METHOD FOR THE QUANTIFICATION OF 227AC IN 223RA COMPOSITIONS

Abstract

A method for the quantification of .sup.227Ac in a .sup.223Ra composition comprising passing the composition through a first solid phase extraction column A, wherein said column comprises a thorium specific resin, passing the eluate of column A through a second solid phase extraction Column B, wherein said column comprises an actinium specific resin and recovering the .sup.227Ac absorbed onto the resin in column B and determining the amount thereof.

Claims

1-21. (canceled)

22. Apparatus for the quantification of .sup.227Ac in a .sup.223Ra composition comprising a first solid phase extraction column A, wherein said column A comprises a thorium specific resin, and a second solid phase extraction column B, wherein said column B comprises an actinium specific resin.

23. The apparatus of claim 22, wherein column A and column B are arranged in series.

24. The apparatus of claim 22, wherein the thorium specific resin comprises a phosphonate extractant.

25. The apparatus of claim 24, wherein the phosphonate extractant is an alkyl phosphonate extractant.

26. The apparatus of claim 22, wherein the thorium specific resin comprises a dialkyl alkyl phosphonate extractant of Formula I: ##STR00005## wherein each of R.sub.1-R.sub.3 is independently a C.sub.3-C.sub.8 straight or branched chain alkyl group.

27. The apparatus of claim 26, wherein the dialkyl alkyl phosphonate extractant is a dipentyl pentylphosphonate extractant

28. The apparatus of claim 22, wherein the actinium specific resin comprises a tetra-alkyl diglycolamide extractant of Formula II: ##STR00006## wherein R.sub.1-R.sub.4 are independently C.sub.3-C.sub.12 straight or branched chain alkyl groups.

29. The apparatus of claim 28, wherein the tetra-alkyl diglycolamide extractant is a N,N,N,N-tetra-n-octyldiglycolamide (DGA) extractant.

Description

FIGURES

[0085] FIG. 1Decay scheme of .sup.227Ac to stable .sup.207Pb. Branches with less than 2% probability are omitted.

[0086] FIG. 2Process flow-chart for actinium, thorium and radium separation and purification using the method of the inventionextraction is shown using aqueous HNO.sub.3 of particular concentrations by way of example only.

[0087] FIG. 3The decay scheme of .sup.225Ac and daughter radionuclides to stable .sup.209Bi.

[0088] FIG. 4HPGe ?-spectrum of .sup.225Ac daughters .sup.221Fr and .sup.213Bi

[0089] FIG. 5HPGe ?-spectrum of in-growth of .sup.227Th from .sup.227Ac 24 hours after separation from .sup.223Ra-chloride drug substance

[0090] FIG. 6Linearity of measured versus theoretical .sup.227Th amount in .sup.223Ra chloride drug substance

EXAMPLES

[0091] The .sup.223Ra composition utilised in the Examples is RaCl.sub.2 (Alpharadin), hereinafter referred to as Ra-chloride drug substance.

[0092] Gamma-spectra were measured with a High-Purity Germanium detector (HPGe) of 50% efficiency (relative to a 3 inch?3 inch NaI detector for a .sup.60Co source at a distance of 25 cm from the detector surface) coupled to a 8192-channel Multi Channel Analyser (MCA). Spectra were analysed using GammaVision Software (GammaVision-3.2 software, v 6.01, Ortec, Oak Ridge, USA). Calibration of the energy dependent efficiency of the HPGe detector at two fixed positions was performed with a reference source (?-mixed standard from Eckert & Ziegler) with an overall uncertainty below 4%. The fixed calibration positions were 5 and 20 cm from the detector. The HPGe detector was energy calibrated in an energy range from 59-1400 keV.

[0093] In order to evaluate .sup.223Ra with an activity in the MBq range, an ionization chamber (dose calibrator, Capintec-CRC15-R) was used. Accurate activity measurements of radionuclides using commercial ionization chambers require that the correct calibration setting (dial setting) must be applied. For many nuclides, the manufactures of dose calibrators, recommend those calibration settings.

[0094] .sup.223Ra is a relatively novel radionuclide in nuclear medicine and a calibration setting for the radionuclide is therefore not available from commercial manufactures of ionization chambers. A primary standardization of .sup.223Ra to establish dial settings was performed by the National Institute of Standards and Technology (NIST). The reason is to assure quality-controlled measurements of the radioactivity of .sup.223Ra during production, quality control and preparation of patient doses.

[0095] Measurements with .sup.223Ra Standard Reference Material (SRM) from NIST were performed in Capintec dose calibrators. The determined calibration setting (dial setting) for the dose calibrator used in the present invention is presented in Table 2.

TABLE-US-00002 TABLE 2 .sup.223Ra dose calibration setting .sup.223Ra Dose Calibration Capintec CRC-15R/serial no: Setting (dial setting) Serial no 157623, B-lab, Algeta 262

[0096] The instruments were qualified, which means verification that the instrument is installed correctly and is capable of operating as intended according to the specifications. Control of the HPGe instrument was performed daily before use by measuring the long-lived radionuclide .sup.226Ra. The radionuclides .sup.57Co and .sup.137Cs are used for daily control of the dose calibrator. The purpose of the quality control of the instruments is to ensure that the instrument provides reliable and consistent results.

[0097] Detection of .sup.227Ac is difficult due to the low energy of its ?-radiation and no useful ?-rays. Therefore, in order to easily obtain rapid information concerning the efficiency of the separation of .sup.227Ac from .sup.227Th and .sup.223Ra using the method of the invention the process was carried out using .sup.225Ac as a radioactive tracer in place of .sup.227Ac. .sup.225Ac can be presumed to have the same resin absorption properties as .sup.227Ac but is easier to detect. Examples 1-3 were carried out with .sup.225Ac and Examples 4 and 5 with .sup.227Ac. .sup.225Ac may be quantified by the in-growth of the daughter .sup.213Bi via ?-spectrometry.

[0098] Uncertainties were calculated as follows:

Uncertainty in the Recovery (Examples 1, 2, 3 and 5)

[0099] There is an uncertainty ?A in the activity of the known (spiked) samples, A.

[0100] There is an uncertainty ?.sub.B in the activity of the found sample (eluate), B.

[00002]$R=\mathrm{Recovery}.Math..Math.\left(\%\right).Math..Math.R=\frac{B}{A}.Math.100.Math..Math.{?}_R=\sqrt{\frac{{?}_B^2}{B^2}+\frac{{?}_A^2.Math.B^2}{A^4}}.$

Uncertainty in the Deviation (Example 4)

[0101] There is an uncertainty ?A in the calculated activity, A.

[0102] There is an uncertainty ?B in measured activity, B.

[00003]$y=\frac{A-B}{A}.Math.100.Math..Math.{{?}^{}}_y=\sqrt{{\left({?}_A\right)}^2+{\left(\frac{A}{B^2}.Math.{?}_B\right)}^2}$

Calculation of the Combined Uncertainty (Example 5)

[0103] There is an uncertainty ?.sub.A in the activity of the ingrowth of .sup.227Th after 24 hours, A.

[0104] There is an uncertainty ?.sub.B in the activity of the ingrowth of .sup.227Th after 48 hours, B.

[0105] Calculation of the combined uncertainty:

f=?{square root over (?.sub.A.sup.2+?.sub.B.sup.2)}

Example 1Separation of .SUP.225.Ac from .SUP.227.Th and .SUP.223.Ra Using Solid-Phase Extraction Columns

Sample Preparation, .SUP.225.Ac

[0106] The .sup.225Ac used was supplied from the Institute for Transuranium Elements, Karlsruhe, Germany. The solution had a nominal total activity of 6 MBq .sup.225Ac at the day of receipt and the activity was diluted in 4 mol/L HNO.sub.3 to an activity of 3.1 Bq/?L.

[0107] A known amount of .sup.227Th and .sup.225Ac was added to 15 MBq .sup.223Ra-chloride drug substance. The activity corresponded to approximately the specification limits of .sup.227Th and .sup.227Ac in a .sup.223Ra-chloride drug substance. The specification states that the activity of .sup.227Th should be less than 0.5% of .sup.223Ra activity and the activity of .sup.227Ac should be less than 0.004% of .sup.223Ra. Table 3 gives the activities of .sup.225Ac and .sup.227Th used in the experiments.

TABLE-US-00003 TABLE 3 Activities (Bq) of .sup.227Th and .sup.225Ac in experiment I and II. Nuclide Experiment I Experiment II .sup.227Th (kBq) 66.9 ? 2.2.sup.1) 74.6 ? 2.4.sup.1) .sup.225Ac (Bq) 497.7 ? 23.9.sup.1) 667.8 ? 33.4.sup.1) .sup.1)Uncertainty in the activity (2 ?)
Activities were determined using the following methods:

[0108] .sup.223Ra-chloride drug substance was transferred to two 20 mL vials (each with an activity of 15 MBq) and measured in a dose calibrator in dial setting 262. The ?-rays of .sup.225Ac and .sup.227Th solutions were measured with an HPGe detector in the calibrated position 5 cm and 20 cm, respectively.

[0109] For determining the areas of the ?-peaks in the energy spectrum, ORTEC GammaVision software was used. The energy and photon yield data are taken from Evaluated Nuclear Data File (ENSDFavailable from http://www.nndc.bnl.gov/nudat2/21 Jun. 2013). The most abundant ?-lines, given in Table 4, are used for the activity calculation.

TABLE-US-00004 TABLE 4 ?-ray energies and emission probabilities (percent) used for the determination of radionuclide activities. Data was taken from ENSDF. Library: Ra_223_DS_Ac_225_ENSDF_20_sep_2012.Lib Nuclide Energy Percent Half-Life Ac-225 99.80 1.0000 10 Days Th-229 124.55 0.6900 7932 Yrs. Th-229 136.99 1.1800 7932 Yrs. Ra-223 144.24 3.2700 11.43 Days Ra-223 154.21 5.7000 11.43 Days Th-229 156.00 1.1900 7932 Yrs. Th-229 204.70 0.6000 7932 Yrs. Th-229 210.85 2.8000 7932 Yrs. Fr-221 218.12 11.4000 4.9 Min. Th-227 235.96 12.9000 18.68 Days Ra-223 269.46 13.9000 11.43 Days Ra-223 323.87 3.9900 11.43 Days Th-227 286.09 2.4000 18.68 Days Ra-223 338.28 2.8400 11.43 Days Bi-211 351.06 12.9200 11.43 Days Rn-219 401.81 6.6000 11.43 Days Pb-211 427.09 1.7600 11.43 Days Bi-213 440.45 25.9400 45.59 Min. Ra-223 445.03 1.2900 11.43 Days Th-229 454.00 0.0100 7932 Yrs. Pb-211 704.64 0.4620 11.43 Days Pb-211 832.01 3.5200 11.43 Days

[0110] Production of .sup.225Ac is based on a .sup.229Th generator from which .sup.225Ac is eluted. The ?-lines from .sup.229Th used for activity calculation are seen in Table 4. No traces of .sup.229Th were found in the sample.

Preparation and Conditioning Procedure of the UTEVA and DGA Columns

[0111] The extraction-chromatographic resins as well as the prefilter material were packed in 2 ml disposable polystyrene plastic gravity-feed columns (obtained from Fisher Scientific). The following steps were carried out: [0112] Weigh in approximately 100 mg of UTEVA resin and 50 mg of DGA resin. [0113] Transfer the resin to two 20 ml plastic vials and add approximately 3 ml of 4 mol/L HNO.sub.3 to each. Swirl to mix. [0114] Prior to packing of the two columns, transfer a filter to the columns. Push the filter down to the base of the columns. [0115] Transfer the solution into the reservoir. Place a filter on the top of the UTEVA resin and DGA resin. [0116] Push the filter and the resin down. [0117] Discard the acid above the top filter. Add 2-3 ml of 4 mol/L HNO.sub.3. Discard the acid again. [0118] Remove the bottom plug from the columns. [0119] Add 2-3 ml 4 mol/L HNO.sub.3. Allow to drain.

Two-Column Separation Procedure

[0120] Place one UTEVA resin column and one DGA resin column in the column rack in series, i.e., solutions from the UTEVA resin column (on top) will drain into the DGA resin column (on bottom) (see FIG. 1). [0121] Based on .sup.223Ra-chloride drug substance radioactivity concentration, MBq/ml, pipette accurately a volume corresponding to 15 MBq into a 20 ml vial. Add an equal amount of 8 mol/L HNO.sub.3. [0122] Known activities of .sup.227Th and .sup.225Ac were added to the .sup.223Ra-chloride solution, according to Table 3. [0123] Use a plastic pipette to transfer the sample to the top of the UTEVA column. .sup.225Ac and .sup.223Ra will elute from the UTEVA column into the DGA column while .sup.227Th will absorb to the UTEVA column. [0124] Wash the columns with 5 ml of 4 mol/L HNO.sub.3. [0125] Disconnect the UTEVA column from the DGA column. [0126] Transfer 5 ml of 4 mol/L HNO.sub.3 onto the top of the DGA column. .sup.223Ra will elute from the column while .sup.227Ac will stay trapped on the DGA column. [0127] Elute the DGA column with 5 ml 0.05 mol/L HNO.sub.3. This will remove .sup.227Ac from the column.
It will be appreciated that in routine analysis (separation of .sup.227Ac, .sup.227Th and .sup.223Ra) no addition of .sup.227Th and .sup.225Ac to the .sup.223Ra -chloride solution was carried out, the solution was transferred directly to the UTEVA column.

[0128] Two separation experiments with .sup.225Ac were conducted. After complete separation, the columns and eluates were counted the day after separation with an HPGe detector. .sup.225AC has no suitable ?-rays and therefore it is quantified through its daughter .sup.213Bi. .sup.225Ac is in secular equilibrium with is daughters after 24 hours. .sup.225AC was identified through the measurement of the daughters .sup.221Fr and .sup.213Bi. A spectrum of the highly purified .sup.225Ac eluate is shown in FIG. 4. The ?-ray energies and intensities, used for the identification and quantification of .sup.225Ac and .sup.227Th, are presented in Table 4.

[0129] Percentage recovery for .sup.225Ac obtained from experiments I and II was 97% and 86% respectively. Results are reported in Table 5. These results show that the extraction resins UTEVA and DGA give an effective, reproducible, robust and rapid separation of .sup.225Ac from .sup.227Th and .sup.223Ra. The DGA resin shows a strong retention of .sup.225Ac in nitric acid and efficient stripping of .sup.225Ac in dilute nitric acid. As seen from Table 5, the breakthrough (recovery in Table 5) of .sup.227Th and .sup.223Ra in the final eluate was less than 2.Math.10.sup.?3 and 8.10.Math..sup.4, respectively. Moreover, only traces of the initial amounts of .sup.223Ra and .sup.227Th were detected in the eluates. This shows that the separation procedure of .sup.225Ac from .sup.227Th and .sup.223Ra with UTEVA and DGA column is highly effective. Thus, the method demonstrates that separation of .sup.227Ac, .sup.227Th and .sup.223Ra will also be effective.

TABLE-US-00005 TABLE 5 Activity (Bq) of .sup.225Ac, .sup.227Th and .sup.223Ra on the UTEVA and DGA column after separation and in the eluate obtained by ?-ray spectrometry. Uncertainty in the activity (2 ?). Activities added at the start of the experiment are seen in Table 3. Experiment I Experiment II .sup.225Ac .sup.227Th .sup.223Ra .sup.227Th .sup.223Ra (Bq) (kBq) (Bq) .sup.225Ac (Bq) (kBq) (Bq) UTEVA N.D. 68.4 ? 2 N.D. <11.6.sup.1 68.4 ? 2 N.D. DGA 23.3 ? 3 N.D. 20.5 ? 1 63.5 ? 1 N.D. <12.9.sup.1 Eluate.sup.2 483.6 ? 20 <0.2.sup.1 41.5 ? 4 74.7 ? 6 N.D. 84.4 ? 16 423.0 ? 20 <1.6.sup.1 21.6 ? 1 73.1 ? 5 <0.3.sup.1 8.4 ? 1 Recovery 97 ? 6 3 .Math. 10.sup.?4 2.8 .Math. 10.sup.?4 86 ? 6 2.1 .Math. 10.sup.?3 7.6 .Math. 10.sup.?4 (%) .sup.1minimum detectable amount (MDA) .sup.2In Experiment II, three fractions of 1.7 ml of the eluate were collected and counted

Example 2Testing of the Robustness of the Method

[0130] Robustness of the method was tested by using new batches (new lot number) of UTEVA and DGA resin. The robustness of an analytical procedure is a measure of its capacity to remain unaffected by small variations in method parameters and provides an indication of its reliability during normal usage.

[0131] A total of three experiments (I, II and III) were conducted with 15 MBq .sup.223Ra-chloride drug substance spiked with known activity of .sup.225Ac. The activity of .sup.225Ac is shown in Table 6. In addition, approximately 75 kBq of .sup.227Th was added. The separation was performed according to the procedure given in Example 1.

[0132] Percentage recovery of .sup.225Ac obtained after eluting with 5 ml 0.05 mol/L HNO.sub.3 was 68, 81 and 77%. The recoveries obtained were lower than for the UTEVA and DGA batches used in Example 1. It was therefore decided to increase the eluting volume (in addition to the 5 ml already added) by adding 2 times a volume of 2.5 ml of 0.05 mol/L HNO.sub.3. The results are presented in Table 6. The percent recovery of .sup.225Ac increased to 81, 85 and 84% and this is comparable to the previously obtained result. This shows that the method is robust with regards to new batches (new lot) of resins. However, the eluting volume may require adjustment to obtain sufficient recoveries (>70%).

TABLE-US-00006 TABLE 6 Measured .sup.225Ac activity (Bq). Testing of robustness of the method by using new batch of UTEVA and DGA resins. Uncertainty in the activity (2 ?). Experiment I Experiment II Experiment III .sup.225Ac spike 652.4 ? 47 609.4 ? 52 712.6 ? 50 Eluate 442.7 ? 35 491.2 ? 39 547.2 ? 43 (5 ml) Eluate 56.4 ? 1 23.7 ? 4 37.5 ? 1 (2.5 ml) Eluate 27.5 ? 8 <1.9 11.7 ? 6 (2.5 ml) Recovery (%) 81 ? 9 85 ? 10 84 ? 9

Example 3Testing of the Range of the Method

[0133] Analytical methods developed by a pharmaceutical company must be validated. The method should be validated in the range from reporting limit to at least 120% of the specification limit. In the previous experiments, the amount of .sup.225Ac has been added related to the specification limit. As the specification limit is 0.004% relative to .sup.223Ra, approximately 600 Bq of .sup.225Ac has been added to 15 MBq .sup.223Ra. The purpose of Example 3 was to add lower and higher activities of .sup.225Ac. The reason was to cover the range of amounts of .sup.227Ac which would be used during validation. Three experiments were conducted with various amounts of .sup.225Ac.

Results

[0134] Percent recovery of .sup.225Ac obtained after elution of 10 ml 0.05 mol/L HNO.sub.3 were 91, 74 and 92%. The results are presented in Table 7. This shows that the recovery is acceptable (>70%) from 46 to 172% of the specification limit.

TABLE-US-00007 TABLE 7 Measured .sup.225Ac activity (Bq). Three experiments with an .sup.225Ac activity ranging from 46-172% of the specification limit of .sup.227Ac. The uncertainties are given as 2 ?. Experiment I Experiment II Experiment III .sup.225Ac Spike 1033.3 ? 64 644.1 ? 32 278.3 ? 15 Eluate 872.8 ? 58 459.9 ? 38 241.0 ? 26 (5 ml) Eluate 70.5 ? 14 15.7 ? 6 15.7 ? 6 (5 ml) Recovery (%) 91 ? 8 74 ? 7 92 ? 12

Example 4Determination of Counting Conditions

[0135] Separation of 227Ac, .sup.227Th and .sup.223Ra was performed by UTEVA and DGA columns as described in Example 1, with the difference that a sample of 534?21 Bq (2?) .sup.227Ac was added to the column rather than the .sup.225Ac spike and supplemental .sup.227Th. Prior to separation, the sample was counted with a HPGe detector and quantified via its daughter .sup.227Th as .sup.227Ac was in equilibrium with its daughters for this sample.

Results

[0136] .sup.227Ac was eluted from the DGA column with 0.05 mol/L HNO.sub.3 and counted in the calibrated position 5 cm from the detector surface for 10000 s after approximately 1 and 2 days. Results are given in Table 8.

TABLE-US-00008 TABLE 8 Results from ?-ray spectrometry ingrowth of .sup.227Th from .sup.227Ac. The uncertainties are given as 2 ?. Activity Measured Deviation Days Calculated (Bq) activity (Bq) from Calculated (%) 1.08 21.0 ? 0.8 21.3 ? 1.2 ?1.4 ? 0.8 2.08 39.7 ? 1.5 39.2 ? 2.8 1.3 ? 1.5
As seen from Tabl, satisfactory counting uncertainties (<7%, 2?) were achievable for a sample counted for 10000 s in position 5 cm from the detector. There was no difference between calculated and measured activity.

Example 5Validation of the Method

[0137] Analytical methods developed by a pharmaceutical company must be validated. Analytical method validation is the process to confirm that the analytical procedure employed for a specific test is suitable for its intended use, i.e. to ensure reliability, consistency and accuracy of the analytical data. In order to demonstrate the applicability of the methods of the invention to commercial applications it was validated according to ICH Harmonized Guideline. Prior to a formal method validation it is mandatory to set up a protocol with test parameters to be evaluated and appropriate acceptance criteria.

[0138] The method was validated in terms of selectivity, accuracy, precision (repeatability/intermediate precision), linearity, range, limit of detection (LOD) and limit of quantification (LOQ). Robustness of the method was performed in Example 2 in terms of different lots of resins and was hence was not repeated in this Example.

[0139] The ICH guideline says nothing about acceptance criteria for the different parameters. However, accuracy in terms of recovery between 80-120% and a precision of ?20% is normally regarded as acceptable. This is applicable for impurities >0.1% of the active ingredient. As the specification for the impurity .sup.227Ac in .sup.223Ra-chloride drug substance is set as low as 0.004% relative to .sup.223Ra, a broader acceptance was required.

[0140] Samples of .sup.223Ra-chloride drug substance were spiked with known amounts of .sup.227Ac and .sup.227Th. Validation parameters and corresponding acceptance criteria for the method validation are given in Table 9.

TABLE-US-00009 TABLE 9 Validation parameters and acceptance criteria Validation Parameter Acceptance Criteria Selectivity The energies of the ?-rays of .sup.227Th are clearly resolved from the energies of the radionuclides potentially present in the matrix. Accuracy of .sup.227Ac as % recovery 70-130% Repeatability 60% of <30% (% RSD) specification (n = 3) 100% of <30% specification (n = 3) 140% of <30% specification (n = 3) Linearity, Correlation coefficient, r >0.95 LOQ (Bq) NA.sup.1 LOD (Bq) NA.sup.1 .sup.1NA = Not applicable

Experimental Parameters

[0141] According to ICH, accuracy and repeatability should be assessed using a minimum of 9 determinations over a minimum of 3 concentration levels covering the specified range (e.g. 3 concentrations / 3 replicates). The recommended range for validation of an impurity method is from reporting limit to at least 120% of the specification. Samples from 60-140% of the specification limit were prepared according to Table 10.

TABLE-US-00010 TABLE 10 Overview of samples used in method validation. .sup.223Ra is spiked with .sup.227Ac and .sup.227Th. The .sup.227Ac activity is from 60-140% of the specification limit. .sup.227Ac .sup.227Th .sup.223Ra Activity Activity Activity % of specification (Bq) (kBq) (MBq) Replicates 60 360 75 15 3 80 480 75 15 1 100 600 75 15 3 120 720 75 15 1 140 840 75 15 3

[0142] The method stipulates a sample size of 15 MBq. According to specifications, the amount of .sup.227Ac and the amount of .sup.227Th should be less than 0.004% and less than 0.5% relative to .sup.223Ra activity, respectively. As the decided validation range was from 60 to 140% of the specification, spikes of .sup.227Ac from 360 to 840 Bq were prepared. The content of .sup.227Th was held constant i.e. 75 kBq (0.5% of the specification). The .sup.227Ac and .sup.227Th stock solutions were both made in 4 mol/L HNO.sub.3 and the activities were approximately 5 Bq/?L and 0.5 kBq/?L, respectively. To determine the exact activity of .sup.227Ac spike solutions, counting with a HPGe detector in position 5 cm for 1000 s was performed. The counting time was selected to give a counting uncertainty (1?) of less than approximately 3%, which was regarded as adequate. Counting of .sup.227Th stock solutions was performed in position 20 cm for 300 seconds.

[0143] .sup.223Ra from three .sup.223Ra-chloride drug substance batches were pooled. Contamination of .sup.227Ac in the pooled batch was determined. An aliquot of 15 MBq was taken from the pooled sample and analyzed according to the method described in Example 1. This was done to ascertain if any correction needed to be made to the above results on account of background contamination of the samples with 227Ac. No traces of .sup.227Ac in the pooled .sup.223Ra-chloride drug substance were found. Hence, no correction was performed.

ResultsSelectivity

[0144] Selectivity is the ability of the measurement to assess the analyte without any interference from other components in the matrix. At the time of analysis, radionuclides present in .sup.223Ra-chloride drug substance have been separated from .sup.227Ac according to the method presented in Example 1. ?-decay of .sup.227Ac does not produce emissions of ?-rays that are appropriate for ?-detection. Traces of .sup.223Ra and its daughters can remain in the sample after purification and the selectivity of the method is demonstrated by comparing the energies of the ?-rays of .sup.227Th with the energies of .sup.223Ra and its ?-emitting daughters, .sup.219Rn, .sup.211Pb and .sup.211Bi, and by showing that they are clearly separated and identifiable. The ?-ray used for quantification of .sup.227Th is 236.0 keV, which is the most abundant ?-line of .sup.227Th (12.9%).

[0145] The ?-ray energies characteristic of .sup.227Th, .sup.223Ra and daughters are shown in Table 11. A spectrum acquired 24 hours after separation of .sup.227Ac from .sup.223Ra-chloride drug substance is shown in FIG. 5.

TABLE-US-00011 TABLE 11 ?-ray energies of .sup.223Ra and its daughters and .sup.227Th (Data taken from Evaluated Nuclear Structure Data File (ENSDF) database) .sup.223Ra .sup.219Rn .sup.211Pb .sup.211Bi .sup.227Th 144.2 154.2 210.6 236.0 256.2 269.5 271.2 286.1 300.0 304.5 323.9 329.9 338.3 351.1 401.8 404.9 427.1 704.6 832.0

[0146] As seen from FIG. 5, the ?-ray used for quantification of .sup.227Th is distinctly and visibly separated from the energies of other nuclides. There are no interferences from the matrix. The method is considered specific and the acceptance requirement was fulfilled.

ResultsAccuracy

[0147] The accuracy of the method was determined by performing recovery experiments on .sup.223Ra-chloride drug substance spiked with five levels of .sup.227Ac at 60%, 80%, 100%, 120%, and 140% of the specification limit of .sup.227Ac. The solutions were additionally spiked with .sup.227Th amounts corresponding to the .sup.227Th specification limit. For the 60%, 100% and 140% levels the solutions were prepared in three-fold. For the 80% and 120% level the solutions were prepared once.

[0148] Solutions were analyzed as described in Example 1. Each solution was measured twice. First measurement was performed 24?1 hour after sample preparation, the second measurement was performed 48?1 hour after sample preparation.

[0149] Accuracy as the percent recovery was determined using the measured .sup.227Ac content calculated as described in Equation 4.

[00004]$\begin{array}{cc}\mathrm{Recovery}=\frac{\mathrm{Measured}.Math..Math.\mathrm{content}}{\mathrm{Nominal}.Math..Math..Math.\mathrm{content}}?100& \left(4\right)\end{array}$

Results are presented in Table 12.

TABLE-US-00012 TABLE 12 Results for the accuracy (as recovery) of the method. Samples in the range of 60-140% of the specification limit for .sup.227Ac relative to .sup.223Ra. Nominal Calculated Diff between activity .sup.227Th after .sup.227Th after activity measurement .sup.227Ac 24 hours 48 hours .sup.227Ac 1 and 2 Recovery Level [%] [Bq].sup.1 [Bq].sup.1 [Bq].sup.1 [Bq].sup.2 [Days] [%].sup.2 60 323 ? 22.0 20.4 ? 2.7 31.6 ? 3.0 303 ? 4.0 1.0 94.0 ? 6.5 322 ? 22.5 10.9 ? 2.6 25.7 ? 2.9 397 ? 3.9 1.0 123.2 ? 8.7 359 ? 23.0 16.9 ? 2.0 26.5 ? 2.3 263 ? 3.0 1.0 73.3 ? 4.8 80 460 ? 27.6 16.9 ? 2.5 30.5 ? 2.9 369 ? 3.8 1.01 80.3 ? 4.9 100 540 ? 30.2 22.1 ? 2.0 42.7 ? 2.8 578 ? 3.4 0.98 107.0 ? 6.0 591 ? 31.9 24.0 ? 2.1 43.8 ? 2.9 556 ? 3.6 0.97 94.1 ? 5.1 527 ? 29.5 21.1 ? 3.2 40.1 ? 4.1 535 ? 5.2 0.97 101.5 ? 5.8 120 715 ? 35.8 26.7 ? 2.3 53.8 ? 2.9 655 ? 3.7 1.14 91.6 ? 4.6 140 868 ? 39.9 36.0 ? 4.9 62.8 ? 5.1 735 ? 7.1 1.00 84.7 ? 4.0 861 ? 41.3 31.0 ? 3.2 56.7 ? 3.6 706 ? 4.8 1.00 82.0 ? 4.0 803 ? 38.5 36.8 ? 4.6 61.0 ? 4.5 662 ? 6.4 1.00 82.4 ? 4.0 Mean recovery [%] (n = 11) 92.2 Relative standard deviation of the recovery [%] (n = 11) 15.5 Confidence interval (95%) of recovery [%] 82.2-102.1 .sup.1Uncertainty in the activity (2 ?). .sup.2Combined and recovery uncertainty

[0150] As seen from Table 12, the single percent recovery and the mean (n=11) is all within the criteria of acceptance (70 to 130%, see Table 9). The method is considered sufficiently accurate for the determination of .sup.227Ac content in the range from 60% to 140% of the specification limit which corresponds to 0.002% - 0.006% of .sup.227Ac in .sup.223Ra-chloride drug substance at release. The requirement is thereby fulfilled.

[0151] The uncertainties in recovery were in the range from 4-8.7%, comparable to 2? in the counting statistics, and it was the lowest activity which gave rise to the largest uncertainties. A contribution to the uncertainties is the uncertainty in the spiked value. This is not relevant for analyses of normal .sup.223Ra-chloride drug substance samples and hence the real uncertainties are lower.

ResultsPrecision

[0152] The repeatability of the method was determined by calculating the relative standard deviation (RSD) for three replicates of .sup.227Ac in .sup.223Ra-chloride drug substance at three different levels at 60% (corresponding to 0.002% of .sup.227Ac), 100% (0.004% of 227Ac), and 140% (0.006% of .sup.227Ac) of the specification limit. For each level, the solutions were prepared in triplicate and analyzed as described in Example 1. Results are presented in Table 13.

[0153] As seen from Table 13, the mean relative standard deviations were <30% for all three levels. The method is considered sufficiently precise and the acceptance requirement was fulfilled (see Table 9).

TABLE-US-00013 TABLE 13 Results for the precision of the method Recovery Mean RSD Level .sup.227Ac (n = 3) (n = 3) [%] [%] [%] [%] 60 94.0 96.8 25.9 123.2 73.3 100 107.0 100.9 6.4 94.1 101.5 140 84.7 83.0 1.7 82.0 82.4

ResultsIntermediate precision

[0154] Intermediate precision expresses within-laboratory variations in terms of e.g. different days, different analysts and different equipment. The intermediate precision was in this case determined in terms of different days. Separation was performed on four different days and the results are given in Table 14.

TABLE-US-00014 TABLE 14 Intermediate precision. Measured activity Level Ac-227 [Days] [Bq] 1 94.0 1 123.2 2 73.3 3 107.0 3 94.1 3 101.5 4 84.7 4 82.0 4 82.4 Mean recovery [%] (n = 9) 93.6 RSD [%] (n = 9) 16.3

[0155] As seen from Table 14, the mean relative standard deviation was <30% for all four days. Data show that the results from different days are comparable and that the acceptance requirement are thereby fulfilled.

ResultsLinearity

[0156] Linearity is the ability to generate a response which is directly proportional to the concentration of an analyte in a sample. To demonstrate the linearity of the method, a sample with an activity of 359?23 Bq .sup.227Ac was used. The sample was separated according to the procedure described in Example 1. The in-growth of .sup.227Th from .sup.227Ac was measured 6 times in a period from 1 to 6 days after separation. The corresponding theoretical activity was calculated using Equation 1. Results are presented in Table 15 and the plot of signals is displayed in FIG. 6.

TABLE-US-00015 TABLE 15 Results for the linearity of the method Time from Theoretical Measured activity separation .sup.227Th .sup.227Th [Days] [Bq] [Bq] 1.0 13.2 ? 0.8 16.9 ? 2.0 2.0 25.8 ? 1.6 26.5 ? 2.3 2.4 30.2 ? 2.0 30.5 ? 2.4 3.1 39.1 ? 2.5 43.7 ? 1.5 4.3 53.0 ? 3.4 48.0 ? 3.0 5.9 70.6 ? 4.5 67.0 ? 3.5 Regression line Slope 0.8629 Intercept [Bq] 5.4606 Correlation coefficient (r) 0.9797
The linearity curve was measured with .sup.227Th activities ranging from 17-67 Bq. This range covers the .sup.227Th activities which are measured from decay of .sup.227Ac in specification levels of 78-156% (100% gives an activity of 21.9 Bq after 24 hours and 42.9 Bq after 48 hours, see Table 1). The measured .sup.227Th activity is plotted as a function of the theoretical .sup.227Th activity. The correlation coefficient was determined to be r=0.98 and is well above the criteria of acceptance (>0.95). The method gives a linear response and the requirement is fulfilled (see Table 9).

ResultsRange

[0157] The method is validated in the specific range of .sup.227Ac content from 0.002% to 0.006% relative to .sup.223Ra with respect to activity (Bq). Linearity, accuracy, and precision of the method were demonstrated over a range of .sup.227Ac amounts as listed in Table 16.

TABLE-US-00016 TABLE 16 Tested range for linearity, accuracy, and precision of the method .sup.227Ac Validation characteristics [%] Linearity 0.003% to 0.006% Accuracy 0.002% to 0.006% Precision 0.002%, 0.004%, and 0.006%

ResultsLimit of Quantification and Limit of Detection

[0158] A part of a formal validation of the method is to determine limit of detection (LOD) and limit of quantification (LOQ). Limit of blank (LOB) is the highest apparent analyte concentration expected to be found when replicates of a blank sample containing no analyte are tested. LOD is the lowest analyte concentration likely to be reliably distinguished from the LOB and at which detection is feasible. LOQ is the lowest amount one can quantify with sufficiently good (and preselected) accuracy and precision. The ?-peak, 236 keV, which is the most abundant .sup.227Th peak, is used for quantification of the ingrowth of .sup.227Th from .sup.227AC. The LOD and LOQ were determined as described using the equations:

[00005]$\begin{array}{cc}\mathrm{LOD}=2.71+3.29.Math..Math.\left({B?\left(1+\frac{n}{2.Math.m}\right)}^{\frac{1}{2}}\right)& \left(5\right)\\ \mathrm{LOQ}=50.Math..Math.\left(1+{\left(1+\frac{\mathrm{nB}}{25.Math..Math.m}\right)}^{\frac{1}{2}}\right)& \left(6\right)\end{array}$

Where:

[0159] n=Number of channels in the peak region
m=Number of channels used for the background estimation
B=Background correction

[0160] Calculated LOD and LOQ from equations 5 and 6 are given in counts and the corresponding activity (Bq) was calculated using the following equation:

[00006]$\begin{array}{cc}{A}_E=\frac{N_E}{{\mathrm{.Math.}}_E.Math.t.Math.?}& \left(7\right)\end{array}$

A.sub.E : The activity in Bq of a nuclide based on a ?-peak with energy E
N.sub.E: the net peak area for a ?-peak at energy E (counts)
?.sub.E: the detector efficiency at energy E
?: emission probability
t: counting time

[0161] The LOD was calculated to be 1.8 Bq. This corresponds to 8% of the specification limit as the ingrowth after 24 hours from 100% of specification (600 Bq) corresponds to 22 Bq. The method is suitable to detect an .sup.227Ac content of 0.0003% (LOD). The LOQ was calculated to 7 Bq of .sup.227Th this corresponds to 32% of the specification limit. The method is suitable to quantify an .sup.227Ac content of 0.0013% (LOQ).

[0162] A summary of the validation results is presented in Table 17. Accuracy and precision were assessed on drug substance sample solutions spiked with .sup.227Ac activity ranging from 60 to 140% of the specification limit. 100% of the specification limits corresponds to 0.004% .sup.227Ac relative to .sup.223Ra.

TABLE-US-00017 TABLE 17 Summary of validation results Samples (% of Acceptance Validation Parameters specification) Criteria Results Accuracy as % recovery Spiked samples 70-130% 92.2% (average, n = 11) from 60-140% Correlation coefficient, r Samples from >0.95 0.98 78-156% Repeatability (% 60% of Spiked samples <30% 25.9% RSD) specification (60%) (n = 3) 100% of Spiked samples <30% 6.4% specification (100%) (n = 3) 140% of Spiked samples <30% 1.7% specification (140%) (n = 3) LOQ (Bq) Spiked samples NA 7 LOD (Bq) Spiked samples NA 2

[0163] As seen from Table 17, LOD is 2 Bq and LOQ is 7 Bq. This corresponds to approximately 8% and 32% of the specification limit, respectively. Investigation of the specificity shows that the ?-ray energy for quantification of .sup.227Th from .sup.227Ac is clearly resolved from interfering ?-ray energies. There are no interferences from the matrix.

[0164] All validation parameters met the pre-specified acceptance criteria. The method is considered suitable for its intended use.