18F-FDG PET/CT findings of sinonasal inverted papilloma with or without coexistent malignancy: comparison with MR imaging findings in eight patients
- 283 Downloads
Sinonasal inverted papilloma (IP) is known for high rate of associated malignancy. The purpose of this study was to identify 18F-FDG PET/CT findings of sinonasal IPs. We also tried to compare the PET/CT findings with the MR imaging findings.
We retrospectively reviewed PET/CT and MR images of eight patients with sinonasal IP with (n = 6) or without (n = 2) coexistent squamous cell carcinoma (SCC). Particular attention was paid to correlate the PET/CT findings with the MR imaging findings in terms of area distribution of standard uptake values (SUVs) and a convoluted cerebriform pattern (CCP).
In two benign IPs, the maximum SUVs measured 8.2 and 7.8, respectively (mean, 8.0). In both tumors, MR images demonstrated a diffuse CCP. In six IPs with coexistent SCC, the maximum SUVs ranged from 13.3 to 31.9 (mean ± SD, 20.2 ± 6.6). In these tumors, MR images demonstrated a diffuse CCP in two, a partial CCP in three, and no CCP in one. A wide discrepancy was noted between MR imaging and PET/CT in terms of area distribution of a CCP and SUVs.
In sinonasal lesions with MR imaging features of IP, 18F-FDG PET/CT demonstrating avid FDG uptake does not necessarily imply the presence of coexistent malignancy. In our small series, although IPs containing foci of SCC had consistently higher SUVs than IPs without SCC, the limited literature on this subject suggests that PET cannot be used reliably to make the distinction.
KeywordsPositron-emission tomography/CT Magnetic resonance imaging Sinonasal neoplasm Inverted papilloma
Inverted papilloma (IP) is an unusual benign epithelial neoplasm of the sinonasal cavity, accounting for 0.5% to 4.0% of all primary nasal neoplasm [1, 2]. Although histologically benign, the incidence of association with synchronous or metachronous malignant tumor, mostly squamous cell carcinoma (SCC), is reportedly high and ranges from 2% to 53% [3, 4, 5, 6]. It would be useful for determining the therapeutic plans if we know the presence or absence of coexistent malignancy preoperatively, because more aggressive therapeutic management would be necessary in cases of coexistent malignancy [6, 7, 8, 9]. However, differentiation between IPs with and those without SCC on imaging studies is often difficult, and even the histologic examination of biopsied tissue may not disclose the presence of SCC residing in a small portion of IP. One previous study with MR imaging suggested that a focal loss of a convoluted cerebriform pattern (CCP), which seemed to be a reliable feature of sinonasal IP, might be an additional sign that might indicate the presence of coexistent malignancy .
Positron-emission tomography (PET) using 18F-fluorodeoxyglucose (FDG) has emerged as a powerful imaging technique for the diagnosis, staging, and surveillance of various kinds of malignancies of the head and neck with promising results [11, 12, 13]. Furthermore, the integrated PET/CT system makes it possible to fuse anatomic images with functional images, which results in better anatomic localization and an improvement of diagnostic accuracy [14, 15]. To our knowledge, however, a very limited number of reports, based mainly on a small number of case series, have described the PET or PET/CT features of sinonasal IP [16, 17, 18]. The purpose of this study was to investigate the 18F-FDG PET/CT findings of sinonasal IPs with or without coexistent malignant tumor. We also tried to compare the PET/CT findings with the MR imaging findings.
Materials and methods
The institutional review board of our hospital approved this retrospective study. From January 1995 to March 2008, search of the electronic data base of our institution revealed a total of 163 patients with sinonasal IP proved on pathologic examination. Of these 163 patients, both MR and PET/CT examinations were performed in eight patients. There were five men and three women, ranging in age from 52 to 83 years, with a mean age of 67 years. All of the patients underwent surgical excision of the tumor, which disclosed that the tumor primarily involved the nasal cavity (n = 3), maxillary sinus (n = 4), and both ethmoid and frontal sinuses (n = 1). Two patients had IP alone and six patients had coexistent SCC on histologic examination. In six patients with IP with coexistent SCC, the presenting complaints related to the sinonasal tumor were nasal obstruction in four patients and anterior cheek swelling in two patients. All of these six patients with IP with coexistent SCC underwent MR imaging prior to PET/CT with a mean interval of 24 days (range, 3–61 days). PET/CT was ordered because of the MR imaging findings which were highly indicative of malignancy (i.e., bone destruction or adjacent tissue invasion) and/or the histologic evidence of malignancy on biopsy specimens. In the remaining two patients with IP alone, the sinonasal tumor was incidentally detected on PET/CT scans obtained for the metastatic work-up of cancer (one in the cervix and the other in the esophagus). Both of these patients with benign IP underwent MR imaging for the characterization of the sinonasal lesion at 22 and 27 days after PET/CT examination, respectively.
All of the MR examinations were performed on a 1.5-T (Signa Advantage Horizon, GE Healthcare, Milwaukee, WI, USA) or 3-T (Intera Achieva, Philips, Best, the Netherlands) scanner. In all of the patients, precontrast T1-weighted spin-echo images (TR/TE/NEX, 400–560 ms/10–14 ms/2) and T2-weighted fast spin-echo images (TR/TE/NEX, 2,500–4,500 ms/80–110 ms/1) with or without fat saturation were obtained, followed by contrast-enhanced T1-weighted spin-echo images with fat saturation after the intravenous injection of 0.1 mmol/kg of gadopentetate dimeglumine. Images were obtained in at least two planes with 3- to 4-mm section thickness, 0- to 0.4-mm intersection gap, 256 × 192 matrix, and 22-cm FOV.
All of the patients fasted for at least 6 h before the PET/CT study. PET/CT scanning was performed using a Discovery LS PET/CT scanner (GE Healthcare, Milwaukee, WI, USA). Whole-body CT was performed by a continuous spiral technique using an eight-slice helical CT with a gantry rotation speed of 0.8 s. The CT data were collected using the following parameters: 40–120 mAs, 140 keV, 5-mm section width, and 5-mm table feed per rotation. No intravenous or oral contrast material was used. Following the CT scans, 370 MBq 18F-FDG was intravenously injected, and an emission scan was obtained from thigh to head for 5 min per frame, for a total of 45 min. The duration of the uptake phase was 45 min. Attenuation-corrected PET images using CT data were reconstructed by an ordered-subsets expectation maximization algorithm (28 subsets, two iterations). The images were displayed in a 128 × 128 matrix (pixel size = 4.29 × 4.29 mm) with a slice thickness of 4.25-mm. The separate CT and PET scan data were coregistered accurately using commercial software (eNTEGRA, Elgems, Haifa, Israel).
One nuclear medicine physician who was experienced in interpreting head and neck imaging for 7 years (J.Y.C.) reviewed all of the PET/CT images without knowing the MR imaging and final histopathologic findings. On PET/CT images, the sinonasal lesions was identified by increased FDG uptake on PET images, and these 18F-FDG-avid lesions were matched anatomically on the coregistered PET/CT images. The patterns of FDG uptake were categorized visually as homogenous or heterogeneous. The tumor was considered homogeneous when a uniform FDG uptake was seen throughout the lesion. The tumor was considered heterogeneous when two or more areas of different FDG uptake were seen within the lesion. We also measured the maximum standard uptake value (SUV) within the tumor by using the attenuation-corrected images, the amount of injected 18F-FDG, the body weight of each patient, and the cross-calibration factors between 18F-FDG PET and the dose calibrator. As for the tumors showing heterogeneous FDG uptake on visual inspection, we measured the maximum SUVs from each of the areas with visually different FDG uptake. We compared the maximum SUVs between IPs with and those without coexistent SCC.
Interpretation of MR images was performed by one dedicated head and neck radiologist with clinical experience of 19 years (H.-J.K.) without knowing the PET/CT and final histopathologic findings. On MR images, the presence or absence of a CCP within the sinonasal tumor was determined. We defined a CCP as a mix of linear or curvilinear hyperintense and hypointense striations partially or diffusely seen in the solid components of the tumor on T2-weighted or contrast-enhanced T1-weighted MR images . When present, the pattern of a CCP was categorized as diffuse or partial according to its distribution within the tumor. We also tried to find the associated valuable MR imaging features to help suggest coexistent malignancy. The size of the tumor measured at the greatest diameter was also recorded.
After the independent image interpretation session, the head and neck radiologist and the nuclear medicine physician together had a conference for correlating the findings of PET/CT with those of MR imaging, and the final conclusion was reached by consensus. Particular attention was paid to correlate the PET/CT findings with the MR imaging findings in terms of area distribution of SUVs and a CCP.
MR imaging and PET/CT findings in eight patients with sinonasal IP with or without squamous cell carcinoma.
MR imaging findings
Pattern of FDG uptake
IP with microscopic SCC
Predominant SCC with small focus of IP
Predominant SCC with small focus of IP
Predominant SCC with small focus of IP
Frontal and ethmoid sinuses
IP with microscopic SCC
Predominant SCC with small focus of IP
Aggressive bone destruction with (n = 4) or without (n = 1) extrasinonasal extension was seen in five of six malignant IPs on MR images (Fig. 3). The CT component of PET/CT confirmed the bone destruction noted on MR images in these five tumors. Intratumoral necrosis, defined as nonenhancing low-signal-intensity area within the tumor on contrast-enhanced T1-weighted images, was also noted in four of these five tumors with bone destruction.
Sinonasal papillomas originate from the ectodermally derived ciliated respiratory epithelium, so-called Schneiderian mucosa, and can be classified into three histologic subtypes: inverted, oncocytic, and exophytic papillomas. As a group, these papillomas are uncommon, consisting of only 0.4% to 4.7% of all sinonasal tumors, among which IP is most common, accounting for 62% of all sinonasal papillomas .
The term inverted papilloma is derived from the characteristic histologic feature of the tumor, that is, inversion of the surface epithelium into the underlying stroma instead of exophytic proliferation [7, 20]. Although histologically benign, IP has a propensity for a high recurrence rate, local aggressiveness, and multicentricity [3, 4, 5, 6, 8, 9, 20, 21, 22, 23]. Moreover, the incidence of malignant change in an individual series of sinonasal IPs has been reportedly high, ranging from 2% to 53%, and SCC is the most commonly associated malignant tumor [3, 4, 5, 6]. Carcinomas may originate from the IP or it may merely be associated with a histologically bland IP. Patients with IPs that are associated with carcinomas fall into three groups: group 1, those who have primarily an IP with only a small focus of carcinoma; group 2, those who have primarily a carcinoma with only a small focus of IP; and group 3, those who had a history of histologically documented IP and subsequently develop a carcinoma in the area in which the IP arose [3, 5]. The first two groups are deemed synchronous, and group 3 is metachronous. Of all the carcinomas associated with IPs, approximately 61% are synchronous and 39% metachronous . Needless to say, it would be useful for determining the therapeutic plans if we could distinguish the IPs with coexistent SCC from those without it preoperatively, because more aggressive therapeutic management would be necessary for the former [6, 7, 8, 9]. However, differentiation of benign IPs from the IPs with coexistent SCC as well as other malignant sinonasal tumors on radiologic imaging is often difficult, because there is a significant overlap between those diseases [8, 10, 21].
Recently, a CCP has been reported as a reliable MR imaging feature of sinonasal IPs [8, 10]. It is created by the juxtaposed epithelial and stromal layers and results in a peculiar pattern on MR imaging, manifesting as the alternating hypointense and hyperintense bands on T2-weighted and contrast-enhanced T1-weighted images. It also has been described under the different terms, including septate-striated appearance  and columnar pattern . Recently, Jeon et al  reported that a CCP was demonstrated in all 30 (100%) of the IPs and 17 (13%) of the 128 malignant sinonasal tumors with the overall sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of CCP for the diagnosis of IP at 100%, 87%, 64%, 100%, and 89%, respectively. They also suggested that a focal loss of a CCP might be an additional sign that might indicate the presence of coexistent malignancy.
PET with FDG has been used as a powerful imaging technique in the diagnosis and surveillance of head and neck cancer, and recent studies have shown a relative superiority of FDG PET over CT and MR imaging for detection of head and neck cancer [11, 12, 25, 26]. Combined with CT, it can offer detailed anatomic information as well as metabolic data [14, 15]. Malignant tumor evokes increased glucose metabolism and is demonstrated as an area of increased radioactivity on FDG PET/CT. However, FDG is neither tumor-specific nor specific to malignant processes, and its uptake is also observed in benign conditions that cause an increased rate of glycolysis, such as benign tumors and infectious processes [17, 27].
There have been a few studies describing the FDG PET or PET/CT findings of sinonasal IP [16, 17, 18]. In the study of the patients with benign or malignant sinonasal diseases, Ninomiya et al  reported that SUVs of FDG were lower for IP (n = 5; range 1.98–4.65, median 3.49) than for malignant tumor (n = 8; range 1.78–19.36, median 12.6). However, their study showed no statistical significance between the SUVs for these two diseases. Lee et al  reported a case of IP of the maxillary sinus which demonstrated high SUVs (9.0 at 1 h and 18.1 at 2 h after injection of FDG) that were indistinguishable from those of malignant tumors. Only one study has described the value of FDG PET for predicting malignancy of sinonasal IP. Shojaku et al  reported that the SUVs of two cases of IP with coexistent SCC were 8.9 and 20.9, respectively, and those of three cases of benign IP ranged from 4.9 to 7.3. The mechanism of high SUVs displayed by some of the sinonasal IPs may partly be explained by the associated inflammatory response within the tumor . Although benign, oncocytic tumors such as Warthin tumors, oncocytomas, and oncocytic papillomas have been observed with high SUVs. It is hypothesized that oncocytic tumors possess a high number of mitochondria causing increased metabolism and an increased uptake of metabolites such as FDG . Likewise, benign tumors which show a tendency for aggressive growth as seen in sinonasal IP might also be expected to cause enhanced glucose metabolism and exhibit high FDG uptake .
Although the SUVs for benign IP (n = 2; range 7.8–8.2, mean 8.0) were lower than those for IP with coexistent SCC (n = 6; range 13.3–31.9, mean 20.2), our results showed that the SUVs for benign IP were sufficiently high to make a false diagnosis of malignant tumor possible, as reported by Lee et al . In both these benign IPs, however, MR imaging demonstrated a diffuse CCP, which supported the diagnosis of benign IP quite likely. We have no appropriate explanation of the reason of the heterogeneous FDG uptake on PET/CT in one of the two benign IPs (case 2). The different growth rate and inflammatory response within the different portions of the tumor might be represented by the heterogeneous metabolic activity on PET/CT images.
Our study also showed a wide discrepancy between PET/CT and MR imaging for the detection and extent estimation of coexistent SCC associated with IP. In two IPs with coexistent SCC, while MR images demonstrated a diffuse CCP suggestive of benign IP, PET/CT demonstrated a homogeneous high FDG uptake with maximum SUVs of 14.9 and 31.9, respectively, highly suggestive of malignant tumor (cases 3 and 7). The histology in these two tumors revealed multiple foci of microscopic SCC scattered within the tumor. Our results did not support the presumption that a focal loss of a CCP on MR imaging might be an additional sign that might indicate the presence of coexistent malignancy . In all of the three IPs with coexistent SCC which demonstrated a partial CCP on MR images, the extent of coexistent malignancy predicted on MR images was smaller than that predicted on PET/CT (cases 4, 5, and 6). The histology in these three tumors revealed a small focus of IP residing in a predominant SCC. At this moment, because of lack of an exact 1:1 MR- and PET/CT-pathologic correlation, we are not sure that our results indicate the superiority of PET/CT over MR imaging for the diagnosis of sinonasal IPs with coexistent SCC. The discordance between the findings on PET/CT and MR imaging may be attributed to either an exaggeration of FDG uptake on PET/CT images, so-called blooming, or the lack of specificity of a CCP on MR images. However, based on the PET/CT findings demonstrating homogeneous high FDG uptake noted in the majority of the IPs with coexistent SCC as well as the pathologic diagnosis reported at the time of surgery in our study, it is likely that MR imaging underestimated the actual extent of coexistent malignancy in these cases. Nevertheless, by providing the detailed, high-quality anatomic images, such as involvement of the important anatomic structures, we believe that MR imaging is still a powerful method for the evaluation of sinonasal IP with or without coexistent malignancy.
In sinonasal lesions with MR imaging features of IP, 18F-FDG PET/CT demonstrating avid FDG uptake does not necessarily imply the presence of coexistent malignancy. In our small series, although IPs containing foci of SCC had consistently higher SUVs than IPs without SCC, but the limited literature on this subject suggests that PET cannot be used reliably to make the distinction.
Conflict of interest statement
We declare that we have no conflict of interest.
- 15.Antoch G, Saoudi N, Kuehl H et al (2004) Accuracy of whole-body dual-modality fluorine-18-2-fluoro-2-deoxy-d-glucose positron emission tomography and computed tomography (FDG-PET/CT) for tumor staging in solid tumors: comparison with CT and PET. J Clin Oncol 22:4357–4368. doi:10.1200/JCO.2004.08.120 PubMedCrossRefGoogle Scholar
- 19.Barnes L, Tse LLY, Hunt JL (2005) Schneidrian papillomas. In: Barnes L, Eveson JW, Reichert P et al (eds) World Health Organization classification of tumours. Pathology and genetics of head and neck tumours. ARC, Lyon, pp 28–32Google Scholar
- 25.Nowak B, Di Martino E, Janicke S et al (1999) Diagnostic evaluation of malignant head and neck cancer by F-18-FDG PET compared to CT/MRI. Nucl Med (Stuttg) 38:312–318Google Scholar