Gastrointestinal Imaging
April 2000

FDG PET Evaluation of Mucinous Neoplasms: Correlation of FDG Uptake with Histopathologic Features


OBJECTIVE. Our goal was to assess the sensitivity of positron emission tomography (PET) with 18F-fluorodeoxyglucose (FDG) for the detection of mucinous carcinoma and to determine the histologic features of these tumors that may affect lesion detectability.
MATERIALS AND METHODS. A retrospective review of all patients with mucinous carcinoma who had undergone FDG PET at our institution from 1995 through 1998 identified 25 patients with new or recurrent mucinous carcinoma at the time of PET. In 22 of these patients, tissue specimens available from either core biopsy or surgical resection allowed detailed histologic analysis.
RESULTS. FDG PET revealed mucinous carcinoma in only 13 (59%) of 22 patients, resulting in an unusually high percentage (41%) of false-negative results. Two histologic features were found to be predictive of FDG PET results: tumor cellularity (p = 0.011) and the amount of mucin within the tumor mass (p = 0.009). There was a positive correlation between tumor FDG uptake and cellularity but a negative correlation with the amount of mucin.
CONCLUSION. FDG PET is limited in the evaluation of mucinous tumors, particularly in hypocellular lesions with abundant mucin.


Positron emission tomography (PET) with 18F-fluorodeoxyglucose (FDG) has gained acceptance in clinical oncology for the detection and staging of a variety of tumors [1]. It is known that most malignant tumors have a higher rate of glucose metabolism than normal tissues. However, there is a wide variation in tumor glucose metabolic rate, depending on the histologic type and the aggressiveness of the tumor [2, 3]. Among malignant tumors, less aggressive and slow-growing lesions exhibit a lower glucose metabolic rate than do more aggressive, rapidly growing lesions. Conversely, benign lesions, such as some inflammatory processes, may show a high rate of glucose metabolism. Knowledge of the potential limitations of FDG PET with particular histologic subtypes is essential for proper application of this technique.
We recently examined a group of patients with mucinous carcinomas that exhibited poor accumulation of FDG. Mucinous carcinomas are commonly found in the gastrointestinal tract and represent approximately 17% of colonic tumors [4]. These tumors contain clear, gelatinous fluid (mucin), which may be intracellular or extracellular. We postulated that FDG PET might be insensitive in showing mucinous carcinomas because of the low cellularity of these tumors caused by the presence of mucin. To test this hypothesis, we retrospectively reviewed our experience with mucinous carcinomas and correlated the results of FDG PET with the histologic features of these tumors.

Materials and Methods


Review of the records of all patients who had undergone clinical FDG PET examinations at our institution from 1995 through 1998 identified 25 patients (18 men and seven women; mean age, 63.7 years; age range, 30-78 years) with mucinous carcinomas. Three patients were excluded because of a lack of concurrent tissue sampling (recurrent colorectal cancer from documented mucinous primary tumors, n = 2) or inadequate tissue sampling to permit detailed histologic analysis (primary lung cancer, n = 1). We included patients in whom mucin was explicitly identified in the pathologic description of the neoplasm. For colorectal carcinomas, tumors had to contain at least 50% mucin to be considered a mucinous carcinoma [5]. For breast carcinoma, tumors had to contain large amounts of extracellular epithelial mucus, sufficient to be visible grossly and recognizable microscopically surrounding and within tumor cells (World Health Organization classification criteria for mucinous breast tumors) [6]. For diagnosis of mucinous tumors of other origins, no specific percentage of mucin has been established as a criterion, to our knowledge, in the literature; for the purpose of this study, all tumors that were characterized as mucinous tumors by histopathologic analysis were included.
Twelve of the 22 patients were examined for new primary carcinomas and the remaining 10 patients were examined for suspected recurrent disease. Either a core biopsy or a surgical specimen was available for these 22 patients, thus allowing detailed histologic analysis.


PET was performed using an ECAT EXACT scanner (Siemens-CTI, Knoxville, TN) beginning approximately 40 min after IV administration of 10-15 mCi (370-555 MBq) of FDG. A series of three to five overlapping 47-slice emission and transmission PET images was obtained to include the region from the middle of the neck to the upper thighs. Activity in kidneys, ureters, and bladder was minimized by IV hydration, diuretic administration, and bladder catheterization during the study, as previously described [7]. The PET images in 12 patients were reconstructed by filtered back projection using a Hanning filter (frequency cutoff, 0.6 × Nyquist value). In the remaining 10 patients, the PET images were reconstructed using an ordered-subset estimation-maximization (OSEM) iterative algorithm and a Butterworth filter (frequency-cutoff range, 0.2-0.4 cycles per pixel). The emission images were corrected for measured attenuation using a local threshold for segmented attenuation [8]. Images were displayed in three orthogonal projections and as whole-body, maximum-pixel-intensity reprojection images for visual interpretation.
All images were interpreted using all available clinical information and correlative anatomic studies (CT, MR imaging, or both). All PET images were evaluated qualitatively by the consensus of at least two observers. For the purpose of this study, the original clinical PET interpretation was used. Image interpretation criteria were those routinely used in scintigraphic imaging. On the basis of knowledge of the normal distribution of FDG, lesions were identified as abnormal or as representing tumor, if the accumulation of FDG was moderately to markedly increased relative to comparable normal contralateral structures or surrounding soft tissues. Mildly increased activity or no increased activity (in the case of an abnormality identified on CT or MR imaging with no corresponding PET abnormality) was considered normal or benign disease.
In addition to qualitative visual analysis of FDG uptake, appropriate regions of interest were drawn around the lesions and around the comparable normal tissues for determination of the simple ratio of tumor-to-normal tissue FDG uptake. We were unable to determine the standardized uptake values of lesions because a recent software upgrade of our computer system made reprocessing of most of the older studies, to allow calibration of the images in standardized uptake value units, impossible.

Pathologic Analysis

Tissue specimens were processed by standard technique (all tissues were fixed in formaldehyde solution, embedded in paraffin, and sectioned; the sections were then stained with H and E). All surgical pathology specimens (i.e., core biopsy or surgical specimens) for all 22 patients with adequate tissue samples were reevaluated in detail for tumor grade on the basis of mitotic figures, nuclear enlargement, and nuclear pleomorphism; cellularity; and the amount of mucin using standard pathologic criteria [9]. The appropriate H and E—stained slides (tissue sections from the tumor mass alone) for each patient were reevaluated by one observer who was unaware of the results of the PET studies. The number of slides reviewed from each patient ranged from one to 10 with a mean of four slides per patient. Each slide was reviewed at low- and high-power objectives (2×, 4×, and 10×) for assessment of the percentage of surface area occupied by solid tumor (tumor cells) and mucin. Each slide was then reexamined at 4× and 10× power objectives, focusing specifically on the mucinous component. After all relevant slides had been examined, tumor grade, the percentage of tumor cellularity, and mucin content for each patient were calculated and tabulated.

Statistical Analysis

To determine the relationship between tumor FDG uptake and the histologic features of the tumor, the tumor-to-normal tissue activity ratio was compared with tumor cellularity, mucin content, and grade using a nonparametric Spearman's rank correlation test. A p value of less than 0.05 was considered indicative of a statistically significant correlation. Sensitivity of FDG PET for detection of mucinous carcinoma was calculated in the standard manner.


All patients had active disease proven by histopathologic examination of concurrent core biopsy or surgical specimen.

PET Results

By qualitative evaluation, FDG PET enabled correct identification of tumor in 13 of the 22 patients. In the remaining nine patients with malignant lesions (recurrent colorectal cancer, n = 5; primary cancer of the esophagus and gastroesophageal junction, n = 2; primary lung cancer, n = 1; and metastatic breast cancer, n = 1), FDG PET showed no or minimal uptake relative to corresponding normal tissues (Figs. 1A, 1B and 2A, 2B). The mean tumor-to-normal tissue ratio was 5.3 ± 2.5 (range, 1.5-9.0) in the 13 lesions detected on PET and 1.2 ± 0.3 (range, 0.5-1.8) in the nine lesions not detected. All the false-negative tumor sites measured at least 1 cm in diameter (range, 1.0-5.0 cm) on either correlative anatomic imaging studies or pathologic specimens; only three of the nine lesions were smaller than 2.0 cm.
Fig. 1A. —Recurrent rectal carcinoma in 35-year-old man. CT scan of lower pelvis shows prominent presacral soft tissue (solid arrow), which was subsequently confirmed as recurrent tumor. Note enlarged prostate gland with irregular margins (open arrow), which is suggestive of prostatitis. B = bladder.
Fig. 1B. —Recurrent rectal carcinoma in 35-year-old man. Corresponding transaxial positron emission tomograph reveals very mildly increased 18F-fluorodeoxyglucose (FDG) uptake in perirectal soft tissues (solid arrow) that was incorrectly interpreted as representing posttherapeutic changes. Note mildly increased FDG uptake in prostatic bed (open arrow), consistent with prostatitis. B = bladder.
Fig. 2A. —Recurrent colonic cancer with hepatic metastases in 55-year-old man. Contrast-enhanced CT scan through upper abdomen shows multiple low-attenuation hepatic lesions, some of which are calcified, which are consistent with metastases.
Fig. 2B. —Recurrent colonic cancer with hepatic metastases in 55-year-old man. Corresponding transaxial positron emission tomograph reveals a few areas of mildly increased uptake (arrows) in liver, roughly corresponding to some abnormalities in right hepatic lobe seen on A. Most lesions seen on A exhibit 18F-fluorodeoxyglucose uptake similar to that of normal hepatic parenchyma.
In our group of 22 patients, the overall sensitivity of FDG PET for the detection of mucinous neoplasm was 59%.

Pathologic Results

Of the 22 cases in which tumors were pathologically reevaluated, 12 represented primary resections, and 10 were biopsies of recurrent or metastatic lesions. Ten of 12 primary mucinous adenocarcinomas originated in the gastrointestinal tract (distal esophagus—gastroesophageal junction, n = 8; pancreas, n = 2) and two originated in the lung. In one of the 12 patients, pulmonary biopsy revealed mucinous adenocarcinoma, but whether this was a primary lung cancer or a metastasis from an unknown primary carcinoma could not be determined. Biopsy sites of metastatic tumors included the liver (n = 3), soft tissue (n = 3), lung (n = 2), lymph node (n = 1), and bone (n = 1). Nine of these metastatic lesions had originated in the gastrointestinal tract (colorectal) and one in the breast.
The failure of PET to reveal tumor foci significantly correlated with low tumor cellularity (p = 0.011) and overall abundance of mucin (p = 0.009). There was no significant correlation between FDG uptake and tumor grade (p = 0.620).


Mucinous carcinoma is an epithelial neoplasm commonly found in the gastrointestinal tract. Mucins are high-molecular-weight glycoproteins. Except for colorectal cancer, for which clinical studies have shown that the presence of mucin is associated with a lower rate of survival [10, 11], mucin production has not been considered an indicator of the aggressiveness of other types of tumors and can be seen in both invasive and noninvasive tumors. However, recently, a relationship between the expression of special types of mucin core proteins and prognosis for some gastrointestinal cancers has been reported [12, 13]. Immunohistochemical studies of the structure of several mucin core proteins in gastric carcinoma, cholangiocarcinoma, and ductal carcinoma of the pancreas have shown that the presence of MUC1 mucin is highly predictive of invasive growth and poor prognosis. In contrast, the presence of MUC2 mucin is indicative of noninvasiveness and a lower potential for metastasis [12, 13].
FDG PET has been shown to be of great value in the detection, staging, and monitoring response to treatment in a variety of tumors. FDG uptake has been correlated with the number of viable tumor cells as well as the grade and the differentiation of some tumors [14, 15]. In recent reports, researchers have shown the limitations of FDG PET in revealing certain tumors with specific histologic characteristics, such as bronchioloalveolar cell carcinoma [2, 16]. Kim et al. [2] showed that bronchioloalveolar cell carcinoma, which often contains abundant mucin, exhibits significantly lower peak standardized uptake values compared with squamous cell carcinoma, adenocarcinoma, and other cell types. In their study population, Kim et al. found that bronchioloalveolar carcinoma was more often well differentiated, showed moderate degrees of nuclear atypia, and had infrequent mitotic figures. Similar results have been reported by Higashi et al. [16]; these investigators found a correlation between FDG uptake and the degree of cell differentiation in adenocarcinoma of the lung. In their study population, Higashi et al. reported that FDG uptake was significantly lower in bronchioloalveolar carcinoma than other carcinomas and that negative FDG PET results were seen in four of seven patients with bronchioloalveolar carcinoma.
In this study, FDG PET had a lower sensitivity for detection of primary and recurrent mucinous carcinoma than that generally reported for tumors of the gastrointestinal tract and lung [17]. False-negative results were seen in nine of 22 patients with mucinous carcinoma. The apparent insensitivity of FDG PET to show mucinous carcinoma is not surprising given the mechanism by which this tracer localizes in tumors. For example, our results indicate that the relative cellularity of the tumor is important in the detection of disease using FDG PET. In an in vitro study of an adenocarcinoma cell line, Higashi et al. [14] showed that differences in FDG uptake strongly correlate with the presence of viable tumor cells. Although FDG uptake by some tumors has been strongly correlated with grade such as tumors of the brain, musculoskeletal system, and breast, we found that tumor grade was not predictive of tumor detectability using FDG PET [15, 18, 19].
This study suffers from the inherent limitations of a retrospective study performed at a single institution, and our results may not necessarily be generalizable to the patient populations in other institutions. However, most of the tumors in this study originated in the gastrointestinal tract, which is significant for two reasons. First, this reflects the fact that mucinous carcinoma is relatively more common in the gastrointestinal tract. Second, FDG PET has proved particularly useful and is frequently applied in both detecting and staging recurrent disease in tumors of the gastrointestinal tract [1, 17, 20].
In addition to qualitative analysis, semiquantitative analysis of FDG uptake has been widely used for characterization of indeterminate lesions. Although semiquantitative analysis has been shown to be slightly superior to qualitative analysis in differentiating benign from malignant lesions, no statistically significant difference has, to our knowledge, been reported [21-23]. In this study, both the qualitative and semiquantitative (simple tumor-to-normal tissue activity ratio) analyses were used for evaluation of the PET images. The semiquantitative results paralleled the qualitative results in this patient population.
Despite the limitations of this study, the results may help to better define the application of FDG PET in patients with mucinous carcinoma. In particular, FDG PET may not be ideal for evaluation of recurrent or metastatic disease in patients with known mucinous carcinoma. Although this information is not always available, knowledge of the limitations of FDG PET in a particular histologic subtype, mucinous carcinoma, may assist in proper application of this technique and in avoiding misdiagnosis.


We thank Andrea Sykes for her assistance in preparing this manuscript and we greatly appreciate the technical assistance of Renee J. Burney, Delynn K. Silvestros, and Martin A. Schmitt.


Address correspondence to F. Dehdashti.


Rigo P, Paulus P, Kaschten BJ, et al. Oncological applications of positron emission tomography with fluorine- 18 fluorodeoxyglucose. Eur J Nucl Med 1996; 23:1641-1674
Kim BT, Kim Y, Lee KS, et al. Localized form of bronchioalveolar carcinoma: FDG PET findings. AJR 1998; 170:935-939
Jadvar H, Segall GM. False-negative fluorine-18-FDG PET in metastatic carcinoid. J Nucl Med 1997; 38:1382-1383
Cohen AM, Minsky BD, Schilsky RL. Colon cancer. In: Devita VT Jr, Hellman S, Rosenberg SA, eds. Cancer: principles and practice of oncology, 4th ed. Philadelphia: Lippincott, 1993: 776-817
Connelly JH, Robey-Cafferty SS, Cleary KB. Mucinous carcinomas of the colon and rectum: analysis of 62 stage B and C lesions. Arch Pathol Lab Med 1991; 115:1022-1025
World Health Organization. Histologic typing of breast tumors. Tumori 1982; 68:181-198
Ogunbiyi OA, Flanagan FL, Dehdashti F, et al. Detection of recurrent and metastatic colorectal cancer: comparison of positron emission tomography and computed tomography. Ann Surg Oncol 1997; 4:613-620
Xu M, Luk WK, Cutler PD, Digby WM. Local threshold for segmented attenuation correction of PET imaging of the thorax. IEEE Trans Nucl Sci 1994; 41:1532-1537
Rosai J. Special techniques in surgical pathology. In: Rosai J, ed. Ackerman's surgical pathology, 8th ed. St. Louis: Mosby-Year Book, 1996: 31-62
Green JB, Timmcke AE, Mithcel WT, Hicks TC, Gathright JB Jr, Ray JE. Mucinous carcinoma—just another colon cancer? Dis Colon Rectum 1993; 36:49-54
Sun XF, Cartensen JM, Stal O, Zhang H, Boeryd B, Nordenskjold B. Interrelations of clinicopathologic variables and their prognostic value in colorectal adenocarcinoma. APMIS 1996; 104:35-38
Baldus SE, Zirabes TK, Engel S, et al. Correlation of the immunohistochemical reactivity of mucin peptide cores MUC1 and MUC2 with the histopathologic subtype and prognosis of gastric carcinomas. Int J Cancer 1998; 79:133-138
Yonezawa S, Taira M, Osaka M, et al. MUC-1 mucin expression in invasive areas of intraductal papillary mucinous tumors of the pancreas. Pathol Int 1998; 48:319-322
Higashi K, Clavo AC, Wahl RL. Does FDG uptake measure proliferative activity of human cancer cells? In vitro comparison with DNA flow cytometry and tritiated thymidine uptake. J Nucl Med 1993; 34:414-419
Adler LP, Blair HF, Makley JT, et al. Noninvasive grading of musculoskeletal tumors using PET. J Nucl Med 1991; 32:1508-1512
Higashi K, Nishikawa T, Seki H, et al. Comparison of fluorine-18-FDG PET and thallium-201 SPECT in evaluation of lung cancer. J Nucl Med 1998; 39:9-15
Delbeke D, Vitola JV, Sandler MP, et al. Staging recurrent metastatic colorectal carcinoma with PET. J Nucl Med 1997; 38:1196-1201
Adler LP, Crowe JP, Al-Kaisi NK, et al. Evaluation of breast masses and axillary lymph nodes with [F-18] 2-deoxy-2-fluoro-D-glucose PET. Radiology 1993; 187:743-750
Di Chiro G, Brooks RA, Bairamian D, et al. Diagnostic and prognostic value of positron emission tomography using [18F]fluorodeoxyglucose in brain tumors. In: Reivich M, Alavi A, eds. Positron emission tomography. New York: Alan Lyss, 1985: 291-309
Flanagan FL, Dehdashti F, Siegel BA, et al. Staging of esophageal cancer with 18F-fluorodeoxyglucose positron emission tomography. AJR 1997; 168:417-424
Lowe VJ, Hoffman JM, Delong DM, Patz EF, Coleman RE. Semiquantitative and visual analysis of FDG-PET images in patients with pulmonary abnormalities. J Nucl Med 1994; 43:7-23
Vansteenkiskiste JF, Stroobant SG, De Leyn PR, et al. Lymph node staging in non-small cell lung cancer with FDG-PET scan: a prospective study on 690 lymph node stations from 68 patients. J Clin Oncol 1998; 16:2142-2149
Avril N, Bense S, Ziegler SI, et al. Breast imaging with fluorine-18-FDG PET: quantitative image analysis. J Nucl Med 1997; 38:1186-1191

Information & Authors


Published In

American Journal of Roentgenology
Pages: 1005 - 1008
PubMed: 10749239


Submitted: May 3, 1999
Accepted: September 22, 1999



Kevin L. Berger
Edward Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 S. Kingshighway Blvd., St. Louis, MO 63110.
Siobhan A. Nicholson
Department of Pathology, Washington University School of Medicine, St. Louis, MO 63110.
Farrokh Dehdashti
Edward Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 S. Kingshighway Blvd., St. Louis, MO 63110.
Barry A. Siegel
Edward Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 S. Kingshighway Blvd., St. Louis, MO 63110.

Metrics & Citations



Export Citations

To download the citation to this article, select your reference manager software.

Articles citing this article

View Options

View options


View PDF

PDF Download

Download PDF







Copy the content Link

Share on social media