Combined [18F]Fluorodeoxyglucose PET and [123]] Iodometomidate-SPECT for diagnostic evaluation of indeterminate adrenal neoplasias-the cross-sectional diagnostic test accuracy study FAMIAN
Check for updates
Stefanie Hahner,* Philipp Hartrampf, Felix Beuschlein,d,e Matthias Miederer,f.g.h, Konstanze Miehle, Wiebke Schlötelburg, Carmina Teresa Fuß, Thomas Pfluger,k Christian Fottner,” Anke Tönjes, Ken Herrmann,™ Holger Amthauer,” Martin Reincke,e Mathias Schreckenberger,“ Osama Sabri,“ Johanna Werner,ª Miriam Reuter,ª Stefan Kircher,P Wiebke Arlt,9.” Martin Fassnacht,“ Andreas Konrad Buck,b Hans-Helge Müller,S,t and Andreas Schirbel,b,t for the FAMIAN investigators
oa OPEN ACCESS
ªDivision of Endocrinology and Diabetes, Department of Internal Medicine I, University Hospital, University of Würzburg, Würzburg, Germany Department of Nuclear Medicine, University Hospital, University of Würzburg, Würzburg, Germany
‘Klinik für Endokrinologie, Diabetologie und Klinische Ernährung, UniversitätsSpital Zürich (USZ) und Universität Zürich (UZH), Zurich, Switzerland
dThe LOOP Zurich - Medical Research Center, Zurich, Switzerland
eMedizinische Klinik und Poliklinik IV, LMU Klinikum, Ludwig-Maximilians-Universität München, Munich, Germany
‘Department of Translational Imaging in Oncology, National Center for Tumour Diseases (NCT/UCC) Dresden, Faculty of Medicine and University Hospital Carl Gustav Carus, University of Technology Dresden (TUD), Dresden, Germany
9German Cancer Research Center (DKFZ) Heidelberg, Heidelberg, Germany
hHelmholtz-Zentrum Dresden-Rossendorf (HZDR), Dresden, Germany
¡Department of Nuclear Medicine, University Medical Center of the Johannes Gutenberg University Mainz, Mainz, Germany Medical Department III- Endocrinology, Nephrology, Rheumatology, University of Leipzig Medical Center, Leipzig, Germany Department of Nuclear Medicine, University Hospital, Ludwig-Maximilians-Universität Munich, Germany
“Department of Endocrinology and Metabolism, I Medical Clinic, University Medical Center of the Johannes Gutenberg University Mainz, Germany “Department of Nuclear Medicine, West German Cancer Center (WTZ), University Hospital Essen, University of Duisburg-Essen; German Cancer Consortium (DKTK), Partner Site University Hospital Essen, Essen, Germany
“Charité - Universitätsmedizin Berlin, Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Department of Nuclear Medicine, Berlin, Germany
“Department of Nuclear Medicine, University Hospital Leipzig, Germany
PInstitute of Pathology, University of Würzburg, Würzburg, Germany
ªMedical Research Council Laboratory of Medical Sciences, London, UK
“Institute of Clinical Sciences, Faculty of Medicine, Imperial College London, London, UK
SInstitute of Medical Bioinformatics and Biostatistics, Philipps University of Marburg, Germany
Summary
Background Adrenal tumours are frequently detected by conventional imaging. However, computed tomography and magnet resonance imaging have limited specificity in classifying the most prevalent tumour type, adrenocortical adenoma (ACA), which typically does not require surgery. We proposed that combined molecular imaging with [18F] Fluorodeoxyglucose-positron emission tomography (FDG PET) and [123I]Iodometomidate-single photon emission tomography (IMTO SPECT) improves non-invasive classification of ACA.
Methods This cross-sectional, multicentre diagnostic study included patients (≥30 years) with non-functioning indeterminate adrenal masses (>3 cm or increase >1 cm, Hounsfield units [HU] >10 on unenhanced computed tomography [CT]) scheduled for surgery. Using histopathology as the reference, we assessed the accuracy of FDG/ IMTO imaging as an ACA-test, assuming that low FDG with high IMTO uptake is indicative of ACA, with a focus on high specificity and moderate to high sensitivity. We also investigated its accuracy in detecting or excluding adrenocortical carcinoma (ACC) and evaluated FDG and unenhanced CT in assessing malignancy. Trial-registration: EudraCT 2012-003604-13; ClinicalTrials.gov-identifier NCT02010957.
eBioMedicine 2025;116: 105735
Published Online xxx https://doi.org/10. 1016/j.ebiom.2025. 105735
E-mail address: hahner_s@ukw.de (S. Hahner).
‘Equal contribution.
Findings From July 2015 to December 2020, 85 patients were enrolled, with 77 included in the final analysis (53 benign, 30 ACA, 9 ACC). FDG/IMTO-imaging classified ACA with high specificity (95-7% [95% CI 85.2%-99.47%]), high positive predictive value (87.5% [95% CI 61.7%-98-4%]) and high positive likelihood ratio (11.1 [95% CI 3-2-122]). However, sensitivity was low (48.3% [95% CI 29-4%-67.5%]) due to moderate/high FDG uptake in 14 of 30 ACA. Malignant masses were classified with high sensitivity but low-to-moderate specificity by both unenhanced CT (cut-off HU ≥20 sensitivity 100% [95% CI 85.8%-100%], specificity 26-4% [95% CI 15.3%-40.3%]) and FDG (visual analysis sensitivity 95.8% [95% CI 78.9%-99.9%], specificity 62.3% [95% CI 47.9%-75.2%]). All four study-related AEs were grade 1, the seven serious AEs were not study-related.
Interpretation Combined FDG/IMTO-imaging classifies ACA with high specificity, potentially reducing unnecessary surgery. A sub-group of FDG-positive ACA lowers sensitivity.
Funding German Research Foundation and EU-FP7.
Copyright @ 2025 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Keywords: Adrenal incidentaloma; Molecular imaging; Fluorodeoxyglucose positron emission tomography; Computed tomography; Iodometomidate SPECT; Adrenocortical adenoma; Adrenocortical carcinoma; Indeterminate adrenal masses
Introduction
Adrenal tumours are detected with increasing frequency in the context of an ageing population and broadening use of cross-sectional imaging, posing a growing public health challenge. The prevalence of adrenal incidenta- lomas in imaging studies is around 3% in adults over 50 years and up to 10% in patients over 80 years of age.1 Adrenocortical adenoma (ACA) is the most common underlying entity, followed by metastases, adrenocor- tical carcinoma (ACC), phaeochromocytomas and rarer entities.2-4 Phaeochromocytomas and adrenocortical masses causing hormone excess can be identified bio- chemically. However, most adrenal masses do not cause signs and symptoms of hormone excess. The clinical challenge lies in determining the optimal therapeutic approach: adrenalectomy is performed for tumours causing overt hormone excess. It is also considered in non-functioning tumours depending on the perceived risk for malignancy based on imaging findings. Esti- mating the malignant potential of an adrenal tumour can be challenging, leading to unnecessary surgery, as a significant number of tumours are ultimately found to be benign on histopathology.5,6 Assessment of tissue density by determination of the tumour attenuation value in Hounsfield Units (HU) in unenhanced computed tomography (CT) is well established, whereby a cut-off value ≤10 is highly indicative of a benign tumour.3,6,7 However, more than 40% of benign tumours exceed this threshold.3,6,8,9 Raising the threshold from 10 to 20 HUs significantly improves specificity for detection of malignancy with minimal loss of sensitivity.6,8,9 Nevertheless, even in tumours with an attenuation value >20 HU, ACA is still the most common entity.3,10
As complementary imaging options to estimate the malignant potential of adrenal lesions, magnetic
resonance imaging (MRI) with chemical shift, analysis of contrast washout in serial CT and [18F]FDG PET imaging have been utilised.1 However, data regarding the diagnostic utility of these imaging methods in differentiating benign from malignant adrenal lesions is generally scarce.7 Furthermore, the additional benefit for wash-out CT was found to be limited.10-12 Current evidence supports FDG PET imaging,13-19 showing high sensitivity for classification of malignant tumours while reaching moderate to high specificity.
However, most studies included patients who were either already surgical candidates due to hormone hypersecretion or in whom surgery was not required due to unenhanced tumour attenuation <10 HU. Consequently, the number of individuals reported in whom FDG imaging was relevant for therapeutic guid- ance has been limited. Considering the high proportion of patients with ACA even after preselection by HU cutoffs, a non-invasive method to classify ACA with high validity would be of major clinical value and avoid unnecessary adrenalectomy. As management strategies for ACC differ from those for non-adrenocortical (non- AC) malignant adrenal masses, non-invasive character- isation of these two tumour categories would be of additional value.
Metomidate and its analogous tracer compounds specifically bind to the cytochrome P450 enzymes 11ß- hydroxylase (CYP11B1) and aldosterone synthase (CYP11B2) in adrenocortical cells.20-28
We aimed at evaluation of combined use of [18F]FDG PET and [123I]Iodometomidate-(IMTO) SPECT imaging for non-invasive classification of such adrenal masses in a multicentre cross-sectional diagnostic study. The primary objective was to determine the accuracy of a combined imaging test for classification of ACA, aiming to achieve high specificity (above 90%) and a high
Research in context
Evidence before this study
Tumours of the adrenal gland are common and often discovered incidentally. Current guidelines recommend additional diagnostic tests to determine whether surgical removal of the tumour is required on the basis of hormone excess or malignant potential. In the vast majority of cases, however, incidentally discovered adrenal masses are benign adrenocortical adenomas (ACA) that do not necessitate further intervention. Non-invasive methods for a reliable and highly specific classification of these ACA have not yet been established. Before initiation of the FAMIAN study, we searched Pubmed, the Cochrane library and Science citation index for publications from January 1st 1970 to January 1st 2015 using combinations of the search terms “adrenal incidentaloma”, “adrenal imaging”, “metomidate”, “etomidate”, “FDG PET”, “molecular imaging”. In addition, in preparation of the study, we organised an ENSAT (European Network for the Study of Adrenal Tumours) Imaging Study Group meeting at the European Conference on Adrenal Imaging, bringing together leading experts in the field of adrenal imaging from several European countries. While FDG PET/CT diagnostics revealed to be the best-established functional imaging method so far to differentiate benign from malignant processes, molecular imaging with the CYP11B1/2 ligand metomidate had been shown to discriminate between adrenocortical and non-adrenocortical lesions in several studies. Two studies, using both FDG and [11C]metomidate imaging included a total of 35 patients. Both revealed that [11C]metomidate distinguishes between adrenocortical and non-adrenocortical lesions, whereas FDG proved valuable in discriminating benign from malignant tumours. The available series investigating FDG- or metomidate-based molecular imaging in adrenal tumours included tumours for which the question of malignant potential or adrenocortical origin could already be approximated based on conventional imaging results such as unenhanced CT tumour attenuation values ≤10 Hounsfield Units (HU) or biochemical and/or clinical evidence of adrenal steroid excess. In addition, final diagnosis based on histopathology was often only available for a subgroup of investigated patients. Analysing the clinical utility of FDG and metomidate tracers for adrenal molecular imaging and particularly the significance of their combined use requires a
prospective study in the most relevant patient group, i.e. patients with adrenal incidentalomas without relevant steroid hormone excess but with uncertain malignant potential based on the results of conventional cross-sectional imaging, i.e. CT or MRI. We hypothesised that adrenal lesions with low FDG uptake and high metomidate uptake would have a high probability of representing an ACA and, thus, not require adrenalectomy.
Added value of this study
In this multicentre, investigator-initiated diagnostic study, we assessed the diagnostic accuracy of combined [18F] Fluorodexoyglucose-positron emission tomography (FDG PET) and [123I]Iodometomidate-single photon emission tomography (IMTO SPECT) imaging in patients with indeterminate adrenal masses. [123I]IMTO was chosen because of its broader applicability compared to [11C]metomidate, which is limited by the need for local availability of a cyclotron due to half-life of only about 20 min of carbon-11. No previous study has investigated this and there is also no other study on any tool for non-invasive classification of ACA in indeterminate adrenal masses. We demonstrate that combined FDG-/IMTO-imaging is of clinical value. However, sensitivity of the combined molecular imaging test was limited by a sub-group of ACA with high FDG uptake, which warrants further investigation. Furthermore, we provide information on the distribution of underlying histopathological diagnoses in this specific group of indeterminate non-functioning adrenal lesions and additional information on the diagnostic accuracy of unenhanced CT and FDG in assessing malignancy in adrenal masses, observing high sensitivity but only low specificity for both FDG and particularly for unenhanced CT.
Implications of all the available evidence
This cross-sectional diagnostic study systematically evaluated the use of molecular imaging to differentiate indeterminate adrenal masses, which have traditionally been treated with surgical removal despite most being benign adrenocortical adenomas (ACA). Implementing this innovative diagnostic method may help to reduce unnecessary surgeries, thereby improving health outcomes and lowering healthcare costs.
likelihood ratio (above 9). We proposed to reach a positive predictive value >90%, indicating high clinical utility.
Methods
Study design and participants
The Combined 18F-Fluorodeoxyglucose (FDG) Positron Emission Tomography (PET) And Metomidate Imaging
for Adrenal Neoplasia (FAMIAN) study was designed as a prolective (in the sense that the data acquisition takes place after the study planning), cross-sectional, multi- centre diagnostic test study assessing the accuracy of combined FDG-/IMTO-imaging in classifying an inde- terminate adrenal mass as ACA or non-ACA (primary diagnostic test). Low FDG uptake (FDG-negative) with high IMTO uptake (IMTO-positive) was considered diagnostic for a benign adrenocortical adenoma (ACA).
We also assessed the accuracy of FDG and unenhanced CT tumour attenuation regarding the diagnosis of malignancy.
Participants presenting at participating study centres were eligible for recruitment if they fulfilled the following inclusion criteria: diagnosis of an adrenal mass of more than 3 cm diameter (or increase in tumour diameter by more than 1 cm upon follow-up imaging) and indeterminate imaging findings defined by a tumour attenuation ≥10 HU on unenhanced CT. Recruitment was restricted to adults ≥30 years both for reasons of radiation protection and as large and/or growing adrenal masses in younger patients are invari- ably surgically removed.
Exclusion criteria were biochemical confirmation of phaeochromocytoma, primary aldosteronism or adrenal Cushing’s syndrome, the latter defined by fulfilment of all three of the following criteria: 1) morning serum cortisol after 1 mg dexamethasone >5 µg/dl (140 nmol/ L); 2) Plasma ACTH <5 ng/L; 3) urinary free cortisol levels twice the upper limit of normal AND/OR bedtime salivary cortisol thrice the upper limit of normal. Details in Appendix p 3.
Study procedures
Study recruitment took place between July 20th 2015, and December 14th 2020 at six tertiary care endocrine specialist centres in Germany (Berlin, Essen, Leipzig, Mainz, Munich, Würzburg). All participants provided written informed consent before inclusion. The study was approved by the ethics committee of the University of Würzburg (No. 282/13_ff) and additionally at the local ethics committees at each trial site as well as by the Federal Institute for Drugs and Medical Devices, BfArM, No. 61-3910-4040442 and the Federal Institute for Radiation Protection, BfS, No. Z5-22463/2-2014- 016). The study was registered at the European Union Drug Regulating Authorities Clinical Trials Database (EudraCT No. 2012-003604-13) and at ClinicalTrials.gov (identifier NCT02010957).
Four study visits were scheduled: screening visit (visit 1), diagnostic evaluation by FDG and IMTO im- aging (visits 2 and 3), and follow-up visit for safety and adverse event assessment 2-4 weeks after visit 3 or preoperativley (visit 4). An optional visit 5 was foreseen 6-12 months after visit 3 in case no surgical interven- tion was performed. Adverse events (AE) were reported until 30 days after second imaging procedure or surgery- whichever came first. Details in Appendix p 3 and 4.
Molecular imaging procedures
[18F]FDG PET was performed using state-of-the-art hybrid PET/CT devices. Static emission imaging was performed 1 h after i.v .- injection of FDG according to PET/CT scanners and guidelines.
[123]]Iodometomidate was prepared as previously described.28 IMTO SPECT was performed with state-
of-the-art gamma cameras and acquisition settings adequate for I-123. After receiving 185 MBq [123I]IMTO i.v., planar scans of the whole body were acquired 4 h post injection using a standard technique. All patients also underwent SPECT/CT imaging between 4 and 6 h post injection. Transmission was measured by low-dose CT. Details in Appendix p 5.
Centralised imaging review
The central imaging board was located at the University Hospital Würzburg. Review of imaging was undertaken by a multidisciplinary panel consisting of two endocrine specialists and two nuclear medicine physicians for all FDG PET scans, all IMTO whole-body planar scans, and all IMTO SPECT scans. All panel members were blin- ded to clinical information including histopathology. Decisions on visual image interpretation were made by consensus using a semi-quantitative scoring system ranging from 1 to 5: Visual scores of 1 and 2 indicate no or low tracer uptake (classified negative regarding tracer uptake), whereas scores 4 and 5 indicate moderate or high tracer uptake (positive). Tumours with intermediate tracer uptake were rated 3 (indeterminate). Regarding the management of lesions classified as indeterminate in the visual image analysis, the following criteria were estab- lished for subsequent statistical evaluation: FDG PET (aiming at high sensitivity for malignancy): FDG inde- terminate was classified as malignant. IMTO SPECT (aiming at high specificity for adrenocortical origin): IMTO indeterminate was classified as non-adrenocortical. Quantitative analyses for tracer accumulation were per- formed by placing a volume-of-interest (VOI) around the adrenal mass and the reference (background) regions (contralateral adrenal, prevertebral region and liver). SUVmax, SUVmean and SUVpeak and respective tumour-to-background ratios (TBR) were determined. Details in Appendix p 5.
Reference standard
The reference standard for ACA, non-adrenocortical benign, ACC, and non-adrenocortical malignant tu- mours was based on tumour histopathology. Paraffin blocks of the resected tumour tissue were re-analysed by one of two central reference pathologists without knowledge of the functional imaging results (Wolfgang Saeger, Hamburg, and Stefan Kircher, Würzburg). Details in Appendix p 5.
Statistical analysis
We assumed that in at least half of patients with inde- terminate adrenal masses, these masses are ACA, where surgery is unnecessary. We evaluated an ACA index test based on combined FDG PET and IMTO SPECT imaging for non-invasive classification of indeterminate adrenal masses. We conducted a cross-sectional evalu- ation to determine the test’s accuracy. The primary objective was to assess the accuracy of the combined
imaging test in classifying ACA (ACA+ and ACA-) and to demonstrate high specificity (>90%) and a high positive likelihood ratio (>9). We proposed to reach a positive predictive value >90%.
The confirmatory statistical analysis tested four null- hypotheses hierarchically, strongly maintaining a type I error level of a = 5% (2.5% for each side, the positive and the negative direction of the alternative hypotheses). The first two tests consider the accuracy measures of the diagnostic ACA-test, its specificity and likelihood ratio of a positive ACA-test (ACA+), respectively: The positive directions of the alternative hypotheses are specificity P (ACA-|non-ACA) > 90% (establishing a specificity of more than 90%) and positive likelihood ratio P (ACA+| ACA)/P (ACA+|non-ACA) >9, respectively. The third and fourth test focused on the ACA-test’s ability to distinguish between ACC and non-ACC tumours, detecting ACC clearly as ACA- and detecting ACC considerably less likely than non-ACC as ACA+. The positive directions of the alternative hypotheses are P (ACA-|ACC) >80% and likelihood ratio P (ACA+|ACC)/ P (ACA+|non-ACC) <7/19.
Rejection of the null-hypotheses in the positive directions indicates, after significant first test, a speci- ficity over 90%, after significant second test-assuming a pre-test probability for ACA of at least 50%-a positive predictive value P (ACA|ACA+) over 90%, after signifi- cant third test a probability P (ACA-|ACC) of more than 80% that ACC is detected as ACA-, and after significant fourth test-assuming a pre-test probability for ACC of at most 12.5%-a predictive value P (ACC|ACA+) for ACC after positive ACA-test below 5%.
Statistical methods included exact binomial tests and Clopper-Pearson confidence intervals (CIs) for speci- ficity, sensitivity, and probabilities. Likelihood ratios were tested and estimated with exact tests and CIs. ROC analyses determined the area under the curve (AUC) with 95% CIs. All two-sided 95% CIs are the in- tersections of two one-sided 97.5% CIs.
Analyses were performed with SAS, version 9-4 (SAS Institute, Cary, NC), StatXact of Cytel Studio, version 6.3.0 (Cytel Studio; Cytel Corporation, Cambridge, MA), GraphPad PRISM 6, IBM SPSS28, and the R Project for Statistical Computing.
Sample size calculation for the first two primary hi- erarchically structured two-sided level 5% tests on the accuracy of the diagnostic ACA-test assumed a 60% a priori probability for ACA. Detection of the first positive alternative hypothesis for the ACA-test with 80% power at 98% specificity, required about 195 participants, including at least 70 with confirmed non-ACA neoplasia for the exact statistical test. For the second positive alternative hypothesis for the ACA-test with 80% power, at 98% specificity, and 92% sensitivity, about 187 assessed participants were required.
Accounting for a 10% incomplete assessment rate, 220 participants were planned for enrolment, with a
screening target of 550 patients assuming a 40% recruitment rate.
For the third and fourth hypotheses, power calcula- tions based on 195 assessed participants, a 60% proba- bility for ACA, and a 10% probability for ACC, indicated about 70% power for the third positive alternative hypothesis if P (ACA-|ACC) = 98% and nearly 80% power for the fourth positive alternative hypothesis if P (ACA-|ACC) = 98% and P (ACA+|non-ACC) = 61.5%.
Role of the funding source
The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.
Results Patient cohort
A total of 85 patients were recruited (Fig. 1). Due to slow recruitment and the discontinuation of GMP-compliant IMTO deliveries during the COVID19-pandemic, the study ended early. Eight patients were excluded from the final analysis: one screen failure, two withdrew consent before study procedures, one declined further in- vestigations after FDG PET, two had HU <10 at central imaging reading, one died before adrenalectomy, and one declined adrenalectomy as functional imaging indicated a benign non-adrenocortical tumour (follow up imaging showed decreased tumour diameter but he was excluded from final analysis due to missing histo- pathology) (Fig. 1).
The final analysis cohort comprised 77 patients with available histopathology (Fig. 1, Table 1). All underwent FDG PET imaging, whereas IMTO SPECT imaging was performed in 73 patients. Four patients did not undergo IMTO SPECT due to delivery problems during the pandemic or radiotracer quality control failures.
Histopathology revealed a wide spectrum of differ- ential diagnoses: 53 (69%) were benign and 24 (31%) were malignant, including 30 ACA (39.0%), 9 ACC (11.7%), 23 non-AC benign tumours (29.9%), and 15 non-AC malignant tumours (19.5%) (Fig. 1, Tables 1 and 2; details in Appendix p 7).
Imaging findings
Combined FDG/IMTO imaging
The primary objective was to classify ACA versus non- ACA by combined FDG/IMTO imaging (FDG-nega- tive/IMTO-positive for ACA) (Fig. 2). If FDG PET alone excluded ACA (due to high FDG uptake, n = 2) or ACC (due to low FDG uptake, n = 2), those results were included in the analysis of the ACA- or ACC-test, even without IMTO imaging.
Of 46 non-ACA included in the ACA test analysis, two were incorrectly characterised as ACA based on visual interpretation, resulting in a specificity of 95.7% (95% CI 85.2%-99-47%), although this did not reach
Articles
376 patients with newly diagnosed adrenal tumours screened for those with a non-functioning adrenal mass >3cm diameter (or increase >1cm in 12 months), an unenhanced CT tumour attenuation value ≥10 HU, and scheduled for surgical removal
294 not approached as they did not meet inclusion criteria
85 patients enrolled
3 patients excluded prior to any study-related procedures
· Screen failure (n=1)
· Withdrawal of consent (n=2)
1 patient declined further investigations and surgery after FDG-PET
4 patients underwent both FDG- and IMTO-imaging but were excluded from final analysis due to
77 patients with imaging (FDG-PET n=77, IMTO-SPECT n=73) and histopathology2 results included in final analysis:
· central imaging reading established HU<10 (n=2)
Adrenocortical Adenoma (ACA) (n=30)
Non-adrenocortical benign tumour (n=23)4
Adrenocortical Carcinoma (ACC) (n=9)
Non-adrenocortical malignant tumour (n=15)3
· death before adrenalectomy,
no histopathology (n=1)
· Surgery declined, no histopathology (n=1)
2 FDG negative tumours with missing IMTO excluded from ACA test; 2 FDG positive tumours with missing IMTO excluded from ACC test
75 included in evaluation of ACA test:
75 included in evaluation of ACC test:
Adrenocortical Adenoma (n=29)
Non-adrenocortical benign tumour (n=22)
Adrenocortical Carcinoma (n=9)
Non-adrenocortical malignant tumour (n=15)
Adrenocortical Adenoma (n=29)
Non-adrenocortical benign tumour (n=23)
Adrenocortical Carcinoma (n=9)
Non-adrenocortical malignant tumour (n=14)
statistical significance for >90% specificity (p = 0-15). The positive likelihood ratio was 11.1 (95% CI 3.19-122.2). Among 29 ACA included in the analysis of the ACA-test (28 with combined imaging result), 14 were correctly characterised as ACA based on visual interpretation of combined imaging. Twelve ACA were both IMTO- and FDG-positive and one ACA was IMTO- positive and FDG-indeterminate and misclassified as ACC (Fig. 3a). One ACA with only 30% vital adenoma fractions was both IMTO- and FDG-negative and one ACA with missing IMTO imaging was FDG positive and misclassified as non-ACA. Thus, sensitivity for classification of ACA was only 48.3% (95% CI 29.4%-67.47%).
The secondary objective was classification of ACC versus non-ACC. Of nine ACC, five were correctly char- acterised as ACC (FDG-positive and IMTO-positive),
whereas four were FDG-positive but IMTO-negative, resulting in a sensitivity of 55-6% (95% CI 21-2%- 86.3%). Among 66 non-ACC, 13 ACA and 3 non-AC tu- mours (one malignant, two benign) were misclassified as ACC, resulting in a specificity of 75.8% (95% CI 63-6%- 85.5%). Using quantitative image analyses, the test per- formance for the ACA test, the ACC test, as well as for the malignancy test and the classification as adrenocortical versus non-adrenocortical could be improved. Results and test performance of combined imaging are shown in Table 2, Fig. 3 and in the Appendix p 7, Supplementary Figure S4.
[18F]FDG PET
In visual analysis of all 77 patients, FDG PET revealed high sensitivity (95.8%; 95% CI 78.9%-99.89%) and a high negative predictive value (97.1%; 95% CI 84-7%-
| All benign (n = 53) | ACA (n = 30) | Non-AC benign (n = 23) | All malignant (n = 24) | ACC (n = 9) | Non-AC malignant (n = 15) | |
|---|---|---|---|---|---|---|
| Age (years) | 57 0 (33-79) | 57.5 (33-79) | 56 0 (36-79) | 57.5 (30-78) | 53 0 (32-76) | 60-0 (30-78) |
| [52.5; 65.0] | [53.0; 65.25] | [49.0; 64-0] | [51.25; 65.0] | [42.0; 72.5] | [52.0; 65.0] | |
| Sex | F 25, M 28 | F 17, M 13 | F 8, M 15 | F 13, M 11 | F 7, M 2 | F 6, M 9 |
| History of previous malignancies, na | 5 | 4 | 1 | 5 | 1 | 4 |
| Adrenal tumour characteristics | ||||||
| Max. diameter (cm)b | 4.5 (2.0-18.1) | 4.4 (2.7-10) | 6.6 (2-0-18.1) | 7.2 (3-4-22) | 7.4 (3-4-14.9) | 6.9 (3.7-22) |
| [3.5; 6.9] | [3.7; 5.2] | [3.4; 9.3] | [4.5; 9.9] | [4.1; 11.2] | [4.5; 9.7] | |
| Unenhanced CT attenuation (HU) | 25-3 (12-53) | 25.6 (12-42) | 25.0 (13-53) | 33 (20-52) | 36-0 (26-42) | 30.0 (20-52) |
| [19-0; 32.5] | [19-8; 31.3] | [19.0; 34.0] | [28.3; 37.5] | [31.0; 37.0] | [28.0; 38.0] | |
| HU > 20, n (%) | 36 (68%) | 20 (67%) | 16 (70%) | 23 (96%) | 9 (100%) | 14 (93%) |
| Heterogeneous appearance, n (%) | 26 (49%) | 14 (47%) | 12 (52%) | 15 (63%) | 6 (67%) | 9 (60%) |
| Diameter > 4 cm + (HU > 20 or heterogeneous appearance) n (%) | 31 (59%) | 16 (53%) | 15 (65%) | 22 (92%) | 8 (89%) | 14 (93%) |
Data are presented as median (minimum-maximum) [IQR], as indicated. In four patients, tumour size was below 3 cm (2 phaeochromocytomas, 1 ACA and 1 desmoidfibromatosis) but they were operated due to increase in size by at least 1 cm). F, female; M, male. ªPatients with history of malignancy: ACA category: history of basalioma (n = 1), breast cancer (n = 2), oesophageal cancer (n = 1); non-AC benign tumour category: history for renal cell carcinoma (n = 1); ACC category: history of adenocarcinoma of the lung (n = 1); non-AC malignant tumour category: history of rectal cancer (n = 1), breast cancer (n = 1), hepatocellular carcinoma (n = 1), renal cell carcinoma (n = 1). b26 tumours were ≤4 cm in size, of these, four were malignant (2 ACC, 1 metastasis of rectal cancer, 1 GIST).
Table 1: Patient characteristics and imaging results from 77 patients with available histopathology.
99.93%) for classifying malignant versus benign lesions. However, specificity was only 62.3% (95% CI 47.9%- 75.2%) (Table 2, Fig. 3a). One low-grade gastrointestinal stroma tumour was misclassified as benign. The low specificity was due to 14 of 30 ACA being categorised as malignant (13 FDG-positive, 1 FDG indeterminate) (Table 2, Fig. 3a). Additionally, one phaeochromocy- toma, one haematoma, one neurofibroma, and two schwannomas were FDG-positive, and a benign myelo- lipoma was indeterminate (examples in Supplementary Figure S5).
Quantitative FDG PET analysis results are presented in Fig. 3c and e, Supplementary Table S3 and Supplementary Figure S7A. Analysis of the 13 FDG- positive and the FDG indeterminate ACA (categorised as malignant) revealed that five FDG-positive ACA were oncocytic tumours. Two FDG-positive ACA were initially diagnosed as ACC by the local pathologist. Comparison of FDG-positive/indeterminate with FDG- negative ACA regarding imaging, demographic and histopathological parameters is shown in Supplementary Table S4.
[123]]lodometomidate SPECT
In visual analysis of the 73 patients with IMTO SPECT, the sensitivity for classifying adrenocortical versus non-adrenocortical tumours was 86.5% (95% CI 71-2%-95-5%) and specificity was 86-1% (95% CI 70.5%-95.3%) (Table 2). Twenty-seven of 28 ACA and five out of nine ACC were correctly classified as adrenocortical (Fig. 3a, Table 2). False positives included two phaeochromocytomas, one haematoma and one desmoidfibroma among benign non- adrenocortical tumours, and one rectal cancer metas- tasis among malignant non-adrenocortical tumours (Supplementary Figure S5).
Quantitative imaging for IMTO was available for 58 patients due to equipment and imaging hard-/software differences (Fig. 3d and Supplementary Figure S7B).
Computed tomography
The maximum tumour diameter was >4 cm in 51 tu- mours (20 malignant, 31 benign) and ≤4 cm in 26 tu- mours (4 malignant, 22 benign). Four tumours were <3 cm but included due to a >1 cm growth on follow-up imaging. Malignant lesions were larger with higher HU values compared to benign lesions, although there was a clear overlap (Table 1, Supplementary Figure S6).
All malignant lesions had attenuation values ≥20 HU (range 20-54). Using a cut-off ≥ 20 HU for ma- lignancy, the sensitivity was 100% (95% CI 85-8%- 100%). However, 39 of 53 benign lesions also had attenuation values ≥20 HU (20-29 HU (n = 19); ≥30 HU (n = 20)), resulting in a specificity of only 26-4% (95% CI 15.3%-40.3%) (Table 2, Fig. 3f). Compared to FDG-PET, unenhanced CT appears to be of less diag- nostic value. ROC curve analysis reveals an AUC of 0.826 (95% CI 0-730-0-922) for visual FDG analysis, compared to 0-727 (95% CI 0-615-0-840) for unen- hanced CT, with a difference of 0-099 (95% CI -0-046 to 0.243, p = 0.1811). Quantitative FDG-PET analysis showed a ROC-AUC of 0.842 (95% CI 0-747-0.938) for SUVmax, the most common quantitative measure, and up to 0.869 (95% CI 0-788-0.949) for the target to background ratio (tumour SUVpeak/prevertebral region SUVmax), increasing the difference to 0-142 compared to unenhanced CT (95% CI (0-003-0-280), p = 0-0459) (Fig. 3f), Appendix p 6, Supplementary Figure S7.
Adverse events
Twentyfour AEs occured in a total of 19 patients, including four grade 3 events in three patients and 20
| Imaging test | Histopathology (Reference Standard) n (%) | ||||||
|---|---|---|---|---|---|---|---|
| FDG-PET classification | All | ACA | Benign non-AC | ACC | Malignant non-AC | ||
| IMTO-SPECT classification | |||||||
| Combined classification and tests | 77 (100) | 30 (39-0) | 23 (29.9) | 9 (11-7) | 15 (19.5) | ||
| FDG-PET malignancy | n (%) | 77 (100) | 30 (100) | 23 (100) | 9 (100) | 15 (100) | |
| + Positive (5 or 4) | malignant+, benign- | 40 (51.9) | 13 (43.3) | 5 (21-7) | 9 (100) | 13 (86-7) | |
| o Indeterminate (3) | indeterminate | 3 (3.9) | 1 (3-3) | 1 (4-3) | 0 (0-0) | 1 (6-7) | |
| - Negative (2 or 1) | malignant-, benign+ | 34 (44-2) | 16 (53.3) | 17 (73.9) | 0 (0-0) | 1 (6-7) | |
| IMTO-SPECT AC | n (%) | 77 (100) | 30 (100) | 23 (100) | 9 (100) | 15 (100) | |
| + Positive (5 or 4) | AC+ | 37 (48.1) | 27 (90-0) | 4 (17-4) | 5 (55-6) | 1 (6-7) | |
| o Indeterminate (3) | indeterminate | 3 (3.9) | 0 (0-0) | 3 (13.0) | 0 (0-0) | 0 (0-0) | |
| - Negative (2 or 1) | AC- | 33 (42.9) | 1 (3.3) | 15 (65.2) | 4 (44-4) | 13 (86-7) | |
| Missing | 4 (5.2) | 2 (6-7) | 1 (4-3) | 0 (0-0) | 1 (6-7) | ||
| FDG | IMTO | Combined Classification n (%) | 77 (100) | 30 (100) | 23 (100) | 9 (100) | 15 (100) |
| + | + | ACC+ | 19 (24-7) | 12 (40-0) | 2 (8-7) | 5 (55-6) | 0 (0-0) |
| + | ○ | malignant+, AC indeterminate | 0 (0-0) | 0 (0-0) | 0 (0-0) | 0 (0-0) | 0 (0-0) |
| + | – | malignant+, AC- | 19 (24-7) | 0 (0-0) | 3 (13.0) | 4 (44-4) | 12 (80-0) |
| + | . | malignant+ | 2 (2·6) | 1 (3.3) | 0 (0-0) | 0 (0-0) | 1 (6-7) |
| ○ | + | AC+, malignancy indeterminate | 2 (2·6) | 1 (3.3) | 0 (0-0) | 0 (0-0) | 1 (6-7) |
| ○ | ○ | both indeterminate | 0 (0-0) | 0 (0-0) | 0 (0-0) | 0 (0-0) | 0 (0-0) |
| ○ | – | AC-, malignancy indeterminate | 1 (1-3) | 0 (0-0) | 1 (4-3) | 0 (0-0) | 0 (0-0) |
| ○ | . | malignancy indeterminate | 0 (0-0) | 0 (0-0) | 0 (0-0) | 0 (0-0) | 0 (0-0) |
| − | + | ACA+ | 16 (20-8) | 14 (46-7) | 2 (8.7) | 0 (0-0) | 0 (0-0) |
| − | ○ | benign+, AC indeterminate | 3 (3.9) | 0 (0-0) | 3 (13.0) | 0 (0-0) | 0 (0-0) |
| − | − | benign+, AC- | 13 (16.9) | 1 (3.3) | 11 (47-8) | 0 (0-0) | 1 (6-7) |
| − | . | benign+ | 2 (2·6) | 1 (3.3) | 1 (4-3) | 0 (0-0) | 0 (0-0) |
| ACA-testª | n (%) | 75 (100) | 29 (38-7) | 22 (29-3) | 9 (12-0) | 15 (20-0) | |
| ACA+ (ACA-Test positive) | 16 (21-3) | 14 (48.3) | 2 (9-1) | 0 (0-0) | 0 (0-0) | ||
| PPV (95% CI) | 87.5% (61-7%-98.45%) | 16 (100) | 14 (87.5) | 2 (12.5) | 0 (0-0) | 0 (0-0) | |
| ACA- (ACA-Test negative) | 59 (78-7) | 15 (51-7) | 20 (90-9) | 9 (100) | 15 (100) | ||
| NPV (95% CI) | 74.6% (61.6%-85-0%) | 59 (100) | 15 (25.4) | 20 (33.9) | 9 (15.3) | 15 (25-4) | |
| Diagnostic test-accuracy | (95% CI) | ||||||
| Sensitivity | 48.3% (29-4%-67.5%) | ||||||
| Specificity | 95.7% (85-2%-99-469%) | ||||||
| LR (ACA+) | 11.1 (3.19-122) | ||||||
| LR (ACA-) | 0.541 (0-350-0-736) | ||||||
| ACC-testb | n (%) | 75 (100) | 29 (38-7) | 23 (30-7) | 9 (12-0) | 14 (18-7) | |
| ACC+ (ACC-Test positive) | 21 (28-0) | 13 (44-8) | 2 (8-7) | 5 (55-6) | 1 (7-1) | ||
| PPV (95% CI) | 23.8% (8.2%-47-2%) | 21 (100) | 13 (61-9) | 2 (9.5) | 5 (23.8) | 1 (4-8) | |
| ACC- (ACC-Test negative) | 54 (72.0) | 16 (55-2) | 21 (91-3) | 4 (44-4) | 13 (92.9) | ||
| NPV (95% CI) | 92.6% (82.1%-97.94%) | 54 (100) | 16 (29-6) | 21 (38.9) | 4 (7-4) | 13 (24.1) | |
| Diagnostic test-accuracy | (95% CI) | ||||||
| Sensitivity | 55.6% (21-2%-86-3%) | ||||||
| Specificity | 75.8% (63.6%-85.5%) | ||||||
| LR (ACC+) | 2.29 (0-755-4.50) | ||||||
| LR (ACC-) | 0.587 (0-197-1-06) | ||||||
| FDG-PET malignancy-test | n (%) | 77 (100) | 30 (39-0) | 23 (29.9) | 9 (11-7) | 15 (195) | |
| Malignancy+ (FDG Malignancy-Test positive) | 43 (55.8) | 14 (46-7) | 6 (26-1) | 9 (100) | 14 (93.3) | ||
| PPV (95% CI) | 53.5% (37.7%-68.8%) | 43 (100) | 14 (32-6) | 6 (14-0) | 9 (20·9) | 14 (32.6) | |
| Malignancy- (FDG Malignancy-Test negative) | 34 (44.2) | 16 (53.3) | 17 (73.9) | 0 (0-0) | 1 (6-7) | ||
| NPV (95% CI) | 97.1% (84.7%-99.926%) | 34 (100) | 16 (47-1) | 17 (50-0) | 0 (0-0) | 1 (2·9) | |
| Diagnostic test-accuracy | (95% CI) | ||||||
| Sensitivity | 95.8% (78.9%-99.895%) | ||||||
| Specificity | 62.3% (47.9%-75.2%) | ||||||
| LR (Malignancy+) | 2.54 (1.79-3.82) | ||||||
| LR (Malignancy-) | 0.0669 (0-00213-0-313) | ||||||
| (Table 2 | continues | on next page) | |||||
| Imaging test | Histopathology (Reference Standard) n (%) | |||||
|---|---|---|---|---|---|---|
| (Continued from previous page) | ||||||
| IMTO-SPECT AC-testd | n (%) | 73 (100) | 28 (38.4) | 22 (30-1) | 9 (12-3) | 14 (19-2) |
| AC+ (IMTO AC-Test positive) | 37 (50-7) | 27 (96-4) | 4 (18-2) | 5 (55·6) | 1 (7-1) | |
| PPV (95% CI) | 86.5% (71-2%-95.46%) | 37 (100) | 27 (73-0) | 4 (10-8) | 5 (13.5) | 1 (2-7) |
| AC- (IMTO AC-Test negative) | 36 (49-3) | 1 (3-6) | 18 (81.8) | 4 (44-4) | 13 (92.9) | |
| NPV (95% CI) | 86.1% (70.5%-95.3%) | 36 (100) | 1 (2-8) | 18 (50-0) | 4 (11-1) | 13 (36-1) |
| Diagnostic test-accuracy | (95% CI) | |||||
| Sensitivity | 86.5% (71.2%-95.46%) | |||||
| Specificity | 86.1% (70.5%-95.33%) | |||||
| LR (AC+) | 6-23 (3-01-17-0) | |||||
| LR (AC-) | 0.157 (0-0575-0-325) | |||||
| Unenhanced CT TA malignancy-teste | n (%) | 77 (100) | 30 (39-0) | 23 (299) | 9 (11-7) | 15 (19.5) |
| Malignancy+ (CT TA Malignancy-Test positive) | 63 (49-4) | 23 (76-7) | 16 (69.6) | 9 (100) | 15 (100) | |
| PPV (95% CI) | 38.1% (26-1%-51-2%) | 63 (100) | 23 (36.5) | 16 (25-4) | 9 (14.3) | 15 (23.8) |
| Malignancy- (CT TA Malignancy-Test negative) | 14 (50-6) | 7 (23.3) | 7 (30-4) | 0 (0-0) | 0 (0-0) | |
| NPV (95% CI) | 100% (76-8%-100%) | 14 (100) | 7 (50-0) | 7 (50-0) | 0 (0-0) | 0 (0-0) |
| Diagnostic test-accuracy | (95% CI) | |||||
| Sensitivity | 100% (85-8%-100%) | |||||
| Specificity | 26.4% (15.3%-40.3%) | |||||
| LR (Malignancy+) | 1.36 (1-15-1.68) | |||||
| LR (Malignancy-) | 0-0000 (0-0000-0-508) | |||||
n, number of participants; FDG, fluorodesoxyglucose; PET, positron emission tomography; IMTO, iodometomidate; SPECT, single photon emission computed tomography; CT, computed tomography; TA, tumour attenuation; AC, adrenocortical (tissue); ACC, adrenocortical carcinoma; ACA, adrenocortical adenoma; HU, Hounsfield units; CI, confidence interval; CP, Clopper-Pearson CI (1934); BSC, Blyth-Still- Casella CI (1983, 1986); PPV, positive predictive value; NPV, negative predictive value; LR(X), likelihood ratio of test result X. Based on the histopathology of 77 participants, the estimated pre-test probabilities with two-sided 95% Cls of the four histopathological groups of incidentalomas were for ACA 39-0% (CP 28-0%-50-8%, BSC 28-0%-50-8%), for benign non-AC incidentalomas 29.9% (CP 20-0%-41-4%, BSC 20.5%-40-8%), for ACC 11-7% (CP 5-5%-21-0%, BSC 6-0%-20.5%), and for malignant non-AC incidentalomas 19-5% (CP 11.3%-30-1%, BSC 11-8%-29-2%). For evaluation of functional imaging (FDG-PET and IMTO-SPECT), results of visual image analysis were used. The visual interpretation criteria included a semi-quantitative scoring system ranging from 1 to 5: Visual scores 1 and 2 indicated no or low tracer uptake, whereas scores 4 and 5 indicate moderate or high tracer uptake. Tumours with intermediate tracer uptake were rated 3 (indeterminate). In the case that FDG-PET alone already excluded imaging diagnosis of an ACA (n = 2 because of high FDG uptake) or an ACC (n = 2 because of low FDG uptake), results were included in the analysis of the ACA- or ACC-test, respectively, even if no IMTO-SPECT imaging was available. The n (%) for calculation of the presented predictive values (post-test probabilities) are presented in italic. The results of the cross-sectional analysis for the diagnostic test accuracy (primary and secondary outcome) are presented for the test positive (+) and the test negative (-) outcome by classification groups from histopathology as n (%) and as sensitivity and specificity with exact two-sided 95% CP by classification into the histopathological positive and negative group, respectively. Alternatively, for calculation of predictive values based on assumed pre-test probabilities of the histopathological positive group, the likelihood ratios LR(test+) and LR(test-) are provided, each with exact two-sided 95% CI. Each two-sided 95% CI is provided as the intersection of two one-sided 97.5% Cls. ªThe ACA-Test was defined focussing on high specificity: FDG o was also classified as Malignancy+, and IMTO o also as AC -. “The ACC-Test was defined focussing on high sensitivity: FDG o was also classified as Malignancy+, and IMTO o also as AC+. ‘The FDG-PET Malignancy-Test was defined focussing on high sensitivity for malignancy: FDG o was also classified as Malignancy+. ªThe IMTO-SPECT AC-Test was defined focussing on high specificity for AC origin: IMTO o was also classified as AC -. “The Unenhanced CT TA Malignancy-Test based on the TA measured in HU was defined focussing on very high sensitivity using a cut-off of 20 HU: >20 HU was classified as Malignancy+ and <20 HU as Malignancy -.
Table 2: Diagnostic test accuracy of visual FDG-PET or CT imaging as malignancy test, of visual IMTO-SPECT as adrenocortical test, and combined for diagnosing adrenal incidentalomas.
grade 1 or 2 events in 17 patients. Four AEs in four patients were judged as study-related: pain at the injec- tion site after IMTO injection (n = 3), metallic taste (n = 1), all of them were grade 1 and self-limiting. Seven serious AEs occurred in three patients, including one patient who died from heart failure and sepsis before scheduled adrenalectomy. All SAEs were unrelated to study treatment. Details on AEs are provided in the Appendix p 9.
Discussion
The FAMIAN study evaluated the diagnostic accuracy of combined FDG/IMTO imaging in classifying ACA. The aim was to develop an ACA-index test based on visual assessments of FDG and IMTO, aiming for high spec- ificity while maintaining adequate sensitivity. This functional imaging approach reached high specificity (95.7%). However, the sensitivity was only 48.3% due to
a high number of FDG-positive ACA, leading to their misclassification as ACC. Despite this, utilizing FDG-/ IMTO-imaging to guide treatment decisions could have avoided unnecessary surgery in nearly half of the patients with indeterminate lesions that turned out to be ACA on histopathology. This proportion could be increased by using quantitative measurements.
The study group had a lower proportion of ACA (39%) than expected, reducing the pre-test probability boundary from the assumed 50%. However, ACA remained the largest subgroup.2,4,9,29 Thus, while excluding malignancy is a primary goal, a method that additionally allows non-invasive entity assignment as ACA can be beneficial. In the FDG-negative tumours, additional IMTO imaging classified ACA with 93% sensitivity and 88% specificity. Importantly, none of the tumours classified as ACA by combined FDG/IMTO imaging were malignant.
CT
18F-FDG-PET/CT 123|-IMTO SPECT/CT
ACA
Non-AC benign
ACC
Non-AC malignant
The secondary objective of our study was non- invasive classification of ACC. All nine ACC were correctly classified as malignant by FDG PET, but four ACC were IMTO-negative, likely due to tumour dedif- ferentiation.28 Accordingly, combined FDG/IMTO im- aging showed limited sensitivity and, due to the large number of ACA falsely classified as ACC, also limited specificity for detection of ACC, making this approach less valuable for preoperative ACC classification.
In terms of differentiating malignant from benign tumours, recent studies suggest raising the cut-off HU value for malignancy suspicion from 10 HU to 20 HU, keeping high sensitivity.3,6,9 This is affirmed by our study, with all malignant tumours showing HUS ≥20. However, 23 out of 30 ACA (76-7%) had HU ≥20, resulting in a specificity of only 26-4% in our group. It is, however, important to note that tumours with HU <10 were excluded.
In our study, both FDG PET imaging and unen- hanced CT imaging were found to have high to very high sensitivity for detection of malignancy, but speci- ficity was low. FDG-PET provided higher specificity compared to unenhanced CT, at a similar sensitivity level. However, improving specificity while maintaining high sensitivity remains a challenge.
Current guidelines suggest stratifying adrenal tu- mours based on size and the degree of heterogeneity or HU in unenhanced CT imaging. Four categories with different malignancy suspicion levels were defined with
corresponding recommendations for further manage- ment.1 In our study, 53 tumours (69%) fall into the category with the highest suspected malignancy and the recommendation for prompt surgical removal. Howev- er, these 53 tumours still included 31 benign masses, including 16 ACA. Of these 31 benign tumours, FDG PET would have correctly classified 18 lesions (58%) as benign. Thus, functional imaging was useful in differ- entiating between benign and malignant lesions in this specific group, without, however, capturing all benign lesions as such.
Remarkably, about half of ACA in our study showed high FDG uptake, raising questions about their malig- nant potential. Further analysis revealed that five FDG- positive ACA were oncocytic tumours, known for increased FDG uptake.30 Weiss scores were higher in FDG-positive adenomas, and steroid analysis revealed increased androgen production in three FDG-positive ACA, whereas age, sex, tumour size, Ki67 index or unenhanced HU were comparable between FDG- positive and negative ACA. These findings warrant further investigation into the nature of FDG-positive ACA and their need for surgical removal. Clinical follow-up and molecular analyses, such as metab- olomics, transcriptomics, and genetic characterisation, comparing FDG-positive tumours with FDG-negative adenomas and ACCs would be of interest.
Our study was limited by the inability to reach the intended number of cases. This clearly reduces
a
FDG visual analysis
c
FDG quantitative analysis (SUVpeak)
e
FDG ROC analyses
30-
1
100-
100-
9
2
3
4
5
4
75.
20
4
10-
1
Sensitivity%
1
1
5
n
5
50
12
12
10-
11
1.
8
25
FDG SUVmax (n 77), AUC 0-842 (0-747-0-938)
FDG SUVpeak (n 77), AUC 0-852 (0-761-0-942)
- FDG TBR TUpeak/prevertebralmax (n 77), AUC 0-869 (0-788-0-949)
1
-. FDG visual analysis (n 77), AUC 0-826 (0-730-0-922)
1
0.1
- CT attenuation (Hounsfield Units) (n 77), AUC 0.727 (0-615-0.840)
0
1
1
non-AC
ACA
non-AC benign
0
ACA
non-AC benign
ACC
ACC
25
50
75
100
malignant
non-AC malignant
100% - Specificity%
b
IMTO visual analysis
d
IMTO quantitative analysis (SUVpeak)
f
IMTO ROC analyses
30-
1
100-
100-
26
2
3
4
5
V
75
1
20
3
10
Sensitivity%
3
n
50
1
14
1
2
10
11
1.
4
25
IMTO SUVmax (n 58), AUC 0-926 (0-860-0-992)
- IMTO SUVpeak (n 58), AUC 0-931 (0-867-0-995)
1
- IMTO TBR TUpeak/prevertebralmax (n 58), AUC 0-919 (0-842-0-995)
2
- IMTO visual analysis (n 58), AUC 0-896 (0-807-0-986)
1
2
0
··· IMTO visual analysis (n 73), AUC 0-917 (0-847-0-988)
1
0.1
non-AC benign
non-AC malignant
0
ACA
non-AC benign
ACC
non-AC
ACA
ACC
25
50
75
100
malignant
100% - Specificity%
statistical power. Nonetheless, to our knowledge the study remains the largest prolective study using inno- vative molecular imaging approaches in indeterminate adrenal lesions. A major strength is that histopathology served as reference standard in all cases. Our results align well with previous analyses of the individual tracers.13-17,19,23,24,27 Patients with a history of malignancy were not excluded but represented a minority. Of note, 50% of the patients with history of malignancy were diagnosed with a benign adrenal lesion in our study and
in cases of malignant adrenal tumours identified, these were not always related to previous extra-adrenal ma- lignancies. Most patients were from the lead centre, Würzburg, a reference centre for adrenal tumours, potentially introducing selection bias. However, the prevalence of adrenocortical lesions was comparable to other prospective studies.14,15,17
In summary, combined FDG and IMTO imaging non-invasively characterises ACA, the most common tumour entity, with high specificity. Sensitivity is
limited due to a notably high number of FDG-positive ACA, an intriguing subgroup deserving further inves- tigation. Current preoperative diagnostic methods, FDG-PET and especially CT, have limitations in dis- tinguishing benign from malignant lesions due to reduced specificity. A significant proportion of tumours can be managed by applying unenhanced HU cut-off values of 10 or 20 HU. Functional imaging with FDG and IMTO helps to further reduce unnecessary sur- geries by more effectively classifying adrenal lesions compared to conventional imaging tools. It may be assumed that other CYP11B1-targeting radiotracers would demonstrate comparable applicability to IMTO for the question addressed in this study.20-23
Contributors
SH, AKB and AS prepared the study protocol and all other study relevant documents. HHM was in charge of the biometric concept and the sta- tistical analysis planning for the study protocol. Statistical analyses were performed by HHM and SH. FB, MM, KM, TP, CTF, CF, AT, KH, HA, MR, MS, and OS were responsible for implementation of the study at the respective centres, for patient recruitment and execution of the study procedures. MF, WS, PH, JW and MR contributed to central data evaluation and visualisation. PH, AKB, MR and SH organised central image evaluation. SK performed additional histopathological evaluation beneath reference pathological examination. MR and SH accessed and verified the underlying data. SH wrote the manuscript together with PH, AS, CTF, WS and WA and all authors edited the report. All authors read and approved the final version of the manuscript.
Data sharing statement
We will consider sharing deidentified, individual participant-level data that underlie the results reported in this article on receipt of a request detailing the study hypothesis and statistical analysis plan. All requests should be sent to the corresponding author. The corresponding author and lead investigators of this study will discuss all requests and make decisions about whether data sharing is appropriate based on the scientific rigour of the proposal. All applicants will be asked to sign a data access agreement.
Declaration of interests
SH and AS are inventors on patents for radiotracers targeting CYP11B enzymes for diagnostic evaluation of primary aldosteronism or for theranostics of adrenocortical tumours/carcinoma EP2575899A1, WO/ 2014/048568 and WO/2018/141541. MM reports grants from AstraZeneca, consulting fees from Novartis, Veraxa and Roche, payments from Telix and GWT-TUD; OS reports funding by the DFG (223465017), grants from Siemens, Life molecular imaging, ABX, honorarium paid to institution from Siemens, advisory board Positrigo (no payments); MF reports grants from Enterome, DSMB Bayer Pharma (payments to institution) and ExCo membership European Network for the Study of Adrenal Tumours and European Society of Endocrinology (unpaid positions), CF reports honoraria and travel support from IPSEN Pharma GmbH, KH reports grants from Novartis, Sofie Biosciences, consulting fees from advanced accelerator applications, Amgen, AstraZeneca, Brain Capital, Bayer, Boston Scientific, Convergent, Curium, Debiopharm, EcoR1, Fusion, GE Healthcare, Imme- dica, Isotopen Technologien München, Janssen, Merck, Molecular Partners, NVision, POINT Biopharma, Pfizer, Radiopharm Theranostics, Rhine Pharma, Siemens Healthineers, Sofie Biosciences, Telix, and Theragnostics, ymabs, Honorarium from PeerVoice, travel support from Janssen, Data Safety Monitoring Board/advisory board participation for Fusion, GE Healthcare, relationships with Sofie Biosciences, Pharma 15, NVision, Convergent, Aktis Oncology, AdvanCell, JW reports travel support by IPSEN Pharma GmbH, WA reports incoming presidency of European Society of Endocrinology. All other authors declare no competing interests related to the manuscript.
Acknowledgements
This work was supported by the Deutsche Forschungsgemeinschaft (DFG AL 203/1-1 to SH, AB and AS and CRC/Transregio 205 “The Adrenal: Central Relay in Health and Disease” to SH and AS) and the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement nº 259735 ENS@T-Cancer to FB, SH and Bruno Allolio, by the Clinical Research Priority Program of the University of Zurich for the CRPP HYRENE to FB, Clinician-Scientist program grant (RISE) from Else Kroner Fresenius Stiftung and Eva-Luise and Horst Köhler Stiftung to JW, and MRC Programme Grant MC_UP_1605/15 to WA.
The manuscript is dedicated to Professor Bruno Allolio, who was the main initiator of this study but sadly passed away in August 2015.
Appendix A. Supplementary data
Supplementary data related to this article can be found at https://doi. org/10.1016/j.ebiom.2025.105735.
References
1 Fassnacht M, Tsagarakis S, Terzolo M, et al. Management of ad- renal incidentalomas- a European society of endocrinology clinical practice guideline in collaboration with the European network for the study of adrenal tumors. Eur J Endocrinol. 2023;189:G1-G42.
2 Mantero F, Terzolo M, Arnaldi G, et al. A survey on adrenal inci- dentaloma in Italy. Study group on adrenal tumors of the Italian society of endocrinology. J Clin Endocrinol Metab. 2000;85(2): 637-644.
3 Ebbehoj A, Li D, Kaur RJ, et al. Epidemiology of adrenal tumours in Olmsted County, Minnesota, USA: a population-based cohort study. Lancet Diabetes Endocrinol. 2020;8(11):894-902.
4 Jing Y, Hu J, Luo R, et al. Prevalence and characteristics of adrenal tumors in an unselected screening population : a cross-sectional study. Ann Intern Med. 2022;175(10):1383-1391.
5 O’Neill CJ, Spence A, Logan B, et al. Adrenal incidentalomas: risk of adrenocortical carcinoma and clinical outcomes. J Surg Oncol. 2010;102(5):450-453.
6 Bancos I, Taylor AE, Chortis V, et al. Urine steroid metabolomics for the differential diagnosis of adrenal incidentalomas in the EURINE-ACT study: a prospective test validation study. Lancet Diabetes Endocrinol. 2020;8(9):773-781.
7 Dinnes J, Bancos I, Ferrante di Ruffano L, et al. Management of endocrine disease: imaging for the diagnosis of malignancy in incidentally discovered adrenal masses: a systematic review and meta-analysis. Eur J Endocrinol. 2016;175(2):R51-R64.
8 Hong AR, Kim JH, Park KS, et al. Optimal follow-up strategies for adrenal incidentalomas: reappraisal of the 2016 ESE-ENSAT guidelines in real clinical practice. Eur J Endocrinol. 2017;177(6): 475-483.
9 Iniguez-Ariza NM, Kohlenberg JD, Delivanis DA, et al. Clinical, biochemical, and radiological characteristics of a single-center retrospective cohort of 705 large adrenal tumors. Mayo Clin Proc Innov Qual Outcomes. 2018;2(1):30-39.
10 Kahramangil B, Kose E, Remer EM, et al. A modern assessment of cancer risk in adrenal incidentalomas: analysis of 2219 patients. Ann Surg. 2022;275(1):e238-e244.
11 Schloetelburg W, Ebert I, Petritsch B, et al. Adrenal wash-out CT: moderate diagnostic value in distinguishing benign from malig- nant adrenal masses. Eur J Endocrinol. 2021;186(2):183-193.
12 . Corwin MT, Badawy M, Caoili EM, et al. Incidental adrenal nodules in patients without known malignancy: prevalence of malignancy and utility of washout CT for characterization-A multiinstitutional study. AJR Am J Roentgenol. 2022;219(5):804-812.
13 Tessonnier L, Sebag F, Palazzo FF, et al. Does 18F-FDG PET/CT add diagnostic accuracy in incidentally identified non-secreting adrenal tumours? Eur J Nucl Med Mol Imaging. 2008;35(11):2018- 2025.
14 Groussin L, Bonardel G, Silvera S, et al. 18F-Fluorodeoxyglucose positron emission tomography for the diagnosis of adrenocortical tumors: a prospective study in 77 operated patients. J Clin Endo- crinol Metab. 2009;94(5):1713-1722.
15 Ansquer C, Scigliano S, Mirallie E, et al. 18F-FDG PET/CT in the characterization and surgical decision concerning adrenal masses: a prospective multicentre evaluation. Eur J Nucl Med Mol Imaging. 2010;37(9):1669-1678.
16 Nakajo M, Jinguji M, Fukukura Y, et al. FDG-PET/CT and FLT- PET/CT for differentiating between lipid-poor benign and malig- nant adrenal tumours. Eur Radiol. 2015;25(12):3696-3705.
17 Guerin C, Pattou F, Brunaud L, et al. Performance of 18F-FDG PET/CT in the characterization of adrenal masses in noncancer patients: a prospective study. J Clin Endocrinol Metab. 2017;102(7):2465-2472.
18 Salgues B, Guerin C, Amodru V, et al. Risk stratification of adrenal masses by [(18) FJFDG PET/CT: changing tactics. Clin Endocrinol (Oxf). 2021;94(2):133-140.
19 He X, Caoili EM, Avram AM, Miller BS, Else T. 18F-FDG-PET/CT evaluation of indeterminate adrenal masses in noncancer patients. J Clin Endocrinol Metab. 2021;106(5):1448-1459.
20 Silins I, Sundin A, Lubberink M, et al. First-in-human evaluation of [(18)F]CETO: a novel tracer for adrenocortical tumours. Eur J Nucl Med Mol Imaging. 2023;50(2):398-409.
21 Heinze B, Schirbel A, Nannen L, et al. Novel CYP11B-ligand [(123/ 131)IJIMAZA as promising theranostic tool for adrenocortical tumors: comprehensive preclinical characterization and first clin- ical experience. Eur J Nucl Med Mol Imaging. 2021;49(1):301-310.
22 Wadsak W, Mitterhauser M, Rendl G, et al. [18F]FETO for adre- nocortical PET imaging: a pilot study in healthy volunteers. Eur J Nucl Med Mol Imaging. 2006;33(6):669-672.
23 Hennings J, Lindhe O, Bergstrom M, Langstrom B, Sundin A, Hellman P. [11C]metomidate positron emission tomography of
adrenocortical tumors in correlation with histopathological find- ings. J Clin Endocrinol Metab. 2006;91(4):1410-1414.
24 Bergstrom M, Juhlin C, Bonasera TA, et al. PET imaging of adrenal cortical tumors with the 11beta-hydroxylase tracer 11C-metomidate. J Nucl Med. 2000;41(2):275-282.
25 Minn H, Salonen A, Friberg J, et al. Imaging of adrenal inci- dentalomas with PET using (11)C-metomidate and (18)F-FDG. J Nucl Med. 2004;45 (6):972-979.
26 Wu X, Senanayake R, Goodchild E, et al. [(11)C]metomidate PET- CT versus adrenal vein sampling for diagnosing surgically curable primary aldosteronism: a prospective, within-patient trial. Nat Med. 2023;29(1):190-202.
27 Hahner S, Kreissl MC, Fassnacht M, et al. Functional character- ization of adrenal lesions using [123I]IMTO-SPECT/CT. J Clin Endocrinol Metab. 2013;98(4):1508-1518.
28 Kreissl MC, Schirbel A, Fassnacht M, et al. [(1)(2)(3)I]Iodometo- midate imaging in adrenocortical carcinoma. J Clin Endocrinol Metab. 2013;98(7):2755-2764.
29 Ichijo T, Ueshiba H, Nawata H, Yanase T. A nationwide survey of adrenal incidentalomas in Japan: the first report of clinical and epidemiological features. Endocr J. 2020;67(2):141-152.
30 Kim DJ, Chung JJ, Ryu YH, Hong SW, Yu JS, Kim JH. Adreno- cortical oncocytoma displaying intense activity on 18F-FDG-PET: a case report and a literature review. Ann Nucl Med. 2008;22(9): 821-824.