Utility of 18F-FDG PET/CT in the Surveillance of Patients With Neurofibromatosis Type 1

Abstract

Background and Objectives

Neurofibromatosis type 1 (NF1) is an autosomal dominant disorder that predisposes affected individuals to benign and malignant tumors throughout their lifetime. Early detection of those lesions is critical for effective management. PET/CT with ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) has emerged as a potentially efficient modality for detecting malignant tumors in patients with NF1. The aim of this study was to investigate this hypothesis by assessing the presence of incidental findings of malignant and potentially malignant tumors using ¹⁸F-FDG PET/CT in an adult NF1 cohort.

Methods

In this retrospective observational study, clinical data were collected from medical records, including sex, age, family history, clinical manifestations, and tumor diagnosis. Data on imaging and diagnostic modalities that led to the diagnosis of NF1-related tumors or other malignant and potentially malignant tumors were collected. These modalities included ¹⁸F-FDG PET/CT, MRI, CT, mammography, scintigraphy, and pathology reports.

Results

A total of 79 adult patients with NF1 were included in this study, comprising 29 male and 50 female patients, all of whom had received an NF1 diagnosis. Of these, 51 patients (64.6%) underwent a ¹⁸F-FDG PET/CT scan during their follow-up. A total of 31 malignant or potentially malignant tumors were diagnosed in addition to 2 pheochromocytomas and 24 gliomas. Notably, ¹⁸F-FDG PET/CT scans incidentally detected 12 significant findings: 4 gastrointestinal stromal tumors, 3 carcinomas and premalignant tumors of the thyroid gland, 2 malignant peripheral nerve sheath tumors, 1 serous borderline ovarian tumor, 1 testicular seminoma, and 1 pheochromocytoma. Most incidental findings required surgery and could have significantly affected disease management if diagnosed at a later stage.

Discussion

In this cohort of 79 adult patients with NF1, ¹⁸F-FDG PET/CT proved as a valuable complementary imaging modality to conventional imaging by enabling the detection of incidental malignant or potentially malignant tumors. Our results highlight ¹⁸F-FDG PET/CT's beneficial impact on disease management by suggesting that the incorporation of ¹⁸F-FDG PET/CT into screening protocols could improve early detection of NF1-related cancers in asymptomatic adults, potentially offering early treatment options to improve clinical outcomes.

Introduction

Neurofibromatosis type 1 (NF1), also known as von Recklinghausen disease, is a dominantly inherited monogenetic disorder characterized by aberrant cell growth that causes multisystemic manifestations and an increased risk of developing various cancers.^1,2 It occurs with an estimated prevalence of 1:3,000.¹ In patients without affected parents, NF1 is diagnosed when 2 or more of the following criteria are met: at least 6 café-au-lait spots over 5 mm in prepubertal individuals and 15 mm in postpubertal individuals, axillary or inguinal freckling, 2 or more neurofibromas of any type or 1 plexiform neurofibroma (pNF), optic pathway glioma, 2 or more iris Lisch nodules or choroidal abnormalities, bone dysplasia, and a heterozygous pathogenic variant in the NF1 gene.^3,4 However, if 1 parent meets the diagnostic criteria, only 1 criterion is required to diagnose NF1 in the child.³ It was estimated that 97% of patients with NF1 meet these diagnostic criteria by the age of 8 and 100% by the age of 20.⁵

NF1 is caused by germline pathogenic variants in the tumor suppressor gene NF1 (17q11.2), which encodes neurofibromin, a negative regulator of RAS oncoprotein activity.^4,6,7 Over 3,000 different pathogenic variants in the NF1 gene have been reported, with 50% of instances being considered sporadic because of de novo variants.^6,8 NF1 pathogenic variants significantly increase the risk of developing cancers.^7,9 More specifically, patients with NF1 commonly develop a range of benign tumors, including but not limited to cutaneous neurofibromas (cNFs), pNFs, and intracranial gliomas, as well as other typical malignancies such as malignant peripheral nerve sheath tumors (MPNSTs).^6,9 MPNSTs, rare and aggressive tumors characterized by high local recurrence and metastatic potential, represent a severe manifestation of NF1 with a poor prognosis, often requiring aggressive treatment.^10,11 Other cancers are also related to pathogenic variants in the NF1 gene, such as gastrointestinal stromal tumors (GISTs), thyroid gland tumors, pheochromocytomas, rhabdomyosarcomas, and breast cancers.^2,9 NF1 significantly reduces life expectancy by 8–15 years in both sexes, primarily due to malignant tumor development and associated systemic complications, underscoring the critical need for enhanced surveillance in affected individuals.^9-13

The clinical heterogeneity of NF1, combined with the unpredictability of its manifestations, poses significant challenges for health care providers in counseling patients about potential complications and long-term prognosis.¹⁴ An annual comprehensive evaluation is recommended to detect and manage serious pathologic complications,¹⁴ with medical imaging playing a crucial role in detecting tumors and guiding the management and follow-ups of patients with NF1.^6,15 However, the choice of imaging modality must be carefully weighed against the increased radiation risks associated with repeated ionizing imaging.¹

Whole-body (WB) MRI is a reliable, radiation-free technique, used to assess the distribution, size, and shape of neurofibromas in patients with NF1.^8,16 It provides continuous coverage of lesions across anatomical boundaries, making it especially useful for detecting extensive infiltrative tumors that may be missed by localized MRI.¹⁷ However, its limited availability, higher cost compared with localized MRI, and several technical limitations—including difficulty imaging certain body regions, limited spatial resolution for small or infiltrative tumors, poor tumor-to-background contrast, and limited ability to differentiate malignant features in NF1-related tumors—may hinder its widespread adoption as a routine clinical evaluation tool.^17-20 PET/CT with ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) injection enables the detection of cells with a high metabolic activity and energy demand such as tumor cells.^19,21,22 This technique has been increasingly used, with its cost-effectiveness depending on multiple factors, including tumor type, the oncological stage, and the potential for long-term cost savings through improved patient management.^23,24 Of interest, several studies have demonstrated the effectiveness of ¹⁸F-FDG PET/CT in detecting tumors, especially MPNSTs in patients with NF1.^18,25-30 Nevertheless, additional evidence is required to more accurately delineate the efficacy of ¹⁸F-FDG PET/CT in enhancing medical surveillance in NF1.

NF1 is one of the most common predisposing conditions for benign and malignant tumors. However, the need for optimized surveillance strategies, the challenges in lesion detection, and the high variability in tumor manifestations complicate diagnosis and prognosis. We hypothesize that ¹⁸F-FDG PET/CT is a valuable complementary imaging modality to conventional imaging by potentially facilitating the detection of incidental malignancies in asymptomatic NF1 patients. The objective of this retrospective observational study was to evaluate the utility of ¹⁸F-FDG PET/CT in detecting significant incidental findings in a cohort of patients with NF1 followed at a university hospital clinic.

Methods

Patient Population

This retrospective observational clinical study was conducted within a single medical center and included 79 adult patients aged 18 years or older diagnosed with NF1. Patients were followed at the Neurogenetic Diseases Clinic of the Centre Hospitalier Universitaire (CHU) de Québec–Université Laval. Data were collected between January 1, 2018, and December 31, 2023, via Cristal-Net, an electronic medical record system, and included various types of information: age, biological sex, age at diagnosis, family history of NF1, and NF1 pathogenic variants, as well as the presence of NF1-specific manifestations such as café-au-lait spots, cNF, pNF, Lisch nodules, gliomas, bone dysplasia, and scoliosis, which is a common skeletal abnormality manifestation of vertebral dysplasia (Table 1).^1,3,6 Additional information, such as disease-related comorbidities (hypertension, cardiopathies, vasculopathies, learning disabilities, psychiatric disorders, polyneuropathies, and thyroid diseases) and smoking as a risk factor, was also collected (Table 1).

Table 1.

Characteristics of Patients With Neurofibromatosis Type 1 (NF1)

Patient characteristics	n = 79 No. (%)
Biological sex
Male	29 (36.7)
Female	50 (63.3)
Age at diagnosis (y)
0–9	43 (54.4)
10–19	10 (12.7)
20–29	10 (12.7)
30–39	6 (7.6)
40–49	6 (7.6)
50–59	2 (2.5)
60–69	2 (2.5)
Family history of NF1	43 (54.4)
Genetic testing for NF1	75 (94.9)
NF1 pathogenic variant	73 (92.4)
Clinical manifestations of NF1
Café-au-lait spots	79 (100.0)
Cutaneous neurofibromas	78 (98.7)
Plexiform neurofibroma	40 (50.6)
Lisch nodules	66 (83.5)
Glioma	24 (30.4)
Bone dysplasia	8 (10.1)
Scoliosis	17 (21.5)
Comorbidities
Hypertension	18 (22.8)
Cardiopathy	8 (10.1)
Vasculopathy	7 (8.9)
Learning disability	18 (22.8)
Psychiatric disorder	12 (15.2)
Polyneuropathy	2 (2.5)
Thyroid disease	19 (24.1)
Risk factor
Smoking	15 (19.0)
Medical imaging
18F-FDG PET/CT	51 (64.6)

Abbreviations: ¹⁸F-FDG = ¹⁸F-fluorodeoxyglucose; NF1 = neurofibromatosis type 1.

Data are presented as the number of patients affected and the corresponding prevalence of each feature in percentage.

Patients were referred to our clinic either on reaching adulthood or after relocating to the CHU de Québec–Université Laval catchment area. ¹⁸F-FDG PET/CT was used, based on clinical judgment, as the baseline imaging modality for newly referred patients to assess for potential malignancies and was performed in 51 of the 79 adult patients with NF1 (Table 1). NF1 pathogenic variants, tumor types, and the series of all imaging/diagnostic modalities combined (¹⁸F-FDG PET/CT, CT, MRI, mammography, scintigraphy, pathology, or endoscopy) used for the diagnosis of malignant and NF1-related tumors are summarized in eTable 1.

¹⁸F-FDG PET/CT Acquisition

¹⁸F-FDG PET/CT scans were performed using Siemens Biograph Vision, GE Discovery MI, and Siemens Biograph 6 scanners. Patients fasted for 6 hours before the examination. A dose of 2 MBq of ¹⁸F-FDG per kilogram was administered, followed by a 60-minute uptake period. CT images without contrast were acquired from the skull base to the proximal third of the femurs. The acquisition time was 1 minute 45 seconds per 25-cm bed position, unless otherwise specified by the institutional protocol for each scanner. For image reconstruction, Siemens Biograph Vision used ordered subset expectation maximization (OSEM) with point-spread function (PSF) modeling and time-of-flight (TOF), using 4 iterations and 5 subsets, a 440 × 440 matrix, and a Gaussian filter with a full width at half maximum of 4 mm. GE Discovery MI used OSEM with TOF, combined with Bayesian penalized likelihood reconstruction (Q.Clear, β = 500) and resolution recovery (SharpIR), reconstructed on a 256 × 256 matrix. Finally, Siemens Biograph 6 used OSEM with PSF modeling, reconstructed on a 168 × 168 matrix. All other reconstruction parameters were applied according to vendor specifications or institutional standards.

Each ¹⁸F-FDG PET/CT scan was independently reviewed by 1 of 6 board-certified nuclear medicine physicians at the CHU de Québec, all of whom had completed dedicated fellowships in ¹⁸F-FDG PET/CT and possess more than 50 years of combined experience in nuclear oncology. The nuclear medicine division at the CHU de Québec performs over 7,000 ¹⁸F-FDG PET/CT examinations annually, most of which are for oncological indications.

Lesion interpretation followed established clinical practice: a lesion was considered hypermetabolic when its ¹⁸F-FDG uptake was visually greater than that of surrounding tissue and/or hepatic parenchyma. No absolute standardized uptake value (SUV) cutoff was applied, as values vary across scanners and depend on clinical conditions. Furthermore, a considerable overlap has been demonstrated between SUV values for ¹⁸F-FDG uptake of MNPSTs and benign lesions.^27,28,31 Instead, each lesion was assessed individually regarding uptake intensity, morphological characteristics, location, distribution, evolution when available, and the patient's clinical context and risk factors. In patients with multiple lesions, discordant behavior of 1 lesion relative to the others could also raise suspicion for malignant transformation.

Statistical Analysis

Statistical analysis was performed using Microsoft Excel version 16.89. Basic descriptive statistics are presented as mean ± SD and interquartile range (IQR) when applicable.

Standard Protocol Approvals, Registrations, and Patient Consents

This study and access to medical records were approved by the institutional research board of the CHU de Québec–Université Laval Research Center (IRB approval number 2012-1286). Participant consent was not required.

Data Availability

The anonymized data set supporting the findings of this study is provided in eTable 1 accompanying this article. Additional data that support the findings of this study are available from the corresponding author on reasonable request.

Results

Cohort Description

Among the 79 patients included in the study cohort, 29 were male (37%) and 50 were female (63%) (Table 1). Based on patient-reported medical history taking, the age at NF1 diagnosis is distributed as follows: 43 patients (54%) before the age of 10, 10 patients (13%) between 10 and 20 years, 10 patients (13%) between 20 and 30 years, 6 patients (8%) between 30 and 40 years, 6 patients (8%) between 40 and 50 years, 2 patients (3%) between 50 and 60 years, and 2 patients (3%) between 60 and 70 years. Most patients had a family history of NF1 (43; 54%). The clinical manifestations of NF1 are presented in Table 1. The most common manifestations were café-au-lait spots, cNF, and Lisch nodules, with a prevalence of 100%, 99%, and 84%, respectively. Genetic testing was conducted in 75 patients (95%), identifying an NF1 pathogenic variant in 73 (92%). In addition, 51 patients (65%) underwent at least 1 ¹⁸F-FDG PET/CT scan (Table 1).

Use of ¹⁸F-FDG PET/CT in the Detection of Malignant Tumors

In nearly half (n = 24) of the 51 patients who underwent at least 1 ¹⁸F-FDG PET/CT scan during their clinical follow-up, ¹⁸F-FDG PET/CT led to the identification of 1 or more hypermetabolic lesions requiring further investigation to rule out malignancy (Figure 1A). In 10 of these 24 cases, at least 1 precancerous or cancerous tumor was identified (Figure 1B). Altogether, in 51 patients, 12 malignant and NF1-related tumors were diagnosed using ¹⁸F-FDG PET/CT as the first imaging modality: 4 GISTs, 3 thyroid tumors (papillary thyroid carcinoma, atypia of unknown significance), 2 MPNSTs, 1 serous borderline ovarian tumor, 1 testicular seminoma, and 1 pheochromocytoma (Figure 1C).

Figure 1. Use of 18F-FDG PET/CT as the First Diagnostic Modality in the Detection and Diagnosis of Malignant and NF1-Related Tumors.

(A) Among the 51 of 79 patients who underwent at least 1 ¹⁸F-FDG PET/CT imaging, the proportion (n) of patients with at least 1 ¹⁸F-FDG PET/CT scan showing a hypermetabolic lesion that required further investigations is illustrated. (B) Among the patients who underwent at least 1 ¹⁸F-FDG PET/CT scan showing a hypermetabolic lesion, the proportion (n) of patients whose scan(s) revealed a significant finding of malignant or NF1-related tumors is illustrated. (C) Detailed illustrations of the 12 malignant and NF1-related tumors first detected by ¹⁸F-FDG PET/CT are presented. Female genital organ tumors refer to a serous borderline ovarian tumor, male genital organ tumors refer to a testicular seminoma, and thyroid gland tumors include 2 papillary thyroid carcinomas and 1 atypia of unknown significance. Tumor details are available in eTable 1. ¹⁸F-FDG = ¹⁸F-fluorodeoxyglucose; GIST = gastrointestinal stromal tumor; MPNST = malignant peripheral nerve sheath tumor; NF1 = neurofibromatosis type 1.

Overall Tumors Detected in Our Patient Population

A total of 31 premalignant or malignant tumors were diagnosed in our cohort of 79 patients, as well as 2 pheochromocytomas and 24 gliomas when the results of all diagnostic modalities were combined (Figure 2). More specifically, 7 GISTs (12%) were diagnosed, 5 thyroid or parathyroid gland malignancies (9%; papillary thyroid carcinoma, atypia of undetermined significance, parathyroid carcinoma), 4 MNPSTs (7%), 4 breast cancers (7%), 2 female genital organ malignancies (4%; smooth muscle tumor of uncertain malignant potential, serous borderline ovarian tumor), 1 male genital organ malignancy (2%; testicular seminoma), 7 skin cancers (12%; melanoma, squamous cell carcinoma, basal cell carcinoma), 1 malignant lipomatous tumor (2%; atypical lipomatous tumor), 2 pheochromocytomas (4%), and 24 gliomas including benign tumors under regular monitoring to glioblastoma (42%). Specific details of all diagnosed tumors are provided in eTable 1, excluding gliomas, which were intentionally removed because data about their diagnosis were not all known in this retrospective study. In several patients, multiple malignancies or NF1-related tumors were identified.

Figure 2. Detailed Proportion of All Malignant and NF1-Related Tumors Diagnosed in Patients With NF1 Using All Imaging Modalities Combined.

Diagnostic modalities used in the 79 patients with NF1 are detailed in Figure 4. Tumor details are listed in eTable 1. GIST = gastrointestinal stromal tumor; MPNST = malignant peripheral nerve sheath tumor; NF1 = neurofibromatosis type 1.

With the exclusion of gliomas, among the 33 malignancies and pheochromocytomas diagnosed, the average age at diagnosis for each tumor type was 59 years for GISTs (SD = 12.9, IQR = 20.5), 49 years for thyroid malignancies (SD = 12.0, IQR = 8.0), 47 years for MPNSTs (SD = 19.7, IQR = 29.0), 52 years for breast cancers (SD = 16.4, IQR = 10.0), 46 years for female genital organ malignancies (SD = 4.2, IQR = 3.0), 50 years for the male genital organ malignancies, 55 years for skin cancers (SD = 10.5, IQR = 13.0), 53 years for the lipomatous tumor, and 41 years for pheochromocytomas (SD = 8.5, IQR = 6.0) (Figure 3).

Figure 3. Age at Diagnosis of Each Tumor Type Among All Malignant and NF1-Related Tumors Diagnosed in Patients With NF1.

Diagnostic modalities used in the 79 patients with NF1 are illustrated in Figure 4. Data are presented as mean ± SD. GIST = gastrointestinal stromal tumor; MPNST = malignant peripheral nerve sheath tumor; NF1 = neurofibromatosis type 1.

Diagnostic Modalities

A total of 12 malignant and NF1-related tumors were first detected by ¹⁸F-FDG PET/CT scan, which includes 4 GISTs, 3 thyroid gland carcinomas, 1 pheochromocytoma, 1 borderline ovarian tumor, 1 testicular seminoma, and 2 MPNSTs. Four malignancies were detected by MRI: 1 atypical lipomatous tumor, 1 breast cancer, and 2 MPNSTs. CT imaging led to the diagnosis of 2 GISTs, 1 papillary thyroid carcinoma, 1 pheochromocytoma, and 1 uterine smooth muscle tumor of uncertain malignant potential. Three breast cancers were diagnosed by mammography. In addition, 1 parathyroid carcinoma was first detected by scintigraphy, and 7 skin cancers were diagnosed in pathology by biopsy or surgical excision. One GIST was diagnosed through another modality in a patient who presented to the emergency department with symptoms of abdominal pain, upper gastrointestinal bleeding, and melena, necessitating immediate upper gastrointestinal endoscopy. A biopsy performed during this procedure initially diagnosed the GIST, which required further imaging for investigation (Figure 4).

Figure 4. First Diagnostic Modalities That Guided the Diagnosis of Malignant and NF1-Related Tumors in Patients With NF1.

Data presented refer to initial diagnostic modalities that first detected and led to the diagnosis of malignant and NF1-related tumors. The “Other” category refers to upper gastrointestinal endoscopy and biopsy, which led to a GIST diagnosis that was further investigated by imaging. ¹⁸F-FDG = ¹⁸F-fluorodeoxyglucose; GIST = gastrointestinal stromal tumor; MPNST = malignant peripheral nerve sheath tumor; NF1 = neurofibromatosis type 1.

Discussion

Early cancer screening is a key point of NF1 management, and medical imaging plays a considerable role in early detection.^14,15,32,33 The most recent European and American guidelines address the use of imaging modalities for surveillance of NF1-associated tumor types but differ in their specific recommendations.^32,33 Taken together, these guidelines recommend brain and orbital MRI to evaluate optic pathway gliomas, while abdominal MRI or CT may be indicated when a GIST is clinically suspected. Annual screening for breast cancer with breast MRI or mammography is recommended in women beginning at the age of 30, and MPNSTs are typically investigated using regional MRI, ¹⁸F-FDG PET/CT, or ¹⁸F-FDG PET/MRI.^32,33 Furthermore, 2 widely used WB imaging approaches are currently employed for the clinical assessment of adults with NF1 in developed countries, WB-MRI and ¹⁸F-FDG PET/CT.^14,15,32,33 The aim of this study was to specifically examine the utility of ¹⁸F-FDG PET/CT in the surveillance of adult patients with NF1, particularly for detecting clinically asymptomatic malignancies that directly influenced subsequent management.

Several studies have reported on the efficacy of MRI and ¹⁸F-FDG PET/CT as essential diagnostic tools in the assessment of tumors in the NF1 population.^14,15 However, there is currently no consensus on the best clinical practices and consolidated surveillance strategies across countries and NF1 specialists.^15,34 Regarding asymptomatic patients, a survey administered in 2019 to medical practitioners caring for patients with NF1 at the Response Evaluation in Neurofibromatosis and Schwannomatosis revealed that 57% of respondents' practitioners would not routinely obtain imaging in an asymptomatic adult without pNF, only 20% would perform a WB-MRI, and only 3% would perform a ¹⁸F-FDG PET/CT.³⁴ Timely use of imaging modalities and recommendations in the management of patients with NF1 vary internationally, mostly reflecting local preferences and available resources.^15,32,33

Among preferred choices for baseline imaging for NF1, WB-MRI is used for its effectiveness in assessing internal pNF burden in patients with NF1, as high internal neurofibroma burden is often associated with an increased risk of developing MPNST.^14,15,17 Concurrently, ¹⁸F-FDG PET/CT provides high diagnostic accuracy for detecting malignant transformation of pNFs^{15,18,28,30,35} and is frequently recommended to guide biopsies and support the diagnosis of MPNST, as it can detect the abnormal metabolic activity associated with these tumors.^15,18,27,31 Indeed, multiple studies reported that ¹⁸F-FDG PET/CT can differentiate MPNSTs from benign lesions with higher sensitivity than MRI, which influenced treatment by reducing the number of surgical interventions for benign lesions in the NF1 adult population.^18,25,27-30 A retrospective study of 493 patients found that ¹⁸F-FDG PET/CT provided additional specific features compared with MRI, such as detection of metastatic lesions, suggesting the utility of ¹⁸F-FDG PET/CT imaging technique in high-grade bone and soft-tissue sarcomas.²¹ Overall, WB-MRI is recognized as a valuable tool for assessing baseline internal tumor burden and pNFs, thereby guiding future surveillance in patients with NF1, while ¹⁸F-FDG PET/CT is particularly useful for characterizing MPNSTs and plays a direct role in clinical management. A recent study explored combining both imaging methods into ¹⁸F-FDG PET/MRI instead of CT, focusing on screening asymptomatic individuals at higher risk of developing MPNST.²⁶ However, taking into account patient-specific factors as well as the broader context, a multidisciplinary and individualized approach to monitoring patients with NF1 must be considered.

Despite advantages such as no radiation exposure, WB-MRI has limitations regarding image acquisition and accessibility. Some studies reported that MRI features such as intratumoral lobulation, ill-defined margins, and irregular enhancement are associated with malignancy, but these classic specific morphological features are not always present in MPNSTs, leading to missed diagnoses.¹⁸ In addition, WB-MRI faces several challenges that complicate accurate assessment and may result in undetected tumors, including difficulty imaging certain body regions, limited spatial resolution for small or infiltrative tumors, poor tumor-to-background contrast, incompatibility with metallic implants, and longer examination times with protocol variability.^17,19,20,36 There are also limited data on WB-MRI reproducibility in patients with NF1, and further studies are needed to compare its diagnostic performance with that of regional MRI.¹⁷ Finally, local socioeconomic context is an important factor when implementing cancer screening guidelines in patients with NF1 using imaging.¹⁵ In the United Kingdom and several European countries, routine WB-MRI is not currently recommended for monitoring asymptomatic tumors.¹⁵

In recent years, substantial efforts have focused on identifying cost-effective imaging modalities with high diagnostic validity. Assessing cost-effectiveness, however, remains challenging because it depends on multiple factors, including tumor type and oncological stage.^23,24 Both WB-MRI and ¹⁸F-FDG PET/CT are increasingly established in oncological diagnostics, enabling tumor-node-metastasis staging within a single examination and reducing the total number of radiologic acquisitions.²³ One study reported lower overall costs and higher effectiveness for ¹⁸F-FDG PET/CT compared with MRI in the initial staging workup of oral squamous cell carcinoma, although both modalities were considered cost-effective.²⁴ Conversely, other studies based on simple full cost analysis have estimated similar or higher costs for ¹⁸F-FDG PET/CT compared with WB-MRI.^20,23 These discrepancies highlight the need for future cost-effectiveness studies, particularly in light of ongoing technological advances that continue to improve the resolution and diagnostic accuracy of both modalities.²⁴

Regarding NF1-associated tumor types, ¹⁸F-FDG PET/CT has demonstrated valuable diagnostic contributions, supporting its key role in NF1 surveillance. ¹⁸F-FDG PET/CT offers a high diagnostic accuracy in the detection of malignant transformation of pNFs into MPNSTs.^{15,18,28,30,35} Specifically for identifying MPNSTs, ¹⁸F-FDG PET/CT demonstrated higher sensitivity (100%) compared with WB-MRI (66.7%), but lower specificity (74.4% vs 97%).¹⁸ Regarding the incidence of MPNSTs among patients with NF1, the literature reports a higher detection rate (24%) compared with our study (7%), among all tumors observed, which might be due to different imaging modalities and a larger population sample.⁹ Other ¹⁸F-FDG PET/CT studies on MPNSTs were conducted under conditions of high clinical or radiologic suspicion for malignant transformation, which resulted in higher detection rates of MPNSTs.^18,25,28 However, 1 ¹⁸F-FDG PET/MRI study found a similar proportion of MPNSTs in both asymptomatic and symptomatic NF1 participants, at 5% and 7%, respectively.²⁶

Moreover, consistent with our findings, other studies have also highlighted the value of ¹⁸F-FDG PET/CT as a valuable tool for imaging and diagnosing cancers beyond nervous system tumors. In a retrospective study examining patients with NF1 with multidetector CT screening, 6 patients were diagnosed with GISTs among a total of 95 patients with NF1 (6%) and 5 of the 6 were asymptomatic, suggesting that a significant number of patients with NF1 could potentially have subclinical intestinal GISTs.³⁷ In 3 recent studies, incidental GISTs were identified among patients with NF1 who were screened, with 4 GISTs among 156 patients (3%), 4 GISTs among 108 patients (4%), and 6 GISTs among 95 patients with NF1 (6%) using scintigraphy, ¹⁸F-FDG PET, or multidetector CT.^37-39 Our results indicate a slightly higher occurrence of incidental GISTs, with 4 GISTs among 51 patients (8%) who underwent ¹⁸F-FDG PET/CT.

Regarding thyroid cancer, a prospective study using ¹⁸F-FDG PET/CT imaging analysis was conducted in 36 patients with NF1 and tumor-related symptoms in known pNF to differentiate it from MPNSTs and found 9 lesions other than neurogenic tumors: 5 thyroid lesions (14%) and 3 malignant lesions including 1 liposarcoma and 2 GISTs.²⁸ In another study involving ¹⁸F-FDG PET/CT performed on 69 patients with NF1 with a focus on the thyroid, 1 patient was diagnosed with thyroid metastasis and another with medullary thyroid cancer.⁴⁰ These results were similar to our study, where 3 thyroid cancers were detected using ¹⁸F-FDG PET/CT among the 51 patients (6%) screened.

Pheochromocytomas and paragangliomas are mostly benign but remain important to detect because of increased cardiovascular risk, and their prevalence is higher in patients with NF1 than in the general population, estimated between 2% and more than 20%.^38,39,41 However, the actual diagnostic method for pheochromocytoma in adults with NF1 in the United States, as recommended by the American College of Medical Genetics and Genomics, relies on biochemical testing of plasma free metanephrines.³³ Two retrospective cohort studies on pheochromocytoma and paraganglioma detection in NF1 patients—using biochemical screening combined with imaging—reported incidental tumor rates of 31% and 47%. Of interest, these lesions are described as mainly asymptomatic, emphasizing the importance of regularly monitoring patients with NF1 for such findings.^41,42 Among the 2 cases of pheochromocytoma detected in our study, 1 was detected incidentally by ¹⁸F-FDG PET/CT. Although ¹⁸F-FDG PET-CT was not the primary diagnosis modality, a small clinical trial investigating dynamic genotype-specific differences in 2 patients with NF1 pathogenic variants found that ¹⁸F-FDG PET/CT pharmacokinetics demonstrated utility in evaluating tumor metabolism in pheochromocytomas and paragangliomas, with high sensitivity but limited specificity for detecting malignant tumors, underscoring the importance of a multimodal imaging approach for accurate tumor characterization in patients with NF1.²²

Globally, considering all imaging modalities and all tumors observed, we observed higher detection rates for certain tumors in our cohort compared with a registry study that investigated more than 1400 patients with NF1 in the Finnish population: GISTs (12% vs 1%), thyroid tumors (9% vs 1%), female genital tumors (4% vs 1%), skin cancers (12% vs 2%), CNS tumors (mostly gliomas) (42% vs 37%), and pheochromocytoma tumors (4% vs 1%). Conversely, we noted lower detection rates for breast cancers (7% vs 13%), MPNSTs (7% vs 24%), and male genital tumors (2% vs 4%).⁹ Overall, while the detection rates are of the same order of magnitude, some notable differences remain that could be attributed to variability in imaging modalities used within the records and the number of individuals in the cohort. This study was based on national registry data, and imaging modalities were not reported.⁹ In other oncological settings, ¹⁸F-FDG PET/CT has proven effective in accurately detecting various malignancies, including lymphoma, melanoma, lung cancer, colorectal cancer, and head and neck tumors.^19,21,32-34 This highlights the potential usefulness of WB screening, particularly with ¹⁸F-FDG PET/CT, in detecting other cancers in the NF1 population, as demonstrated by our study's significant findings of premalignant and malignant lesions initially first identified using this technique. Almost all of these significant findings eventually required surgery.^{43 18}F-FDG PET/CT, as a preoperative tool in patients with NF1, helps to prevent unnecessary surgeries for benign tumors and facilitates tailored treatment based on the severity or malignancy of the lesions.²⁹ Finally, ¹⁸F-FDG PET/CT offers the unique advantage of identifying metabolically active tumors, enabling early detection of malignancies among incidental findings in the nervous system and beyond.

Although ¹⁸F-FDG PET/CT is associated with radiation risks that are not present with MRI, millisievert effective doses can be managed and kept as low as reasonably achievable while maintaining efficiency.⁴⁴ Radiation exposure must be carefully discussed, particularly in the pediatric population. In pediatric syndromes predisposing to multiple cancers, such as Li-Fraumeni syndrome, hereditary paraganglioma-pheochromocytoma syndrome, and rhabdoid tumor syndrome, WB-MRI is considered a safe and effective technique for cancer screening in children.^45,46 Another study compared ¹⁸F-FDG PET/CT and WB-MRI in a young adult population with malignant lymphomas and sarcomas, and while the overall results were similar, ¹⁸F-FDG PET/CT detected more different types of cancers compared with WB-MRI (lymph nodes, bone lesions, splenic lesions, liver lesions, and lesions in Waldeyer ring).²⁰ It is desirable to minimize the radiation risk of ¹⁸F-FDG PET/CT for secondary malignancies in children while preserving its effectiveness to detect MPNSTs.^31,47 A retrospective study found that the cumulative absolute risk of brain tumors and leukemia in children after radiation imaging exposure is small and outweighed by clinical benefits.⁴⁸ In adults, radiation risks are generally considered lower than in children because of their cellular maturity and shorter life expectancy.^49,50 However, minimizing radiation exposure remains essential to reduce the risk of cancer.⁴⁹ Differences in exposure levels, radiation sensitivity, types of cancer studied, and confounding factors due to advancements in technology and lifestyle changes over the past century must be considered. A right balance must be considered between radiation risk and the potential risk of missing a malignant tumor, which could significantly decrease life expectancy in patients with NF1.

This study is limited by its retrospective and observational nature. Collected data could be incomplete or biased by the quality of the medical records and available information. For some patients referred to our clinic from more distant regions, medical history may be incomplete because of the lack of fully digitized medical records or the absence of initial follow-up by an NF1 neurology specialist. This explains, for example, the wide variation in age at diagnosis, which is also influenced by recall bias, as well as the missing information on glioma diagnoses, most of which were made during childhood. In addition, some patients continued their follow-up outside the CHU de Québec, resulting in a loss of follow-up. Moreover, no systematic criteria were established for prescribing an ¹⁸F-FDG PET/CT scan, and some patients did not undergo the procedure. One confounding factor is the use of different ¹⁸F-FDG PET/CT devices, with reduced imaging time and radiation exposure from the newer devices acquired during the observation time of our study, resulting in nonreproducibility of our results in other centers. Furthermore, our study did not compare ¹⁸F-FDG PET/CT with other WB imaging modalities, such as WB-MRI, because of its unavailability in our region's public health system. Therefore, further research is needed to compare both imaging modalities for their potential value in detecting malignancies in the NF1 population and their optimal clinical use.

This study highlights the unique advantages of ¹⁸F-FDG PET/CT in identifying metabolically active tumors and enabling early detection of malignancies or potential malignancies among incidental findings in adults with NF1. Our findings support its role as a valuable complementary imaging modality to conventional imaging and suggest its integration into baseline assessments for selected patients with NF1, as it enabled the detection of clinically asymptomatic malignancies in our cohort that directly influenced subsequent management. This could pave the way toward establishing consensus on medical surveillance for NF1-related cancers in adult patients to ultimately improve clinical practices. Alongside other WB imaging techniques, ¹⁸F-FDG PET/CT could significantly enhance clinical management and patient outcomes in the NF1 population.

Glossary

¹⁸F-FDG

¹⁸F-fluorodeoxyglucose

CHU

Centre Hospitalier Universitaire

cNF

cutaneous neurofibroma

GE

General Electric

GIST

gastrointestinal stromal tumor

IQR

interquartile range

MI

molecular imaging

MPNST

malignant peripheral nerve sheath tumor

NF1

neurofibromatosis type 1

OSEM

ordered subset expectation maximization

pNF

plexiform neurofibroma

PSF

point-spread function

RAS

rat sarcoma

SUV

standardized uptake value

TOF

time-of-flight

WB

whole-body

WB-MRI

whole-body magnetic resonance imaging

Author Contributions

M.G. Tanase: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data. L. Djilani: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data. R. Lamontagne: major role in the acquisition of data. R. Derbel: major role in the acquisition of data. M. Baril: drafting/revision of the manuscript for content, including medical writing for content. V. Roy: drafting/revision of the manuscript for content, including medical writing for content. M. Blais: drafting/revision of the manuscript for content, including medical writing for content. M. Leclerc: drafting/revision of the manuscript for content, including medical writing for content. F.A. Buteau: drafting/revision of the manuscript for content. H.T. Khuong: study concept or design. F. Gros-Louis: drafting/revision of the manuscript for content, including medical writing for content. N. Dupre: drafting/revision of the manuscript for content, including medical writing for content.

Study Funding

The authors report no targeted funding.

Disclosure

The authors report no relevant disclosures. Go to Neurology.org/NG for full disclosures.