Differentiation of Renal Oncocytoma From Renal Cell Carcinoma Using 99mTc-Sestamibi SPECT/CT.


 Background: 99mTc-Sestamibi Single Photon Emission Computed Tomography/Computed Tomography (SPECT/CT) contributes to the non-invasive differentiation of renal oncocytoma (RO) from renal cell carcinoma (RCC). We investigated whether standard uptake value (SUV) SPECT, has a beneficial role in differentiating renal oncocytoma (RO) from renal cell carcinoma (RCC) besides visual assessment. As a secondary aim, we evaluated the mitochondrial content of 19 oncocytic tumours arranged in a tissue microarray, by immunohistochemistry, using succinate dehydrogenase complex subunit B (SDHB) protein expression. In addition to visual evaluation of 99mTc-Sestamibi uptake by characterizing renal tumours as Sestamibi positive or Sestamibi negative regarding their uptake compared to the non-tumoral renal parenchyma, SUVmean and SUVmax measurements were performed in the renal tumour and the non-tumoral renal parenchyma. Intra Class Correlation calculated to assess the intra-reader reliability of SUV measurements. ROC-analysis demonstrated an optimal cut off SUV value that differentiates RO from RCC. SDHB score was analysed using Cochran–Armitage test for trend.Results: 57 renal tumours from 52 patients were evaluated. Visual evaluation of 99mTc-Sestamibi SPECT/CT examination resulted in a sensitivity of 83%, whereas quantitative evaluation showed a sensitivity of 64% regarding the differential of RO from RCC. A significant trend (p=0,0328) of increased SDHB score found in the Sestamibi positive group.Conclusion: Quantitative evaluation with SUV SPECT measurements did not improve the performance of 99mTc-Sestamibi SPECT/CT in differentiating RO from RCC. 99mTc- Sestamibi SPECT/CT identified a larger Sestamibi-positive tumour group containing RO, Hybrid Oncocytic Chromophobe Tumours and the majority of chromophobe RCCs. Thus, Sestamibi-negative renal tumours, that are possibly malignant, should be considered for surgery. Patients with Sestamibi-positive tumours can be suited for biopsy and follow up according to active surveillance protocols. Low Grade Oncoytic Tumour, a provisional renal entity with no proven recurrence or metastatic potential, appears to be positive on 99mTc- Sestamibi SPECT/CT examination.


Introduction
Benign and malignant solid renal tumours overlap in their standard radiological imaging features [1,2]. Currently, it is not possible to differentiate preoperatively in a de nite way between various solid renal tumours, using either ultrasound (US), contrast-enhanced ultrasound (CEUS), contrast-enhanced multiphasic computed tomography (CECT), radiomics, perfusion CT scans (CTp) or magnetic resonance imaging (MRI) examinations [3][4][5] The inability to accurately diagnose benign renal neoplasia results in up to 30% of unnecessary partial nephrectomies according to a study by Kim et al. [6]. The introduction of 99m Tc-Sestamibi however SPECT/CT improves the characterisation and the distinction of the solid renal tumours, as demonstrated in the recent meta-analysis by Wilson et al. [7]. In particular, 99m Tc-Sestamibi SPECT/CT, followed by renal biopsy improved the management of benign kidney tumours such as renal oncocytomas (RO) by reducing the unnecessary surgeries [8] and also by introducing cost-saving e ciencies, compared with the established clinical routines [9].
The histopathological and molecular characterisation of renal neoplasms is continuously evolving, as re ected in the 2016 World Health Organization (WHO) classi cation, and in the introduction of potentially new and emerging renal entities in the post-WHO era [10,11]. For example, clear cell papillary RCC (ccpRCC), a tumour of low malignant potential, is now classi ed as a separate entity [12], while the so called "hybrid chromophobe oncocytic tumour" (HOCT), showing features that overlap between RO and chromophobe renal cell carcinoma (chRCC), considered a subtype of chRCC in the 2016 WHO classi cation, has been found to demonstrate molecularly distinct pro le, with genomic features intermediate between those of RO and chRCC [13,14]. Other entities, e.g. fumarate hydrase (FH)-de cient RCCs, as well as emerging entities, such low grade-oncocytic tumours (LOT) and high grade-oncocytic tumours [14][15][16][17] are also expanding the spectrum of renal oncocytic neoplasia [11].
Our group has previously published the preliminary experiences in visual differentiation of RO from RCC utilising 99m Tc-Sestamibi SPECT/CT [19]. During the past year more scienti c groups have published similar results [20][21][22] underlining the increasing interest in this eld of renal oncologic imaging. According to the initial hypothesis of Rowe et al. [23], renal tumours, such as RO that demonstrate high number of mitochondrial content exhibit an increased uptake of 99m Tc-Sestamibi, given that Sestamibi acts as a mitochondrial agent. On this basis, renal tumours with increased 99m Tc-Sestamibi can be characterised as Sestamibi-positive. In contrast, renal tumours with decreased 99m Tc-Sestamibi uptake are called Sestamibi-negative. In our pilot study we found that 11 of 12 RO (91.6%) and 3 of 3 (100%) HOCT were Sestamibi positive, in line with the previous ndings by Gorin et al. [24], while only one papillary RCC (pRCC) had slightly increased 99m Tc-Sestamibi uptake.
In the present study, we sought to evaluate if implementation of quantitative tools, in addition to visual assessment of 99m Tc-Sestamibi SPECT/CT examination in patients with solid renal tumours, could improve the diagnostic performance in differentiating RO from other RCCs. In particular, we investigated if standardised uptake value (SUV) SPECT, could be used to strengthen the preoperative characterisation of the solid renal tumours prospectively. The secondary aim was to study the differences in the 99m Tc-Sestamibi uptake within tumour subgroups, including RO, HOCT, and chRCC, and the implications on tumour entity characterisation. Thus, we additionally performed an immunohistochemical investigation to correlate the tumour mitochondrial content with the 99m Tc-Sestamibi SPECT/CT results.

Material And Methods
The MIDOR Project Sweden's innovation agency VINNOVA nancially supported this study as part of the Molecular Imaging for Differentiation of Oncocytomas from Renal cancer (MIDOR) study (Reference nr 2015 − 01080) and Hermes Medical Solutions AB Stockholm, Sweden provided the software used for the evaluation of SPECT/CT images. The MIDOR study was initiated at the Department of Nuclear Medicine, Karolinska University Hospital, following an approval by the Regional Ethical Review Board and the local Radiation Safety Committee (Reference nr 2015/923 − 31/4). We also acquired written informed consent by all MIDOR project participants.

Study Design and Image Acquisition
Patients with T2 + tumours, tumours > 7 cm in maximum diameter and/or patients with metastatic disease were excluded. The participants underwent a 99m Tc-Sestamibi SPECT/CT examination before nephrectomy or renal biopsy following the same imaging protocol as previously described by our group [25]. In brief, a CT scan was performed after a SPECT acquisition with projections reconstructed using Hermes SUV SPECT Hybrid Recon™ Oncology software (v.1.2) (HERMES Medical Solutions AB, Stockholm, Sweden). To enable quantitative evaluation, we collected the available data on syringe activity prior to administration, syringe residual activity post-administration, patient weight, time points of injection, and scan start, which were all used in the reconstruction software.
Evaluation of 99m Tc-Sestamibi SPECT/CT examination Two readers independently performed the visual evaluation: a Consultant in Radiology and Nuclear Medicine (SG) and a Consultant in Radiology (AT). To mitigate the effects of interobserver variation, we rede ned the uptake of 99m Tc-Sestamibi in the examined renal tumour, as follows: a tumour was classi ed as Sestamibi-positive if the 99m Tc-Sestamibi uptake was visually higher compared to the ipsilateral normal renal parenchyma. If the 99m Tc-Sestamibi uptake in the tumour was visually equal to, or lower than the ipsilateral renal parenchyma, the tumour was classi ed as a Sestamibi-negative. In cases with a disagreement between the readers, a third reader (RA; Consultant in Nuclear Medicine and Radiology) was added and all three readers (S.G., A.T. & R.A.) independently performed a blind reevaluation of these cases.
The two main readers also performed additional quantitative measurements of SUV mean and SUV max, both in the renal tumour and in the ipsilateral non-tumoral parenchyma. Freehand regions of interest (ROIs) of the complete tumours were drawn in the axial CT images of the SPECT/CT examination. Volumes of interest (VOIs) were automatically generated, based on the manually drawn ROIs. The VOIs were copied onto the SPECT-images and SUV-parameters were annotated. Fixed 1cm 3 VOIs spheres were likewise placed in the regions of high renal 99m Tc-Sestamibi uptake to obtain SUV mean and SUV max values of the ipsilateral non-tumoral parenchyma. In the current study, we excluded the SUV peak , because many tumours were too small (volume approximately 1cm 3 ) to be evaluated using this parameter. By de nition, the SUV peak in the HERMES software represents the maximum activity concentration in 1cm 3 volume.
In cooperation with the Department of Clinical Pathology-Cytology, Karolinska University Hospital, the diagnostic tissue material (i.e., surgical resections and/or biopsies) was re-evaluated. Two Consultant Histopathologists (TP & WW) independently reviewed the haematoxylin & eosin (H&E), as well as the immunohistochemical (IHC) slides of all tumours in a blinded manner, and without any knowledge of the prior visual or quantitative evaluation of the 99m Tc-Sestamibi SPECT/CT dataset. The con rmed histopathological diagnoses and/or updated diagnoses, based on consensus, were used as the gold standard to correlate with the results from 99m Tc-Sestamibi SPECT/CT examinations. A third expert urologic pathologist (K.T) was subsequently asked to blindly evaluate all chRCCs included in this study, to aid in correlating the observed in situ metabolomic differences in the Sestamibi-positive versus Sestamibi-negative chRCCs, as shown in our previous study [26].
To evaluate the mitochondrial content of the renal oncocytic tumours, we used succinate dehydrogenase complex subunit B (SDHB) protein expression, which can be detected by ICH. SDHB and SDHA subunits constitute the catalytic domain of succinate dehydrogenase (SDH), a hetero-tetrameric enzyme complex anchored to the inner mitochondrial membrane [27]. We utilized samples from 19 tumours

Statistical analysis
Intra Class Correlation (ICC) was calculated to assess the intra-reader reliability of SUV measurements. We calculated the average value between the two separate measurements (SUV max, respectively SUV mean SPECT/CT) from each reader per case of renal tumour. This average value of SUVmax respectively SUVmean between readers based on the average measurements from each reader was used in our analysis as well as the ratio of the relative 99m Tc-Sestamibi uptake (rate of SUV max respectively SUV mean uptake from the renal tumour / SUV max respectively SUV mean uptake from ipsilateral non-tumoral renal parenchyma). To illustrate and understand the trade-off in sensitivity and speci city of Sestamibi-positive vs Sestamibi-negative renal tumours, ROC-analysis was performed to identify an optimal cut off SUV value. To facilitate the illustration of the results, each tumour category was designated with a letter (A-J) ( Supplementary Fig. 1). SDHB score was analysed as categorical ordinal variable and compared to 99m Tc-Sestamibi uptake/no uptake from the above-mentioned 19 renal tumours using Cochran-Armitage test for trend.
Five renal tumours were re-classi ed upon re-evaluation by consensus, as follows: 4 RO were re-classi ed as chRCC and HOCT (2 cases each), while one pRCC (that in our previous pilot study exhibited slightly increased 99m Tc-Sestamibi uptake) was re-classi ed as chRCC. Another tumour previously characterised as HOCT corresponded to normal renal parenchyma and was excluded from further analysis reducing the total number of tumours to 57. The patient and tumour characteristics are presented in supplementary Table 1. All amended diagnoses were immediately communicated to the referral urologists for appropriate clinical management and follow up, whenever necessary. Accordingly, the nal diagnoses were established (in decreasing order of frequency), as follows: 13 clear cell RCCs, 11 chRCCs, 11 RO (one with adjacent) papillary adenoma, 9 papillary RCCs, 5 HOCTs, 4 clear cell papillary RCCs (one with adjacent papillary adenoma), 1 "collision tumour" comprising chromophobe RCC with adjacent papillary RCC, 1 B cell non-Hodgkin's (follicular) lymphoma, 1 metanephric adenoma and 1 angiomyolipoma.
Visual evaluation of 99m Tc-Sestamibi SPECT/CT examinations Agreement on the visual evaluation of 99m Tc-Sestamibi uptake was observed in 51 out of 57 (89.5%) solid renal tumours. In 6 (10.5%) cases, the two readers were in disagreement; however, a complete agreement and a nal consensus was reached on all 6 cases when a third reader was included in the assessment (as previously explained). The results of the visual assessment are presented in Table 1. Table 1 Visual evaluation of 99m Tc-Sestamibi uptake on 57 solid renal tumours resulting in 82% sensitivity and 76% speci city in detecting RO. Quantitative SUV SPECT measurements Similar to our previously published study, the ICC for SUV max and SUV mean measurements on the examined tumours between the two readers showed a high agreement of 88% and 94%, respectively, Table 2. Likewise, the ICC between the readers, when measuring SUV parameters in the ipsilateral nontumoral renal parenchyma showed moderate agreement, Table 2.   Fig. 2).

Re-evaluation of all chRCCs
An expert urological pathologist (K.T.) re-evaluated all chRCCs resulting in 5 Sestamibi-negative tumours being classi ed 4 chRCCs of classic type (Fig. 2a-e) and 1 as an eosinophilic variant of chRCC. Of the 6 Sestamibi-positive tumours, 3 were re-classi ed as low-grade oncocytic tumours (LOTs) (Fig. 2f-j), 2 as an eosinophilic variant of chRCC and 1 as chRCC classic type. However, all these tumours were included in the chRCC subgroup, because LOT represents a recently proposed and provisional renal entity, awaiting formal acceptance and potential inclusion in the WHO classi cation [28].

Mitochondrial content of 19 renal oncocytic tumours
The most common SDHB score in the Sestamibi positive group was 4 vs 3 in the Sestamibi negative group. A Cochran-Armitage test for trend, keeping SDHB as categorical variable but taking the order of values into account, suggested a signi cant trend (p = 0,0328) of increased SDHB score in the Sestamibi positive group (supplementary Table 2).

Discussion
Quantitative evaluation with SUV measurements performed in HERMES Hybrid Viewer PDR v2.5 did not improve the performance on a 99m Tc-Sestamibi SPECT/CT examination in differentiating RO, a benign renal tumour from RCC. 99m Tc-Sestamibi uptake was visually evident in 5 out of 5 HOCTs (100%), 9 out of 11 RO (82%) (Fig. 3), and 6 out of 11 chRCCs (55%) in decreasing order of frequency. All HOCTs and approximately half of the chRCC cases were classi ed as Sestamibi-positive, raising the question of the clinical importance of misclassifying tumour entities with an indolent clinical course [29]. This nding could be eventually incorporated into modern active surveillance programs [30] resulting in a multimodality imaging approach concerning the non-invasive characterisation of renal neoplasia.
A noteworthy observation is that 3 of 11 chRCCs were re-classi ed as LOTs upon expert review, and these 3 tumours were all classi ed as Sestamibi-positive. Our ndings suggest that Sestamibi-positive renal tumours are of low malignant potential and thus could be considered in active surveillance programs utilising renal biopsy and longer-term imaging follow up. The emerging question, however, concerns the diagnostic accuracy of the biopsy obtained from those renal tumours that are Sestamibi-positive.
The limitations of imaging methods for accurate differentiation of renal tumours Sestamibi-positive on 99m Tc-Sestamibi SPECT/CT (RO, HOCT, chRCC) are valid not only on hybrid molecular and conventional imaging but also considering the complexities on histopathologic evaluation. In the current study, 4 diagnoses of RO were amended, either into HOCT or chRCC following a combined morphological and immunohistochemical assessment of three biopsies and one resection. These reclassi cations are consistent with the ndings from systematic review and meta-analysis [31], suggesting that core biopsy can often be unreliable for the diagnosis of RO. The diagnosis on renal biopsy can be problematic also due to the regional tumour heterogeneity and for these reasons many pathologists do not even issue a de nitive diagnosis of RO on biopsy. In particular, Patel et al. [31] demonstrated that 1 in 4 RO cases was misdiagnosed with 12.5% and 6.3% of tumours re-classi ed as chRCCs or HOCTs following excision, respectively. It is also important to note that one of the most challenging areas in the routine renal pathology practice is the diagnosis of tumours with overlapping or equivocal features between RO and chRCC, as well as differentiating RO and chRCC from other oncocytic renal tumours that do not completely ll the diagnostic criteria. However, current consensus is that these di cult to classify "hybrid" oncocytic tumours have exceedingly low risk for developing metastatic disease.
Immunohistochemistry plays a major role in renal tumour diagnostics due to its widespread availability; a panel comprising cytokeratin 7 (CK7)/carbonic anhydrase IX (CAIX)/alpha-methyl acyl-CoA racemase (AMACR)/KIT (CD117) may be used for screening [32]. Other emerging markers such as FOXI1, RHCG, and LINC01187 that appear to be lineage-speci c for renal epithelial neoplasms arising from intercalated cells in the distal nephron segment, may also play a role in the future [33]. Nevertheless, limitations are well known in routine practice, and slight differences in immunohistochemical pro les have also been identi ed among HOCT subtypes (i.e. Birt-Hogg-Dubé syndrome, renal oncocytosis/oncocytomatosis and sporadic form) [13,34].
Other approaches to potentially identify unique oncocytic tumour-speci c features include (i) whole scale approach of computer-assisted morphometry [35], (ii) molecular genetic approaches, including gene expression, microRNA, single-nucleotide polymorphism (SNP) as well as array comparative genomic hybridisation (array-CGH) pro ling analyses [36,37] and (iii) metabolomic approach [38,39]. Currently, however, these diagnostic tools are generally restricted to the research setting and are not part of the routine clinical practice.
In this study, a semi-quantitative immunohistochemical analysis of the mitochondrial content did not demonstrate any signi cant differences between Sestamibi-positive and Sestamibi-negative renal tumour subsets, possibly attributed either to the functional status of the mitochondria or to the intratumoral heterogeneity, given the limited tumour sampling. Despite increased mitochondrial density documented in various oncocytic tumours, electron microscopy assessment has shown morphological differences in the degree of mitochondrial pleomorphism and/or structure of cristae, in parallel with distinct genomic and metabolomic alterations [40]. Of interest is that Rowe et al. demonstrated a marginal association of mitochondrial staining with 99m Tc-sestamibi uptake ratio, which was improved by incorporation of multidrug resistance pump expression via subtractive normalisation [41].

Competing interests
The authors declare no competing interests.

Ethics approval and consent to participate
This study was approved by the Regional Ethical Review Board and by the local Radiation Safety Committee (Reference no. 2015/923-31/4). Written informed consent was required from all study participants.

Funding
This study was nancially supported by Sweden's innovation agency VINNOVA as part of the Molecular Imaging for Differentiation of Oncocytomas from Renal cancer (MIDOR) study (Reference nr. 2015-0180).
Hermes Medical Solutions AB Stockholm, Sweden, provided the software used for the evaluation of SPECT/CT images.
Email addresses to co-authors: