Production of actinium-225 for alpha particle mediated radioimmunotherapy

Abstract

The initial clinical trials for treatment of acute myeloid leukemia have demonstrated the effectiveness of the alpha emitter ²¹³Bi in killing cancer cells. Bismuth-213 is obtained from a radionuclide generator system from decay of 10-days ²²⁵Ac parent. Recent pre-clinical studies have also shown the potential application of both ²¹³Bi, and the ²²⁵Ac parent radionuclide in a variety of cancer systems and targeted radiotherapy. This paper describes our five years of experience in production of ²²⁵Ac in partial support of the on-going clinical trials. A four-step chemical process, consisting of both anion and cation exchange chromatography, is utilized for routine separation of carrier-free ²²⁵Ac from a mixture of ²²⁸Th, ²²⁹Th and ²³²Th. The separation of Ra and Ac from Th is achieved using the marcoporous anion exchange resin MP1 in 8 M HNO₃ media. Two sequential MP1/NO₃ columns provide a separation factor of ∼10⁶ for Ra and Ac from Th. The separation of Ac from Ra is accomplished on a low cross-linking cation exchange resin AG50-X4 using 1.2 M HNO₃ as eluant. Two sequential AG50/NO₃ columns provide a separation factor of ∼10² for Ac from Ra. A 60-day processing schedule has been adopted in order to reduce the processing cost and to provide the highest levels of ²²⁵Ac possible. Over an 8-week campaign, a total of ∼100 mCi of ²²⁵Ac (∼80% of the theoretical yield) is shipped in 5–6 batches, with the first batch typically consisting of ∼50 mCi. After the initial separation and purification of Ac, the Ra pool is re-processed on a bi-weekly schedule or as needed to provide smaller batches of ²²⁵Ac. The averaged radioisotopic purity of the ²²⁵Ac was 99.6 ± 0.7% with a ²²⁵Ra content of ⩽0.6%, and an average ²²⁹Th content of (4₋₄⁺⁵)×10⁻⁵%.

Introduction

Radioisotopes which decay with the emission of alpha particles are rapidly becoming of greater interest for short range and site-specific therapy of cancers and micrometastic disease. Within the past five years, the investigation of targeted cancer therapy using alpha-emitters has considerably developed, and recent clinical trials have generated promising results and have attracted an increasing number of researchers and clinicians. The initial clinical trials in patients with acute myeloid leukemia have demonstrated the effectiveness of the alpha emitter ²¹³Bi in killing cancer cells (Jurcic et al., 2002). The 8-MeV alpha particles from ²¹³Bi can penetrate 6–10 nearby cell layers, killing everything in their short path (∼100 μm) including tumor cells. Recent preclinical studies have also demonstrated the potential application of both ²¹³Bi, and the ²²⁵Ac parent radionuclide, in a variety of cancer systems and targeted radiotherapy (McDevitt et al., 1999, McDevitt et al., 2001; Behr et al., 1999a, Behr et al., 1999b; Davis et al., 1999; Kennel et al., 1999; Kennel and Mirzadeh, 1998; Deal et al., 1999; Nikula et al., 1999; Chen et al. 1998; Kock et al., 1998; Kaspersen et al., 1995; Geerlings et al., 1993). A comprehensive review of the applications of alpha emitters is beyond the scope of present work and potential readers are referred to two recent review articles (McDevitt et al., 1998; Hassfjell and Brechbiel, 2001). The latter provides a useful account of the development of the bi-functional chelating ligands for the attachments of bismuth and actinium radioisotopes to proteins. Beside ²¹³Bi and ²²⁵Ac, the list of potential radionuclides for these applications includes only eight α-emitting radioisotopes, namely ¹⁴⁹Tb, ²¹¹At, ²¹²Bi, ²¹²Pb, ²²³Ra, ²²⁴Ra, ²²⁵Ra, and ²⁵⁵Fm (See for examples, Henriksen et al., 2002; Mirzadeh 1998; Howell et al., 1997).

In addition to the in vitro and in vivo stability of the radiolabeled biomolecules, there are a number of other factors that must be considered for potential clinical use of any radioisotope. These factors include availability, cost, nuclear characteristics, and chemical properties. This paper focuses on the availability and describes our five years of experience in production of ²²⁵Ac in partial support of the on-going clinical trials at the Sloan Kettering Memorial Cancer Center (SKMCC), where ²¹³Bi is generated in-house from the decay of ²²⁵Ac in a generator system (Ma et al., 2001; Bray et al., 2000; Mirzadeh, 1998; Wu et al., 1997; Boll et al., 1997a).

Actinium-225 is currently obtained from the decay of 7340-years ²²⁹Th, through one intermediate radionuclide, 15-days ²²⁵Ra (Fig. 1) (Apostolidis et al., 2001; Boll et al., 1997a, Boll et al., 1997b; Khalkin et al., 1997). After the initial separation of ²²⁵Ra and ²²⁵Ac from ²²⁹Th, a stock of purified ²²⁵Ra further provides a continuous but, of course, declining (with a 15 day half life) supply of ²²⁵Ac over a 45-day period. While the bulk of ²²⁵Ac, freshly separated from ²²⁹Th (up to 50 mCi), have been typically provided to SKMCC, smaller batches of ²²⁵Ac (20 mCi and less) from the decay of ²²⁵Ra has been supplied to a number of research institutes and universities throughout the US, as well as Europe and Australia.

Thorium-229 originates from ²³³U as the first alpha decay daughter, and both nuclides are members of the extinct Neptunium series (Fig. 1, Friedlander et al., 1981). Uranium-233, which has a fission characteristic similar to that of ²³⁵U, was produced as part of the US molten salt breeder reactor research program and is currently stored at ORNL. Irradiation of a natural thorium target in a nuclear reactor simply transmutes a non-fissile target to a fissile product; the neutron capture product of ²³²Th is short-lived ²³³Th, which beta decays quickly to ²³³Pa, then to ²³³U by the ²³²Th[n,γ]²³³Th(${\beta }^{-}$, $t_{1/2}=22.3\min$)²³³Pa(${\beta }^{-}$, $t_{1/2}=27\mathrm{days}$)²³³U reaction. Although ²³³U is presently the only viable source for high purity ²²⁹Th, the anticipated growth in demand for ²²⁵Ac may soon exceed the levels of ²²⁹Th present in the aged ²³³U stockpile. It is estimated that only ∼37 g or ∼8 Ci of ²²⁹Th (the theoretical specific activity of ²²⁹Th is 0.213 Ci/g based on a 7340 years half life) can be extracted from the entire ORNL ²³³U stockpile. This is about 50 times the current recovered ORNL ²²⁹Th stock. Considering the rather low annual in-growth production rate of ²²⁹Th from ²³³U and the increasing difficulties associated with ²³³U safeguards, routine processing of ²³³U is not presently feasible. Other means of production are being considered in order to meet the future demand for ²²⁵Ac and ²¹³Bi (Garland and Mirzadeh, 2003; Apostolidis et al., 2001; Mirzadeh, 1998).

The current ORNL ²²⁹Th supply is about 150 mCi. During the 1995–1996 period, ∼90 mCi of ²²⁹Th (batch “Th–W”) was extracted from the pre-existing waste material which was stored at ORNL for about 35 years. In the past three years, ∼60 mCi of ²²⁹Th (batch “Th–U”) has also been recovered from chemical separation of stored ²³³U oxides. In the 1960s, after recovery of ²³³U, the processed Th target (which now includes ²²⁹Th) was stored in stainless steel waste tanks. The initial purification of the ²²⁹Th from the waste material began at ORNL in 1995 and was solely supported by internal ORNL funding. In 1994, about 30% of the sludge from the waste tank was purchased by PharmActinium, Inc. (Alphamedical Holding BV, Amsterdam, The Netherlands) and was sent to the Institute for Transuranium Elements (ITU, Karlsruhe, Germany) where the Th was purified. Since then, the ²²⁵Ac from ITU has been routinely sent to MSKCC also in partial support of the clinical trials (Jurcic et al., 2002; Apostolidis et al., 2001).

This paper describes a sequence of chemical steps consisting of both anion and cation exchange chromatography for routine separation of carrier-free Ra radioisotopes and carrier-free ²²⁵Ac from macro amounts of Th (a mixture of ²²⁸Th, ²²⁹Th and ²³²Th). The paper further outlines the separation of ²²⁵Ac from a mixture of ²²⁴Ra and ²²⁵Ra.