High doses of the radiopeptide [¹¹¹In-DTPA0]-octreotide were first applied for the radionuclide therapy in NETs (Krenning et al. 1994). However, due to the short range (nanometers to micrometers) of the emission and therefore the short tissue penetration of the emitted Auger electrons, objective responses were rare. Furthermore, leukemia or myelodysplastic syndrome has been reported in three patients from the group receiving the highest cumulative dose (Valkema et al. 2002). ¹¹¹In-pentetreotide is still used by some departments in the USA, mainly due to lack of accessibility to ß-emitting radiotracers. In fact, a recent study reported safety and survival benefit in more than 100 patients treated with ¹¹¹In-pentetreotide (Delpassand et al. 2012).

The most frequently used radionuclides in PRRT are ⁹⁰Y and ⁷⁷Lu. ⁹⁰Y is produced as a carrier-free radionuclide from the parent radioisotope ⁹⁰Sr. The radioisotope ¹⁷⁷Lu is produced by nuclear reactors either via the direct or the indirect pathway; the direct pathway involves the n-gamma irradiation of an enriched ¹⁷⁶Lu target. The indirect pathway involves the irradiation of Ytterbium-176 (¹⁷⁶Yb) to produce the short-lived ¹⁷⁷Yb (half life, t1/2 1.9 h) that decays to ¹⁷⁷Lu. The latter method avoids the production of the long-lived ^177mLu (t1/2 60 days).

These radionuclides have different physical characteristics, for e.g., beta particles are emitted at different energies, resulting in various tissue penetration ranges. A pronounced “cross-fire effect” is found when using ⁹⁰Y, since the beta particles are emitted with a relatively high energy (maximum energy, 2.27 MeV), resulting in a tissue penetration range of up to 12 mm. This may be preferable for relatively uniform irradiation of larger tumors, which may show heterogeneous uptake. Due to its pure β-emission, dosimetric analysis has to be performed using a surrogate like ¹¹¹In. The use of ⁸⁶Y and ⁹⁰Y PET or bremsstrahlung imaging has also been reported for dosimetry (Delpassand et al. 2012). ¹⁷⁷Lu on the other hand, emits intermediate-energy beta particles (maximum energy 0.5 MeV), resulting in a tissue penetration range of up to 2 mm, which may be preferable for smaller tumors. In addition, ¹⁷⁷Lu has two gamma peaks at 113 and 208 keV with 10 and 6 % abundance respectively, which makes it suitable for imaging with a gamma camera for post-therapeutic dosimetry with the same radiopharmaceutical. Another difference between these radionuclides is their physical half-life. Although the influence of the physical half-life is not yet fully understood, it is very likely that it influences the therapeutic as well as the secondary effects. For ⁹⁰Y, it is 2.7 days, whereas the physical half-life of ¹⁷⁷Lu is more than double (6.7 days). These differences in characteristics in turn influence the decision on the choice of radionuclide to be used for PRRT.

Chelators are essential for the radiolabeling of SST analogs. The macrocyclic chelating agent DOTA (1, 4, 7, 10-tetraazacyclododecane-N, N′, N″, N‴-tetraacetic acid) forms the most widely used complexes for diagnosis and therapy of NETs, namely [DOTA⁰,Tyr³] octreotide (DOTATOC) and [DOTA⁰,Tyr³] octreotate (DOTATATE) for complexing trivalent radiometals, i.e., ⁶⁸Ga (PET) and ¹⁷⁷Lu/⁹⁰Y (PRRT) for the theranostics of NETs. The radiochemical properties of diethylenetriaminepentaacetic acid (DTPA) are favorable for chelation with ¹¹¹In; however, ⁹⁰Y-DTPA has poor in vivo stability (Heppeler et al. 2000). ⁹⁰Y-DOTA, on the other hand, is a highly stable complex with enhanced thermodynamic and kinetic stability relative to an open-chain analog (Moi and Meares 1988; Deshpande et al. 1990). Due to the high affinity for SSTR 2, DOTATOC has favorable potential for therapeutic use (Otte et al. 1997; Paganelli et al. 1999). Moreover, it has high hydrophilicity, stability, and is suitable for labeling with ⁹⁰Y and ¹¹¹In (de Jong et al. 1998).

DOTATATE has a nine-fold higher affinity for the SSTR 2 compared to DOTATOC (Reubi et al. 2000). DOTANOC ([DOTA 0–1-Nal 3]octreotide) exhibits the highest affinity to SSTR 3 and 5, and a high affinity to the SSTR 2 (Wild et al. 2003). Wehrmann et al. compared the kinetics and dosimetry of ¹⁷⁷Lu-DOTATATE and ¹⁷⁷Lu-DOTANOC and concluded that whole-body and normal tissue uptake was higher for DOTANOC, resulting in a higher mean absorbed dose to the whole body. Hence though the kidney and tumor doses were comparable for the two radioligands, ¹⁷⁷Lu-DOTATATE has characteristics more favorable for PRRT than ¹⁷⁷Lu DOTANOC, because a lower whole-body dose implies a lower potential bone marrow toxicity (Wehrmann et al. 2007). Esser et al., in their study comparing ¹⁷⁷Lu-DOTATATE and ¹⁷⁷Lu-DOTATOC in seven patients, had found a longer residence time in the kidneys with DOTATATE, but the ratio of residence times of DOTATATE to DOTATOC in the tumor (2.1) was higher than that for the kidneys; hence, DOTATATE was found to deliver a higher dose to the tumor (Esser et al. 2006). However, another recent study comparing the two peptides in 22 patients concluded that the tumor-to-kidney ratio was comparable for both peptides (Kulkarni et al. 2013). Though uptake, residence time and mean absorbed renal dose per unit administered activity showed significantly higher values for DOTATATE in 19 (86 %) patients, both the peptides were found to be safe in terms of potential for renal toxicity. On the other hand, Forrer et al. demonstrated no difference in the tumor uptake of ¹¹¹In-DOTATATE and ¹¹¹In-DOTATOC, whereas a higher tumor-to-kidney absorbed dose ratio for ¹¹¹In-DOTATOC (Forrer et al. 2004).

Curative treatment of NETs usually requires the possibility of complete surgical resection of the primary tumor and perhaps regional lymph node metastases. Surgery and cold somatostatin analogs (i.m. or s.c. octreotide or lanreotide) are most commonly used as the first line of treatment; chemotherapy is used as the last option or in high-grade tumors. Recent years have seen the development of PRRT as a highly effective treatment option for inoperable or metastasized progressive well-differentiated NETs. The WHO grade of the tumor determined by the Ki-67 tumor proliferation index is essential for patient selection (Ezziddin et al. 2011). G1 and G2 tumors are known to respond well to PRRT. Also, high somatostatin receptor expression of tumors must be established before therapy, in order to ensure therapeutic efficacy. This requires semiquantitative interpretation (Krennings’ score) with somatostatin receptor scintigraphy using ¹¹¹In-Octreotide, or with SSTR PET/CT using ⁶⁸Ga-DOTA-peptides. Although dosimetry is the best way to “individualize” PRRT, semiquantitative evaluation of standardized uptake values (SUVs) of the tumors on SSTR PET/CT, is an excellent measure of the degree of SSTR expression (Kaemmerer et al. 2011). A recent study established a correlation between the SUV on pre-PRRT SSTR PET and tumor dose during PRRT (Ezziddin et al. 2012). A trend in terms of the tumor dose–response relationship during PRRT was demonstrated by another preliminary study (Kulkarni et al. 2011). Hence, pretherapeutic assessment of the tumor SUV could potentially help to predict the response to PRRT based on the knowledge of this dose–response relationship and lead to better selection of the patients for PRRT. However, further studies are required to examine this early evidence.

Severi et al. further determined the prognostic value of ¹⁸F-FDG PET in 52 patients with progressive advanced well-differentiated grade 1/2 NETs treated with ¹⁷⁷Lu-DOTATATE (Lu-PRRT). They found that none of PET-negative patients had progressed at the first follow-up examination after ¹⁷⁷Lu-PRRT, whereas grade 2 NET and ¹⁸F-FDG PET positivity (arbitrary SUV cutoff >2.5) were frequently associated with more aggressive disease, suggesting perhaps the requirement of a more intensive treatment regimen for the latter group (Severi et al. 2013).

In a recent study by Campana et al., 69 NET patients (37 men and 32 women; 45 pancreatic and 24 gastrointestinal; 22 G1 and 41 G2) were treated with ⁹⁰Y or ¹⁷⁷Lu. Partial response (PR) was noted in 27.5 % patients, while 50.7 % had stable disease and 23.2 % had progressive disease. Low tumor burden and a low proliferation index represent independent prognostic factors for long PFS while stage IV, NET G2, and previous TACE were found to be significant factors for tumor progression at multivariate analysis (Campana et al. 2013).

The challenge for internal radiation therapy is to deliver the highest possible dose to the tumor, while sparing normal organs from damage. In addition, the interindividual differences in dose delivery should be considered, particularly on the account of varying metabolism or receptor density in organs and tumor lesions. This makes individual patient dosimetry absolutely necessary in PRRT. The advantage of ¹⁷⁷Lu is the concomitant emission of low-energy γ-rays at 208 and 113 keV with the therapeutic β-radiation, which allows not only assessment of biodistribution with post-therapy scans, but also enables individualized dosimetry. It is imperative to calculate the absorbed radiation doses to dose-limiting organs such as the kidneys and bone marrow to better influence the decision for administering multiple cycles of PRRT, taking into account the cumulative administered activity. Taking into consideration the highly variable SSTR densities on tumor cells and other factors like tumor size and viability, individualized dosimetry helps in patient selection and therapy planning.

The data required as input for dosimetry are blood and urine samples, and whole-body scans at adequate time intervals up to at least 3 days post PRRT. Regions of interest (ROI) on planar images are useful to obtain the time–activity curves. However a much better organ-specific three-dimensional activity distribution is provided by SPECT or SPECT/CT images, although the acquisition is time consuming. The dosimetry estimation is based on the MIRD scheme (Siegel et al. 1999). Using dedicated software like OLINDA/EXM, mean absorbed dose estimates can be derived based on the calculated residence time of radiopeptide used (Stabin et al. 2005).

The lack of gamma emission by ⁹⁰Y is a drawback so as to enable direct dosimetric analysis. Quantification using bremsstrahlung images is difficult and requires complex corrections. An excellent alternative is use of the positron emitter ⁸⁶Y (Walrand et al. 2011). The advantage of using ⁸⁶Y-DOTATOC is the accurate mimicking of the in vitro pharmacokinetics and a much superior resolution and accurate quantification by PET. However, the physical half-life (14.7 h) and low positron abundance of ⁸⁶Y does not permit delayed imaging, and the high production cost and low availability considerably limits its routine utilization. Due to the half-life of ¹¹¹In and the pharmacokinetics, ¹¹¹In-DOTATOC is used in clinical practice as a surrogate for ⁹⁰Y-DOTATOC. ¹¹¹In-pentetreotide scintigraphy, however, is not suitable for accurate dosimetric calculation due to its different kinetic behavior and receptor affinity profile. The short physical half-life of ⁶⁸Ga also makes pretherapeutic dosimetry unsuitable. The use of long-lived positron emitters, e.g., ⁴⁴Sc or ⁶⁴Cu for receptor PET/CT could enable pretherapeutic organ and tumor dosimetry by comparison of these results with those obtained from ⁶⁸Ga-SSTR PET/CT results. The appropriate peptide and radionuclide for individualized therapy could then be selected by pretherapeutic measurement of organ and tumor doses.

Due to the small size, radiopeptides are filtered through glomerular capillaries in the kidneys and subsequently reabsorbed and retained in the proximal tubular cells. Kidneys are the dose-limiting organs for PRRT, due to their marked radiosensitivity (Bodei et al. 2004; Valkema et al. 2005). While the cross-fire effect is beneficial to overcome the inhomogeneity of receptor expression in cancer cells, allowing irradiation of tumor cells which are not directly targeted by the radiopharmaceutical, the long range of the ⁹⁰Y beta particles appears to further increase the potential for kidney toxicity. Cases of severe, end-stage renal disease have been reported with high activities of ⁹⁰Y at early stage of introduction of PRRT (Cybulla et al. 2001). However, adequate renal protection can minimize the risk of renal damage. Positively charged molecules such as l-lysine and/or l-arginine competitively inhibit the proximal tubular reabsorption of the radiopeptide and hence are co-administered with PRRT, having shown to reduce the renal dose by 9–53 % (Bernard et al. 1997; Rolleman et al. 2003). The dose reduces further up to 39 % when the infusion is prolonged over 10 h and up to 65 % by prolonging it over 2 days after radiopeptide administration, thereby covering almost the entire elimination phase through the kidneys (Bodei et al. 2004). The patients should also be well hydrated. Gelofusine, in addition to the amino acids, also inhibits the reabsorption of these peptides and has been shown to reduce the renal uptake significantly (Rolleman et al. 2008). The role of radioprotective drugs has also been investigated in preclinical studies. Amifostine has been demonstrated to alleviate the radiation-induced renal damage during therapy with ¹⁷⁷Lu-DOTATATE (Rolleman et al. 2007). The same group also recently established significant reduction in the mean absorbed renal dose with ¹¹¹In-DOTATATE when amifostine was co-administered, which could have mitigated the renal damage (Melis et al. 2012).

Despite renal protection, there is a risk of renal toxicity (usually mild) months after irradiation. A decline in creatinine clearance with time has been demonstrated, more pronounced with ⁹⁰Y-labeled peptides (median 7.3 % per year), but also for ¹⁷⁷Lu (median 3.8 % per year) (Valkema et al. 2005). Cumulative and per-cycle renal absorbed dose, age, hypertension, and diabetes were found to be contributing factors to the decline of renal function after PRRT. The maximum tolerated renal dose according to the studies with external beam therapy was considered to be 23–27 Gy (Cassady 1995). However, in internal radiation therapy, which involves different dose rates, biologically equivalent dose (BED) was found to more accurately predict the renal toxicity (Dale 2004). Barone et al. showed that the individual renal volume, dose rate, and fractionation, which are taken into account by BED, were able to predict renal function impairment (Barone et al. 2005). A much lower threshold of (BED) to the kidney, i.e., 28 Gy in patients with risk factors (mainly hypertension and diabetes), than in patients without risk factors (40 Gy) predicts renal toxicity after PRRT (Bodei et al. 2008).

A large retrospective study of 1,109 patients treated with ⁹⁰Y-DOTATOC by the Basel group found a 9 % prevalence of permanent renal toxicity (Imhof et al. 2011). This could have been due to the rather high activities of ⁹⁰Y administered per cycle (3.7 GBq/sq. m body surface area). Fractionation of PRRT, as in the case of external beam radiotherapy may be safer since internal radiation therapy leads to a continuous radiation delivery at a low and decreasing dose rate over time, which may allow the normal tissues like kidney to repair the radiation damage (Dale and Carabe-Fernandez 2005). With a long-term low-dose protocol, administering multiple (up to 8) cycles per patient and adequate nephroprotection (co-administering gelofusine additionally in case of ⁹⁰Y), no significant renal toxicity, i.e., end-stage renal insufficiency has been reported by the Bad Berka group in more than 1,000 patients treated with ¹⁷⁷Lu and/or ⁹⁰Y (>1,500 treatment cycles using ⁹⁰Y) (Baum and Kulkarni 2012). In 2009, two patients on hemodialysis due to end-stage renal insufficiency and with progressive NET after exhausting all other treatment options, were treated with fractionated low-dose PRRT using ¹⁷⁷Lu at our center (first ever worldwide experience with such a condition to the best of the authors’ knowledge). No significant hematotoxicity was observed in either of these patients on dialyses and both showed a good clinical and objective therapy response. A recent study of a cohort of 51 patients treated with up to 29 GBq of ¹⁷⁷Lu also revealed no major acute or delayed renal toxicity, demonstrating the safe use of ¹⁷⁷Lu (Bodei et al. 2011). PRRT should only be performed at specialized centers as NET patients need highly individualized interdisciplinary treatment and long-term care (Hörsch et al. 2011).