Post-radioembolization yttrium-90 PET/CT - part 1: diagnostic reporting

Background Yttrium-90 (90Y) positron emission tomography with integrated computed tomography (PET/CT) represents a technological leap from 90Y bremsstrahlung single-photon emission computed tomography with integrated computed tomography (SPECT/CT) by coincidence imaging of low abundance internal pair production. Encouraged by favorable early experiences, we implemented post-radioembolization 90Y PET/CT as an adjunct to 90Y bremsstrahlung SPECT/CT in diagnostic reporting. Methods This is a retrospective review of all paired 90Y PET/CT and 90Y bremsstrahlung SPECT/CT scans over a 1-year period. We compared image resolution, ability to confirm technical success, detection of non-target activity, and providing conclusive information about 90Y activity within targeted tumor vascular thrombosis. 90Y resin microspheres were used. 90Y PET/CT was performed on a conventional time-of-flight lutetium-yttrium-oxyorthosilicate scanner with minor modifications to acquisition and reconstruction parameters. Specific findings on 90Y PET/CT were corroborated by 90Y bremsstrahlung SPECT/CT, 99mTc macroaggregated albumin SPECT/CT, follow-up diagnostic imaging or review of clinical records. Results Diagnostic reporting recommendations were developed from our collective experience across 44 paired scans. Emphasis on the continuity of care improved overall diagnostic accuracy and reporting confidence of the operator. With proper technique, the presence of background noise did not pose a problem for diagnostic reporting. A counter-intuitive but effective technique of detecting non-target activity is proposed, based on the pattern of activity and its relation to underlying anatomy, instead of its visual intensity. In a sub-analysis of 23 patients with a median follow-up of 5.4 months, 90Y PET/CT consistently outperformed 90Y bremsstrahlung SPECT/CT in all aspects of qualitative analysis, including assessment for non-target activity and tumor vascular thrombosis. Parts of viscera closely adjacent to the liver remain challenging for non-target activity detection, compounded by a tendency for mis-registration. Conclusions Adherence to proper diagnostic reporting technique and emphasis on continuity of care are vital to the clinical utility of post-radioembolization 90Y PET/CT. 90Y PET/CT is superior to 90Y bremsstrahlung SPECT/CT for the assessment of target and non-target activity.


Background
The science of yttrium-90 ( 90 Y) radioembolization is rapidly evolving. In suitable patients, 90 Y radioembolization is an effective treatment modality for inoperable liver malignancies, and its expansion into organs other than the liver has been described [1][2][3][4][5]. Contemporary techniques and clinical outcomes are well described in recent literature [6][7][8][9][10][11]. 90 Y radioembolization is complex and requires close multidisciplinary coordination across all stages of its workflow, described as a continuum of care [7]. 90 Y radioembolization is brachytherapy delivered by 90 Y-labelled microspheres implanted permanently in the target vascular bed following intra-arterial injection. The microsphere biodistribution within the target arterial territory is dependent on the locoregional flow environment distal to the point of injection. This in turn is influenced by a myriad of inter-related biophysical variables such as the catheter tip cross-sectional spatial location, injection rate and timing interval, proximity to branching daughter vessels, extent of shunting, cardiovascular status, particle load and microembolization [12][13][14]. Such complexity means that technical success should not be assumed without some form of microsphere imaging. A technically unsuccessful 90 Y radioembolization may have prognostic implications for tumor response or non-target radiation injury.
In recent years, 90 Y bremsstrahlung single-photon emission computed tomography with integrated computed tomography (SPECT/CT) has been the modality-of-choice for post-radioembolization microsphere imaging, for the lack of a better alternative [8]. 90 Y bremsstrahlung scintigraphy suffers from a low spatial resolution, ranging between 11 to 15 mm depending on the choice of energy window, collimator, and image processing [15][16][17]. This is due to its inherent technical limitation as a form of indirect imaging of a continuous scatter radiation spectrum without a pronounced photopeak, resulting in a coarse representation of the microsphere biodistribution. 90 Y bremsstrahlung activity on SPECT/CT often appears smooth and diffuse, with ill-defined margins which are difficult to distinguish from closely adjacent activity foci. As a result, assessment of activity within subcentimeter tumors or tumor vascular thrombosis are often suboptimal on 90 Y bremsstrahlung SPECT/CT. Quantitatively, 90 Y bremsstrahlung scintigraphy is largely inaccurate despite compensation techniques for attenuation, scatter, and collimator-detector response, making it unsuitable for dose-response analysis [18][19][20].
Positron emission tomography with integrated computed tomography (PET/CT) represents a technological leap from 90 Y bremsstrahlung SPECT/CT, both qualitatively and quantitatively, allowing direct imaging of 90 Y microspheres. Although 90 Y is predominantly a β − emitter, coincidence imaging of 90 Y is possible because of a minor decay branch to the 0 + first excited state of zirconium-90 followed by β − β + internal pair production at a very low branching ratio of 31.86 ± 0.47 × 10 −6 [21][22][23][24]. However, the clinical application of 90 Y PET/CT has been fraught with challenges. Today's time-of-flight PET/CT scanners use lutetium-based crystals due to its superior timing resolution. Such crystals, such as lutetiumyttrium-oxyorthosilicate (LYSO), contain 2.6% of naturally occurring lutetium-176 ( 176 Lu) which results in a background activity of intrinsic true and random coincidences which cannot be distinguished from coincidences by an external radioactive source [22,25]. This issue was previously thought to be clinically irrelevant until recent interest in 90 Y PET/CT. In addition, the low positron activity of 90 Y and the intrinsic activity of 176 Lu can have an impact on scatter correction. For time-of-flight PET, several authors have computed the scatter based on the PET activity [26][27][28]. Since the scatter correction algorithm depends on the PET activity map, which is poor due to low signal and intrinsic noise, scatter correction will be negatively affected. The unfavorable combination of intrinsic background activity and a very low positron fraction results in reconstructed 90 Y PET/CT images which at the outset, seem uninterpretable due to high background noise.
Despite this, several institutions have published encouraging early experiences using 90 Y time-of-flight PET for post-radioembolization imaging [29][30][31]. These early reports suggest that despite background noise, it is possible to obtain high-resolution images of microsphere biodistribution using conventional time-of-flight PET scanners, with only minor adjustments to scan technique and image reconstruction such as PET acquisition time, number of bed positions, reconstruction iterations, and subsets. 90 Y phantom studies of lutetiumbased time-of-flight PET/CT systems found the spatial resolution to be 9.3 mm, and the minimum detectable 90 Y radioconcentration of a 10-mm phantom sphere was approximately 1 MBq/ml [23,32]. Count rate linearity was maintained when tested up to 4.7 GBq, and detector saturation from 90 Y bremsstrahlung radiation was not an issue [22,23,[32][33][34].
For tumor dose quantification, several phantom studies using various protocols on lutetium-based time-of-flight PET/CT systems have recently been published. Carlier et al. found activity quantification of phantom spheres to be feasible at 90 Y radioconcentrations greater than 2 MBq/ml, with recovery coefficients of approximately 0.6 and 0.7 for spheres of diameters 17 and 28 mm respectively [32]. Willowson et al. showed a phantom sphere of 90 Y radioconcentration 3.9 MBq/ml and diameter 37 mm to have a recovery coefficient of approximately 0.8, and a trend was observed for recovery coefficients to approach 1 with increasing sphere size [33]. Goedicke et al. achieved a mean quantitative error of +2.4% ± 3.4% standard deviation in a 350 ml phantom, and 90 Y radioconcentrations greater than 2.9 MBq/ml had quantitative errors less than 2% [34]. Elschot et al. found the absorbed dose quantitative error of a 37-mm diameter phantom sphere of 90 Y radioconcentration 2.4 MBq/ml to be −11% uncorrected and +6% corrected for partial volume effects [35]. These phantom studies suggest that tumor dose quantification by 90 Y PET may be feasible, and complements clinical reports of 90 Y PET tumor dose quantification correlating well with treatment response [24,36,37].
On the basis of superior image resolution, quantitative capability, and encouraged by favorable early experiences [29][30][31], we recently introduced post-radioembolization 90 Y PET/CT as an adjunct to 90 Y bremsstrahlung SPECT/ CT in diagnostic reporting. This is a two-part retrospective report reviewing our experience over a 1-year period: Part 1 focuses on qualitative analysis and diagnostic reporting; Part 2 analyzes tissue dose-response by 90 Y PET quantification and evaluates the accuracy of tumor predictive dosimetry [38].

Methods
Institutional review board approval was obtained for the conduct and publication of this retrospective report (CIRB 2010/781/C, SingHealth, Singapore), and all research is in compliance with the Helsinki Declaration. We recently introduced 90 Y PET/CT as an adjunct to 90 Y bremsstrahlung SPECT/CT for high resolution imaging of microsphere biodistribution. Pairing 90 Y PET/CT with 90 Y bremsstrahlung SPECT/CT enabled the trouble shooting of indeterminate findings on either modality, and facilitated the learning of our nuclear medicine physicians in the diagnostic reporting of 90 Y PET/CT. As there were no 90 Y PET/CT protocols specific to our scanner make and model in prior literature, our early scan protocols were based on publications of other scanner types [22,29]. In the first few months after implementation, our 90 Y PET/CT protocol was continually refined by empirical adjustment of acquisition and reconstruction parameters, with a primary emphasis on visual image quality. The aim was to qualitatively achieve a balance between image resolution and background noise, of sufficient clarity for diagnostic reporting purposes. Quantitative accuracy was not a consideration during this early optimization phase.
Our PET/CT is the GE Discovery 690 (General Electric Medical Systems, GEMS, Milwaukee, WI, USA). It uses cerium-activated LYSO crystals, and its image reconstruction is a fully three-dimensional (3D) ordered subset expectation maximization (3D-OSEM) algorithm which incorporates time-of-flight and point spread function information (GEMS name: VUE-point FX; Sharp-IR) [25]. The resulting image is corrected for attenuation, scatter, randoms, dead time, and normalization [25]. 90 Y PET was acquired using positron fraction 3.186 × 10 −5 , half-life 64.1 h, 15 min per bed position, 1 to 2 bed positions from the diaphragm downwards to cover the entire liver, reconstructed with 3 iterations and 18 subsets. Our full Table 1 Recommendations for diagnostic reporting of post-radioembolization 90 Y PET/CT

Item Recommendation
Continuity of care Post-radioembolization 90 Y PET/CT is best reported by the same attending nuclear medicine physician who has followed through the entire planning-therapy continuum from exploratory angiography, predictive dosimetry, to 90 Y radioembolization.
PET display threshold setting for non-target 90 Y activity detection For detection of non-target 90 Y activity, the operator should actively adjust the upper PET visual display threshold setting to deliberately increase the background noise to moderate levels.
Criteria for technical success 1. 90 Y activity present in the majority of targeted tumors, or good overall activity coverage of large targeted tumors; and 2. The absence of clinically significant non-target 90 Y activity; and 3. All findings are in keeping with pre-therapy radiation planning expectations. protocol is presented in Additional file 1. 90 Y bremsstrahlung SPECT/CT was performed on a Philips Precedence (Philips, Amsterdam, The Netherlands), which has a dualhead gamma camera and 16-slice CT. A medium energy general purpose collimator was used. Based on our phantom studies, we centered our energy window at 80% ± 15% keV to capture the most prominent photopeak amidst a continuous bremsstrahlung spectrum. Scans were acquired in 128 frames over 360°, 20 s per frame; 3D-OSEM reconstruction using Philips Astonish software; 3 iterations and 8 subsets; matrix size 128 × 128; low-dose CT for anatomical correlation and attenuation correction. All paired 90 Y PET/CT and 90 Y bremsstrahlung SPECT/CT scans were performed within 1 h of each other. Diagnostic reporting for all paired scans were done in consensus by an experienced team of nuclear medicine physicians with cumulative experience of more than 200 90 Y radioembolization cases over 5 years. As no prior guidelines existed in literature for the diagnostic reporting of post-radioembolization 90 Y PET/CT, a gestalt interpretation was applied to all paired scans based on a thorough knowledge of the case-specific treatment intent, angiographic and dosimetric factors. 90 Y resin microspheres (SIR-Spheres, Sirtex Medical Limited, New South Wales, Australia) were used. Our predictive dosimetric protocol was previously described, which involves catheter-directed CT angiography for arterial territory delineation, quantitative 99m Tc macroaggregated albumin (MAA) SPECT/CT of the target organ, and Medical Internal Radiation Dose (MIRD) macrodosimetry to achieve artery-specific personalized radiation planning [11]. As our treatment paradigm deviates considerably from conventional semi-empirical methods, our nuclear medicine team emphasizes continuity of care and an in-depth knowledge of case-specific angiography, in close collaboration with our interventional radiologists.
One of the qualitative study objectives was to compare the ability of 90 Y PET/CT versus 90 Y bremsstrahlung SPECT/CT in providing conclusive information about the presence or absence of 90 Y activity within targeted tumor vascular thrombosis. We chose to evaluate tumor vascular thrombosis because these are tumorous lesions within small-to-medium calibre vessels which carry prognostic significance [9]. Another qualitative study objective was to compare the ability of 90 Y PET/CT versus 90 Y bremsstrahlung SPECT/CT in detecting non-target activity.
Where these were appropriate, specific findings on 90 Y PET/CT were corroborated by 90 Y bremsstrahlung SPECT/CT, 99m Tc MAA SPECT/CT, follow-up diagnostic sectional imaging (e.g. CT or MRI), or review of clinical records for symptoms or endoscopy. Clinical toxicities attributable to 90 Y radioembolization were graded according to the Common Terminology Criteria for Adverse Events version 4.03 (CTCAE; National Institutes of Health, National Cancer Institute, USA). A retrospective review of clinical records was performed for all patients to exclude any clinically significant non-target activity undetected by both imaging modalities. The accuracies of both modalities were calculated by binary classification testing.
Our methods for the diagnostic reporting of target and non-target activity are presented in Table 1. In the absence of established diagnostic reporting guidelines, our methods were heuristically formulated based on rational angiographic and dosimetric principles from our collective experience across all paired scans. Using these guidelines, we were able to overcome the problem of background noise and other general issues related to the complexity of 90 Y radioembolization. Thus, the presence of background noise did not pose a problem for diagnostic reporting. In Table 1, we propose an unconventional technique for non-target activity detection, which counter-intuitively requires an increase in noise levels by adjusting the visual display threshold, but was effective in overcoming the problem of background noise. As outlined in Table 1, operator adjustment of the PET visual display threshold was a key step to qualitatively distinguish between signal and noise for diagnostic reporting. The lower PET visual display threshold was set to zero for both target and non-target activity detection. For target activity detection, the upper PET visual display threshold setting was adjusted to reduce the background noise to a minimum. On our system, we used the following upper PET visual display threshold settings for target activity detection: median 53% (7,000 kBq/ml), mean 54% (6,900 kBq/ml), 95% confidence interval (CI) 47% to 61% (5,900 to 7,900 kBq/ml).
For non-target activity detection, the upper PET visual display threshold setting was adjusted to deliberately increase the background noise to moderate levels. This counter-intuitive technique was necessary because genuine trace non-target activity may appear visually less intense than noise spikes (see "Discussion"). Therefore, if the upper PET visual display threshold had remained at the settings used to suppress background noise for target activity assessment, it will be unlikely for visually subtle, trace nontarget activity to be detected by the operator. Albeit subjective, this method improved the detection of non-target activity with appropriate diagnostic technique (Table 1). On our system, we used the following upper PET visual display threshold settings for non-target activity detection: median 14% (2,000 kBq/ml), mean 22% (2,200 kBq/ml), 95% CI 4 to 41% (1,200 to 3,100 kBq/ml). Non-target activity was sometimes more easily detected on the rotating maximum intensity projection (MIP) than on the sectional PET or PET/CT images. On the rotating MIP, any activity in a non-random pattern protruding from the smooth outline of targeted tissue amidst background noise was regarded as suspicious and pursued on sectional PET and PET/CT images. The above data is tabulated in Additional file 1: Table S1.

Terms and definitions
The following terms and definitions are used throughout our two-part report. Predictive dosimetry refers to personalized radiation planning guided by modern (d) 90 Y bremsstrahlung SPECT/CT was barely able to detect subtle activity in the larger (arrow) of the four small tumors, but was unable to confirm any focal activity within the other three small tumors due to its low resolution. tomographical techniques of microsphere simulation, where intended absorbed doses are determined a priori based on case-specific clinical, angiographic and dosimetric parameters, in accordance with the achievable intent [11]. Target refers to the arterial territory intended for 90 Y radioembolization, represented by planning target volumes used in artery-specific predictive dosimetry [11]. Non-target refers to unintended 90 Y activity beyond planning target volumes, e.g., stomach and duodenum, which also includes untargeted non-tumorous liver due to microsphere reflux or arterio-portal shunting. Technical success refers to the qualitative assessment of a satisfactory 90 Y activity biodistribution in accordance with radiation planning expectations as seen on post-radioembolization 90 Y PET/CT, not to be confused with 'clinical success' [8]. Our definition of technical success is slightly dissimilar to conventional reporting standards [8] because angiographic appearances alone at the conclusion of 90 Y radioembolization may not necessarily reflect the actual microsphere biodistribution [12,14]. The planning-therapy continuum for 90 Y radioembolization at our institution is a seamless, integrated workflow for nuclear medicine physicians with emphasis on continuity of care, a multi-disciplinary approach, open and continuous interdisciplinary communication, and predictive dosimetry.

Results
A total of 63 90 Y radioembolization procedures were performed over a 1-year period. Of these, 44 (70%) paired 90 Y PET/CT and 90 Y bremsstrahlung SPECT/CT scans were performed. Twenty-one of 44 paired scans were performed using various 90 Y PET/CT protocols during our early optimization phase. After finalization of our scan protocol, a further 23 paired scans were performed. For data consistency, all case examples, statistical and dose-response results presented in this two-part report were limited to the latter 23 patients only.
Of these 23 patients, 19 (83%) had hepatocellular carcinomas (HCC), 2 had intra-hepatic cholangiocarcinomas, 1 had liver metastases from pancreatic neuroendocrine cancer, and 1 had a bulky right adrenal gland metastasis from gastrointestinal stromal tumor (GIST) -a case of 90 Y radioembolization of an organ other than the liver. Median age was 64.5 years (mean 63.5 years; range 40 to 77 years); there were 16 (70%) male patients. Tumor vascular thrombosis was present in seven (30%) patients: portal vein in six patients and inferior vena cava in one; and all were HCCs. The median hepatopulmonary shunt fraction estimated by 99m Tc MAA planar scintigraphy was 5.8% (mean 6.6%; range 1.5% to 24.2%).
A total of 39 arterial territories were treated across 23 patients (mean 2; range 1 to 4), which included the right inferior phrenic, right internal mammary, and right inferior adrenal arteries. The median injected 90 Y activity was 2.2 GBq (mean 2.4 GBq; range 1.1 to 5.0 GBq), and the median time interval from 90 Y radioembolization to 90 Y PET/CT was 21.3 h (mean 19.2 h; range 4.7 to 24.0 h). One patient was deemed technically unsuccessful due to non-target activity in the gastric and duodenal walls resulting in CTCAE Grade 3 toxicity; all others were technically successful. All the above data are summarized in Table 2 and presented in full detail in Additional file 1: Tables S2 and S3. 90 Y PET/CT outperformed 90 Y bremsstrahlung SPECT/ CT in all aspects of qualitative analysis. Its image resolution was consistently superior to 90 Y bremsstrahlung SPECT/CT (Figures 1 and 2), which also improved operator confidence in diagnostic reporting. Among 23 patients, 7 had tumor vascular thrombosis seen on diagnostic sectional imaging (Additional file 1: Table S4). 90 Y PET/CT was able to provide conclusive information about the presence or absence of activity within targeted tumor vascular thrombosis in all seven patients (Figure 3), whereas 90 Y bremsstrahlung SPECT/CT was conclusive in only three of seven patients; the remaining four patients were indeterminate due to its low resolution. In two patients without visually detectable activity on 90 Y PET/CT within targeted tumor vascular thrombosis, follow-up scans confirmed disease progression in these lesions ( Figure 4). Therefore, within the statistical limitations of our small dataset, 90 Y PET/CT had a 100% accuracy for the provision of conclusive information about the presence or absence of activity within targeted tumor vascular thrombosis, compared to only 30% for 90 Y bremsstrahlung SPECT/CT (60% sensitivity and positive predictive value; 0% specificity and negative predictive value). 90 Y PET/CT detected non-target activity in all eight of eight patients, whereas 90 Y bremsstrahlung SPECT/CT detected only four of eight patients initially -a further three were seen only with hindsight of 90 Y PET/CT findings (Additional file 1: Table S5). These eight cases include the stomach ( Figure 5), duodenum, chest wall ( Figure 6), kidney (Figure 7), untargeted liver, and gallbladder. All 23 patients underwent a retrospective review of clinical records to exclude any clinically significant non-target activity undetected by both imaging modalities. Three of 23 patients were nonresidents who were discharged back to their home countries after 90 Y radioembolization and were lost to followup. Of the remaining 20 patients, there were no cases of toxicity greater than or equal to CTCAE Grade 3 attributable to undetected non-target activity in extra-hepatic viscera at a median follow-up of 5.4 months (mean 6.3 months, 95% CI 4.5 to 8.1 months). Therefore, within the statistical limitations of our small dataset, 90 Y PET/CT had a 100% accuracy for non-target activity detection, compared to 80% for 90 Y bremsstrahlung SPECT/CT (50% sensitivity; 100% specificity and positive predictive value; 75% negative predictive value).  Y radioembolization is complex, and continuity of care is paramount to the clinical utility of 90 Y PET/CT. This is because case-specific issues occur throughout the planning- (d) MIP of 90 Y bremsstrahlung SPECT shows concordant non-target activity, but of lower intensity and resolution. By quantification 90 Y PET activity, detailed in Part 2 [38], the non-target mean absorbed doses to the gastric greater curve, gastric outlet, and proximal duodenum were approximately 49, 65, and 53 Gy, respectively. Within weeks, the patient developed persistent epigastric pain. (e) Gastroscopy at 3.3 months revealed extensive inflammation along the stomach and proximal duodenum. A Forrest III ulcer was present in the pylorus (large arrow). The ulcer edge (small arrows) was inflamed and indurated. (f) Photomicrograph (H&E stain; ×200) of the gastric biopsy showed edema, congestion, and diffuse lymphoplasmacytic, eosinophilic, and neutrophilic infiltrates. A purple-staining resin microsphere (arrow) surrounded by a halo was seen in the lamina propria. The surrounding gastric glands show flattening of the epithelium, dilatation, and reactive atypia, consistent with severe cellular injury. therapy continuum which can affect the final microsphere biodistribution, e.g., selective sparing of liver segments, vascular stasis, risk of microsphere reflux, vessels-at-risk, ontable changes to the intended vascular approach. The same attending nuclear medicine physician who follows through the entire planning-therapy continuum already has in mind an expected activity biodistribution when reviewing the 90 Y PET/CT. This case-specific knowledge improves overall diagnostic accuracy and operator confidence, and leads to clinically meaningful reporting.
Despite the superior image resolution of 90 Y PET/CT over 90 Y bremsstrahlung SPECT/CT, parts of extrahepatic viscera which lie closely adjacent to the liver still remain blind spots for non-target activity detection, e.g., gallbladder fundus, gastric lesser curve, pylorus, and proximal duodenum. These often appear anatomically inseparable from the adjacent liver on 90 Y PET/CT, making assessment for non-target activity in these areas very challenging. This issue is further compounded by varying degrees of PET/CT mis-registration due to the relatively (e, f) 90 Y bremsstrahlung SPECT/CT was unable to detect any non-target activity in the chest wall due to its low image resolution.  9. 90 Y radioembolization of an organ other than the liver. KIT-negative GIST with bulky metastasis to the right adrenal gland, refractory to tyrosine kinase inhibitors. (a) Pre-radioembolization digital subtraction angiogram (DSA) with catheter tip in the right renal artery 'R' demonstrates the origin of the right inferior adrenal artery 'A' and the arterial trees of the adrenal tumor and right kidney. The catheter tip was advanced deep into the right inferior adrenal artery for 90 Y radioembolization. (b) Post-radioembolization DSA with no change in catheter tip position demonstrates significant vascular stasis and reflux of contrast down the right inferior adrenal artery and distally into the terminal branches of the right renal artery (arrows). (c, d) 90 Y PET/CT depicts in high resolution, non-target activity conforming to the anatomy of the right renal cortex (arrows). (e, f) On 90 Y bremsstrahlung SPECT/CT, non-target activity was indistinguishable from adjacent activity bloom in the coronal plane, but seen as low-grade, diffuse activity in the saggital plane (arrows). long 90 Y PET acquisition time. However, these issues similarly affect 90 Y bremsstrahlung SPECT/CT and should therefore not be viewed as a comparative disadvantage.
Despite background noise, 90 Y PET/CT had a higher detection rate for non-target activity than 90 Y bremsstrahlung SPECT/CT. Our counter-intuitive technique of deliberately increasing background noise levels has proven to be effective. Its rationale lies in our observation that noise spikes may be of higher visual intensity than nontarget activity, as the latter depends on the quantity of shunted microspheres and its 90 Y radioconcentration within the non-target arterial territory. This means that visually subtle but genuine trace non-target activity may appear less intense than background noise. Quantitative reconstruction methods proposed by Carlier et al. to distinguish signal from noise were not investigated in our study [32]. Therefore, from a qualitative perspective, a diagnosis of non-target activity should not be made on the basis of its visual intensity, but on its pattern and whether it conforms to underlying anatomy. The presence of a plausible vascular etiology will greatly lend support to the diagnosis, although this is not an essential criteria because it may not always be possible to identify a culprit vessel upon retrospective review of angiography.
The presence or absence of correlative clinical signs or symptoms do not feature in our criteria for non-target activity because clinical sequelae is a quantitative function of dose-response radiobiology over time, with no bearing on the qualitative presence of non-target activity at the time of scan. For example, it is common to detect subtle non-target activity in an untargeted liver lobe due to minor microsphere reflux or arterio-portal shunting, with no clinically relevant sequelae. Another example is the presence of mild non-target activity in the gallbladder wall, without ensuing clinical cholecystitis. It follows that the absence of correlative clinical signs or symptoms do not invalidate a 90 Y PET/CT diagnosis of non-target activity.
The detection of non-target activity should immediately be followed by an assessment of the risk of developing clinically significant radiation toxicity. Accurate risk assessment facilitates appropriate action to be undertaken early to mitigate potential radiation injury, even if the patient is asymptomatic at the time of scan. This should be based on 90 Y PET quantification of the non-target tissue absorbed dose, which is superior to subjective visual assessment alone. Exceptions are cases of visually subtle, trace non-target activity where the absorbed doses are unlikely to be clinically relevant. Similarly, it may sometimes be difficult to distinguish noise spikes from genuine 90 Y activity. However, such indeterminate activity foci are usually too mild to result in any clinically relevant toxicity even if genuine, and therefore do not impact post-radioembolization management. The topic of 90 Y PET quantification and its clinical applications in tissue dose-response assessment and tumor predictive dosimetry are presented in part 2 [38].