Abstract
Background/Aim: Positron emission tomography/computed tomography (PET/CT) plays an important role in cancer localization in ectopic Cushing’s syndrome (ECS). However, the choice of the optimal tracer for investigation of this disease is still unclear. We aimed to evaluate the diagnostic feasibility of [18F]fluoro-2-deoxyglucose ([18F]FDG), [18F]fluoro-L-dihydroxyphenylalanine ([18F] FDOPA), and [68Ga]-DOTA-1-Nal3-octreotide ([68Ga]-DOTANOC) in ECS. Patients and Methods: All PET/CT scans of patients admitted to our department for suspected ECS between 2010 and 2020 were retrospectively analysed. Results: Collectively, 30 PET/CT examinations, 11 with [18F]FDOPA, 11 with [18F]FDG and 8 with [68Ga]GaDOTANOC were conducted for 18 patients eligible for analysis. [18F]FDG detected the tumour in 3/6 of the cases, [18F]FDOPA in 3/4, and [68Ga]GaDOTANOC in 3/3. [18F]FDOPA was the only tracer without false positive results. Conclusion: [68Ga]GaDOTANOC and [18F]FDOPA showed superior results compared to [18F]FDG, although the sensitivity of the tracers might be influenced by the aetiology of the tumour underlying the ECS.
Ectopic Cushing’s syndrome (ECS) is a rare form of Cushing’s syndrome (CS), the clinical manifestation of long-term elevated plasma cortisol levels. Endogenous CS can be either adrenocorticotropin-independent (adrenal) or adrenocorticotropin-dependent. ECS accounts for about 10% of adrenocorticotropin (ACTH)-dependent CS that are not caused by pituitary ACTH-secreting tumours (central CS, also called Cushing’s disease) (1-7). ECS can be caused by any non-pituitary ACTH-secreting tissue, most commonly small-cell lung cancer (SCLC) and pulmonary neuroendocrine tumours (NETs), and less commonly other NETs (e.g., pancreas or gut), thymic carcinoids, thyroid carcinoma, pheochromocytoma, and others (8-11). Tumour resection is the ideal curative treatment of ECS, but the variety of possible locations, the occurrence of incidentalomas, and small tumour sizes pose a great challenge in pinpointing the source of ECS. Currently, computed tomography (CT) scan is the first-line imaging modality for identification of the origin of ECS. Its sensitivity is estimated around 55-65%, with the detection of pulmonary NETs being especially difficult (11, 12). The employment of functional imaging as second-line imaging modality in cases of equivocal lesions or no findings in CT has been recommended in several studies (11, 13, 14). So far, the most investigated secondary imaging modalities are Octreoscan or positron emission tomography/CT (PET/CT) with gallium-68 labelled somatostatin analogues and [18F]fluoro-2-deoxyglucose ([18F]FDG) PET/CT scan. For the detection of tumours in ECS, radiolabelled somatostatin-analogues for PET/CT are generally seen as superior to Octreoscan (14, 15) and its proposed sensitivity ranges from 57-82% (11, 13, 16). [18F]FDG PET/CT examination is known to detect tumours with higher proliferation rate (17, 18) and has an estimated overall sensitivity of 52% (13). A third tracer that is sometimes used in patients with ECS is [18F]fluoro-L-dihydroxyphenylalanine ([18F]FDOPA) with a suggested sensitivity of 57% (13). Due to the rarity of the disease and rapid development of PET tracers, the sensitivities mentioned above are mainly based on small-sized studies with non-uniform definitions and diagnostic approaches and therefore, provide only ambiguous insight into the diagnostic capability of the utilized tracers (11, 14). Moreover, tracer evaluation is further complicated by the limited number of studies comparing multiple tracers in the same patient cohort.
Thus, the aim of this study was to assess the feasibility of PET tracers for tumour detection in patients with ECS. We therefore analysed [18F]FDG, [18F]FDOPA and [68Ga] GaDOTANOC ([68Ga]-1-Nal3-octreotide tetra-azacyclododecatatro-acetic-acid) PET/CT examinations and compared the results with the available histological diagnosis to calculate the sensitivity and specificity of each tracer.
Patients and Methods
Patients. This retrospective study of consecutive patients was approved by the Ethics Committee of the Medical University of Vienna (EK: 1853/2018) and included all patients (n=24) who were referred to PET/CT scan at Vienna general hospital in the period of 1st March 2010 to 31st May 2020 due to clinically suspected or laboratory confirmed ECS or CS. The patients had received PET/CT examinations with either [18F]FDG and/or [18F]FDOPA and/or [68Ga]GaDOTANOC on different days. Written informed consent has been obtained from all patients prior to every PET/CT examination. Only the first PET/CT scan from each tracer was included for the calculation of diagnostic metrics in every patient.
Protocol for PET/CT examination. All PET/CT examinations were performed with a 64-row multidetector hybrid system (Biograph TruePoint 64; Siemens, Erlangen, Germany) with an axial field-of-view of 216 mm, a PET sensitivity of 7.6 cps/kBq, and a transaxial PET resolution of 4-5 mm (full-width at halfmaximum, FWHM). PET scan was performed with 4 min/bed position, a 5 mm slice thickness, and a 168×168 matrix. The reconstruction was based on the TrueX algorithm (4 iterations, 21 subsets). CT scan used a collimation of 64×0.6 mm, a 3 mm slice thickness at a 2 mm increment, and a 512×512 matrix. All patients fasted for at least 4 h before the PET examinations and Carbidopa was administered prior to [18F]FDOPA scans. Patients were scanned from the skull to the upper thighs 45 min after intravenous administration of 5 MBq/kg body weight [18F]FDG or 3 MBq/kg body weight [18F]FDOPA as well as [68Ga]GaDOTANOC. Patients with no recent CT examination received an intravenous injection of 100 ml of a tri-iodinated, non-ionic contrast medium at a rate of 2 ml/s with a tube voltage of 120 mA, and a tube current of 230 kV for venous-phase contact-enhanced CT. Lesions were considered tracer-positive if tracer uptake was visually higher than the surrounding physiological uptake or background activity. The evaluation of the performed PET/CT examinations and the gathered data was conducted by two specialists in nuclear medicine and one radiologist (at least 5 years of experience).
Based on biochemical results, four patients with central CS and two patients with adrenal CS were excluded from further analysis, leaving collectively 18 patients eligible for the study. In these patients, ECS was classified as “confirmed” if the presence of a tumour was histologically proven (n=7) or highly suspected in PET/CT scan without histology available (n=1). In patients without any detectable tumour in any imaging modality, ECS was classified as “occult” (n=10). Furthermore, clinical parameters such as symptoms associated with CS, ACTH and serum cortisol levels, high-dose dexamethasone suppression test (HDDST), corticotropin releasing hormone (CRH) test, inferior petrosal sinus sampling (IPSS) as well as pituitary MRI, abdomen and chest MRI/CT prior to the first PET/CT examination and pathology reports of the biopsied/resected lesions were collected.
Statistical analysis. All data analyses have been performed using the software SPSS, version 26.0 (IBM Corp., Armonk, NY, USA). Descriptive statistics are presented as mean value±standard deviation (SD) or, if according to Kolmogorov-Smirnov-Test not-normally distributed, as median and range. Owing to the small sample size of the studied patients, exact Mann-Whitney-U-Test was performed to compare the mean values of the parameters between patients with confirmed and occult ECS. Categorical parameters were evaluated by Pearson’s Chi Square. A p-value <0.05 was considered as statistically significant. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of all PET tracers regarding tumour detection were calculated using a 2×2 contingency table with the above-mentioned criteria for confirmed ECS as the gold standard.
Results
Study population and biochemical results. A total of 18 patients (8 females, 10 males) with a mean age of 51±18 years were clinically diagnosed with ECS. In eight patients (44%), a histological diagnosis of malignant disease was made (confirmed ECS): 6 with pulmonary NETs, one with SCLC, and one with papillary thyroid carcinoma. In the other ten patients (56%), ECS remained occult. Table I provides an overview of the clinical and biochemical characteristics of all patients and those with confirmed or occult ECS, respectively. The prevalence of symptoms associated with CS was about the same in both groups with cushingoid physique as the leading characteristic (circa 60%). Serum cortisol levels were also close to equal [confirmed: median 38.7 (53.3) μg/dl; occult: median 39.5 (105.2) μg/dl], whereas patients with confirmed ECS tended to have higher, but not significantly different, ACTH levels than patients with occult ECS [median 157.5 (159.8) pg/dl vs. 91.10 (70.2) pg/dl; p=0.101].
Patient characteristics according to ectopic Cushing’s syndrome classification.
ECS was indicated by HDDST in 9/11, by CRH test in 10/13, and by IPSS in 4/4 patients. All patients with confirmed ECS presented with concordant ECS-positive HDDST and CRH test with the exception of one HDDST with partial suppression of cortisol. In contrast, patients with occult ECS presented with contradictory test results in four cases. Overall, there was no patient with ECS-negative HDDST and CRH test.
PET/CT results. In total, 30 diagnostic PET/CT examinations were performed for these 18 patients: 11 [18F]FDG, 11 [18F]FDOPA and 8 [68Ga]GaDOTANOC (Table II). Collectively, 6/11 [18F]FDG, 4/11[18F]FDOPA and 3/8 [68Ga]GaDOTANOC scans were conducted in patients with confirmed ECS (n=8). [18F]FDG detected the tumour in 3/6 of these cases (a pulmonary NET G2, a pulmonary NEC SCLC and a papillary thyroid carcinoma), [18F]FDOPA in 3/4 and [68Ga]GaDOTANOC in 3/3 (all pulmonary NETs). Table II displays the results of conventional imaging, PET scans, and pathohistological assessment of each patient. Table III lists the diagnostic indices of the three studied PET tracers.
Results of conventional imaging, functional imaging and histology in each patient.
Diagnostic indices of [18F]FDG, [18F]FDOPA, and [68Ga]GaDOTANOC.
Overall, [18F]FDOPA uptake was accurate in 10/11 patients (90%), [68Ga]GaDOTANOC in 7/8 (88%) and [18F]FDG in 6/11 (50%). [18F]FDOPA failed to detect a pulmonary NET G1 in a patient where recurrence was already suspected and detected due to visibility in CT. Its sensitivity was therefore 3/4 (75%), compared to 3/3 (100%) of [68Ga]GaDOTANOC and 3/6 (50%) of [18F]FDG. However, [18F]FDOPA correctly remained negative in the one case of false positive [68Ga]GaDOTANOC (chronic bronchitis) and one of the two cases of false positive [18F]FDG (infections). The respective specificities of the tracers are therefore 7/7 (100%) for [18F]FDOPA, followed by 4/5 (80%) for [68Ga] GaDOTANOC and 3/5 (60%) for [18F]FDG. Accordingly, the positive predictive value (PPV) was 3/3 (100%) for [18F]DOPA, 3/4 (75%) for [68Ga]GaDOTANOC and 3/5 (60%) for [18F]FDG. The negative predictive value (NPV) was 7/8 (88%) for [18F]GDOPA, 4/4 (100%) for [68Ga]GaDOTANOC and 3/6 (50%) for [18F]FDG. [18F]FDG was unable to detect two pulmonary NET G1 (one of them also with [18F]FDG negative liver metastases) and one NET G2 that were positive for [18F]FDOPA and [68Ga]DOTANOC (Figures 1 and 2).
Representative case of a patient with pulmonary neuroendocrine tumour (NET) G1 accumulating [68Ga]GaDOTANOC and [18F]FDOPA, but not [18F]FDG. (A) No evidence of pathological [18F]FDG uptake. (B) [68Ga] GaDOTANOC revealed the tumour (lower arrow) and a hilar lymph node (upper arrow). (C) [18F]FDOPA positron emission tomography (PET) performed after tumour resection shows tracer accumulation in the known hilar lymph node (arrow).
Computed tomography (CT) and positron emission tomography (PET)/CT images of a patient with pulmonary neuroendocrine tumour (NET) G2 (arrows) with highly suspicious [68Ga]GaDOTANOC PET/CT (left) and previous [18F]FDG PET/CT (right) that was evaluated as inconspicuous. (A) and (C) show the visible lesion in CT scan. It had a high uptake of [68Ga]GaDOTANOC (B) but not of [18F]FDG (D).
Two patients with confirmed ECS, but persistent ACTH elevation even after resection of the histologically confirmed ACTH-positive tumour (both pulmonary NETs G1), and five patients with occult ECS underwent follow-up PET examinations every 1.5±0.8 years for up to 11 years (mean 4.3±3.9), which did not yield new results.
Conventional vs. functional imaging. In all patients with confirmed ECS, tumours were visible in conventional imaging (Table II). However, PET imaging enabled more accurate diagnosis in two cases: One was a papillary thyroid carcinoma that was interpreted as multinodular struma in an initial CT scan, but high uptake of [18F]FDG indicated a neoplastic process. Secondly, in one case of NET G1, conventional imaging detected the tumour but not the liver metastases that were visible in the [18F]FDOPA as well as [68Ga]GaDOTANOC scan (Figure 3).
Computed tomography (CT) and [68Ga]GaDOTANOC positron emission tomography/CT image of a patient with liver metastases of a pulmonary neuroendocrine tumour G1. Metastases do not stand out in CT scan (A), but are clearly visible in [68Ga]GaDOTANOC scan (B, arrows).
Discussion
The curative treatment for patients with ECS is the surgical resection of the tumour. However, small lesion sizes and the diversity of possible tumour locations tend to make detection of the source of ECS difficult. The present analysis focused on the roles and the diagnostic capabilities of 11 [18F]FDG, 11 [18F]FDOPA and 8 [68Ga]GaDOTANOC PET/CT examinations in consecutive patients sent to PET/CT examination with the indication of ECS. A tumour responsible for the ECS was detected in 8 out of 18 patients. [18F]FDOPA had the highest PPV (100%) and [68Ga]GaDOTANOC the highest NPV (100%), whereas [18F]FDG only showed diagnostic ratios ranging from 50-60%. To the best of our knowledge, this is the first study of consecutive patients assessing all three tracers in one clinical centre.
The proportion of pulmonary NETs in this study was 75%, notably higher than the previously reported 30-50% in general ECS studies (8-10). The only non-pulmonary tumour was a papillary thyroid carcinoma, which is only disclosed in relation to CS in a few case reports (19, 20) but not specifically mentioned in larger ECS studies (13, 21, 22). The disparity in the prevalence of identified tumour entities responsible for ECS between our and previous studies might be due to the fact that our study only included data of ECS patients referred to our clinic for a PET/CT examination. Isidori et al. (13) showed a high sensitivity of CT for the most common tumour localizations other than the lung, and thus these patients might have never received PET/CT examination.
The previously reported high sensitivity of about 95% of [68Ga]GaDOTANOC in pulmonary NETs (23, 24) was also observable in our patient cohort. [18F]FDOPA was similar to [68Ga]GaDOTANOC, only missing one pulmonary NET G1 in a patient who never received a [68Ga]GaDOTANOC PET scan. Thus, contrary to studies that reported inferiority of [18F]FDOPA compared to [68Ga]GaDOTANOC (12, 25, 26), our study is in line with findings that reveal the value of [18F]FDOPA in tumours commonly associated with ECS (27-29). Similar to our current results, a case of correctly negative [18F]FDOPA scan in a patient with false positive Octreoscan imaging in Pneumocystis jirovecii infection (11), and two previously described cases of [18F]FDOPA positive but [68Ga]GaDOTANOC negative ACTH-secreting tumours (13, 30) advert the employment of [18F]FDOPA alternatively to [68Ga]GaDOTANOC. Nevertheless, studies on [18F]FDOPA in ECS are still scarce (11, 14) and more research is required to clearly depict the role of [18F]FDOPA in this field.
Concerning the use of [18F]FDG in patients with ECS, previous studies reported contradictory, overall promising (31-33) or unfavourable (12, 34) results. The performance of [18F]FDG strongly depends on the patient cohort. In accordance with its biomolecular background, [18F]FDG is more sensitive in tumours with a higher proliferation index (>15%) (17). This characteristic is reflected by our own results of lower [18F]FDG sensitivity in pulmonary NET G1 than G2 and studies reporting superior sensitivity of [18F]FDG compared to [68Ga]GaDOTANOC in pulmonary NET G2 (23, 35). Since, according to Isidori et al. (13), the source of ectopic ACTH-secretion is more likely to be a pulmonary NET G1 than G2, we conclude that for tumour detection in patients with ECS, [68Ga]GaDOTANOC should primarily be chosen over [18F]FDG.
The present study has some limitations. First, the retrospective design with concomitant lack of data on clinical parameters or histological result in some patients. Second, even though when compared to similar studies the sample size is above average, it is still too small for reliable calculation of diagnostic indices. Further, the cases of confirmed ECS included almost only referred patients with lung tumours, and the study results are therefore not representative of the entire population of patients with ECS. Lastly, not all tracers were used in every patient, resulting in different patient samples for the calculation of diagnostic indices. More and larger prospective studies with equal tracer employment in all the patients are needed for a more reliable comparison of [18F]FDG, [18F]FDOPA, and [68Ga]GaDOTANOC in patients with ECS.
Based on the results of our research, we conclude that [68Ga]GaDOTANOC is a highly recommendable tracer for tumour detection in patients with ECS. [18F]FDOPA has shown promising results, especially regarding specificity, but it still needs further assessment. [18F] FDG has an overall lower sensitivity and specificity but it can be helpful in [68Ga]GaDOTANOC negative cases and should be considered as a complementary diagnostic tool, since the sensitivity of the tracers is highly affected by the pathogenesis of the tumour.
Footnotes
Authors’ Contributions
LZ wrote the manuscript, collected, and researched the data, contributed to discussion. OCK, BI and BF researched the data, reviewed the manuscript. MW contributed to statistical analysis of the data. PM reviewed the manuscript and contributed to discussion. CV and VP contributed to tracer preparation and reviewed the manuscript. AKW and MH contributed to discussion, reviewed the manuscript. GK designed the study, reviewed/edited the manuscript. SR: designed the study, wrote the manuscript, collected, and researched the data, contributed to discussion.
Conflicts of Interest
The Authors have nothing to declare and have no known competing financial interests or personal relationships that could have influenced the work reported in this article.
- Received April 9, 2021.
- Revision received April 16, 2021.
- Accepted April 19, 2021.
- Copyright © 2021 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.