18F-FDG PET-CT postoperative changes after maxillectomy: Findings and pitfalls in interpretation

Abstract

Objective:

We investigated the findings and pitfalls of FDG-PET/CT scanning after maxillectomy with reconstruction/rehabilitation procedures, in patients with head and neck malignancies treated during nine years at one tertiary medical centre.

Methods:

Fourteen patients (10 males), aged 22–84 years, underwent 17 reconstruction/rehabilitation maxillectomy surgeries and 35 PET/CT scans. Postoperative PET/CT findings were correlated with clinical and imaging follow-up.

Results:

Increased FDG uptake, mean SUVmax 2.4 ± 1.4 (range 0.3–4.3), was observed at the postoperative bed following 12 of 17 surgeries (71%; 10 obturators, two mesh reconstructions). Following the remaining 5/17 surgeries (three with a fat flap and two without any reconstructions), abnormal FDG uptake was not observed at the postoperative bed.

CT features of postoperative sites included: non-homogeneous mixed iso/hyperdense structures (hollow or filled) with multiple surrounding and/or inside air bubbles (“sponge appearance”) and mucosal thickening along the postoperative bed wall (in all cases with obturator implants); rich fat density material in reconstructions with a fat flap and in closures without reconstruction, and radiopaque elongated structures in mesh reconstructions.

No correlation was found of the mean SUVmax in initial scans, with the time from the surgery date (10 ± 6 months; r=0.04, P=0.90), or with the mean SUVmax in final scans (at 25± 17 months, P=0.17).

Conclusions: :

Increased FDG uptake, together with corresponding non-specific CT features, may persist for a prolonged period after surgery with obturators and mesh implantations, mimicking malignancy or infection. Awareness of variations in postoperative PET-CT appearance can help avoid false interpretations and redundant invasive procedures.

Background

The annual global incidence of oral cavity and lip cancer was estimated (as of 2018) at 354,800, and deaths at 177,300.¹ Squamous cell carcinoma (SCC) is the most common oral type lesion.^2–5 Ablative surgery is generally recommended as the initial therapy, both for early stage and locally advanced oral cavity cancers. In most cases, simultaneous resection and reconstruction are feasible, with acceptable functional outcomes.⁶ The postablative maxillary defect can be reconstructed by a variety of methods, such as by obturator prostheses for rehabilitation, fat pad graft, regional flap transfer, vascularized free flaps with or without bone segments, and the use of alloplastic implants (such as titanium reconstruction mesh).⁷

PET/CT is recognized as the mainstay of disease evaluation following curative local therapy for head and neck malignancy. However, specificity may be limited in certain clinical scenarios. Notably, increased FDG-uptake in post-surgical cavities due to benign conditions, such as inflammatory responses,^8,9 infectious processes^10,11 and reactions to retained surgical foreign bodies^12–20 may lead to misinterpretation.

FDG uptake due to the implantation of foreign bodies after surgery or medical treatment is a well-documented pitfall of PET/CT interpretation. Catheters, tubes and stomas,¹³ teflon implants,^14,15 retained textiles,^16–18 surgical sutures²⁰ and mesh²¹ are all examples of foreign bodies that can mislead interpretation. The FDG avidity in these pitfalls can persist for extended periods of time, even more than 20 years following surgery.^17,22 To prevent misinterpretations in head and neck cancer patients, radiologists should be familiar with the normal postoperative PET/CT appearances, as well as possible pitfalls at the surgical bed following maxillectomy. However, only a few reports describe PET/CT findings that are related to postsurgical changes after maxillectomy with reconstruction/rehabilitation.

The objective of this study is to present our experience in assessing and recognizing postoperative FDG-PET/CT findings of the oral cavity, and the pitfalls of this modality in head and neck malignancy patients after maxillectomy with reconstruction or with obturator prostheses. In addition, we report changes in serial PET/CT scans, performed at different time points after surgery.

Methods

Study design and setting

This comparative study was approved by the ethics board of our tertiary medical center. Informed consent was waived due to the retrospective design of the study. We reviewed data of head and neck cancer patients with a history of maxilla surgeries who were referred and treated for their disease between January 2008 and January 2017. Clinical and imaging data were obtained from the hospital’s computerized medical records and from the picture archive and communication system (PACS, Carestream Health 11.0, Rochester, NY). We gathered the following clinical data: cancer diagnosis, surgical reports and disease status, according to clinical evaluation. Study inclusion criteria were a diagnosis of head and neck malignancy treated surgically with maxillectomy, with or without reconstruction, and the availability of FDG-PET/CT scans. These scans were performed as part of the initial evaluation and during clinical follow-up. They were reviewed with blinding to the clinical course of the patients by a physician who is an expert in nuclear medicine and radiology (20 years’ experience).

FDG pet/CT

¹⁸FDG PET/CT examinations were performed with a combined PET/CT (Philips Gemini GXL, Philips Medical Systems, Cleveland OH, USA) that includes 16 detector rows helical CT. This scanner enables simultaneous acquisition of up to 45 transaxial PET images with inter-slice spacing of 4 mm in one bed position and provides an image from vertex to the thigh with about 10 bed positions.²³ The transaxial field of view and pixel size of the PET images reconstructed for fusion were 57.6 cm and 4 mm, respectively, with a matrix size of 144 × 144. The technical parameters used for CT imaging were: pitch 0.8, gantry rotation speed 0.5, 120 kVp, 250 mAs, 3-mm slice thickness and specific breath-holding instructions. After 4–6 h of fasting, patients received an i.v. injection of 370 MBq F-18 FDG. About 60 min later, CT images were obtained from the vertex to the mid-thigh for about 32 s. CT scans were obtained 60 s after injection of 2 mL/kg of non-ionic contrast material (Omnipaque 300; General Electric Healthcare). An emission PET scan followed in 3D acquisition mode for the same axial coverage, 1.5 min per bed position. CT images were fused with the PET data and were used to generate a map for attenuation correction. PET images were reconstructed using a line of response protocol with CT attenuation correction, and the reconstructed images were generated for review on a computer workstation (Extended Brilliance Workstation, Philips Medical Systems, Cleveland OH, USA).²³

Image assessment

According to CT findings, postoperative changes at the surgical bed were analysed, including the presence of foreign body (obturator and /or metallic mesh), mucosal thickening, fluid and air bubble collections, soft tissue mass and fatty infiltrations with or without other soft tissues in the surgical bed.

According to FDG-PET/CT findings, the patterns of ¹⁸F-FDG uptake in the corresponding sites at the postoperative bed, or along the postoperative cavity wall, were classified as focal, homogeneous and non-homogeneous/nodular. The intensity¹⁸F-FDG uptake was measured by standardized uptake values (SUV) max. This was calculated by manual placement of regions of interest over the sites of abnormally increased FDG activity. Abnormally increased FDG activity was defined in cases with intensity of uptake higher than the physiological uptake in the surrounding soft tissue, or higher than the physiological uptake in the liver or mediastinal vessel blood pool. In patients with more than one available postoperative PET/CT scan, we assessed the dynamic changes in the postoperative bed in all consecutive imaging studies.

In this study, non-specific postoperative/post-prosthesis changes were determined according to the follow criteria:

1. In cases with available repeat postoperative PET/CT scans, stability in intensity of FDG uptake or in CT appearance (without evidence of a new noticeable mass at the postoperative bed).

2. In case of available biopsy – pathological confirmation.

3. PET/CT data were correlated clinically to exclude grumbling infection around the prosthesis as a cause of increased uptake.

We note that HPV status was not studied. This is because HPV is not associated with cancers included in this study (orbit, maxillary, sinus and jaws), despite its association with SCC of the oropharynx.

Statistical analysis

Data were represented as means ± standard deviations (SDs) for continuous variables, and as percentages for categorical parameters. Spearman’s rank correlation was used to assess correlations of the first postoperative SUV max value with the time interval from surgery to the first postoperative FDG-PET/CT. For analysis of the follow-up scans that were available, a paired sample t-test was used to compare the lesion’s intensity of FDG-uptake between the first and last PET/CT studies. The analysis was performed with the use of SPSS v.21.0 (SPSS, IBM, USA). p < 0.05 was considered statistically significant.

Results

During the study period, 362 head and neck patients were referred and treated for their disease at Chaim Sheba Medical Center. Of them, 37 patients with head and neck malignancy had undergone maxillary operations. Most of these patients did not undergo PET/CT follow-up due to the feasibility of follow-up by direct inspection of the area. Postoperative PET/CT scans were available for only 14 patients: 11 (79%) with squamous cell carcinoma (SCC) of the oral cavity or paranasal sinus, two (14%) with peripheral nerve sheet tumour and one patient (7%) with an adenoid cystic carcinoma. All 14 patients included in the study had been referred for FDG PET/CT examinations as part of routine oncological follow-up. Their mean age was 57 ± 16 years (range: 22–84 years); 10 (71%) were males (Table 1).

Table 1.

Clinical characteristics of the patients included in the study

Patient number	Age (years)	Gender	Type of surgery	Tumour histopathology	Primary tumour site
1	77	Female	Obturator	Squamous cell carcinoma	Oral cavity
2	44	Male	Fat flap	Squamous cell carcinoma	Oral cavity
3	22	Female	Primary closure	Peripheral nerve sheet tumour	Paranasal sinus
4	48	Male	Obturator	Squamous cell carcinoma	Oral cavity
5	59	Male	Obturator	Squamous cell carcinoma	Paranasal sinus
6	66	Male	Obturator	Squamous cell carcinoma	Oral cavity
7	47	Male	Fat flap	Squamous cell carcinoma	Paranasal sinus
8	69	Male	Obturator Mesh	Peripheral nerve sheet tumour	Oral cavity
9	41	Female	Primary closure Mesh Obturator	Adenoid cystic carcinoma	Paranasal sinus
10	68	Female	Obturator	Squamous cell carcinoma	Oral cavity
11	62	Male	Fat flap	Squamous cell carcinoma	Paranasal sinus
12	42	Male	Obturator	Squamous cell carcinoma	Oral cavity
13	70	Male	Obturator	Squamous cell carcinoma	Oral cavity
14	84	Male	Obturator	Squamous cell carcinoma	Oral cavity

For 11 patients (79%), pre-surgical PET/CT images were available. Nine of the 14 patients (64%) had more than one postoperative PET/CT scan performed and five (36%) had only one postoperative PET/CT scan. The mean number of scans per patient was 2.5 ± 1.6. A total of 35 postoperative PET/CT studies were available for review. The mean time interval from the operation to the first postoperative PET/CT was 10 ± 6 months (range: 2–24 months). The mean time interval between the first and the last postoperative PET/CT scans was 25 ± 17 months (range: 4–52 months).

The 14 patients underwent a total of 17 maxillectomy surgeries (Table 2). Two types of surgical closure of the post-ablative bed were observed. Closure with reconstruction was observed in 12 patients and closure without reconstruction in the remaining two. The materials used for reconstruction at the first operation included a fat pad graft in three patients and obturator prosthesis in 9. One of the latter had an additional titanium mesh (patient 8, Table 1). In one patient, the closure was done without reconstruction at the first operation subsequently underwent mesh reconstruction; and later, after failure of the mesh, underwent an additional operation with obturator prostheses (patient 9, Table 1). We analysed each of these 17 operations. The findings of the first postoperative images of these surgeries are described in Table 3.

Table 2.

Types of maxillectomy surgeries

Type of operation	N = 17	%
With reconstruction/rehabilitation: Obturator prostheses Fat pad graft Mesh	15 10 3 2	88 59 18 12
Closure without reconstruction	2	12

Table 3.

FDG -PET-CT findings at the surgical bed in patients with squamous cell carcinoma, following 17 reconstructive maxillectomy procedures

On CT:
Findings	N = 17	%	Type of surgery
Mixed iso/hyperdense material (hollow or filled) with multiple surrounding and/or inside air bubbles	10	59	Obturator prostheses
Mucosal thickening along the postoperative bed wall	10	59	Obturator prostheses
Rich fat density structure	5	29	Fat pad graft Operative closure without reconstruction
Radiopaque (hyperdense) mesh	2	12	Surgical metallic mesh
On PET:
FDG uptake, the mean SUV max was 2.4 ± 1.4 (range 0.3–4.3)
Non-homogeneous nodular	10: 8 2	59	Obturator prostheses  Mesh
Homogeneous	1	6	Obturator prostheses
Focal	1	6	Obturator prostheses
None	5: 3 2	29	Fat pad graft  Operative closure without reconstruction

^aSUV max, standardized uptake values at the surgical site/along the postoperative cavity wall

CT imaging features in the postoperative sites included a combination of the following:

1. Non-homogeneous mixed iso/hyperdense structures (hollow or filled) with multiple surrounding and/or inside air bubbles. These were demonstrated in cases with rehabilitation with obturator prostheses (10/17, 59%) and, therefore, were consistent with foreign bodies (prostheses) (Figure 1).

2. Mucosal thickening along the postoperative fossa wall (10/17, 59%). This finding was observed in all cases with obturator prostheses (Figures 1 and 2).

3. Radiopaque/hyperdense elongated structures in 2/17 (12%) cases. The findings were consistent with surgical metallic mesh (Figure 3).

4. Rich fat density structures, which were present in 5/17 (29%) of the cases: three in reconstructions with a fat pad graft (Figure 4) and two with an operative closure without reconstruction (Figure 5).

Figure 1.

FDG-PET (a) and CT (b) axial slices.

Six months after right-maxillectomy (patient 10, Table 1). (a) High-grade FDG uptake on PET (black arrows), (b) A non-homogeneous mixed density structure with multiple air bubbles, filling the postoperative cavity on CT (white arrows), consistent with obturator use.

Figure 2.

FDG-PET (a) and CT (b) axial slices.

Seventeen months after left-maxillectomy (patient 6, Table 1). (a) High-grade non-homogeneous FDG uptake on PET (black arrows), (b) Mucosal thickening with air bubbles along the walls of the postoperative cavity on CT (white stars), consistent with obturator use.

Figure 3.

FDG-PET (a) and CT (b) axial slices.

Twenty months after right-maxillectomy (patient 8, Table 1). (a) Low-grade FDG uptake on PET (black arrow), (b) The area alongside the corresponding surgical mesh at the postoperative bed on CT (white arrows).

Figure 4.

FDG-PET (a) and CT (b) axial slices.

Forty-two months after right-maxillectomy (patient 2, Table 1). (a) No FDG uptake on PET (black star), (b) Prominent fatty tissue at the postoperative bed (white star), consistent with a fat flap on CT.

Figure 5.

FDG-PET (a) and CT (b) axial slices.

Fifty-two months after right-maxillectomy (patient 3, Table 1). (a) No FDG uptake on PET (black star), (b) Marked filling of fatty tissue (white star) at the postoperative bed on CT, consistent with postoperative primary closure.

Increased FDG uptake at the region of the postoperative bed was demonstrated in 12/17 (71%) of the first postoperative scans; the mean SUV max was 2.4 ± 1.4 (range 0.3–4.3). All the scans with increased FDG uptake were after obturator use (N = 10) or mesh reconstruction (N = 2). No correlation was found between the first postoperative SUV max value and the time interval from surgery to the first postoperative PET/CT (r = 0.04, p = 0.90).

Of the 12/17 postoperative cases with increased FDG uptake, the uptake was non-homogeneous/nodular in 10 (eight with obturator and two with mesh), homogeneous in one (with an obturator) and focal in another (with an obturator). In contrast, in the remaining 5/17 postoperative cases (three with a fat flap and two without any reconstructions), the PET-CT did not show increased FDG uptake at the postoperative bed.

For the nine patients with more than one postoperative PET/CT scan, the mean duration between the first and last scan was 25 ± 17 months (range: 4–52 months).

Five of the nine had obturator prostheses (these comprised 5/10 of the patients with obturator prostheses and increased FDG uptake at the postoperative bed at the initial PET/CT), two patients had fat graft reconstruction and another two were without reconstruction (those without increased uptake at the postoperative bed at the initial PET). The difference in mean SUVmax between the first and the last 18F-FDG scans at the postoperative bed: 2.4 ± 1.4 (range: 0.9–4.5) vs 1.6 ± 1.7 (range: 0.7–4.0) was not statistically significant, p = 0.17 (Figure 6). One patient with a verrucous SCC maxilla had non-homogeneous/nodular uptake at the postoperative bed on the initial postoperative PET scan, whereas a subsequent follow-up scan revealed a new FDG-avid soft tissue lesion in the nasopharynx, which extended to the operation bed. This was consistent with recurrence of the disease, as confirmed by tissue biopsy. Biopsies were performed in four additional patients who had increased FDG uptake. All biopsies showed fragments of fibrotic or granulation tissue, with no evidence of malignancy (including in the FDG focal lesion examined here). For the nine patients with more than one postoperative PET/CT scan, CT findings at the surgical site, along the postoperative cavity wall, were similar between the first and the last postoperative scans.

Figure 6.

FDG-PET (a,c) and CT (b,d) axial slices.

The follow-up scans with 6 months interval (A, B) after right-maxillectomy, (patient 12, Table 1). No marked changes in CT findings (white arrows) with high grade FDG uptake (black arrows) surrounding the postoperative cavity, 12 (A.a,b) and 18 (B.c,d) months after surgery (on different scanners), consistent with obturator use.

Of the five patients with only one postoperative PET/CT, three were lost to follow up; one died due to systemic advancement of the disease and one was followed clinically without evidence of the disease for the duration of the study period.

Discussion

This study highlights typical findings and indicates possible pitfalls in the interpretation of ¹⁸F-FDG PET/CT of head and neck malignancy patients following maxillectomy and reconstruction. Notably, PET/CT findings differed according to whether the operation included reconstruction with a foreign body (mesh or a removable obturator) or without (allogenic fat graft or without any reconstruction).

Postoperative changes may persist for a prolonged period and may be observed as incidental findings on PET/CT exam. When PET/CT scans are done for oncological follow-up, distinguishing between benign findings associated with surgical procedure and oncological recurrence or progression is critical for appropriate clinical management. To the best of our knowledge, this is the first study that describes the postoperative changes and possible pitfalls of ¹⁸F-FDG PET/CT in oncological patients after maxillectomy and reconstruction. Only one case report reviewed the CT findings of obturator prostheses²⁴ and no other studies describing PET/CT findings were found in our literature search. Obturator prostheses may be temporary, as when used immediately at the time of the operation, and later changed to permanent.²⁵ Obturator rehabilitation is composed of various materials such as Trusoft and acrylic resin. Trusoft obturators have a heterogeneous iso- or hyper-attenuated appearance, with foci of air that can be wrongly interpreted as local infection.²⁴

Notably, in the current series, which followed all cases of oral rehabilitation with obturator prosthetics during the study period, the CT scan revealed a characteristic appearance of a non-homogeneous mixed iso/hyperdense structure (hollow or filled), with multiple surrounding and /or inside air bubbles, with a “sponge appearance”. Moreover, in all our patients with an obturator, mucosal thickening along the wall of the open postoperative bed (cavity) was demonstrated. This appearance remained stable, without significant changes, several months after the operation. Of note, among patients with fat reconstruction or with primary closure without any reconstruction, extra fat density was demonstrated at the postoperative bed. This may represent the fat flap itself, after implantation, or probably fat hyperplasia in cases without reconstruction. Additionally, among the patients who underwent mesh reconstruction, the CT image was a radiopaque foreign body on CT. Other studies have described this appearance of mesh in CT.^21,22,26,27

Surprisingly, in all patients with obturator use and mesh reconstruction, increased FDG uptake was demonstrated, even high intensity uptake (mean SUV max 3.4 with obturator and 2.9 with mesh). Uptake was homogenous or non-homogeneous along the wall of the postoperative bed and even focal. Therefore, for these patients who underwent PET/CT as part of their oncological follow-up, the combination of pathological uptake and a “sponge” appearance could mimic either recurrence of the disease or exacerbated infection. The probability is high that the findings represent local tissue reaction or a change in inflammation to a foreign body, with fibrotic or granulation tissue formation.^12,19 Interestingly, these findings were irrespective of the time period between the operation and the PET/CT performed, which sometimes entailed a number of years. Moreover, imaging follow-up, when available, did not demonstrate any noticeable changes in these PET/CT findings at the postoperative bed, thus supporting the findings of benign changes. Nonetheless, it is recommended that PET/CT should be performed no sooner than 2 to 3 months after surgery and chemoradiation, to decrease the number of false-positive results secondary to inflammation.²⁸ In contrast to the above, in patients with fat grafts or patients without reconstruction, increased FDG uptake was not observed in the postoperative bed, concurring with the results of a previous study.²⁹

There are several limitations to this study. First, this is a retrospective study conducted in a single university tertiary centre. Second, correlations of the postoperative FDG-PET/CT findings to pathology were not examined in all cases, and the aetiology of FDG activity was not verified by biopsy. Third, our study includes a relatively small cohort with heterogeneous surgical procedures. This is because the size of the cohort was contingent on the protocol at our institution for examination of cancer recurrence in the maxilla area. Usually, oral and maxillofacial surgeons remove the obturator from the maxilla and make a physical examination of the area. The patient then undergoes chest CT for a systemic evaluation. Generally, only patients with more complicated local or systemic disease undergo FDG PET/CT scans of the head and neck area. Notably, for the patients who were lost to follow-up, only one postoperative PET/CT scan was performed and changes in PET/CT findings over time could not be demonstrated. However, for all but one of the patients with available clinical and imaging follow-up, monitoring showed that the findings at the postoperative bed were not related to an underlying disease. The remaining patient had recurrence of the disease at the region adjacent to the postoperative site.

Conclusions

The current study highlights the importance of correct interpretation of FDG PET/CT findings in head and neck malignancy patients after maxillectomy with reconstruction, to avoid misinterpretation of benign conditions as residual or recurrent lesions. This study shows that postoperative changes may appear during oncological follow-up and may persist for a prolonged period. Familiarity of the possible changes and follow-up scans may distinguish between pathological and incidental benign findings on FDG PET/CT imaging, and can prevent unnecessary procedures including biopsies and even additional operations.