European Journal of Nuclear Medicine and Molecular Imaging https://doi.org/10.1007/s00259-018-3971-x

ORIGINAL ARTICLE

Diagnostic performance and impact on patient management of 68Ga-DOTA-TOC PET/CT for detecting osteomalacia-associated tumours Marie Paquet 1,2 & Mathieu Gauthé 1,3 & Jules Zhang Yin 1,3 & Valérie Nataf 1,4 & Ophélie Bélissant 1,3 & Philippe Orcel 5,6 & Christian Roux 7,8 & Jean-Noël Talbot 1,3 & Françoise Montravers 1,3 Received: 6 October 2017 / Accepted: 4 February 2018 # Springer-Verlag GmbH Germany, part of Springer Nature 2018

Abstract Purpose Oncogenic osteomalacia is an endocrine disorder induced by small benign tumours (TIO) producing excessive fibroblast growth factor-23 (FGF23). The only way of curing oncogenic osteomalacia is surgical resection of the culprit TIO, which is extremely difficult to detect using conventional imaging modalities due to its small size and variable location in the body. Since TIO frequently overexpress somatostatin receptors, a clinical utility of SPECT or PET with radiolabelled somatostatin analogues has been reported. Among them, 68Ga-DOTA-TOC has recently been granted a marketing authorization, facilitating its routine application. We report here the results of the first series evaluating the diagnostic performance of 68Ga-DOTA-TOC PET/CT in detecting TIO and its impact on patient management. Methods 68Ga-DOTA-TOC PET/CT and clinical and imaging data from 15 patients with clinical and biochemical signs of oncogenic osteomalacia were retrospectively reviewed. The 68Ga-DOTA-TOC PET/CT findings were compared with the results of post-surgical pathology and clinical and biochemical follow-up. Results 68Ga-DOTA-TOC PET/CT resulted in the detection of one focus suspicious for TIO in nine of 15 patients (60%), and a tumour was surgically removed in eight. Post-operative pathology confirmed a TIO in those eight patients whose symptoms diminished promptly and biochemical anomalies resolved. 68Ga-DOTA-TOC PET/CT sensitivity, specificity and accuracy were 73%, 67% and 71%, respectively. 68Ga-DOTA-TOC PET/CT findings affected patient management in 67% of cases. In particular, 68Ga-DOTA-TOC PET/CT was able to detect the TIO with a negative or a false-positive result of a previous 111In-pentetreotide SPECT/CT in 5/8 patients (63%) or a previous FDG PET/CT in 7/11 patients (64%). No close relationship was found between the positivity of 68Ga-DOTA-TOC PET/CT and the serum level of a biochemical marker. However, a true-positive result of 68Ga-DOTA-TOC PET/CT was obtained in only one patient with a non-elevated serum level of FGF23. Conclusion 68Ga-DOTA-TOC PET/CT is an accurate imaging modality in the detection of TIO; in particular, it is worthwhile after failure of somatostatin receptor SPECT(/CT) or FDG PET/CT. Keywords

68

Ga-DOTA-TOC . PET/CT . Oncogenic osteomalacia . Diagnostic performance . Impact on management

* Jean-Noël Talbot jean–[email protected] 1

Médecine nucléaire, Hôpital Tenon, AP-HP, 4 rue de la Chine, 75020 Paris, France

2

Médecine nucléaire, Centre Antoine Lacassagne, Nice, France

3

Faculté de Médecine P&M Curie, Sorbonne Université, Paris, France

4

Radiopharmacie, Hôpital Tenon, AP-HP, Paris, France

5

Rhumatologie, Hôpital Lariboisière, AP-HP, Paris, France

6

Université Paris Diderot, Paris, France

7

Rhumatologie, Hôpital Cochin, AP-HP, Paris, France

8

Université Paris Descartes, Paris, France

Eur J Nucl Med Mol Imaging

Introduction Oncogenic osteomalacia (OO) is a rare paraneoplastic disorder characterised by biochemical anomalies such as hypophosphatemia and hyperphosphaturia [1]. It results in osteomalacia by default of primary mineralisation of the bone matrix, and clinically in severe asthenia, myalgia, bone pain and ultimately spontaneous fractures. OO is most often secondary to one small benign tumour of a mesenchymal nature [2] responsible for an inappropriate secretion of proteins called phosphatonins acting directly on the phosphate homeostasis. The most common of these phosphatonins is fibroblast growth factor-23 or FGF23 [3, 4]. Although these tumours are benign and rare, the clinical consequences of OO are particularly disabling for patients. Treatment with oral phosphate supplementation has little effect. Only the resection of the culprit tumour allows a complete and rapid cure of the clinical symptoms and biochemical anomalies [5, 6]. The precise localisation of these mesenchymal tumours is challenging due to their small size and their highly variable anatomical situation [5]. In addition, due to the slow growth of these small tumours, local symptoms are frequently overshadowed by the systemic consequences of osteomalacia [5]. Anatomy-based imaging is often non-contributive [6–8]. Actually TIOs are able to overexpress somatostatin receptors (SR), with a clear predominance of subtype 2, as shown by immunochemistry [9, 10]; a weaker expression of subtype 5 has also been reported [9]. In recent decades, scintigraphy using a somatostatin analogue binding to SR (SRS) has been used in the detection of mesenchymal tumours inducing osteomalacia (TIO) [11–13]. However, the low spatial resolution of scintigraphy and SPECT is a limit for detecting subcentimeter tumours. PET/CT yields a better image resolution. Several somatostatin analogues for somatostatin-receptor PET (SRPET) labelled with gallium-68 using a DOTA cage are available: DOTANOC, DOTA-TATE and DOTA-TOC or edotreotide. Somatostatin receptor PET has been widely documented in neuroendocrine tumours (NET), with superior diagnostic performance compared to SPECT [14–16]. Furthermore, PET yields a complete tomographic imaging over the whole field of view, covering the whole-body from vertex to toes in a reasonable acquisition time, which is hardly feasible with SPECT. Fusion of PET and CT images allows accurate anatomical localisation of the focus suspicious to correspond to culprit lesion, which is solitary according to literature data [2]. In 2007, Hesse et al. reported one case of concordance for TIO detection between SRS and DOTA-NOC PET/CT in a 66year-old woman, while CT alone failed [17]. Case reports and series reporting the interest of DOTA-NOC or DOTATATE in this setting have subsequently been published. 68 Ga-edotreotide (or DOTA-TOC) has been widely used in Europe to detect NETs and has very recently been granted a marketing authorisation in the EU, making it easily available

for routine use. However, no series reporting the performance of DOTA-TOC PET/CT in detecting the culprit TIO (which is not a NET) has currently been published, just few case reports [18–20]. The aim of the present study was to evaluate the diagnostic performance of DOTA-TOC PET/CT in detecting TIO as well as its impact on patient’s management.

Material and methods Patients We retrospectively analysed the data of patients referred from December 2014 to January 2017 to the nuclear medicine department of Hôpital Tenon in Paris, France. All patients underwent DOTA-TOC PET/CT for locating a TIO. The inclusion criteria for each patient were: clinical symptoms and biochemical anomalies compatible with OO (according to the respective normal ranges of medical laboratories, in particular for serum FGF23 assay) and availability of the result of serum FGF23 assay (intact protein or C terminal portion). The serum levels of calcium, parathyroid hormone (PTH) and calcidiol (25OH-vitamin D3) were also noted. It was systematically recorded whether patients previously underwent SRS, FDG PET/CT and/or bone densitometry (BMD). All patients gave their informed consent for using their data for research purposes.

DOTA-TOC PET/CT Acquisition protocol The gallium-68 was eluted from a 68Ge/68Ga generator (GalliaPharm, Eckert and Ziegler Radipharma GmbH) and the labelling of DOTA-TOC with gallium-68 was carried out according to the manufacturer’s instructions. No specific preparation of the patient was required before the radiotracer was injected. The usual recommendations for PET/CT and radiation protection instructions were given to each patient prior to imaging. Then, 1.6 MBq/kg of body mass of 68Ga-DOTATOC were injected intravenously. Whole-body PET acquisitions were performed 60 min after injection on one of our PET/CT machines: Biograph mCT (Siemens Healthineers, Erkangen, Germany) or Gemini TF16 (Phillips, Best, The Netherlands). It started with computed tomography (120kVp and 80 mA.s) for correcting the uneven attenuation of the annihilation photons and for a precise landmarking of the foci of tracer uptake. PET images were acquired in a 3D mode with a 576 mm FOV and a 144 × 144 matrix. The images were reconstructed according to the OSEM protocol (33 subsets, 3 iterations) with correction of attenuation and of diffused and random coincidences.

Eur J Nucl Med Mol Imaging

Masked reading Images were analysed using the SyngoVia console (Siemens Healthcare, Erlangen, Germany). Masked reading was performed by two senior and one junior nuclear medicine physicians. PET was interpreted as positive when at least one focus suspicious for TIO was visualised; foci suspicious for another origin (e.g. inflammation, fracture, meningioma) on basis of the aspect on PET and/or on CT were ignored. Regions of interest were obtained using the software’s isocontour tools by pointing on abnormal foci. The corresponding maximum standardised uptake value (SUVmax) was calculated with reference to body mass. The biological volume of the tumour (BV) was then measured by contouring with a threshold of 40% (BV 40) or 50% (BV 50) of the SUVmax in the focus.

FGF23 was greater than the upper limit of normal range (assaying C-terminal portion of FGF23 in ten patients and the intact protein in two patients). In one single patient, the serum concentration of the intact FGF23 protein was fluctuating with time (patient 4). Eleven patients had bone densitometry that revealed low bone mineral density or very low bone mineral density in eight of them. Ten patients previously underwent SRS which was negative in nine cases and falsepositive in one case corresponding to a thymoma of the mediastinum. Twelve patients previously underwent FDG PET/ CT which was negative in ten cases and showed foci interpreted as non-specific inflammation in two cases. All patients underwent DOTA-TOC PET/CT, with an average time lapse of 5.25 years between the onset of symptoms and the completion of the examination (range 1–16 years). The characteristics of the population are reported in Table 1.

Standard of truth

DOTATOC PET/CT results

Standard of truth (SOT) was determined on the basis of histology when a surgical resection could be performed. Results of long-term clinical and biochemical monitoring and followup imaging were also considered. In particular, recovery was defined according to the literature as the rapid normalisation of biochemical parameters and the disappearance of clinical symptoms after removal of a tumour considered as culprit.

DOTA-TOC PET/CT detection rate and final diagnosis

Statistical analysis The IBM Statistical Package for Social Sciences (SPSS) software was used for statistical calculations. Comparisons between biochemical parameters (phosphatemia, calcemia, FGF23, PTH and calcidiol serum levels), bone mineral density and duration of oncogenic osteomalacia evolution according to DOTA-TOC PET/CT results were performed by using Mann-Whitney test. A p value less than 0.05 was considered to be statistically significant.

Results Patients’ characteristics Fifteen consecutive patients (5 women and 10 men) referred to DOTA-TOC PET/CT in search of TIO were included. The median age at diagnosis was 53 years, ranging from 23 to 83 years. All patients presented with clinical symptoms associating asthenia, myalgia, fractures and with biochemical abnormalities suspicious for OO, with severe hypophosphatemia poorly compensated by oral phosphorus supplementation. The mean serum concentrations were 0.48 mmol/L (range 0.45–0.58) for phosphorous and 39 μg/L (range 18–70) for calcidiol. In eleven of 15 patients, the serum level of the

DOTA-TOC PET/CT resulted in the detection of one focus suspicious for TIO in 9 of 15 patients (60%). In all those nine patients, only one single focus suspicious for TIO was reported and no other suspicious lesion non-concentrating DOTATOC was found. Surgical resection was performed in 8 of them and a tumour was resected that corresponded to the focus on DOTA-TOC PET/CT, located in the skeleton (4/8) or in the soft tissue (two in the pleura and two in the adipose tissue). Pathology confirmed its phosphaturitic mesenchymal nature, with mixed connective tissue variant (PMTMCT) without evidence of malignancy in six cases, a secreting capillary angioma in one case and without a more precise identification in one case. The tumour resection was followed by recovery in all eight patients, i.e. a rapid normalisation of biochemical parameters with complete regression of clinical symptoms (Table 2). In one non-operated patient, DOTA-TOC PET/CT detected a focus suspicious for TIO (patient 2) but the diagnosis of OO was finally challenged by expert physicians: biochemical anomalies were non-typical; particularly, serum concentration of intact FGF23 protein was at the low end of the normal range (26 ng/L, normal range of the laboratory between 25 and 50 ng/L). The DOTA-TOC PET/CT focus was a falsepositive result in relation to an inflammatory granuloma on pathological analysis. In this patient, the hypophosphoremia was finally considered as idiopathic. In six patients, DOTA-TOC PET/CT did not detect foci evocative of TIO (patients 1, 4, 7, 8, 11 and 14). In two of them (patients 8 and 14) the final diagnosis did not correspond to OO. Patient 8 had corticosteroid-induced very low bone mineral density in a context of rheumatoid arthritis, and osteomalacia was considered to be idiopathic by the expert conference. In patient 14, the normalisation of biological

Gender

M M

M M F F M

M M M

M

F F

F M

Patient

1 2

3 4 5 6 7

8 9 10

11

12 13

14 15

1380 28.9 50 259

86.8 37.8 – 32.6

107

247 170 476

– 38.1 31.9 53.6

79 104 125 295 1916

68 –

C-term FGF23 (1–90 U/mL)

35 23 79.9 19 37

70.4 –

Phosphaturia (10–40 mmol/24 h)

2.3 2.16

2.2 2.3

– 284 – 307

2.37

–

2.4 2.31 2.29 2.45 2.38

– – – – – 2.2 2.11 2.3

2.49 2.21

– 26

– – –

Calcemia (2.2–2.6 mmol/L)

iFGF23 (25–50 pg/mL)

35 88

36 26

52

57 61 57

70 26

14 17.7

–

36 20.9 20

56 18 45 55 55.7

46

60.4 25.7

Calcidiol (30–50 μg/L)

10 29.7 74.6 57

19 82

PTH (18–68 ng/L)

Normal Low bone mineral density

– Very low bone mineral density

Low bone mineral density

Very low bone mineral density – Very low bone mineral density

Normal Normal Low bone mineral density Low bone mineral density

–

Low bone mineral density –

BMD

M male, F female, ( - ) normal range, C-term FGF23 serum level of C-terminal portion of fibroblast growth factor 23, iFGF23 serum level of intact protein fibroblast growth factor 23, PTH serum level of parathyroid hormone, calcidiol serum level of calcidiol or 25OH vitamin D3, BMD bone mineral density

0.50 0.44

0.45 0.41

0.48

0.58 0.55 0.58

0.35 0.48 0.56 0.46 0.57

0.37 0.42

Phosphatemia (0.8–1.4 mmol/L)

Patients’ characteristics

Table 1

Eur J Nucl Med Mol Imaging

22 × 17 mm 0.5 3 5.3 5.8 Prevertebral T4 Not done 2

SRS somatostatin receptor scintigraphy, DOTA-TOC edotreotide, SUV standardised uptake value, SUVmax maximal SUV value in a focus of uptake, BV40 and BV50 biological volume of the tumour determined using a SUV threshold set at 40% or 50% of the SUVmax of the focus, TIO tumour inducing osteomalacia, PMTMCT phosphaturic mesenchymal tumour with mixed connective tissue

Inflammatory granuloma

15 × 11 mm 17 × 5 mm 8 × 9 mm

Tissue nodule

TIO (PMTMCT)

TIO (PMTMCT) TIO (PMTMCT) TIO (No available histology)

33 × 11 mm Tissue nodule partially calcified

Tissue nodule Tissue nodule Osteolytic nodule 3.4 0.7 0.5 1.2 1.8 3.3 Fat Pleura Femoral head

6 Major trochanter

False-negative False-positive Not done

10

12 13 15

False-negative

35 4.5 6.6

4.2

1.3

12 × 6 mm 10 × 4 mm 6 × 5 mm

12.5

1.4 1.2 2.6

TIO (PMTMCT)

TIO (PMTMCT) TIO (PMTMCT) TIO (Capillary angioma)

13 × 9 mm Osteoblastic nodule

Tissue nodule Osteolytic nodule Osteoblastic nodule 0.5 0.5 0.4

0.9 0.3

1 0.9 1

0.7

1.6 1.2 1.7

5.1

5.2 4.3 3.9

Gluteal bursae

Pleura Rib Right side of mandible

Note done

False-negative False-negative Note done 5 6 9

3

Final diagnosis (histology) BV40 SUVmax Ga-DOTA-TOC uptake site

68

SRS Case

Table 2

Location and semi-quantitative analysis of foci of DOTA-TOC uptake

BV50

SUVmax ratio to liver

CT finding

Lesion size

Eur J Nucl Med Mol Imaging

anomalies (serum phosphorus and FGF23) was obtained after correction of iron deficiency. In these two patients, DOTATOC PET/CT was considered to be true-negative. In three patients (patients 1, 7, 11), the persistence of clinical and biochemical symptoms of osteomalacia was observed during follow-up, despite well-conducted medical treatment. DOTA-TOC PET/CT was thus considered to be falsenegative. For patient 4, a second DOTA-TOC PET/CT was performed due to the persistence of a high serum level of the FGF23 C-terminal portion (104-155 U/mL); it was also negative. However, clinical symptoms and hypophosphatemia were partially improved by medical treatment. In the absence of SOT, this patient was considered non-evaluable and excluded from the final analysis. Relationship with FGF23 serum level and other biochemical parameters There was no significant difference for any of the biochemical parameters and for the duration of the disease between DOTATOC-positive and DOTA-TOC-negative cases. In particular, no close relationship was observed in our series between the positivity of DOTA-TOC PET/CT and the serum level of FGF23, as illustrated by the case of patient 7 with the greatest FGF23 serum level and a negative DOTA-TOC PET/CT. Nevertheless, the lowest FGF23 serum level in patients with positive DOTA-TOC PET/CT was 79 U/mL, in patient 3, close to the upper limit of normal (90 U/mL). In the three other patients with lower normal values of FGF23 serum level, no TIO could be found, corresponding to one falsenegative (patient 1), one false-positive (patient 2) and one true-negative (patient 14) results. The proportion of positive DOTA-TOC PET/CT was somewhat higher when BMD found low bone mineral density or very low bone mineral density (4/8) than when it was normal (1/3), the number of patients being too small to conclude. Impact of DOTA-TOC PET/CT on patient management The 15 patients were referred in the aim of localising a suspected TIO for surgical resection, as the culprit tumour remained occult after conventional work-up. DOTA-TOC PET/CT induced a change in patient management in eight patients by pinpointing a potential TIO but also in two patients as its negative result shed doubt on the diagnosis of OO, which was finally rejected. The impact rate of DOTA-TOC PET/CT was thus 10/15 = 67% [95% confidence interval: 38–88%]. Visual and semi quantitative analysis Foci, even small sized (Figs. 1 and 2), that corresponded to TIO and one false positive focus (Fig. 3) appeared visually of

Eur J Nucl Med Mol Imaging Fig. 1 This 47-year-old patient 5 has been suspected of clinical and biochemical oncogenic osteomalacia for 7 years, with a moderate elevation of serum concentration of C-terminal FGF23 at 125/U mL [normal range = 1–90]. DOTA-TOC PET/ CT showed an intense 4 mm focus (arrow) in the pleura (SUVmax 5.2). After excision of the suspicious lesion, the histology was evocative of a PMTMCT without sign of malignancy. Immediate postoperative follow-up showed rapid disappearance of clinical symptoms and normalisation of biochemical parameters

a high intensity. The inter-observer concordance was perfect (100%) between the three masked readers for the visual detection of DOTA-TOC foci evocative of TIO. As illustrated in Table 2, the SUVmax of those foci was greater than 40% of that of the liver which takes-up DOTATOC intensely. In the semi-quantitative analysis, the mean values of SUVmax, BV 40 and BV 50 were 9.7 g/mL (range 1.8–35), 1.9 cm3 (range 0.7–6) and 1.3 cm3 (range 0.3–4.2), respectively. No significant correlation was found between any of the biochemical parameters and the functional parameters (SUVmax and its ratio to liver SUVmax, BV, BMD).

examinations were true-negative and the last case was not evaluable (patient 4 negative SRS and positive DOTA-TOC PET/CT). In relation to the results of a previous FDG PET/CT, DOTATOC was true-positive in five patients with false-negative FDG PET/CT (patients 6, 9, 12, 13, 15), true-positive in two patients at the same locations as the FDG foci which had been interpreted as non-specific inflammation (patients 5, 10), falsenegative in three patients with false-negative FDG PET/CT (patients 1, 7, 11), true-negative as FDG PET/CT in one case (patient 14) and the last case was not evaluable (patient 4 negative FDG PET/CT and positive DOTA-TOC PET/CT).

Diagnostic performance

Discussion

DOTA-TOC PET/CT resulted in the detection of an actual TIO in eight of the 14 evaluable patients (57%) included in the final analysis. The sensitivity of DOTA-TOC PET/CT was 73%, specificity 67%, positive predictive value 89%, negative predictive value 40% and accuracy 71% (Table 3). In relation to the results of a previous SRS, DOTA-TOC was true-positive in four patients with false negative SRS (patients 5, 6, 10 and 12) and one patient with false-positive SRS (patient 13), and was also false-negative in three patients with falsenegative SRS (patients 1, 7 and 11); in one case (patient 8), both

What are the criteria for referring a patient to nuclear medicine imaging due to suspected OO? In practice, the diagnosis of OO should be suspected in a patient who has had a long history of hypophosphatemia with bone pain. The key biochemical criterion is hypophosphatemia that results from an excessive renal loss of phosphate. Serum levels of PTH are most frequently normal, but may be low or elevated. In the present series, moderately elevated serum levels of PTH were observed in three patients: one was actually not suffering

Eur J Nucl Med Mol Imaging Fig. 2 This 48-year-old patient 3 has been suspected of clinical and biochemical oncogenic osteomalacia for 2 years despite a serum concentration of FGF23 of 79 U/mL, within the laboratory range for normal values (1–90). DOTA-TOC PET/CT showed an intense focus (arrow) in the right femoral head (SUVmax 5.1). The thoracic foci of uptake corresponded to bone fractures. After excision of the suspicious lesion, the histology was evocative of a PMTMCT without sign of malignancy. Immediate postoperative follow-up showed rapid disappearance of clinical symptoms and normalisation of biochemical parameters

from OO (patient 2) but, in the two others (patients 6, 15) whose FGF23 serum level was high, a TIO was localised on DOTATOC PET and resected. Thus high serum levels of PTH and FGF23 should not prevent referring a patient to SRPET and should not lead to 18F-fluorocholine PET/CT in an attempt to detect hyperfunctionning parathyroid glands [21]. The serum assay of the tumour-derived factor FGF23 adds to the armamentarium for the diagnosis of OO [4]. Although the C-terminal peptide is most frequently assayed, in patient 13 of our series the intact FGF23 serum level was elevated whereas the serum level of C-terminal FGF23 was within the normal range and a TIO was found on DOTA-TOC PET/CT. From our limited series, we derive that the probability of a true-positive result of DOTA-TOC PET/CT is more likely in a patient with an elevated FGF23 serum level. No other biochemical parameter correlated with positivity of DOTATOC PET/CT. In the literature, we did not find any previous attempt to determine such relationship, aiming to propose a threshold of a biomarker to indicate PET/CT in this context.

TIO detection with SRS In this clinical setting, using 111In-pentetreotide that has been on the market for decades, Chong et al. reported a sensitivity of SRS of 95% [7] (Table 4). A less favourable detection rate

of 36% was observed in the comparative study of El-Maouche et al. [30]. Using the more recently introduced technetium99m labelled somatostatin analogue 99mTc HYNIC-TOC, Jing et al. reported on 183 patients [24] that had SRS with a sensitivity of 86%, a specificity of 99% and an accuracy of 93%. A series previously published by the same team that included fewer patients [22] was not taken into consideration to avoid probable resampling of results. This high detection rate of 99mTc HYNIC-TOC was confirmed in the short series of Jadhav et al. [25]. In the patients of our study who had previous SRS, DOTATOC PET/CT confirmed the absence of TIO in one truenegative case (patient 14), and allowed the detection of TIO in five of eight patients whose SRS was non-contributive. In four cases, this failure of SRS can be explained, at least partly, by the small size of the TIOs, the long diameter ranging 17– 10 mm and the short diameter 11–4 mm, as PET is able to detect smaller lesions than SPECT. In the last case, the long diameter of the TIO in the left major trochanter was 33 mm; a small and mild focus was seen at this site on SRS, interpreted as tendinitis and discarded. Another important factor is the binding affinity of the somatostatin analogue to SR type 2 and type 5. The affinity measured by IC50 (nmol/L) of 111In-pentetreotide is roughly ten times lower for SR type 2 and three times lower for SR type 5 than that of 68Ga-DOTATOC [23], whereas this

Eur J Nucl Med Mol Imaging Fig. 3 Patient 2 aged 56 years with a history of aggressive angiomas leading to multiple cementoplasties. He was referred for clinical suspicion of oncogenic osteomalacia despite a serum concentration of the intact FGF23 protein (26 pg/mL) within the normal range of the laboratory (25-50) but fluctuating with time. DOTA-TOC PET/CT showed an intense right prevertebral focus (arrow) at the level of the T4 vertebra (SUVmax = 5.8) which was the site of a cementoplasty. After reviewing the entire clinical, biological and imaging records of the patient, the diagnosis of oncogenic osteomalacia was questioned by expert consensus, which retained idiopathic hypophosphoremia. Finally, the DOTATOC PET/CT result was considered to be false-positive, probably due to a granuloma in relation with the past cementoplasty

ratio is only 1.5 for binding to SR type 2 of 99mTc HYNIC-TOC [26], which contributes to a good sensitivity of this tracer despite the limitations of SPECT. Our results cannot be compared to those of studies of SRS alone or of comparative studies including SRS, since patients with definitely positive SRS were not referred to us for DOTA-TOC PET/CT. SRS is widely available whereas

DOTA-TOC PET/CT is still currently considered in France as a last chance modality, performed in specialised nuclear medicine centres. However our study confirmed that DOTATOC PET/CT is able to detect TIO in patients with noncontributive SRS, as already remarked in small series with the two other somatostatin analogues for SRPET [27, 29].

TIO detection with SRPET Table 3

Diagnostic performance of DOTA-TOC PET/CT

Measure

N or ratios, [95% CI]

Cases

TP

8

FP TN FN Sensitivity Specificity PPV NPV

1 2 3 8/11 = 73% [39–94%] 2/3 = 67% [9–99%] 8/9 = 89% [52–100%] 2/5 = 40% [5–85%]

3 – 5 – 6 – 9 – 10 – 12 – 13 – 15 2 8 – 14 1 – 7 – 11

Accuracy

10/14 = 71% [42–92%]

N number of cases, CI confidence interval, TP true positive, FP false positive, TN true negative, FN false negative, PPV positive predictive value, NPV negative predictive value

Since the publication of the pilot series by Clifton-Blight et al. in 2013 [10] (Table 4), several teams have confirmed that the suspicious foci on DOTA-TATE PET/CT corresponded in almost all cases to TIO [25, 27, 28, 31]. The team in Beijing, which reported the largest series on 54 patients using DOTA-TATE with a very high accuracy [31], had previously observed a slightly lower accuracy with 99mTc-HYNIC-octreotide on 183 patients [24]. Similarly, three shorter series have reported a high detection rate of TIO with DOTA-NOC PET/CT [29, 32, 33]. Compared to the larger study conducted by Zhang et al. using DOTA-TATE [31], our results of diagnostic performance with DOTA-TOC seems less favourable, although the differences are not significant: 73% versus 100% for sensitivity and 67% versus 91% for specificity. However, in the study of Zhang et al. in 54 patients, the 12 patients whose DOTA-TATE PET/ CT did not show a focus suspicious for TIO were excluded

Eur J Nucl Med Mol Imaging Table 4

Literature review of series (N > 2) reporting nuclear medicine imaging for detecting tumour-inducing osteomalacia

Reference

Number of patients

Imaging technique

Diagnostic performance

Jing et al. (2013) [21]

183

SPECT(/CT) 99mTc-HYNIC-octreotide

Chong et al. (2013) [7]

31

SPECT or SPECT/CT 111In-pentetreotide

Se = 69/80 = 86%, Sp = 102/103 = 99%, Acc =171/183 = 93% Se =18/19 = 95%, Sp = 7/11 = 64%, Acc =25/30 = 83% Se = 14/16 = 88%, Sp = 4/11 = 36%, Acc = 18/27 = 67%

PET/CT 18F-FDG Clifton-Bligh et al. (2013) [22]

6

PET/CT 68Ga-DOTA-TATE

Breer et al. (2014) [23] Jadhav et al. (2014) [24]

5 16

Agrawal et al. (2015) [25] Zhang et al. (2015) [26]

6 54 (43 evaluable)

PET/CT 68Ga-DOTA-TATE SPECT (/CT) 99mTc-HYNIC-octreotide PET/CT 68Ga-DOTA-TATE PET/CT 18F-FDG PET/CT 68Ga-DOTA-TATE PET/CT 68Ga-DOTA-TATE

El Maouche et al. (2016) [20]

11

Ho et al. (2015) [27]

3

Bhavani et al. (2016) [28] Singh et al. (2017) [29] Present study

10 17 15 (14 evaluable)

SPECT/CT 111In-pentetreotide PET/CT 68Ga-DOTA-TATE PET/CT 18F-FDG PET/CT 68Ga-DOTA-NOC PET/CT 68Ga-DOTA-NOC PET/CT 68Ga-DOTA-NOC PET/CT 68Ga-DOTA-TOC

Detection rate = 6/6 = 100% Detection rate = 5/5 = 100% Detection rate = 6/6 = 100% Detection rate = 7/7 = 100% Detection rate = 4/8 = 50% Detection rate = 5/6 = 83% Se = 32/32 = 100%, Sp =10/11 = 91%, Acc = 42/43 = 98% Detection rate = 4/11 = 36% Detection rate = 6/11 = 55% Detection rate = 4/11 = 36% Detection rate = 3/3 = 100% Detection rate = 9/10 = 90% Detection rate = 9/17 = 53% Se = 8/11 = 73%, Sp = 2/3 = 67%, Acc =10/14 = 71%

N number of patients, Se sensitivity, Sp specificity, Acc accuracy, evaluable patients for whom a standard of truth could be determined

from the statistical analysis [31]. This option is likely to artificially increase the sensitivity. The binding affinity of DOTATATE is roughly ten times higher for SR type 2 but 50 times lower for SR type 5 than that of DOTA-TOC [23], thus no clear consequence can be anticipated for the detection of TIO.

DOTA-NOC also binds to SR type 3 [23], which is overexpressed by inflammatory cells [34]. Inflammation is usually present in case of fractures which are a frequent complication of OO.

TIO detection with FDG PET/CT Specificity of SR imaging 18

Concerning specificity, somatostatin receptors are ubiquitous and expressed on the surface of different cell types. Thus, the uptake of somatostatin analogue radiotracers, for SRPET or SRS, may be high in other pathologies such as NET [14–16]. On the other hand, an uptake focus of inflammatory origin can be observed in bone fractures that may be confused with an authentic TIO, especially in those fragile patients at risk of fractures by bone insufficiency. Singh et al. [32] insisted on the frequency of non-specific focal tracer-avid skeletal lesions in those patients, in particular fractures or degenerative changes that explained 26 of the 37 lesions detected with DOTANOC. The only false-positive result in our series (patient 2) was observed in a patient with an aggressive angiomatous disease, with multiple episodes of cementoplasty that made interpretation of DOTA-TOC uptake rather difficult (Fig. 3). This better specificity of DOTA-TOC compared to DOTANOC in searching for TIO could be due to different binding profiles: DOTA-TOC essentially binds to SR type 2 whereas

F-fluorodeoxyglucose (FDG) is the most widely used radiopharmaceutical in our nuclear medicine department. It was also the first PET tracer to be proposed for the detection of TIO in some case reports [35, 36] (Table 4). Very soon, its limits were described, in particular its low specificity in this context: 36% in the series of Chong et al. [7]. Jadhav et al. [25] and El Mahouche et al. [30] confirmed its low detection rate. In our series, DOTATOC was able to detect TIO in seven patients whose FDG PET/CT had been interpreted as negative. Actually, in two patients DOTATOC foci matched FDG foci: in one patient a focus in the pleura had a very low FDG SUVmax = 2 vs. 5.2 with DOTATOC; in the other patient with a large TIO in the femur, the calcified FDG focus was interpreted as a hematoma. All the three OO patients with no evocative DOTA-TOC focus had a negative FDG PET/CT. Successful FDG PET has been reported in case of aggressive or poorly differentiated NET with negative or equivocal SRPET [37]. In our series, all TIO were histologically well-differentiated and benign. We performed a survey of literature and were unable to find cases of OO due to

Eur J Nucl Med Mol Imaging

a poorly differentiated tumour. Thus there is no evidence for a subgroup of patients with OO in whom FDG could be preferred to SRPET. Even considering the recruitment bias of our series, the role of FDG PET in this context appears very limited.

Limitations The main limitations of this study are its retrospective design, and its long inclusion period due to the rarity of osteomalaciaassociated tumours. Although our series is the first reporting on the diagnostic performance and impact on patient management of 68Ga-DOTA-TOC PET/CT for detecting TIO, the sample size is still small. In our study, DOTA-TOC PET/CT was performed in patients with a TIO remaining occult after conventional work-up. Thus, DOTA-TOC PET/CT was not performed in patients whose TIO was successfully located with conventional imaging modalities, which constitutes a referral bias. Therefore, we demonstrate that DOTA-TOC PET/CT is able to detect TIO in patients with noncontributive imaging workup, including non-contributive SRS or non-contributive FDG PET/CT, but not that it should be used as the first line imaging modality in all suspected TIO. Differences in imaging performance between our two PET/ CT devices could have introduced a slight variability in our results, although they both belong to the new generation of PET/CT devices with time of flight technology. In our series, only 2/15 patients were scanned on our most recent PET/CT machine installed in 2016, the positivity rate of which was 2/2 = 100% vs. 7/13 = 54% with the less recent one installed in 2010 (p = 0.5 using Fisher’s exact test). Because of these limitations, our results need to be validated in a larger prospective and multicentre study gathering a larger cohort.

Conclusion Our work confirms the utility of PET/CT with a somatostatin receptor ligand for the detection of tumours inducing oncogenic osteomalacia. We report a sensitivity of 73% and a specificity of 67% with DOTA-TOC, which has the advantage of currently benefiting from a marketing authorisation in Europe, unlike the other somatostatin analogues for PET imaging, DOTA-TATE or DOTA-NOC. DOTA-TOC PET/CT led to a change in patient management in 67% of cases either by allowing surgery by targeting the culprit tumour, or by challenging the diagnosis of TIO. We did not perform a head-to-head comparison with SRS or FDG PET/CT including patients with a positive conventional imaging work-up, but we showed that DOTA-TOC was able to detect the TIO in five of eight patients (63%) with a negative or a false-positive result of a previous 111In-pentetreotide SRS and in seven of 11 patients (64%) whose FDG PET/CT was interpreted as negative.

Acknowledgements The authors are grateful to the following physicians for their confidence in referring their patients: from Paris, Dr. Beauvais and Dr. Nion, Hôpital Saint-Antoine, Prof Bertocchio and Prof Houillier, HEGP, Dr. Frazier, Hôpital Lariboisière, Prof Haymann, Hôpital Tenon, Dr. Salcion, Hôpital Cochin, and, from other cities, Dr. Abbad, CH Valenciennes and Dr. Che, CHU Lapeyronie, Montpellier. We also want to acknowledge the commitment and the excellent technical skills of all the team of the nuclear medicine department of Hôpital Tenon where DOTA-TOC PET/CTs were performed.

Compliance with ethical standards Conflicts of interest The authors declare that they have no conflict of interest. Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinski declaration and its later amendments. Research involving human participants and/or animals: for this type of study, formal consent is not required. Informed consent This study was approved by the institutional review board; the patients gave their written consent for a subsequent use of their PET/CT images for research purposes.

References 1. 2.

3.

4.

5.

6. 7.

8.

9.

10.

Carpenter TO. Oncogenic osteomalacia—a complex dance of factors. N Engl J Med. 2003;348:1705–8. Folpe AL, Fanburg-Smith JC, Billings SD, Bisceglia M, Bertoni F, Cho JY, et al. Most osteomalacia-associated mesenchymal tumors are a single histopathologic entity: an analysis of 32 cases and a comprehensive review of the literature. Am J Surg Pathol. 2004;28:1–30. Nelson AE, Bligh RC, Mirams M, Gill A, Au A, Clarkson A, et al. Clinical case seminar: fibroblast growth factor 23: a new clinical marker for oncogenic osteomalacia. J Clin Endocrinol Metab. 2003;88:4088–94. Jonsson KB, Zahradnik R, Larsson T, White KE, Sugimoto T, Imanishi Y, et al. Fibroblast growth factor 23 in oncogenic osteomalacia and X-linked hypophosphatemia. N Engl J Med. 2003;348(17):1656–63. Ledford CK, Zelenski NA, Cardona DM, Brigman BE, Eward WC. The phosphaturic mesenchymal tumor: why is definitive diagnosis and curative surgery often delayed? Clin Orthop. 2013;471:3618–25. Farrow EG, White KE. Tumor-induced osteomalacia. Expert Rev Endocrinol Metab. 2009;4:435–42. Chong WH, Andreopoulou P, Chen CC, Reynolds J, Guthrie L, Kelly M, et al. Tumor localization and biochemical response to cure in tumorinduced osteomalacia. J Bone Miner Res. 2013;28:1386–98. Chong WH, Yavuz S, Patel SM, Chen CC, Collins MT. The importance of whole body imaging in tumor-induced osteomalacia. J Clin Endocrinol Metab. 2011;96:3599–600. Houang M, Clarkson A, Sioson L, Elston MS, Clifton-Bligh RJ, Dray M, et al. Phosphaturic mesenchymal tumors show positive staining for somatostatin receptor 2A (SSTR2A). Hum Pathol. 2013;44(12):2711–8. Clifton-Bligh RJ, Hofman MS, Duncan E, Sim I-W, Darnell D, Clarkson A, et al. Improving diagnosis of tumor-induced osteomalacia with gallium-68 DOTATATE PET/CT. J Clin Endocrinol Metab. 2013;98:687–94.

Eur J Nucl Med Mol Imaging 11.

12. 13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

Jan de Beur SM, Streeten EA, Civelek AC, EF MC, Uribe L, Marx SJ, et al. Localisation of mesenchymal tumours by somatostatin receptor imaging. Lancet Lond Engl. 2002;359:761–3. Jan de Beur SM. Tumor-induced osteomalacia. JAMA. 2005;294: 1260–7. Nguyen BD, Wang EA. Indium-111 pentetreotide scintigraphy of mesenchymal tumor with oncogenic osteomalacia. Clin Nucl Med. 1999;24:130–1. Krausz Y, Freedman N, Rubinstein R, Lavie E, Orevi M, Tshori S, et al. 68Ga-DOTA-NOC PET/CT imaging of neuroendocrine tumors: comparison with 111In-DTPA-octreotide (OctreoScan®). Mol Imaging Biol. 2011;13:583–93. Deppen SA, Blume J, Bobbey AJ, Shah C, Graham MM, Lee P, et al. 68Ga-DOTATATE compared with 111In-DTPA-octreotide and conventional imaging for pulmonary and gastroenteropancreatic neuroendocrine tumors: a systematic review and meta-analysis. J Nucl Med. 2016;57:872–8. Graham MM, Gu X, Ginader T, Breheny P, Sunderland JJ. 68GaDOTATOC imaging of neuroendocrine tumors: a systematic review and metaanalysis. J Nucl Med. 2017;58:1452–8. Hesse E, Moessinger E, Rosenthal H, Laenger F, Brabant G, Petrich T, et al. Oncogenic osteomalacia: exact tumor localization by coregistration of positron emission and computed tomography. J Bone Miner Res. 2007;22:158–62. Haeusler G, Freilinger M, Dominkus M, Egerbacher M, Amann G, Kolb A, et al. Tumor-induced hypophosphatemic rickets in an adolescent boy - clinical presentation, diagnosis, and histological findings in growth plate and muscle tissue. J Clin Endocrinol Metab. 2010;95:4511–7. Maybody M, Grewal RK, Healey JH, Antonescu CR, Fanchon L, Hwang S, et al. Ga-68 DOTATOC PET/CT-guided biopsy and cryoablation with autoradiography of biopsy specimen for treatment of tumor-induced osteomalacia. Cardiovasc Intervent Radiol. 2016;39:1352–7. Kawai S, Ariyasu H, Furukawa Y, Yamamoto R, Uraki S, Takeshima K, et al. Effective localization in tumor-induced osteomalacia using 68Ga-DOTATOC-PET/CT, venous sampling and 3T-MRI. Endocrinol Diabetes Metab Case Rep. 2017;2017 Michaud L, Balogova S, Burgess A, Ohnona J, Huchet V, Kerrou K, et al. A pilot comparison of 18F-fluorocholine PET/CT, ultrasonography and 123I/99mTc-sestaMIBI dual-phase dual-isotope scintigraphy in the preoperative localization of hyperfunctioning parathyroid glands in primary or secondary hyperparathyroidism: influence of thyroid anomalies. Medicine (Baltimore). 2015;94(41):e1701. Jiang Y, Xia W, Xing X, Silva BC, Li M, Wang O, et al. Tumorinduced osteomalacia: an important cause of adult-onset hypophosphatemic osteomalacia in China: report of 39 cases and review of the literature. J Bone Miner Res. 2012;27:1967–75. Johnbeck CB, Knigge U, Kjær A. PET tracers for somatostatin receptor imaging of neuroendocrine tumors: current status and review of the literature. Future Oncol. 2014;10(14):2259–77. Jing H, Li F, Zhuang H, Wang Z, Tian J, Xing X, et al. Effective detection of the tumors causing osteomalacia using [Tc-99m]HYNIC-octreotide (99mTc-HYNIC-TOC) whole body scan. Eur J Radiol. 2013;82:2028–34.

25.

Jadhav S, Kasaliwal R, Lele V, Rangarajan V, Chandra P, Shah H, et al. Functional imaging in primary tumour-induced osteomalacia: relative performance of FDG PET/CT vs somatostatin receptorbased functional scans: a series of nine patients. Clin Endocrinol. 2014;81:31–7. 26. Dong C, Zhao H, Yang S, Shi J, Huang J, Cui L, et al. 99mTclabeled dimeric octreotide peptide: a radiotracer with high tumor uptake for single-photon emission computed tomography imaging of somatostatin receptor subtype 2-positive tumors. Mol Pharm. 2013;10(8):2925–33. 27. Breer S, Brunkhorst T, Beil FT, Peldschus K, Heiland M, Klutmann S, et al. 68Ga DOTA-TATE PET/CT allows tumor localization in patients with tumor-induced osteomalacia but negative 111Inoctreotide SPECT/CT. Bone. 2014;64:222–7. 28. Agrawal K, Bhadada S, Mittal BR, Shukla J, Sood A, Bhattacharya A, et al. Comparison of 18F-FDG and 68Ga DOTATATE PET/CT in localization of tumor causing oncogenic osteomalacia. Clin Nucl Med. 2015;40:e6–10. 29. Ho CL. Ga68-DOTA peptide PET/CT to detect occult mesenchymal tumor-inducing osteomalacia: a case series of three patients. Nucl Med Mol Imaging. 2015;49:231–6. 30. El-Maouche D, Sadowski SM, Papadakis GZ, Guthrie L, CottleDelisle C, Merkel R, et al. 68Ga-DOTATATE for tumor localization in tumor-induced osteomalacia. J Clin Endocrinol Metab. 2016;101:3575–81. 31. Zhang J, Zhu Z, Zhong D, Dang Y, Xing H, Du Y, et al. 68Ga DOTATATE PET/CT is an accurate imaging modality in the detection of culprit tumors causing osteomalacia. Clin Nucl Med. 2015;40:642–6. 32. Singh D, Chopra A, Ravina M, Kongara S, Bhatia E, Kumar N, et al. Oncogenic osteomalacia: role of Ga-68 DOTANOC PET/CT scan in identifying the culprit lesion and its management. Br J Radiol. 2017;90:20160811. 33. Bhavani N, Reena Asirvatham A, Kallur K, Menon AS, Pavithran PV, Nair V, et al. Utility of gallium-68 DOTANOC PET/CT in the localization of tumour-induced osteomalacia. Clin Endocrinol. 2016;84:134–40. 34. Lichtenauer-Kaligis EG, Dalm VA, Oomen SP, Mooij DM, van Hagen PM, Lamberts SW, et al. Differential expression of somatostatin receptor subtypes in human peripheral blood mononuclear cell subsets. Eur J Endocrinol. 2004;150(4):565–77. 35. Dupond JL, Mahammedi H, Prié D, Collin F, Gil H, Blagosklonov O, et al. Oncogenic osteomalacia: diagnostic importance of fibroblast growth factor 23 and F-18 fluorodeoxyglucose PET/CT scan for the diagnosis and follow-up in one case. Bone. 2005;36:375–8. 36. Malhotra G, Agrawal A, Jambhekar NA, Sarathi V, Jagtap V, Agarwal MG, et al. Interesting image. The search for primary tumor in a patient with oncogenic osteomalacia: F-18 FDG PET resolves the conundrum. Clin Nucl Med. 2010;35:896–8. 37. Chan DL, Pavlakis N, Schembri GP, Bernard EJ, Hsiao E, Hayes A, et al. Dual somatostatin receptor/FDG PET/CT imaging in metastatic neuroendocrine tumours: proposal for a novel grading scheme with prognostic significance. Theranostics. 2017;7(5):1149–58.