Open Collections

UBC Faculty Research and Publications

Accuracy and reproducibility of simplified QSPECT dosimetry for personalized 177Lu-octreotate PRRT Del Prete, Michela; Arsenault, Frédéric; Saighi, Nassim; Zhao, Wei; Buteau, François-Alexandre; Celler, Anna; Beauregard, Jean-Mathieu Oct 15, 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-40658_2018_Article_224.pdf [ 2.46MB ]
JSON: 52383-1.0372977.json
JSON-LD: 52383-1.0372977-ld.json
RDF/XML (Pretty): 52383-1.0372977-rdf.xml
RDF/JSON: 52383-1.0372977-rdf.json
Turtle: 52383-1.0372977-turtle.txt
N-Triples: 52383-1.0372977-rdf-ntriples.txt
Original Record: 52383-1.0372977-source.json
Full Text

Full Text

ORIGINAL RESEARCH Open AccessAccuracy and reproducibility of simplifiedQSPECT dosimetry for personalized 177Lu-octreotate PRRTMichela Del Prete1,2, Frédéric Arsenault1,2, Nassim Saighi1,2, Wei Zhao3,4, François-Alexandre Buteau1,2,Anna Celler3,4 and Jean-Mathieu Beauregard1,2** Correspondence: jean-mathieu.beauregard@chudequebec.ca1Department of Radiology andNuclear Medicine and CancerResearch Center, Université Laval,Quebec City, Canada2Department of Medical Imagingand Oncology Branch of CHU deQuébec Research Center, CHU deQuébec – Université Laval, 11 côtedu Palais, Quebec City, QC G1R 2J6,CanadaFull list of author information isavailable at the end of the articleAbstractBackground: Routine dosimetry is essential for personalized 177Lu-octreotatepeptide receptor radionuclide therapy (PRRT) of neuroendocrine tumors (NETs),but practical and robust dosimetry methods are needed for wide clinicaladoption. The aim of this study was to assess the accuracy and inter-observerreproducibility of simplified dosimetry protocols based on quantitative single-photonemission computed tomography (QSPECT) with a limited number of scanning timepoints. We also updated our personalized injected activity (IA) prescription scheme.Methods: Seventy-nine NET patients receiving 177Lu-octreotate therapy (with a total of279 therapy cycles) were included in our study. Three-time-point (3TP; days 0, 1, and 3)QSPECT scanning was performed following each therapy administration. Dosimetry wasobtained using small volumes of interest activity concentration sampling for the kidney,the bone marrow and the tumor having the most intense uptake. Accuracy of thesimplified dosimetry based on two-time-point (2TP; days 1 and 3, monoexponential fit)or a single-time-point (1TPD3; day 3) scanning was assessed, as well as that of hybridmethods based on 2TP for the first cycle and 1TP (day 1 or 3; 2TP/1TPD1 and 2TP/1TPD3,respectively) or no imaging at all (based on IA only; 2TP/no imaging (NI)) for thesubsequent induction cycles. The inter-observer agreement was evaluated for the3TP, 2TP, and hybrid 2TP/1TPD3 methods using a subset of 60 induction cycles(15 patients). The estimated glomerular filtration rate (eGFR), body size descriptors(weight, body surface area (BSA), lean body weight (LBW)), and products of both wereassessed for their ability to predict IA per renal absorbed dose at the first cycle.(Continued on next page)EJNMMI Physics© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 InternationalLicense (, which permits unrestricted use, distribution, and reproduction in any medium,provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, andindicate if changes were made.Del Prete et al. EJNMMI Physics  (2018) 5:25 from previous page)Results: The 2TP dosimetry estimates correlated highly with those from the 3TP datafor all tissues (Spearman r > 0.99, P < 0.0001) with small relative errors between themethods, particularly for the kidney and the tumor, with median relative errors notexceeding 2% and interdecile ranges spanning over less than 6% and 4%, respectively,for the per-cycle and cumulative estimates. For the bone marrow, the errors wereslightly greater (median errors < 6%, interdecile ranges < 14%). Overall, the strength ofcorrelations of the absorbed dose estimates from the simplified methods with thosefrom the 3TP scans tended to progressively decrease, and the relative errors to increase,in the following order: 2TP, 2TP/1TPD3, 1TPD3, 2TP/1TPD1, and 2TP/NI. For the tumor, the2TP/NI scenario was highly inaccurate due to the interference of the therapeutic response.There was an excellent inter-observer agreement between the three observers, inparticular for the renal absorbed dose estimated using the 3TP and 2TP methods,with mean errors lesser than 1% and standard deviations of 5% or lower. TheeGFR · LBW and eGFR · BSA products best predicted the ratio of IA to the renaldose (GBq/Gy) for the first cycle (Spearman r = 0.41 and 0.39, respectively; P < 0.001). Forthe first cycle, the personalized IA proportional to eGFR · LBW or eGFR · BSA decreasedthe range of delivered renal absorbed dose between patients as compared with thefixed IA. For the subsequent cycles, the optimal personalized IA could be determinedbased on the prior cycle renal GBq/Gy with an error of less than 21% in 90% of patients.Conclusions: A simplified dosimetry protocol based on two-time-point QSPECTscanning on days 1 and 3 post-PRRT provides reproducible and more accuratedose estimates than the techniques relying on a single time point for non-initialor all cycles and results in limited patient inconvenience as compared to protocolsinvolving scanning at later time points. Renal absorbed dose over the 4-cycle inductionPRRT course can be standardized by personalizing IA based on the product of eGFRwith LBW or BSA for the first cycle and on prior renal dosimetry for the subsequentcycles.Keywords: Dosimetry, Neuroendocrine tumors, Peptide receptor radionuclidetherapy, Personalized, Quantitative SPECTBackgroundFor patients with metastatic neuroendocrine tumors (NETs), peptide receptor radio-nuclide therapy (PRRT) with 177Lu-octreotate is an effective palliative treatment thatrarely causes serious toxicity [1, 2]. PRRT has been mostly administered as a 4-cycle in-duction course using a fixed injected activity (IA) of not more than 7.4 GBq per cycle,in order to not exceed cumulative absorbed doses of 23 Gy to the kidney and 2 Gy tothe bone marrow (BM) in the majority of patients [1–4]. However, it is well known thatfor these critical organs, and in particular for the kidney which is the dose-limitingorgan for most patients, the absorbed dose per IA is highly variable and usually lowerthan 23 Gy per 4 cycles, resulting in most patients being undertreated with such anempiric PRRT regime [5, 6]. We and others have proposed personalized PRRT(P-PRRT) protocols in which the number of fixed IA cycles or the IA per cycle aremodulated to deliver a safe prescribed renal absorbed dose, with the aim to maximizetumor irradiation while keeping the toxicity low [4, 6]. Such P-PRRT protocols requirecareful dosimetry monitoring, which is often perceived as a complex andresource-consuming process, therefore constituting a barrier for wide clinical adoption.As a result, the clinical practice of “one-size-fits-all” PRRT prevails, at the potential costof delivering a suboptimal treatment to most patients.Del Prete et al. EJNMMI Physics  (2018) 5:25 Page 2 of 20We have been routinely performing post-PRRT dosimetry using quantitativesingle-photon emission computed tomography (QSPECT) combined with thesmall-sphere volume of interest (VOI) activity concentration sampling [5, 6]. Aiming tosimplify the dosimetry process and to reduce the clinical burden thereof, we examinedthe impact of reducing the number of QSPECT sessions on the accuracy and theinter-observer reproducibility of the resulting dose estimates. In parallel, based on alarge dataset from our growing cohort of patients treated with PRRT, we updated ourpersonalized IA determination scheme.MethodsPatients and PRRT cyclesFrom November 2012 to December 2017, 81 patients with progressive metastatic and/or symptomatic NET were treated with PRRT at CHU de Québec—Université Laval.Two patients who underwent only 1 cycle were each excluded because of their incom-plete dosimetric data, and therefore, only data from 79 patients was analyzed. This in-cludes 23 patients who received only empiric PRRT (i.e., fixed IA of approximately8 GBq, occasionally reduced) until March 2016, for whom the requirement for consentwas waived due to the retrospective nature of the analysis. All other patients were en-rolled in our P-PRRT trial (NCT02754297) and gave informed consent to participate(protocol described in [6]). Patient characteristics are reported in Table 1.Two hundred and eighty-four therapy cycles were administered during the studyperiod. Five cycles in five patients were excluded from the analysis because of dosim-etry protocol deviation or missing data. Among the 279 therapy cycles analyzed, 142were empiric (median IA = 7.6 GBq; range, 3.8–9.1 GBq) and 137 were personalized(median IA = 9.0 GBq; range, 0.7–32.4 GBq). Anti-nausea premedication (ondansetronand dexamethasone) and a nephroprotective amino acid solution (lysine and arginine)were administered [1]. We administer a 4-cycle induction course for which the pre-scribed cumulative renal dose is 23 Gy (5 Gy at the first cycle; two-monthly intervals)and, in responders only, we offer consolidation, maintenance, and/or salvage cycles(prescribed renal dose of 6 Gy each; personalized intervals). As previously described,prescribed renal absorbed radiation doses were reduced in patients with renal or bonemarrow impairment [6].Reference dosimetry methodAt each cycle, after therapeutic administration of 177Lu-octreotate, QSPECT/computedtomography (QSPECT/CT) scans were performed at approximately 4 h (day 0), 24 h(day 1), and 72 h (day 3) using a Symbia T6 camera (Siemens Healthcare, Erlangen,Germany) (Fig. 1) [6, 7]. Following the same data processing as described in [7], thedead-time corrected reconstructed images were converted into the positron emissiontomography (PET) DICOM format, which includes a “rescale slope” parameter thatconverts count data into Bq/mL and also enables display of QSPECT images in stan-dardized uptake values normalized for body weight (standardized uptake value; SUV).These three imaging time points were initially selected for the following practicalreasons: (1) the day 0 scan does not incur any additional hospital visit for thepatient and allows capturing early kinetics of the radiopharmaceutical and (2)Del Prete et al. EJNMMI Physics  (2018) 5:25 Page 3 of 20performing scans beyond day 3 would not be easy for logistical reasons (PRRT be-ing administered on Tuesday, day 4 or 5 would fall on the weekend) and would in-convenience patients (in particular those out-of-city patients who would need toprolong their stay).Table 1 Patient characteristicsAll patients (n = 79)Gender, n (%)Female 36 (45.6)Male 43 (54.4)Age at first cycle, median (range) 60.7 (26.1–82.3)Site of primary tumor, n (%)Small intestine 30 (38.0)Pancreas 26 (32.9)Adrenal glanda 6 (7.6)Lung 6 (7.6)Colon 2 (2.5)Stomach 1 (1.3)Esthesioneuroblastoma 1 (1.3)Unknown 7 (8.9)Metastases, n (%)Liver 66 (83.5)Lymph nodes 51 (64.6)Bone 29 (36.7)Lung 9 (11.4)Otherb 25 (31.6)Body size descriptors, mean ± SD (range)Weight (Kg) 72.1 ± 16.6 (42.6–121.0)Lean body weight (Kg) 52.4 ± 9.9 (35.4–81.2)Body surface area (m2) 1.8 ± 80.2 (1.4–2.5)eGFR (ml/min/1.73 m2), mean ± SD (range) 86.3 ± 22.2 (42.0–154.1)Number of cycles, n (%)1 8 (10.1)2 6 (7.6)3 16 (20.3)4 38 (48.1)5 3 (3.8)6 6 (7.6)7 1 (1.3)8 1 (1.3)Type of cycles, n (%)Empiric only 23 (29.1)Personalized only 45 (57.0)Mixed 11 (13.9)eGFR estimated glomerular filtration rate, PRRT peptide receptor radionuclide therapyaThree patients with pheochromocytoma and three patients with paragangliomabPeritoneum, ovary, subcutaneous, pleura, meningesDel Prete et al. EJNMMI Physics  (2018) 5:25 Page 4 of 20As in our clinical practice, we routinely performed dosimetry based on the data ac-quired at these three time points (3TP), this approach constituted the reference methodfor the present analysis. In brief, at each time point, we sampled the activity concentra-tion in tissues of interest (Fig. 1), including both kidneys (areas of representative paren-chymal uptake), the BM (L4 and L5 vertebral bodies, or elsewhere when the latter wereobviously affected by metastases), and the tumor having the most intense uptake(Tumormax), using 2-cm (4.2 cm3) spherical VOIs, as previously described [5, 6]. Thiswas performed using either Hybrid Viewer (Hermes Medical Solutions, Stockholm,Sweden) or MIM Encore (MIM Software Inc., Cleveland, OH, USA) software. As previ-ously described in [6], we also computed the total body retention for the purpose ofcomputing the cross-dose component of the BM absorbed dose (BMcross), which weFig. 1 Post-treatment serial QSPECT/CT was performed at (from left to right) 5, 24, and 70 h after a22.0 GBq 177Lu-octreotate administration in a 55-year-old male with metastatic NET of unknown origin.Small volumes of interest (2-cm diameter) were placed over tissues of interest. Left kidney (red arrows), L5bone marrow cavity (orange arrows), and dominant tumor (green arrows) VOIs are pointed on anteriormaximum intensity projections (top row) and selected transaxial fusion slices (mid and bottom rows).QSPECT images are normalized using an upper SUV threshold of 7. During this consolidation cycle, thepersonalized injected activity allowed the delivery of 6.1 Gy (6.0 Gy prescribed) to the kidneyDel Prete et al. EJNMMI Physics  (2018) 5:25 Page 5 of 20added to the self-dose component (BMself ) to estimate the total BM absorbed dose(BMtotal).Based on these 3TP data, trapezoidal-monoexponential (3TPTM) time-activity curves(TACs) were drawn using the following procedure (Fig. 2). For each organ/tumor, aconstant mean SUV was assumed from the time of 177Lu-octreotate injection until thetime of the day 0 scan (approximately 4 h). This was followed by a linear (trapezoid) fitto the SUV corresponding to the day 0 and the day 1 scans. Then, a monoexponentialcurve was fit using the day 1 and day 3 data, resulting in an effective decay model beingused from day 1 onwards (trapezoidal-monoexponential; 3TPTM; Fig. 2). However, incases when the day 3 SUV was higher than that corresponding to day 1, we assumed alinear SUV variation between days 1 and 3, followed by the physical decay of activity(i.e., λbiol = 0, λeff = λphys) from day 3 to infinity (trapezoidal-constant; 3TPTC).Then, the area under each TAC curve was integrated and multiplied by the appropri-ate activity concentration dose factors (ACDF). The values of these factors have beenderived from OLINDA/EXM software data (Vanderbilt University, Nashville, TN,USA), as previously described [6]: 84 mGy · g/MBq/h for Tumormax and 87 mGy · g/MBq/h for kidneys and BMself. For BMcross, we integrated the total body activity overtime and multiplied it by a dose factor of 1.09 × 10−4 mGy/MBq/h for males or 1.29 ×10−4 mGy/MBq/h for females, i.e., to account for their different gamma fraction of en-ergy deposition from the whole body to the BM [6].Simplified dosimetry methodsFrom our experience and as suggested by others, the day 0 data, although it capturesthe rapid kinetics of the radiopharmaceutical (which includes competing accumulationA BC DFig. 2 Time-activity curves (TACs) of the renal (a), tumor (b), and bone marrow (c) activity concentrationsand of the whole-body retention (d) over time for the patient case illustrated in Fig. 1. TACs in MBq/cc orMBq (red) and SUV or percentage of injected activity (%IA) (blue) are illustrated for the three-time-point(3TP; solid lines) and two-time-point (2TP; dashed lines) methodDel Prete et al. EJNMMI Physics  (2018) 5:25 Page 6 of 20and rapid washout), contributes little to the area under the TAC, which is mostly deter-mined by the slow washout kinetics and tends to follow a monoexponential decay be-yond 24 h [8]. Accordingly, we eliminated the day 0 data from all simplified dosimetryapproaches. A total of five methods were investigated, as detailed below.2TP: Two-time-point (2TP) dosimetry estimates were obtained using VOI data fromday 1 to day 3 scans. From time of administration to infinity, monoexponential (2TPM)effective decay was applied, except in cases of biological accumulation of activity, i.e.,when the SUV of the tissue increased between day 1 and day 3. In such cases, we as-sumed a SUV equal to that of day 3 SUV (λbiol = 0), from time of treatment administra-tion to infinity and thus applied only physical decay (λeff = λphys; 2TPC). The 2TPmethod is the combination of 2TPM and 2TPC.1TPD3: As proposed by Hänscheid et al., we estimated doses using asingle-time-point method based on the day 3 data (1TPD3) [8]. In this method, the ac-tivity concentration (MBq/cc) was multiplied by the time at which the day 3 scan wasperformed (h) and by 0.25 Gy · g/MBq/h (based on Eq. 8 in [8]). To compute BMcross,the total whole-body activity (MBq) was multiplied by imaging time (h) and by3.2 × 10−7 Gy/MBq/h for males or 3.7 × 10−7 Gy/MBq/h for females, i.e., the gammafraction of energy deposition from the whole body to the BM, multiplied by0.25 Gy · g/MBq/h (from [8], as above), divided by 87 mGy · g/MBq/h (ACDF of theBM and kidney).2TP/1TPD1: We evaluated a hybrid dosimetry protocol based on the 2TP method forthe first cycle, as described above, and employed a single scan on day 1 for the subse-quent induction cycles. In this scenario, the absorbed dose to a given tissue during thesecond and the subsequent induction cycles was obtained by applying the monoexpo-nential curve corresponding to the effective decay, as determined for this tissue duringthe first cycle, to the activity concentration observed on day 1 of the subsequent cycle.2TP/1TPD3: This is the same as 2TP/1TPD1, but the single scan on subsequent cycleswas that performed on day 3.2TP/NI: Similar to the two previous methods, this method was also based on 2TPscanning for the first cycle, but no imaging (NI) was performed for the subsequent cy-cles. For the latter, the absorbed dose per IA during the subsequent cycles was simplyassumed to be equal to that delivered during the first cycle.Cumulative renal, BMtotal, and Tumormax absorbed doses were compiled for all pa-tients who received three or four induction cycles (n = 65). Per-cycle and cumulativedoses resulting from each of the simplified dosimetry methods were compared withthose obtained using the reference (3TP) method, and relative errors were calculated.Inter-observer variabilityFor 60 induction cycles in 15 patients, the dosimetry analysis was performed independ-ently by three observers having different backgrounds and purposely varied levels of ex-perience in internal dosimetry. Observer 1 (M.D.P.), a certified endocrinologist, currentPRRT Fellow and Ph.D. student, performed 258 of the 279 primary analyses describedin this paper and, as such, accumulated the most experience with this dosimetry pro-cedure. Observer 2 (F.A.) was a certified nuclear medicine physician and current Nu-clear Oncology fellow who performed 21 primary analyses. Observer 3 (N.S.) was anDel Prete et al. EJNMMI Physics  (2018) 5:25 Page 7 of 20M.D. student who was new to both nuclear medicine and dosimetry and who receivedonly a short training. Relative errors of per-cycle and cumulative absorbed doses be-tween each pair of observers were computed for the reference method (3TP) and thetwo most accurate simplified methods.Personalized 177Lu-octreotate activity prescriptionWe previously derived a model based on the body surface area (BSA) and the estimatedglomerular filtration rate (eGFR; according to the CKD-EPI Creatinine Equation [9]) todetermine the personalized 177Lu-octreotate activity to be administered at the first cycle[6]. Using data from our entire cohort of 79 patients, we aimed to formulate a simplerprescription equation. To this end, we correlated the ratio of IA to the renal absorbeddose estimated from the first cycle (GBq/Gy, obtained by the 3TP or the 2TP methods)with the patient’s weight, lean body weight (LBW), BSA, eGFR, and the products ofeGFR with each of the three body size descriptors. Then, for each of these seven corre-lations, we performed a linear regression forced through the origin (eliminating theintercept) and calculated the relative errors of the predicted renal GBq/Gy using theslope of the linear regression.We also compared the accuracy of predicting the renal GBq/Gy in any givennon-initial cycle with that from the previous cycle or with the average renal GBq/Gy ofthe two previous cycles, as we have initially been doing in our P-PRRT trial [6].Statistical methodsData are presented as median and interdecile range or as mean ± SD according to thedata distribution using D’Agostino-Pearson omnibus normality test. Ranges are also re-ported. Pearson or Spearman correlations were used depending on the normality of thedata. A difference was considered as statistically significant if the P value was below0.05. Correlations and linear regressions were performed using GraphPad Prism soft-ware (version 7, GraphPad Software Inc., La Jolla, CA, USA).ResultsAccuracy of simplified dosimetry methodsTissue-specific effective half-lives derived from monoexponential fitting of the activityconcentrations measured on days 1 and 3 are presented in Table 2. The per-cycle dos-imetry results obtained with the 3TP and the 2TP methods are summarized in Table 3.For the kidney, there was only one patient case during which no biological eliminationof activity between days 1 and 3 was observed. There were 30 such cases for the BMselfand 26 for Tumormax. In these cases, the 3TPTC and 2TPC methods were applied, while3TPTM and 2TPM methods were used for all other cases. The 3TP and 2TP data (i.e.,3TPTM pooled with 3TPTC, and 2TPM pooled with 2TPC) were very highly correlated(Spearman r > 0.99, P < 0.0001 for all tissues). The median relative errors between themethods were small, particularly for the kidney and the tumor (≤ 2%).The results of applying the single-measurement method proposed by Hänscheid et al.[8] to our day 3 QSPECT uptake data (1TPD3) are shown in column 8 of Table 3. Weobtained the same median error for the kidney as Hänscheid (6% at 72 h) with a com-parable interdecile range. Thus, our dosimetry results, based on tomographic dataDel Prete et al. EJNMMI Physics  (2018) 5:25 Page 8 of 20acquisition, validate this practical approach, which was devised using planar imagingdata. Further, despite the different imaging techniques, we obtained a similar medianeffective half-life for the kidney (47 h, Table 2; vs. 51 h in [8]), although we observed awider inter-patient variability. For Tumormax, the 1TPD3 technique was slightly less ac-curate when applied to our data, but the range of errors was comparable.Table 4 shows our results for the hybrid methods. In all cases, the 2TP method wasapplied in the first cycle. In this analysis, 2TP/1TPD3 was found to be more accuratethan both 2TP/1TPD1 and 2TP/NI. The latter method yielded particularly inaccurateresults for Tumormax, due to the interference of therapeutic response. Please note thatfor all tissues, we obtained median errors closer to zero with 2TP/1TPD3 than with1TPD3 (Table 3). For the kidneys, among all the simplified dosimetry methods, 2TP wasfound to be the most accurate when compared to 3TP, on a per-cycle basis (Fig. 3).As the aim of the induction course of our P-PRRT regime is to deliver a given pre-scribed renal absorbed dose, e.g., 23 Gy in patients without significant bone marrow orrenal function impairment, we compared the accuracy of the simplified dosimetrymethods to that of the 3TP method, for the assessment of the cumulative dosimetry inpatients having completed at least three of the four intended induction cycles (Table 5).From all the simplified approaches, the 2TP method was by far the most accurate, inparticular for the kidneys (Fig. 4). Using the latter, the cumulative renal and Tumormaxabsorbed dose for all patients agreed to within only 9% and 5%, respectively, with thecorresponding absorbed doses derived from 3TP scanning (Table 5) confirming thesmall influence of the day 0 scan and early kinetics on the precision of dosimetry esti-mates. Even the total BM dose was quite accurately estimated using the 2TP protocol.When compared with 2TP, median errors increased, and error ranges widened for all1TP-based techniques and even more so if no imaging was done on subsequent cycles.Inter-observer variabilityTable 6 illustrates the inter-observer variability of the per-cycle and cumulative renal,BMtotal, and Tumormax absorbed doses assessed independently by the three observersin 15 patients (60 cycles) using three methods: 3TP, 2TP, and 2TP/1TPD3. There wasan excellent inter-observer agreement between all observers for the kidney using thethree methods, the best agreement being for the cumulative renal dose estimated usingthe 3TP and 2TP methods. The span of errors was larger for Tumormax, owing it toTable 2 Tissue-specific effective half-lives derived from activity concentration at day 1 and day 3,and absorbed doses per injected activity for the 3TP reference method (n = 279)Effective half-life (h)a Absorbed dose per injected activity (Gy/GBq)Kidney 46.6 [36.3–55.7] (24.3–161.0) 0.54 [0.31–0.88] (0.21–4.25)Bone marrowself 72.3 [44.9–161.0] (29.4–161.0) 0.031 [0.014–0.087] (0.004–0.258)Bone marrowcrossb 66.9 [50.3–91.6] (24.6–121.6) 0.0030 [0.0016–0.0059] (0.0005–0.0161)Bone marrowtotal – 0.035 [0.018–0.092] (0.009–0.262)Tumormaxc 100.9 [60.0–158.4] (27.7–161.0) 3.8 [1.0–8.6] (0.1–32.0)Data is presented as median [interdecile range] (range)aIn cases of biological accumulation of activity (kidney, n = 1; bone marrowself, n = 30; tumormax, n = 26), effective half-lifewas assumed to be equal to the physical half-life of 177Lu, i.e., 161 hbBone marrow cross-dose is derived from the gamma contribution of whole-body activity retention over timecn = 278Del Prete et al. EJNMMI Physics  (2018) 5:25 Page 9 of 20Table3Per-cycledosimetryestimatesobtainedwiththree-,two-,andone-time-pointmethods(n=279)Absorbeddose(Gy)Correlationvs.3TP(r)bRelativeerrorvs.3TP(%)BiologicaldecayNobiologicaldecayaCombined3TP TM2TP M3TP TC2TP C3TP2TP1TP D32TP1TP D32TP1TP D3Kidney4.6[2.7–6.6](1.6–10.5)4.7[2.8–6.8](1.6–11.1)[2.7–6.6](1.6–10.5)4.7[2.7–6.8](1.6–11.1)4.8[2.7–7.1](1.4–11.6)0.9970.9902.0[−0.6–4.9](−3.4–16.0)5.8[−0.4–9.2](−37.7–17.0)Bonemarrowself0.26[0.12–0.63](0.05–2.08)0.25[0.11–0.62](0.04–2.18)0.36[0.21–0.90](0.17–3.84)0.34[0.20–0.86](0.16–3.88)0.27[0.12–0.64](0.05–3.84)0.25[0.12–0.63](0.04–3.88)0.22[0.11–0.56](0.04–2.41)0.9960.774−5.2[−11.4–0.6](−21.9–14.1)−10.0[−35.4–1.3](−43.3–17.0)Bonemarrowcross0.024[0.013–0.061](0.004–0.159)0.022[0.011–0.060](0.003–0.155)––0.024[0.013–0.061](0.004–0.159)0.022[0.011–0.060](0.003–0.155)0.022[0.012–0.054](0.003–0.122)0.9980.990−5.0[−12.6to−0.8](−29.2–10.0)−6.7[−16.5–1.3](−35.4–9.7)Tumormaxc30.9[7.5–70.8](0.5–277.6)31.3[7.8–71.9](0.5–271.1)47.0[12.1–81.0](2.9–120.0)47.6[7.4–82.6](2.9–121.7)31.2[7.4–74.1](0.5–277.6)31.7[7.6–76.0](0.5–271.1)27.8[6.8–61.4](0.5–215.3)1.0000.6511.7[−0.4–3.9](−3.5–14.0)−9.6[−30.4–8.0](−35.9–15.6)Cconstant,D3day3,Mmonoexponential,TCtrapezoid-constant,TMtrapezoid-monoexponential,TPtimepoint(s)Dataispresentedasmedian[interdecilerange](range)Medianinjectedactivitywas7.7(range,0.7–32.4)GBqa Kidney,n=1;BMselfn=30;BMcrossn=0;Tumormaxn=26bSpearman’scorrelation,P<0.0001inallcasesc n=278Del Prete et al. EJNMMI Physics  (2018) 5:25 Page 10 of 20Table4Per-cycledosimetryestimatesobtainedwithhybridmethodsbasedontwotimepointsforthefirstcycleandonetimepointornoimagingatallforsubsequentcycles(inductioncyclesonly,n=173)Absorbeddose(Gy)Correlationvs.3TP(r)aRelativeerrorvs.3TP(%)3TP2TP/1TP D12TP/1TP D32TP/NI2TP/1TP D12TP/1TP D32TP/NI2TP/1TP D12TP/1TP D32TP/NIKidney4.8[2.9–6.6](1.6–8.7)5.1[3.0–7.1](1.8–11.3)4.9[3.0–6.9](1.7–9.2)4.7[2.7–7.3](1.1–10.5)0.8870.9850.8073.4[−9.5–26.8](−38.3–86.3)2.2[−2.0–7.4](−17.9–32.1)0.6[−23.5–27.8](−60.2–106.3)Bonemarrowself0.29[0.14–0.65](0.06–3.84)0.22[0.11–0.73](0.06–2.33)0.24[0.13–0.64](0.06–2.83)0.24[0.10–0.65](0.04–3.85)0.7330.9390.707−7.0[−56.6–76.3](−73.9–155.8)−7.2[−33.2–22.4](−44.2–51.0)−13.8[−56.0–65.3](−93.8–370.9)Bonemarrowcross0.023[0.011–0.052](0.005–0.100)0.022[0.011–0.054](0.003–0.113)0.021[0.010–0.050](0.004–0.107)0.022[0.010–0.060](0.003–0.141)0.9660.9910.907−2.2[−21.2–14.9](−55.1–66.4)−4.2[−15.7–2.7](−37.3–22.6)1.3[−35.7–40.8](−52.5–113.0)Tumormaxb29.7[6.2–65.9](1.7–120.0)32.8[7.0–66.6](1.5–104.5)31.2[6.2–66.8](1.9–102.1)45.6[9.9–113.9](3.2–235.1)0.9460.9800.7133.6[−20.4–39.0](−46.7–141.8)2.3[−11.5–21.6](−30.2–65.3)31.0[−13.0–192.5](−49.9–6171.7)D1day1,D3day3,NInoimaging,TPtimepoint(s)Dataispresentedasmedian[interdecilerange](range)Medianinjectedactivitywas7.8(range,0.7–32.4)GBqa Spearman’scorrelation,P<0.0001inallcasesbn=172Del Prete et al. EJNMMI Physics  (2018) 5:25 Page 11 of 20variations in the precise placement of the VOI over the most intense region of thedominant lesion. For both BMtotal and Tumormax, there was a trend towards a lesserinter-observer agreement when the least experienced observer (observer 3) was in-volved. BMtotal reproducibility suffered from the low-level and noisy uptake data in theBM compartment, and consequently BMself, the dominant component of BMtotal, wasmore sensitive to the position of the VOI than were the absorbed doses of the other tis-sues of interest. Nevertheless, the inter-observer agreement on the cumulative BMtotaldose was fair.Accuracy of activity prescription at first and subsequent cyclesCorrelations and linear regression slopes between the body size descriptors, the eGFR,or their products vs. the IA per renal absorbed dose at the first induction cycle are re-ported in Table 7. The strongest correlations were found when using either eGFR · LBWor eGFR · BSA (Fig. 5) as predictors of renal GBq/Gy, and both seem equally appropri-ate for personalized IA prescription at the first cycle (Fig. 6). We therefore elected tocontinue using eGFR and BSA for determining IA at the first cycle and updated ourinitial formula (found in [6]) with this simpler equation:Personalized AI GBqð Þ ¼ K  eGFR mL= min=1:73m2   BSA m2  Prescribed renal absorbed dose Gyð Þ ð1Þwhere K = 0.012 GBq · min · 1.73/mL/Gy, i.e., the slope of the linear regression (Fig. 5).For the subsequent cycles, the median error of the renal GBq/Gy relative to that ofthe previous cycle was − 0.7% (interdecile range, − 21.0 to 20.2%; range, − 49.1 to54.0%; n = 194) for the 3TP method, − 0.8% (interdecile range, − 20.4 to 19.0%; range,− 49.4 to 50.2%; n = 194) for the 2TP method, and − 0.3% (interdecile range, − 20.5 to18.6%; range, − 40.2 to 87.4%; n = 168) for the 2TP/1TPD3 method. When in the ana-lysis of the third and fourth induction cycle the average of the renal GBq/Gy of the twoprior cycles was used, the resulting errors were − 0.9% (interdecile range, − 21.3–19.1%;range, − 49.9–54.4%; n = 192), − 0.6% (interdecile range, − 21.4–18.6%; range, − 48.5–Fig. 3 Comparison of the relative errors of per-cycle renal absorbed dose estimates obtained by thesimplified methods relative to the three-time-point (3TP) method. Boxes represent the interquartile range,and whiskers the interdecile range (2TP and 1TPD3, n = 279; 2TP/1TPD1, 2TP/1TPD3, and 2TP/NI, n = 173)Del Prete et al. EJNMMI Physics  (2018) 5:25 Page 12 of 20Table5Cumulativedosimetryestimatesobtainedinpatientshavingcompletedthreeorfourevaluableinductioncycles(n=65)Absorbeddose(Gy)Relativeerrorvs.3TP(%)3TP2TP1TP D32TP/1TP D12TP/1TP D32TP/NI2TP1TP D32TP/1TP D12TP/1TP D32TP/NIKidney19.3[11.8–23.6](6.5–26.3)19.9[11.8–24.3](6.6–26.7)18.2[10.7–24.4](4.5–27.2)17.7[11.1–24.5](6.1–27.0)17.4[10.6–23.5](4.6–26.6)17.2[9.9–25.3](4.5–29.3)1.9[0.1–3.8](−1.1–8.2)3.8[−24.2–6.3](−39.3–11.1)−0.5[−26.1–9.6](−41.1–34.4)0.4[−28.1–6.4](−40.4–33.2)−7.1[−32.4–18.0](−42.8–38.3)Bonemarrowtotal1.10[0.65–2.23](0.52–9.59)1.07[0.57–2.15](0.48–9.68)0.89[0.56–1.97](0.38–6.9)0.80[0.48–2.59](0.32–7.37)0.86[0.57–2.35](0.38–8.05)0.86[0.50–2.35](0.40–9.51)−5.4[−9.5to−1.4](−19.0–3.9)13.3[−20.5to−4.1](−33.2to−0.7)−15.7[−41.4–23.9](−61.2–68.8)−14.4[−24.7–9.0](−57.7–66.4)−20.3[−45.0–27.5](−72.9–110.3)Tumormax129.0[38.7–268.2](15.3–473.2)131.9[39.9–271.0](15.8–466.2)107.6[34.1–205.3](16.8–390.6)118.8[34.1–216.6](15.4–402.9)118.9[33.6–210.3](13.0–416.1)155.4[38.8–341.6](14.0–575.2)1.4[−0.2–3.3](−1.5–4.3)−16.1[−45.9–3.2](−60.0–12.0)−5.6[−41.0–10.3](−62.7–39.5)−2.9[−40.2–6.4](−63.4–21.9)15.9[−29.2–80.5](−75.6–400.5)D1day1,D3day3,NInoimaging,TPtimepoint(s)Dataispresentedasmedian[interdecilerange](range)Mediancumulativeinjectedactivitywas30.5(11.9–78.6)GBqDel Prete et al. EJNMMI Physics  (2018) 5:25 Page 13 of 2054.5%; n = 192) and − 0.6% (interdecile range, − 19.8–19.6%; range, − 45.7–75.3%; n =166), respectively. Hence, unlike in our initial P-PRRT protocol, these results convincedus that averaging the renal GBq/Gy from the two prior cycles (instead of just one) doesnot significantly improve the precision of the IA prescription. The IA prescription forthe subsequent cycle has been updated in our P-PRRT protocol as follows:Personalized IA GBqð Þ ¼ Prior cycle IA per renal dose GBq=Gyð Þ Prescribed renal dose Gyð Þ ð2ÞDiscussionThe widely adopted one-size-fits-all PRRT protocol, i.e., four induction cycles of7.4 GBq 177Lu-octreotate, has been initially devised in 2001 based on the dosimetrydata from only five patients [3]. Since that time, dosimetry has not been routinely per-formed in most centers (including for PRRT administered in the NETTER-1 trial [2]).This fixed IA regime is known to yield highly variable absorbed doses to critical organs,but because these fall well below conservative safety thresholds (e.g., 23 Gy for the kid-ney) in the vast majority of patients, this confers to PRRT a very favorable safety profile[1, 2]. In parallel, current cure rates are marginal, suggesting that most patients are be-ing undertreated with the empiric PRRT regime [1, 2].Escalating tumor absorbed dose could potentially improve the efficacy of PRRT, al-though realistically, this cannot be done through a conventional empiric IA escalationwithout compromising the excellent safety record of PRRT. To optimize tumor irradi-ation at the patient level, we and others have proposed to optimize the renal absorbeddose by either personalizing the IA per cycle or the number of induction cycles [4, 6].These two approaches resulted in increased cumulative IA and tumor absorbed dose inthe majority of patients [4, 6]. Further, when administering personalized IA, our prelim-inary results suggest a similar short-term side effect and toxicity profile as those ob-served when using the empiric PRRT regime [10]. Although our outcome data is notyet mature enough to document significantly improved clinical outcomes, nevertheless,Fig. 4 Comparison of the relative errors of per-induction course cumulative renal absorbed dose estimatesobtained by the simplified methods relative to the three-time-point (3TP) method. Boxes represent theinterquartile range, and whiskers the interdecile range (n = 65)Del Prete et al. EJNMMI Physics  (2018) 5:25 Page 14 of 20we believe that personalized radionuclide therapy is more faithful to the principles ofradiation oncology, where the absorbed doses are prescribed and monitored. In internalradiotherapy, this could be done through IA personalization and routine dosimetry.Despite the fact that our imaging protocol did not include a late time point, we ob-tained similar median renal absorbed doses per IA (median, 0.54 Gy/GBq) to those ob-served by Sandstrom et al. (medians, 0.62 and 0.59 Gy/GBq for the right and the leftkidneys, respectively), who also used QSPECT and small-VOI sampling, but scannedpatients until day 7 [5]. The concordance between our results is consistent with the factthat the renal activity concentration decays moxoexponentially after 24 h, as demon-strated by Handshied et al. [8].The median BM absorbed dose we obtained when using our QSPECT-based method(0.035 Gy/GBq) is well within the range of estimates published by others using varioustechniques based on imaging, blood, and urine sampling [5, 11]. Furthermore, thisTable 6 Inter-observer variability of dosimetry estimates in 60 induction cycles received by 15patientsRelative error (%)Kidney Bone marrowtotal TumormaxObserver 2 vs. observer 1Per cycle3TP − 0.2 ± 3.4 (− 7.8–10.0) − 2.4 ± 20.4 (− 53.7–75.9) − 0.2 ± 8.6 (− 34.1–40.5)2TP − 0.3 ± 3.4 (− 9.0–8.2) − 2.2 ± 20.9 (− 50.9–7.8) − 0.2 ± 8.2 (− 33.0–38.4)2TP/1TPD3 − 0.5 ± 3.4 (− 8.8–7.8) − 11.2 ± 25.5 (− 98.2–30.6) −1.9 ± 6.2 (− 33.0–6.4)Cumulative3TP − 0.2 ± 2.3 (− 2.9–3.9) − 4.5 ± 8.9 (− 24.0–7.8) − 0.6 ± 4.0 (− 8.3–8.0)2TP − 0.3 ± 2.3 (− 4.2–3.6) − 4.4 ± 8.8 (− 21.5–8.0) − 0.6 ± 3.8 (− 7.8–7.1)2TP/1TPD3 − 0.5 ± 2.3 (− 3.9–4.0) − 12.1 ± 19.7 (− 68.5–11.4) − 6.1 ± 10.1 (− 32.2–4.5)Observer 3 vs. observer 1Per cycle3TP − 0.5 ± 4.7 (− 21.5–9.4) 3.6 ± 29.0 (− 31.2–153.2) − 4.6 ± 9.0 (− 42.2–14.2)2TP − 0.3 ± 4.5 (− 21.4–8.2) 4.2 ± 29.9 (− 33.5–159.3) − 4.3 ± 8.7 (− 39.9–13.5)2TP/1TPD3 − 0.7 ± 5.8 (− 35.7–9.4) − 16.3 ± 35.9 (− 98.0–57.9) − 2.9 ± 8.6 (− 39.9–15.0)Cumulative3TP − 0.4 ± 1.5 (− 2.6–3.5) 1.9 ± 12.0 (− 13.2–31.4) − 3.9 ± 5.2 (− 11.0–10.7)2TP − 0.2 ± 1.2 (− 1.6–3.3) 2.3 ± 12.3 (− 13.7–31.9) − 3.7 ± 5.0 (− 10.7–10.1)2TP/1TPD3 − 0.6 ± 2.8 (− 9.5–3.9) − 17.3 ± 25.8 (− 74.7–3.3) − 5.5 ± 11.0 (− 33.8–13.5)Observer 3 vs. observer 2Per cycle3TP − 0.2 ± 5.1 (− 24.0–13.1) 7.3 ± 23.2 (− 30.0–82.2) − 4.1 ± 9.4 (− 28.4–25.2)2TP 0.0 ± 4.8 (− 22.6–10.0) 7.8 ± 23.9 (− 29.8–92.5) − 3.8 ± 9.1 (− 27.5–23.6)2TP/1TPD3 − 0.2 ± 6.4 (− 38.4–10.3) − 3.7 ± 33.9 (− 93.0–92.5) − 0.9 ± 7.4 (− 14.8–24.0)Cumulative3TP − 0.2 ± 2.0 (− 3.6–3.0) 7.0 ± 11.4 (− 7.9–34.3) − 3.2 ± 7.8 (− 17.5–20.7)2TP 0.1 ± 2.2 (− 3.9–3.6) 7.2 ± 11.0 (− 8.6–29.2) − 2.9 ± 7.4 (− 16.6–19.4)2TP/1TPD3 − 0.1 ± 3.4 (− 10.3–3.3) − 4.4 ± 25.6 (− 68.2–27.9) 1.0 ± 10.4 (− 8.0–27.8)D3 day 3, TP time point(s)`Data is presented as mean ± SD (range)Del Prete et al. EJNMMI Physics  (2018) 5:25 Page 15 of 20result is particularly close to the mean BM absorbed dose reported by Svensson et al.(0.027 Gy/GBq), which was derived from imaging data only and included a later timepoint [12, 13]. The correlation between QSPECT-based BM absorbed dose estimatesand subacute thrombocytopenia provides initial clinical validation of our technique[6]. However, in patients with bone metastases, the BM dosimetry estimates mayTable 7 Correlation between body size predictors, eGFR, and the IA per renal absorbed dose(GBq/Gy) at the first induction cycle (n = 77)Correlation a Linear regression slope Relative error (%)3TP 2TP 3TP 2TP 3TP 2TPr P r PWeight (Kg) 0.13 0.25 0.13 0.26 0.026 0.026 2.3[−46.0–66.6](− 61.5–165.7)2.0[−47.0–66.3](− 62.6–168.8)LBW (Kg) 0.18 0.12 0.17 0.13 0.037 0.036 5.6[−38.4–73.9](− 58.4–190.4)6.8[− 39.4–71.0](− 59.6–193.7)BSA (m2) 0.16 0.17 0.15 0.13 1.09 1.07 7.4[−38.2–73.5](− 53.5–227.7)6.9[− 39.1–76.8](− 54.8–231.5)eGFR (ml/min/1.73m2) 0.34 0.002 0.36 0.001 0.022 0.021 1.0[− 36.2–70.2](− 61.2–164.1)1.5[− 36.2–71.9](− 61.1–173.3)eGFR · weight 0.34 0.002 0.35 0.002 0.00026 0.00026 −1.9[− 46.7–63.8](− 62.9–124.4)−2.7[− 45.9–61.7](− 62.8–123.3)eGFR · LBW 0.40 0.0003 0.41 0.0002 0.00037 0.00036 −0.3[− 43.0–57.1](− 67.6–140.2)0.8[− 43.4–58.3](− 67.5–139.0)eGFR · BSA 0.38 0.0008 0.39 0.0005 0.012 0.012 − 0.0[−36.2–70.2](− 63.8–128.2)− 0.7[− 36.9–70.1](− 63.7–127.3)BSA body surface area, eGFR estimated glomerular filtration rate, LBW lean body weight, TP time pointsRelative error data is presented as median [interdecile range] (range)aSpearman’s correlationFig. 5 Injected activity per renal absorbed dose at the first cycle vs. the product of body surface area andestimated glomerular filtration rate (n = 77). There was a moderate correlation between variables (Spearmanr = 0.39, P = 0.0005). The slope of the linear regression curve forced through origin (solid line; 95% confidenceinterval, dashed lines), which was 0.012 GBq · min × 1.73/mL/Gy, is to be used to adjust the injected activity atthe first cycle in a personalized PRRT protocolDel Prete et al. EJNMMI Physics  (2018) 5:25 Page 16 of 20be less reliable, as even if obvious bone metastases were avoided when placing theBM VOIs, we cannot rule out the influence of non-apparent micrometastases ordiffuse BM infiltration.Dosimetry is an essential component of P-PRRT but is often perceived by the medicalcommunity as being too complex, or by the physics community as not accurateenough. However, SPECT/CT cameras are now widely available, and simple 177Lu cali-bration methods have been proposed [7, 14], facilitating implementation of individual-ized dosimetry based on QSPECT in the clinics. Performing dosimetry calculationbased on simplified activity concentration sampling methods, such as the small-VOImethod used in this study, is more practical to perform than the full organ segmenta-tion while yielding similar results and is more accurate than planar imaging-based dos-imetry [15]. For these reasons, we have routinely been performing dosimetry using a3TP QSPECT scanning schedule along with the small-VOI sampling. But still, manywould see dosimetry as resource-consuming. This opinion would be based on the gen-eral beliefs that a minimum of three measurements are necessary [16] or that the scan-ning protocol must include late time points, such as up to 4 to 7 days [17]. Suchrequirements tend to increase both the clinical burden and the patient inconveniencewhen performing dosimetry, and as such constitute barriers to its wide clinicaladoption. Conversely, simplified dosimetry approaches having a clinically relevantlevel of accuracy could facilitate making dosimetry a standard of care, not just formonitoring purposes, but also for personalizing radionuclide therapy. To overcomethe issues discussed above, the primary objective of this study was to further sim-plify our dosimetry methods.The 2TP method offers an excellent accuracy for both the per-cycle and the cu-mulative absorbed dose estimates relative to the 3TP method, in particular for thekidney and the tumor, which convinced us to abandon the day 0 scan. Further, ourFig. 6 Comparison of renal absorbed dose delivered during the first cycle of fixed injected activity (IA) vs.personalized PRRT regimes (n = 77). In the latter, the prescribed renal absorbed dose is 5 Gy and the IA isadjusted based on weight, lean body weight (LBW), body surface area (BSA), estimated glomerular filtrationrate (eGFR), or the product of eGFR and of a body size descriptor. For comparison, a fixed IA of 9.1 GBqwould yield a median renal absorbed renal absorbed dose of 5 GyDel Prete et al. EJNMMI Physics  (2018) 5:25 Page 17 of 20results validated the 1TPD3 technique proposed by Hänscheid and co-workers [8].While they advocate scanning on day 4 to achieve the best accuracy for both thekidney and the tumor, scanning on day 3 is considered more practical in our set-ting, offers about the same accuracy for the kidney dosimetry and a very reason-able accuracy for the tumor. This 1TPD3 method is an appealing alternative to2TP, although accuracy could be slightly improved, at least for the kidney andTumormax, by simply adding a second imaging time point during the first cycleand then applying the effective decay constant to the one-time samples during sub-sequent cycles (2TP/1TP). This hybrid method is more accurate when, for thenon-initial cycles, the imaging is performed on day 3 (2TP/1TPD3) rather than onday 1 (2TP/1TPD1). This is likely because, in individual patients, day 3 measure-ments are better correlated with the integrated TACs (i.e., absorbed doses) thanare those performed on day 1 and, as such, are less sensitive to small intra-patientcycle-to-cycle differences in tissue uptake and kinetics [8]. However, since in paral-lel we alter the IA prescription based on renal dosimetry, we prefer pursuing ourP-PRRT program with the 2TP protocol, which offers, in our opinion, the best bal-ance between high accuracy and practicality. Importantly, we would not recom-mend not imaging patients at subsequent cycles and extrapolating dosimetry fromthe first cycle (2TP/NI), assuming constant Gy/GBq in tissues (i.e., completely ig-noring any cycle-to-cycle difference in uptake or kinetics), as this approach causesthe uncertainty of the resulting absorbed dose estimates to increase, in particularfor tumor, which can be affected by the therapeutic response. The clinical burdenof the 2TP schedule in terms of the camera and personnel time is reasonable andcomparable to that of performing an 111In-octreotide scan. Furthermore, perform-ing the last scan on the third day limits the inconvenience for the out-of-citypatients.A very good inter-observer agreement has already been reported for renal dosimetry,with median errors of less than 5% for the small-VOI dosimetry method [17]. Our re-sults confirm these observations. Of note, we intentionally chose observers with differ-ent backgrounds and have shown that even dosimetry estimates from our noviceobserver (first-year medical student) were well in agreement with those from more ex-perienced observers, suggesting that a reasonably reproducible activity concentrationsampling technique is easily attainable with a relatively short training. Also, the wholeprocessing of one patient case, including VOI drawing and data transfer to the spread-sheet or database, can be performed in about 15 to 20 min. We are contemplating totrain nuclear medicine technologists to perform PRRT dosimetry under medical super-vision, eventually making them sub-specialized as nuclear dosimetrists.Finally, we revisited our personalized IA prescription scheme for our P-PRRT protocol.For the first cycle, we derived a simpler equation to determine the personalized IA than theone we initially suggested [6]. The latter is still based on the product of eGFR and BSA, buteGFR · LBW would have provided a similar level of predictive accuracy. We acknowledgethat this accuracy is at most moderate and comparable to that of an initial fixed IA in termsof interquartile or interdecile range (Fig. 6). However, the main advantage of personalizingthe first cycle IA is to avoid extreme cases of overdosing, such as delivery 18 Gy to the kid-ney when administering a fixed IA of 9.1 GBq to every patient (Fig. 6). Rather, personalizingIA could limit the renal absorbed dose to 11 Gy, for the same median of 5 Gy. WhenDel Prete et al. EJNMMI Physics  (2018) 5:25 Page 18 of 2068Ga-octreotate PET will be routinely performed in all our PRRT patients, we will exploreadding the pre-treatment tumor sink effect analysis into the prescription scheme, whichcould potentially improve the predictive accuracy of the model [18].ConclusionsWe propose a 177Lu dosimetry protocol based on two-time-point QSPECT imaging andthe small-VOI sampling, which yields accurate dosimetry results, particularly for the kid-ney and the tumor, with a very high inter-observer reproducibility. Performing the lastQSPECT/CT scan no later than on the third day post-PRRT increases patient conveni-ence, particularly for the out-of-city patients who travel to receive PRRT. Pragmatic 177Ludosimetry methods could facilitate the practice of personalized radionuclide therapies, in-cluding the rapidly emerging prostate-specific membrane antigen radioligand therapy.AbbreviationsACDF: Activity concentration dose factor; BSA: Body surface area; CT: Computed tomography; eGFR: Estimatedglomerular filtration rate; IA: Injected activity; LBW: Lean body weight; NET(s): Neuroendocrine tumor(s); NI: Noimaging; PET: Positron emission tomography; P-PRRT: Personalized PRRT; PRRT: Peptide receptor radionuclide therapy;QSPECT: Quantitative SPECT; SPECT: Single-photon emission computed tomography; SUV: Standardized uptake value;TAC(s): Time-activity curve(s); TP: Time point; VOI(s): Volume(s) of interestAcknowledgementsWe are grateful to the nurses and nuclear medicine technologists at the CHU de Québec—Université Laval who providedcare to PRRT patients, as well as to Marc Bazin, Ph.D., for his help with the writing of this manuscript.FundingM.D.P. is supported by a Merit Scholarship for Foreign Students from the Ministère de l’éducation et de l’enseignementsupérieur du Québec. This work was funded by the Canadian Institutes of Health Research (CIHR) operating grant MOP-142233 to J.M.B.Availability of data and materialsPlease contact the corresponding author, Jean-Mathieu Beauregard (, forthe data used in this manuscript.Authors’ contributionsMDP participated in the design of the analysis, collected and analyzed the data, and drafted the manuscript. FA andNS performed dosimetry analyses. FAB and JMB were responsible for the good conduct of the clinical study and dataacquisition. WZ and AC participated in the design of the analysis and contributed to the manuscript editing. JMBdesigned the analysis, supervised the project, analyzed the data, and edited the manuscript. All authors read andapproved the final manuscript.Ethics approval and consent to participateAll procedures performed in studies involving human participants were in accordance with the ethical standards ofthe institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendmentsor comparable ethical standards. Specifically, for the 23 patients who received only empiric PRRT until March 2016, therequirement for consent was waived due to the retrospective nature of the analysis. All other patients were enrolledin our P-PRRT trial (NCT02754297) and gave informed consent to participate.Consent for publicationAll authors read the manuscript and consented for its publication.Competing interestsThe authors declare that they have no competing interests.Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.Author details1Department of Radiology and Nuclear Medicine and Cancer Research Center, Université Laval, Quebec City, Canada.2Department of Medical Imaging and Oncology Branch of CHU de Québec Research Center, CHU de Québec –Université Laval, 11 côte du Palais, Quebec City, QC G1R 2J6, Canada. 3Medical Imaging Research Group, University ofBritish Columbia, Vancouver, Canada. 4Department of Physics and Astronomy, University of British Columbia,Vancouver, Canada.Del Prete et al. EJNMMI Physics  (2018) 5:25 Page 19 of 20Received: 4 April 2018 Accepted: 1 August 2018References1. Kwekkeboom DJ, de Herder WW, Kam BL, van Eijck CH, van Essen M, Kooij PP, et al. Treatment with theradiolabeled somatostatin analog [177Lu-DOTA0,Tyr3] octreotate: toxicity, efficacy, and survival. J Clin Oncol. 2008;26(13):2124–30. Strosberg J, El-Haddad G, Wolin E, Hendifar A, Yao J, Chasen B, et al. Phase 3 trial of 177Lu-Dotatate for midgutneuroendocrine tumors. N Engl J Med. 2017;376(2):125–35. Kwekkeboom DJ, Bakker WH, Kooij PP, Konijnenberg MW, Srinivasan A, Erion JL, et al. [177Lu-DOTA0Tyr3]octreotate: comparison with [111In-DTPA0] octreotide in patients. Eur J Nucl Med. 2001;28(9):1319–25.4. Sundlov A, Sjogreen-Gleisner K, Svensson J, Ljungberg M, Olsson T, Bernhardt P, et al. Individualised 177Lu-DOTATATE treatment of neuroendocrine tumours based on kidney dosimetry. Eur J Nucl Med Mol Imaging. 2017;44(9):1480–9. Sandstrom M, Garske-Roman U, Granberg D, Johansson S, Widstrom C, Eriksson B, et al. Individualized dosimetryof kidney and bone marrow in patients undergoing 177Lu-DOTA-octreotate treatment. J Nucl Med. 2013;54(1):33–41. Del Prete M, Buteau FA, Beauregard JM. Personalized 177Lu-octreotate peptide receptor radionuclide therapy ofneuroendocrine tumours: a simulation study. Eur J Nucl Med Mol Imaging. 2017;44(9):1490–500. Beauregard JM, Hofman MS, Pereira JM, Eu P, Hicks RJ. Quantitative 177Lu SPECT (QSPECT) imaging using acommercially available SPECT/CT system. Cancer Imaging. 2011;11:56–66. Hanscheid H, Lapa C, Buck AK, Lassmann M, Werner RA. Dose mapping after endoradiotherapy with 177Lu-DOTATATE/DOTATOC by a single measurement after 4 days. J Nucl Med. 2018;59(1):75–81. Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, et al. A new equation to estimateglomerular filtration rate. Ann Intern Med. 2009;150(9):604–12.10. Del Prete M, Buteau FA, Beaulieu A, Beauregard JM. Personalized 177Lu-octreotate peptide receptor radionuclidetherapy of neuroendocrine tumors: initial dosimetry and safety results of the P-PRRT trial. J Nucl Med. 2017;58(Suppl. 1):242.11. Cremonesi M, Botta F, Di Dia A, Ferrari M, Bodei L, De Cicco C, et al. Dosimetry for treatment with radiolabelledsomatostatin analogues. A review. Q J Nucl Med Mol Imaging. 2010;54(1):37–51.12. Svensson J, Rydén T, Hagmarker L, Hemmingsson J, Wängberg B, Bernhardt P. A novel planar image-basedmethod for bone marrow dosimetry in 177Lu-DOTATATE treatment correlates with haematological toxicity.EJNMMI Physics. 2016;3:21. Hagmarker L, Svensson J, Rydén T, Gjertsson P, Bernhardt P. Segmentation of whole-body images into twocompartments in model for bone marrow dosimetry increases the correlation with hematological response in177Lu-DOTATATE treatments. Cancer Biother Radiopharm. 2017;32(9):335–43. Uribe CF, Esquinas PL, Tanguay J, Gonzalez M, Gaudin E, Beauregard JM, et al. Accuracy of 177Lu activity quantificationin SPECT imaging: a phantom study. EJNMMI Phys. 2017;4(1):2. Sandstrom M, Garske U, Granberg D, Sundin A, Lundqvist H. Individualized dosimetry in patients undergoingtherapy with 177Lu-DOTA-D-Phe1-Tyr3-octreotate. Eur J Nucl Med Mol Imaging. 2010;37(2):212–25. Lassmann M, Chiesa C, Flux G, Bardies M, Committee ED. EANM dosimetry committee guidance document: goodpractice of clinical dosimetry reporting. Eur J Nucl Med Mol Imaging. 2011;38(1):192–200. Sandstrom M, Ilan E, Karlberg A, Johansson S, Freedman N, Garske-Roman U. Method dependence, observervariability and kidney volumes in radiation dosimetry of 177Lu-DOTATATE therapy in patients with neuroendocrinetumours. EJNMMI Phys. 2015;2(1):24. Beauregard JM, Hofman MS, Kong G, Hicks RJ. The tumour sink effect on the biodistribution of 68Ga-DOTA-octreotate: implications for peptide receptor radionuclide therapy. Eur J Nucl Med Mol Imaging. 2012;39(1):50–6. Prete et al. EJNMMI Physics  (2018) 5:25 Page 20 of 20


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items