestinal uptake in patients without diabetes or those with diabetes not taking metformin (Figure 1, patient 2).
Figure 1. Coronal views of 18F-FDG SPECT/CT scan. Patient 1: Patient with diabetes mellitus taking metformin. Patient 1 has intense segmental uptake (green arrows) in his large and small bowels compared with Patient 2 who does not have diabetes mellitus and is not taking metformin.
The focal or nodular appearance of uptake is similar to the appearance of cancerous or precancerous lesions. Segmental high 18F-FDG uptake may imply inflammation, whereas diffusely high uptake (Figures 2, 3) is not often associated with pathology. The mechanism of increased intestinal FDG uptake in the gut by metformin is not clear; 18 F-FDG and D-glucose are transported into cells by glucose transporters. Many malignant tumors express higher numbers of glucose transporters with a higher affinity for glucose than normal cells, which permits increased glucose intake into the cancerous cells. It has been proposed that the 18F-FDG uptake in the digestive tract caused by swallowed secretions, metabolically active smooth muscle or mucosa or colonic bacterial uptake.
Figure 2. Axial views of 18F-FDG SPECT/CT scan. Patient 3: Patient with diabetes mellitus taking metformin. Patient 3 has intense and diffuse uptake (green arrows) in the transverse colon.
An unknown, reversible condition
Metformin-induced high 18F-FDG uptake in the intestine may result in false-negative and false-positive results. Fortunately, this phenomenon is reversible. A small study of 10 patients with paired 18F-FDG PET scans before and after stopping metformin demonstrates that stopping metformin 2 days before significantly reduces the gut uptake in all segments of the intestines while having no significant changes in blood glucose level. Two of the 10 patients in this small study with paired scans revealed colorectal malignancies that could not be identified while the patient was on metformin.
This observation that metformin significantly increases 18F-FDG uptake in the colon and small intestine is largely unknown to endocrinologists. Because metformin is now the first-line drug for the management of type 2 diabetes, it is important that endocrinologists recognize that metformin may interfere with the interpretation of 18F-FDG uptake in the abdomen.
The intense gut uptake can mask a malignant lesion and lead to a false-negative 18F-FDG PET scan. Similar to the protocol to stop metformin before contrast administration to protect renal function, it also is important to stop metformin before an 18F-FDG scan to prevent increased intestinal 18F-FDG uptake.
Figure 3. Colorized 18F-FDG SPECT PET scan fused with CT scan. Another example of intestinal uptake in a patient with diabetes mellitus taking metformin. Patient 4 has intense and diffuse uptake (green arrows) in the small and large intestines. A. Coronal view. B. Mid-abdomen axial view. C. Lower pelvis axial view.
Images reprinted with permission from: Stephanie L. Lee, MD, PhD
Metformin cessation 48 hours before a PET scan is apparently sufficient to reduce bowel uptake in most patients without a significant increase in glucose levels. It is required that the blood sugar is less than an institutional threshold (200 mg/dL at Boston Medical Center, but may be as low as 140 mg/dL at other PET centers) to prevent a false-negative 18F-FDG scan.
Despite concerns that stopping metformin may allow sugars to rise and compete with 18F-FDG to prevent tumor uptake, limited clinical data suggest this is not a major problem. Future studies are needed to more clearly define this phenomenon.