Renal and Red Marrow Dosimetry in Peptide Receptor Radionuclide Therapy: 20 Years of History and Ahead

The development of dosimetry and studies in peptide receptor radionuclide therapy (PRRT) over the past two decades are reviewed. Differences in kidney and bone marrow toxicity reported between 90Y, 177Lu and external beam radiotherapy (EBRT) are discussed with regard to the physical properties of these beta emitter radionuclides. The impact of these properties on the response to small and large tumors is also considered. Capacities of the imaging modalities to assess the dosimetry to target tissues are evaluated. Studies published in the past two years that confirm a red marrow uptake in 177Lu-DOTATATE therapy, as already observed 20 years ago in 86Y-DOTATOC PET studies, are analyzed in light of the recent developments in the transferrin transport mechanism. The review enlightens the importance (i) of using state-of-the-art imaging modalities, (ii) of individualizing the activity to be injected with regard to the huge tissue uptake variability observed between patients, (iii) of challenging the currently used but inappropriate blood-based red marrow dosimetry and (iv) of considering individual tandem therapy. Last, a smart individually optimized tandem therapy taking benefit of the bi-orthogonal toxicity-response pattern of 177Lu-DOTATATE and of 90Y-DOTATOC is proposed.


Introduction
Peptide receptor radionuclide therapy (PRRT) is a well-established therapy for metastatic cancers expressing somatostatin receptors. A somatostatin analog peptide, such as octreotide or octreotate, is covalently bounded to a chelator, such as DTPA or DOTA. This chelator can be viewed as an empty basket. Before intravenous injection, a radionuclide ( 111 In, 90 Y, 177 Lu) is mixed with the chelator-peptide in a solution at a low pH and at a temperature both optimized to promote the entrance of the radionuclide within the chelator cage. High labelling fractions are easily reached, i.e., less than 2% of residual free radionuclide. The labeled compound exhibits a rather high stability when the pH solution is increased above 6.
The quantity to be injected is expressed in becquerel (Bq), which corresponds to one decay per second, and is named activity. By analogy with pharmaceutical drug, this quantity is often named dose, which is incorrect-the term dose identifies the quantity of energy deposed in a tissue by the particles emitted during the decays. This dose is expressed in gray (Gy), corresponding to 1 joule (J) delivered in 1 kg of tissue.
The first PRRT trials used 111 In-DTPA-octreotide [1], a tracer initially developed for diagnostic intent. Due to the cell internalization of the tracer, the short range 111 In auger electrons were considered suitable for tumor therapy [2]. The observed tumor control appeared promising, but an escalating activity study quickly revealed hematological concerns [3]. Indeed, as most of the 111 In decay energy is converted into gamma rays, the cross-irradiation of the red marrow from the remainder of the body was significant [4].
This issue was overcome using 90 Y-DOTA-octreotide ( 90 Y-DOTATOC) [5], which for therapy intent, can be considered as a pure beta emitter. This was confirmed by the 90 Y-DOTATOC clinical phase I trial SMT-487 [6,7]. With this compound, the first organ at risk was no longer the red marrow, but the kidneys, which is less life threatening [8]. However, due to its long beta range, i.e., maximal 11 mm in water, the dose delivered to small tumors is limited. Thus, relapse of micro-metastases initially not visible on imaging modalities were observed, although all known metastases completely responded.
To improve small tumor control, 177 Lu-DOTA-octreotate ( 177 Lu-DOTATATE) was developed [9], the beta of which has a smaller range, i.e., 2 mm in water. Although the decay energy is mostly brought by the beta, the main tissue at risk identified in the phase III study was the red marrow, with about 10% of the patients undergoing grade 3 or 4 lymphopenia [10].
The aim of this paper is to provide a comprehensive review of the literature evidencing and explaining the different toxicity and response patterns observed between 90 Y-DOTATOC and 177 Lu-DOTATATE PRRT. Last, by combining therapy improvements reported in the literature, a smart optimized tandem therapy design is proposed.

Kidney Dosimetry and Toxicity
The 90 Y-DOTATOC clinical phase I trial 86 Y-SMT-487 [6][7][8] was designed following the FDA requirements. The total injected activity was computed to deliver a maximum dose of 27 Gy to the kidneys, based on a pre-therapy 86 Y-DOTATOC PET using the MIRD pamphlet no. 11 [11], which assumed a standard kidney volume for the dosimetry calculation. The PET scan was reconstructed using a dedicated prompt single gamma ray correction [12].
The trial included a cycle activity escalation, i.e., a reduction in the number of therapy cycles needed to reach 27 Gy to the kidneys, starting from five cycles to a single cycle. The patient kidney follow-up was set as 5 years, with patients exhibiting a creatinine clearance lost per year up to 60%. Figure 1A clearly shows that obviously no toxicity-dose correlation was observed, as all the patients, apart from one receiving an extra cycle for compassionate use, received the same kidney dose.  [11], (B) after rescaling with the individual kidney volume and (C) converted into BED. The dots diameter corresponds to the number of cycles. Reprinted with permission from Ref. [8]. Copyright 2021 SNM. (D) Matching of the NTCPs observed in 90 Y-DOTATOC with that of EBRT. Reprinted with permission from Ref. [13]. Copyright 2021 SNM.
Rather than to conclude that dosimetry was useless, which sometimes happens in nuclear medicine, the patient CT analogic films were scanned and the renal volume measured. At this time, no hybrid SPECT-CT or PET-CT was available, and due to limited data storage capacity, most patient CT slices were only analogically archived. Figure 1B shows that by just rescaling the kidney dose with the standard to individual kidney volume ratio, a toxicity-dose relation became visible.
It was noted that globally, the patients receiving fewer cycles experienced higher creatinine clearance lost per year ( Figure 1B). The absorbed doses were thus converted into a biological effective dose (BED) using the Lea-Catcheside formalism [14], and a clear toxicity-dose correlation appeared ( Figure 1C). This toxicity-dose relation observed in nuclear medicine was the first one matching that observed in EBRT ( Figure 1D): the dose calculation based on fractionation commonly used in EBRT was proved applicable in nuclear medicine.
In all PRRT studies, i.e., 111 In-DTPA-octreotide SPECT [15], 86 Y-DOTATOC PET [12] and 177 Lu-DOTATATE SPECT [16], showed that the kidney uptake is localized in the cortex such as initially observed by Hammond et al. in planar 111 In-DTPA-octreotide scintigraphy [17]. Intra-patient 86 Y-DOTATOC PET studies proved that amino acid infusion significantly reduced the renal reuptake [6]. By competing with the megalin/cubilin complex on the apical membrane of proximal tubular cells (PTCs), basic amino acids, such as L-lysine and L-arginine, can reduce by ≈50% the reuptake of the radiolabeled peptide. The other fraction is taken up by the PTCs by fluid-phase endocytosis, that is not influenced by the presence of high amounts of basic amino acids [18,19]. Studies in knock-out rat provided evidence that megalin is essential for renal tubule reabsorption of the peptide [20].
Ex vivo autoradiography of healthy kidney tissue, from patients who received 111 In-DTPA-octreotide before nephrectomy (Figure 2A), showed an uptake gradient decreasing from the inner to the outer cortex boundary [21]. As the radiosensitive glomerulus is about 6 mm far from the inner boundary ( Figure 2E), its crossfire irradiation by the activity taken up by the tubules will strongly depend on the radionuclide beta range, as clearly shown in the Monte Carlo (MC) isodose simulations ( Figure 2B,C). These MC simulations explain why the first limiting tissue is the kidney with 90 Y-DOTATOC and the red marrow in 177 Lu-DOTATATE PRRT, respectively (see Section 3). nuclear medicine, the patient CT analogic films were scanned and the renal volume measured. At this time, no hybrid SPECT-CT or PET-CT was available, and due to limited data storage capacity, most patient CT slices were only analogically archived. Figure 1B shows that by just rescaling the kidney dose with the standard to individual kidney volume ratio, a toxicity-dose relation became visible.
It was noted that globally, the patients receiving fewer cycles experienced higher creatinine clearance lost per year ( Figure 1B). The absorbed doses were thus converted into a biological effective dose (BED) using the Lea-Catcheside formalism [14], and a clear toxicity-dose correlation appeared ( Figure 1C). This toxicity-dose relation observed in nuclear medicine was the first one matching that observed in EBRT ( Figure 1D): the dose calculation based on fractionation commonly used in EBRT was proved applicable in nuclear medicine.
In all PRRT studies, i.e., 111 In-DTPA-octreotide SPECT [15], 86 Y-DOTATOC PET [12] and 177 Lu-DOTATATE SPECT [16], showed that the kidney uptake is localized in the cortex such as initially observed by Hammond et al. in planar 111 In-DTPA-octreotide scintigraphy [17]. Intra-patient 86 Y-DOTATOC PET studies proved that amino acid infusion significantly reduced the renal reuptake [6]. By competing with the megalin/cubilin complex on the apical membrane of proximal tubular cells (PTCs), basic amino acids, such as Llysine and L-arginine, can reduce by ≈50% the reuptake of the radiolabeled peptide. The other fraction is taken up by the PTCs by fluid-phase endocytosis, that is not influenced by the presence of high amounts of basic amino acids [18,19]. Studies in knock-out rat provided evidence that megalin is essential for renal tubule reabsorption of the peptide [20].
Ex vivo autoradiography of healthy kidney tissue, from patients who received 111 In-DTPA-octreotide before nephrectomy (Figure 2A), showed an uptake gradient decreasing from the inner to the outer cortex boundary [21]. As the radiosensitive glomerulus is about 6 mm far from the inner boundary ( Figure 2E), its crossfire irradiation by the activity taken up by the tubules will strongly depend on the radionuclide beta range, as clearly shown in the Monte Carlo (MC) isodose simulations ( Figure 2B,C). These MC simulations explain why the first limiting tissue is the kidney with 90 Y-DOTATOC and the red marrow in 177 Lu-DOTATATE PRRT, respectively (see Section 3).

Red Marrow Dosimetry and Toxicity
A highly variable red marrow inter-patient uptake in PRRT was evidenced early in 2005 during the 86 Y-DOTATOC phase I 86 Y-SMT-487 trial, which used state-of-the-art PET imaging [23]. Furthermore, this red marrow uptake was intra-patient, correlated to that measured with 111 In-DTPA-octreotide by SPECT [24]. This was overlooked by the nuclear medicine community, who argued that RM uptake was not visible in 90 Y bremsstrahlung or 177 Lu SPECT [25,26]. The case appeared to be solved in 2017 when the LutatheraTM insert package, which states that establishing red marrow dosimetry is useless in predicting observed toxicity, was agreed upon by the FDA and the EMA [27].
These past few years, several teams using state-of-the-art SPECT/CT demonstrated clear red marrow uptake of 177 Lu in 177 Lu-DOTATATE therapy [16], the dosimetry based on enabling some toxicity prediction [28][29][30]. The observed red marrow dosimetry was about fourfold that of the blood-based method used in the Netter study [10], explaining why about 10% Grade 3-4 hematological toxicity was observed in this study.
Such red marrow uptake could appear surprising, given the high DOTA affinity and stability for yttrium and for lutetium. However, in the transchelation competition with transferrin, DOTA is just a passive and naïve chelator stroke by an active and cunning thief. Iron is a vital compound for mammalians and evolution spent hundreds of millions of years to improve transferrin, versus a few decades for chemists.
Transferrin is a protein having two active lobes ( Figure 3A) [31]. In the iron-free state, i.e., apo-transferrin, the N-lobe and the C-lobe are open, ready to catch a metallic ion. When a metallic ion enters a lobe, the lobe closes and surrounds the ion as a result of the Van der Waals forces. In this state, no other external force can easily remove the ion from the locked lobe. When an appropriate anion binds to the corresponding active site, the lobe opens and releases the metallic ion. This key-padlock mechanism ensures that iron will be released to the appropriate cells and not elsewhere in the body. Previously irradiated vertebra by EBRT, which has eradicated active marrow, have a similar behavior as the blood pool. Reprinted with permission from Ref. [24]. Copyright 2021 Springer.
The ion transferrin release is an active process, along with the ion catching one. Inorganic chelators, such as DOTA, have to wait for the release of a metal ion by another chelator in order to trap it, which is rare from a good competitor. In contrary, kinetics and spectrometry studies evidenced that transferrin can create a ternary (or mixed) complex with the initial metal-chelator complex [32]. During the life of this ternary complex, the metal ion is transferred from the chelator to the transferrin metal binding site, likely by the Van der Waals forces. Many in vitro competition studies between transferrin and DOTA evidenced the transferrin superiority [24,33]. In vitro data using plasma from subjects injected with 111 In-pentetreotide demonstrated that after 7 days, the radioactive metal was effectively bound to transferrin [34].
A recent study in humans [35] showed that only 23% and 2% of 177 Lu-DOTATATE remain intact after 24 h and 96 h post-injection, respectively. Figure 3B clearly illustrates the uptake difference between the tumor receptor-based mechanism and the transferrin transport to red marrow mechanism observed in the 86 Y-SMT-487 study [24]. Directly during the first pass, the 86 Y-DOTATATE binds to the tumor receptors, resulting in a strong uptake within 4 h pi, while it takes a much longer time for the transferrin to catch the metal and to deliver it to the red marrow cells. The 4 and 24 h images proved the red marrow activity behavior in the opposition phase versus the blood pool, which makes the blood-based red marrow dosimetry method unlikely to be physiologically adequate. Table 1 shows red marrow dosimetry assessments and dose-toxicity correlations reported in the recent literature. Using 4 SPECT/CT performed after the first and second cycles, Santoro et al. [28] evaluated the organs at risk in 12 patients treated with 177 Lu-DOTATATE. The mean dosimetry for kidney and red marrow was 0.43 ± 0.13 mGy/MBq and 0.04 ± 0.02 mGy/MBq, respectively. As the maximal tolerated dose for red marrow is about tenfold lower than that of kidney [24], this explains why hematological toxicity is the limiting factor in 26% of individually optimized 177 Lu PRRT [36].
In 200 patients treated with 177 Lu-DOTATATE, Garsk et al. [36] observed fourfold lower mean blood-based red marrow dosimetry than that observed in the two reported SPECT studies, i.e., 0.016 mGy/MBq, which was unable to predict the hematological toxicity observed in 40 patients. Page et al. [37] compared blood-based and SPECT/CTbased red marrow dosimetry in 11 patients treated with 177 Lu-DOTATATE and observed the same fourfold ratio. The results of these six studies [28][29][30][35][36][37] urge for using post-cycles SPECT/CTbased red marrow dosimetry for all 177 Lu therapies. With optimized imaging and reconstruction methods of SPECT/CT, these studies provide evidence that direct RM uptakebased dosimetry clearly outperforms the blood-based estimations, as advocated in the Luthatera package insert. One must keep in mind that the maximal red marrow tolerable dose is about tenfold lower than that of other tissues. Thus, a long duration, i.e., >30 min, dual-head SPECT is required to accurately assess the red marrow uptake at several timepoints. Red marrow activity should be measured on the thoracic vertebras: the region with the lowest attenuation and without surrounding tissues with high uptake, such as the liver, spleen or kidneys, that can produce long tail artefacts in reconstructed slices. Fortunately, nowadays, a single SPECT/CT scan encompass the thorax and the kidneys. Clearly, state-of-the-art imaging capacities as available now can help improve this paradigm of RM dosimetry based on imaging, rather than on blood activity.

177 Lu-90 Y PRRT Tandem Therapy: A Bi-Orthogonal Synergy
Hobbs et al. [38] introduced the "orthogonal radionuclides" concept to promote tandem therapy with radionuclides taken up by different tissues and thus presenting different toxicities. Such tandem allows increasing the tumor dose by splitting the unwanted irradiation between different organs, as similarly performed in EBRT by rotating the beam source around the targeted tumor.
Comparing toxicities of 177 Lu-DOTATATE and 90 Y-DOTATOC therapies is very challenging considering the sparse information provided in the clinical trials publications. In the Netter phase III study [10], in 116 patients receiving four repeated injections of 7.4 GBq every 8 weeks, 0.4% and 11% underwent a renal and hematological toxicity of grade 3 or 4, respectively. In the phase II 90 Y-DOTATOC study [39] in 1109 patients receiving in average 2.5 cycles of 3.7 GBq/m 2 , 9.2% and 12.8% experienced a renal and hematological toxicity of grade 3 or 4, respectively. However, the frequency of this hematological toxicity cannot be directly compared; indeed, assuming similar uptakes, the 90 Y protocol delivered a much higher tissue dose than that of 177 Lu ( Table 2). The hematological to renal toxicity ratio is about twentyfold higher for the 177 Lu-DOTATATE therapy. Regarding dosimetry-based studies, in individualized 177 Lu-DOTATATE study aiming to deliver 23 Gy to the kidney in 154 patients, 26% presented a hematological toxicity, halting the therapy [36], while this event occurs only in 1 patient out of 60 in the individualized 90 Y-DOTATOC study aiming to deliver 27 Gy to the kidney [7]. Again, a twentyfold higher prevalence of hematological toxicity is observed in 177 Lu-DOTATATE therapy.
This much higher prevalence of hematological toxicity in 177 Lu-DOTATATE therapy is in line with the low in vivo stability of this compound: only 23% and 2% of 177 Lu-DOTATATE remain intact after 24 h and 96 h post-injection in patients [35], whereas 90% of 86 Y-DOTATOC remains intact 5 h post-injection in primates [40]. Intuitively, it makes sense that the larger electron orbital of 177 Lu makes it more sensitive to external Van der Waals forces coming from an open transferrin lobe. Note that, as transferrin is not excreted by the kidneys, radionuclide to protein binding analysis in urine is not appropriate to evaluate the in vivo compound stability. In vitro transchelation competition with apo-transferrin could also be biased, considering that in vivo, apo-transferrin is continuously renewed by the release of the metal ion in the red marrow. Such in vitro studies thus have to be performed with a huge apo-transferrin overload. Last, caution is also needed with in vivo rodent stability studies, considering that experiments proved different iron binding and release properties compared to humans [41]; this could also hold true for yttrium and lutetium.
All these facts clearly show that the orthogonal toxicity concept applies to the couple 177 Lu-DOTATATE and 90 Y-DOTATOC for which the prime tissue at risk is the red marrow and the kidney, respectively. This feature alone should be sufficient to promote the tandem approach as a new standard for PRRT therapy.
Last, but not least, this tandem is also orthogonal regarding the tumor response (justifying the bi-orthogonal appellation): by its short beta range, 177 Lu is efficient to deliver high doses in sub-centimetric tumors, whereas the 90 Y beta, with its one cm range, is more efficient to cross-irradiate low vascularized regions often present in larger tumors [42,43].
This dose-response synergy was already observed in 2005 in a preclinical model [44]. Pancreatic cancer cells were successively grafted, i.e., spaced by 3 weeks, in the two opposite flanks of rats, resulting in rats bearing a small (≈0.5 cm 2 ) and a large (≈8 cm 2 ) tumor. Rats were split into four groups of about 12 individuals. Survival was impressively higher in rats treated with the 177 Lu-DOTATATE and 90 Y-DOTATOC tandem than that of rats treated with only one radionuclide ( Figure 5A). Kunikowska et al. [45] compared overall survival (OS) between single 90 Y-DOTATOC therapy and 90 Y-177 Lu tandem therapy. The patients were not randomly drawn: the first 25 consecutive patients were treated with 90 Y alone (7.4 GBq/m 2 ), and the following 25 consecutive patients were treated with the 90 Y-177 Lu tandem (3.7-3.7 GBq/m 2 ). However, the same patient enrollment criteria were chosen for both groups. As in the rat study, the OS was impressively better in the tandem group, especially that, when assuming similar uptake, the tandem delivers absorbed doses 30% lower than that of the 90 Y alone using this activity protocol.
In this study, the tandem was given together in two or three cycles, which is not the right strategy: the well-vascularized large tumor region takes fewer radionuclides in the following cycle due to the additional 177 Lu irradiation eradicating cells which uptake the peptide. The best strategy is to first perform the 90 Y cycles to benefit from the high uptake of vascularized regions to cross-irradiate the other tumor regions, then to end up with a 177 Lu cycle to efficiently irradiate, and hopefully eradicate, small metastases [46].

Individual PRTT Optimization Planning
Multi-cycle PRRT is well adapted to individualized therapy planning as the first cycle delivered dose is safe for all patients. Afterwards, post-cycle dosimetry can be performed to optimize the activity to be injected in the following cycles. Note that if 98 Ga-DOTATATE PET is of prime interest to select a candidate for the PRRT therapy, it is useless for dosimetry estimation given its too short half-life. In 90 Y-DOTATOC therapy, phantom and patient studies proved 90 Y PET imaging providing an accurate kidney dose estimation [47], while medium energy collimator SPECT/CT is well adapted for 177 Lu-DOTATATE [16], provided that the acquisition time is long enough.
PRRT dosimetry requires at least two imaging time points, as tissue activity curves exhibit a bi-exponential asymmetric bell shape. This feature renders proposed single timepoint estimates inaccurate [48], which is only for a single exponential starting from t = 0, the model used to propose it [49]. One time point after the maximal uptake is needed, e.g., 24 h post-injection, and another time point somewhere around one effective half-life, i.e., 48 or 72 h post 90 Y injection and 5 to 8 days post 177 Lu injection.
Only the two tissues at risk have to be imaged, i.e., the thoracic vertebrae and the kidneys, which can be performed with two PET/CT positions for 90 Y and with one single SPECT/CT position for 177 Lu. Planar imaging and combined planar and hybrid SPECT/CT should be avoided given the superimposition of the residual blood-pool activity for red marrow and of the liver and of the spleen activity for the kidneys. Many physicians think it is mandatory to use state-of-the-art TOF-PET/CT to perform tumor staging or follow-up! It is time to tackle claims that short-duration, whole-body planar scan is sufficient to perform dosimetry. By always trying to simplify dosimetry to satisfy this claim, medical physics experts (MPE) are left with poor quality data for accurate dosimetry, which reinforces the common belief that dosimetry is useless.
Menda et al. [50] conducted a prospective post-cycle renal dosimetry using 90 Ybremsstrahlung SPECT/CT in 25 patients with neuroendocrine tumors. A 90 Y TOF-PET/CT [47] was used at the first time for bremsstrahlung SPECT/CT calibration purposes. The study confirmed the very high variability of inter-patient renal dosimetry, as already observed using 86 Y-DOTATOC PET [8], advocating the interest to individually optimize the injected 90 Y activity.
Garske-Roman et al. [36] observed an impressive progression-free survival (PFS) and overall survival (OS) improvement of patients who were able to reach the 23 Gy renal dose limit (n = 114) versus those who could not, due to hematological toxicities (n = 40) ( Figure 6).

Smart Optimized Tandem Therapy Design: A Proposal
As in tandem therapy, the potential toxicity is split between two different tissues, and the activity fractionation can be reduced versus a single radionuclide therapy, which has the benefit to reduce aggressive tumors regrowth between cycles. As a result, a smart full optimized tandem therapy, fractionated in three cycles, can be proposed (Figure 7). It requires only four patient visits to the hospital, three amino acid infusions and five imaging sessions. At day 0, a fixed activity of 90 Y-DOTATOC/TATE is injected with amino acid infusion. The following day, before releasing the patient from the hospital, a 90 Y kidney 30 min-PET/CT is performed.
At day 3, a 90 Y kidneys 30 min-PET/CT is performed to compute the kidney dosimetry in order to assess the residual 90 Y activity needed to reach a BED of 15 Gy to the kidneys. Afterwards, this 90 Y-DOTATOC/TATE activity complement is injected to the patient together with a fixed 177 Lu-DOTATATE activity along with amino acid infusion, and a blood sample is withdrawn to obtain the cell count base line. As this session corresponds to the last 90 Y cycle, co-injection of 177 Lu does not impact the 90 Y cross irradiation. Note that with regard to the effective kidney half-life (≈30 h, [8]), the initial dose rate is already reduced by a factor ≈5, corresponding to an effective dose fractionation. The following day, before releasing the patient, a 177 Lu thorax-abdomen 30 min dual-head SPECT/CT is performed. Additionally, for a highly valuable scientific point of view, a blood sample could be taken to assess the relative binding of 177 Lu and 90 Y to transferrin in the same patient at the same time using size-exclusion chromatography, with appropriate MW standards.
At day 8, a 177 Lu thorax-abdomen SPECT/CT is performed, and the kidney and RM dosimetry is computed. The kidney dosimetry should be computed using the Sfactors taking into account the 90 Y and 177 Lu beta range for the cross irradiation of the glomerulus by the taking up tubules [21]. At day 31, a blood sample is taken by the treating team or by the general practitioner to estimate the blood cell count nadir.
At day 41, a blood sample is withdrawn to check the cell count recovery, and depending on the values at nadir and recovery, a 177 Lu-DOTATATE activity complement is injected, satisfying, for the whole therapy, the two limits: D < 2 Gy to the red marrow and BED < 31 Gy to the kidneys. Note that the physician has the competence to modulate these limits according to the patient status and to his cell count recovery. The following day, before releasing the patient, a thorax-abdomen 177 Lu SPECT/CT is performed.
Afterwards, the patient undergoes the normal follow-up.

Conclusions
In recent years, many studies provided evidence of the huge benefit for the patient outcome to individually optimize the activity to be injected. Furthermore, although not performed along an optimized planning, recent studies have also established the importance to profit of the bi-orthogonal toxicity-response patterns of the tandem 90 Y-DOTATOC-177 Lu-DOTATATE. There is no doubt that the combination of these two approaches, as proposed in Figure 7, will still improve the patient outcome and will push PRRT to a real curative intent.
However, last year, the EANM published a position paper [53] on article 56 of the Council Directive 2013/95 Euratom, which required an individualized optimization planning in all radiotherapies. This position paper aims to provide guidance on how to interpret the Directive and states that 177 Lu-DOTATATE, used according to the package insert, is a standard therapy not requiring any individualized planning, which is scientifically questionable considering the present review.
However, let us recall that legally, the MPE responsibility is framed by the corresponding national transposition of the directive. However, the MPE or practitioner cannot be prosecuted for having followed Directive 2013/95 rather than the national implementation.
Indeed, as the Directive 2013/95 provisions are unconditional and sufficiently clear and precise, the directive has direct effects, and any individual can invoke its provisions in front of any national court [54]. The European Court of Justice's jurisprudence has extended this principle to cases with incorrect implementation of a directive [55].