Otto is yet another example of the outstanding performance achieved by modern 68Ge/68Ga generators. The 100% increase in 68Ge activity at calibration when compared to any other reported 68Ge/68Ga generator, did not led to any measurable increase of 68Ge content, radiolysis products or other metal contaminants in the elution. In fact, the purity of the elution of this particular 68Ge/68Ga generator outperformed any other reported generator in the literature (Amor-Coarasa et al. 2016, 2017, 2018; McElvany et al. 1984; Roesch 2013; Greene and Tucker 1961). The 100% increase in 68Ge activity at calibration, resulted in only a 20–30% increase in eluted 68Ga activity, due to a decreased elution yield when compared to previously published reports of similar generators from the same manufacturer (Amor-Coarasa et al. 2016, 2017). These elution yields did not appear to change significantly over time, eluting more than 1 TBq (≈27 mCi) of 68Ga a year after calibration (Fig. 2). In our clinical setting, this increase in overall eluted activity allowed us to prepare multiple doses of [68Ga]Ga-PSMA-11 out of a single clinical production run, thus reducing the overall cost of drug production as well as the labor involved.
The decrease observed in both the 68Ge content and the elution yield cannot be explained since we had no part in the production of this generator, but we could speculate that it maybe is the result of incorporating an enlarged column to accommodate the higher initial activity. Interestingly, and contrary to previous reports, the amount of 68Ge breakthrough increased along with the generator age, as shown in Fig. 2. Nevertheless, the maximum 68Ge breakthrough observed (1.32 kBq or 0.036 μCi), accounts for only 8 × 10− 5% of the total 68Ge activity present in the generator at the time of elution, which is almost 100-fold lower than the one observed with previous generators at their purest 68Ge elution levels.
The elution profiles for both 68Ga and the 68Ge impurity changed over time. The highest activity concentration was initially found in fraction 3, and later moved to fraction 2. This change is again contrary to what was reported before for smaller generators from the same manufacturer, for which the elution profile was extended with time. While the elution profiles were determined in the 6 ml quality control elution, the elution yields were determined with all elutions (performed with both 6 and 4 ml). Hence, this initially extended profile could have reduced the overall yield measured when eluting with 4 ml 0.05 M HCl for labeling (Fig. 3a). This change in profile can also be partially responsible for the “stable” elution yield observed over time, as well as the minor elution yield variabilities here reported (Fig. 2).
The extended ICP-MS metal contamination study performed here revealed: i) the amounts of Iron contaminant found (main interference in the labeling of [68Ga]Ga-PSMA-11) were 10 times lower than the ones reported for previous 68Ge/68Ga generators from this manufacturer (Amor-Coarasa et al. 2016), ii) the Zinc contaminant was found in similar quantities to previously reported data for previous 68Ge/68Ga generators - most likely the direct result of accumulation due to 68Ga decay and iii) Of all other metals studied, Aluminum concentrations were always found to be the most prominent, however never exceeding 30 ng/ml (ppb). The amounts of metal contaminants did not change significantly during the studied period (p > 0.05) and did not showed a marked elution profile (p > 0.5, between fractions for all metals), which indicates that fractioning should perhaps be avoided as a purification method for this generator, given that there will not be a reduction in 68Ge amounts either (Fig. 3), and valuable 68Ga activity will be lost. Another important consideration is that the determination of metal contaminants presented in this report was based exclusively to quality control elution samples collected without the 24 h pre-elution that routinely precedes the clinical production runs of [68Ga]Ga-PSMA-11. Therefore, the concentrations reported herein for metal contaminants represent the “worst case scenario” and are estimated to be significantly lower in production elutions.
[68Ga]Ga-PSMA-11 syntheses were reproducibly performed with activity eluted from the 4.04 GBq (109 mCi) 68Ge/68Ga Generator and with an average radiochemical yield of 89 ± 4%. A few lower yield outliers could most likely be linked to operator manipulation errors. As stated before, the 68Ge breakthrough in the final drug product was found < 50 Bq/ml (< 5·10− 6% of 68Ge activity in the generator) at all instances, which is > 200 times below the acceptance criteria of 0.001% for [68Ga]Ga-PSMA-11. The waste vial from [68Ga]Ga-PSMA-11 production was found to contain the bulk of the 68Ge breakthrough from the elution. No radio or UV impurities were noticed in any of the 68GaPSMA chromatograms, and all batches showed > 99% radiochemical and chemical purity. The pure and reliable 68Ga produced by Otto resulted in a year of reproducible drug production for clinical use. Although typically the manufacturer specified shelf life of 68Ge/68Ga generators is set to 1 year due to the decrease of 68Ga elution yield and the parallel increase in 68Ge breakthrough (Amor-Coarasa et al. 2017), this type of 68Ge/68Ga Generators (Containing approximately 3.7 GBq or 100 mCi, Otto-like) could easily surpass it while still performing according to specifications.