Journal of Nuclear Medicine, published on August 17, 2023 as doi:10.2967/jnumed.123.

265907
EDITORIAL

Is Actinium Really Happening?

Richard Zimmermann

Chrysalium Consulting, Lalaye, France; MEDraysintell, Louvain-la-Neuve, Belgium; and Oncidium Foundation, Mont-Saint-Guibert,
Belgium

T he most recent survey (1) related to a-radiotherapy develop-

ment showed that 27 molecules labeled with 225Ac are presently
Russian sources. In the United States, the company TerraPower
obtained access to larger amounts of 229Th that still need to be puri-
under development, among which 13 have already reached human fied and plans stepwise progress over the next 10 y, although the
test level. The first 225Ac-labeled molecule has entered the clinical output will still be only a small fraction of the future need. If alter-
phase III stage (2) and might reach the market by 2028. These mole- native routes become successful, this generator route will not
cules cover the most important indications that are studied with remain competitive, but for the time being, it remains the largest
b-emitting radionuclides, but it is obvious that each single 177Lu- source of very clean 225Ac.
labeled drug will be explored as a 225Ac-labeled analog. Among the
35 177Lu-labeled molecules that have already reached the clinical Technology B
stage, an estimated dozen have a high chance to reach the market The 232Th activation programs (United States and Canada) have
before 2030, not even taking into account all the generics. Actinium- progressed well. A very high capacity can be reached, and the tech-
labeled drugs will follow the same trend, with a delay of about 5 y. A nology might allow production of several terabecquerels per year.
global target of half a million patients represents only 1% of the 5-y Unfortunately, this product remains contaminated with 227Ac (half-
prevalence of cumulated cancers (Global Cancer Observatory; https:// life, 21.8 y). The mixture can be used for development purpose up to
gco.iarc.fr), which remains realistic in terms of share of the market clinical phase II without limitation, but cleaner forms of 225Ac will
compared with surgery, external radiotherapy, or chemotherapy. be preferred in routine applications and as marketed forms. Specifi-
Evaluation of further needs is based on today’s average patient cations limit the 227Ac threshold to 2%, a level that was demon-
dose of 100 kBq/kg. At least 10–12 MBq at end of bombardment strated not to affect patients (5,6). Release of radioactive waste from
must be produced per dose, taking into account losses during han- patients in the waste tanks of hospitals is the real issue. Although
dling and transport and labeling yields. On the basis of an average European authorities will probably recommend, if not constrain,
of 3 doses for a full treatment, each patient will need a total of users to avoid 227Ac, in the United States this radionuclide needs to
30–36 MBq of 225Ac at end of bombardment. In other words, be added to the radioactive waste and will directly affect the level of
3,000 GBq at end of bombardment would be sufficient to treat the decommissioning financial assurance that is supposed to be in
100,000 patients each year. Industry will have to guarantee capac- place at the user’s end. A similar situation was experienced in the
ity for 5–6 times this yearly amount by 2032. past when the industry was given a choice between no-carrier-added
Over the past few years, several large investments were made in 177
Lu and 177mLu-contaminated carrier-added 177Lu. 225Ac/227Ac
5 different technologies to develop large-scale production of might find better applications in the development of 225Ac/213Bi gen-
225
Ac. The different routes have been described in the literature erators, provided that the industry becomes interested in 213Bi.
(3,4). Already-operating sites and new-development units for tech-
nologies A–E are summarized in Table 1 with their production Technology C
capacities. In the thorium activation process B, 225Ra as a by-product can
easily be separated from the mixture, allowing indirect access to
THE 5 TECHNOLOGIES clean 225Ac through its decay. Unfortunately, yields are limited to
only 10% compared with the 225Ac/227Ac mix, generating high
Technology A
levels of waste and limiting financial attractiveness.
Carrier-free 225Ac has been produced through the natural decay
of 229Th. The 3 sites that can presently produce high-quality 225Ac Technology D
(United States, Russia, and Germany) will not significantly increase Accelerator production is possible by irradiating 226Ra targets
their production capacity. Only Russia is planning such capacity using small cyclotrons. Several large-scale production sites are now
improvement, but the additional amount will remain insignificant under construction, supported by companies in the United States
compared with the future demand. Fortunately, the war has not and Europe. Eventually, large amounts of 225Ac might be produced
(yet) altered access of 225Ac to U.S. patients in clinical trials from per week, in theory more than 4 TBq a month (7), but realistically
a tenth of this figure would allow us to stay on the safe side, with
Received May 1, 2023; revision accepted Jul. 25, 2023. cooling of larger targets becoming the limiting factor.
For correspondence or reprints, contact Richard Zimmermann (richard.
[email protected]). Technology E
Published online Aug. 17, 2023.
COPYRIGHT ß 2023 by the Society of Nuclear Medicine and Molecular Imaging. More recent developments have shown that photoconversion
DOI: 10.2967/jnumed.123.265907 technology not only is a way to generate very clean 225Ac but also

IS ACTINIUM REALLY HAPPENING?
Zimmermann 1

 TABLE 1
Technologies Under Development and Operating Sites for 225Ac Production, Including Estimated Present
and 2032 Capacities

Yearly production capacity (GBq/y/site)

Total (GBq/y)
Technology Source 2023 2032 in 2032 Comment

233 229 225

A: [( U!) Th! Ac] ORNL, United States 26 26 Up to 3,000 [80] Highest quality of nca
225
(generator) Ac; may enter price
competitivity
IPPE, Russia 37 (est.) 150 (est. 2025) to
300 (est. 2030)
JRC-ITU, Germany 11 11
TRIUMF, Canada 0.4 0.4
TerraPower, United .10 #2,700
States
Pantera, Belgium 0 .70 TerraPower source
B: [232Th(p,x)225Ac1227Ac] BNL/ORNL LANL; 16.7 Potential, .3,700 .9,000 [.240] Contaminated with
227
(high-energy accelerator) Tri-Lab, United States Ac (0.2% EOB -
1.5% at calibration);
not suitable for large
scale routine use
CNL/TRIUMF, Canada; .1 Idem .3,700
BWXT/ITM, United
States/Germany
INR, Russia #1,000
SpectronRx, United .1 .200
States
Others: Arronax, France; First GBq Potential,
IsoDar, Japan; in 2024 .200 each
CIAE, China
C: [1 … 1225Ra!225Ac] 10% of above; 0.3 .370 (theory) .370 [.10] High level of waste -
(as side product) CNL/TRIUMF, Canada expensive
226 225
D: [ Ra(p,2n) Ac] First GBq .4,500 [.120] Additional sites under
(cyclotron) evaluation in other
countries (Asia)
SpectronRx, United 2023 .500
States
Ionetix, United States 2023 1,900
Eckert&Ziegler, Germany 2024 550
Alfarim, Netherlands 2025 450–850
N-MediPhysics, Japan .2023 .500
KIRAMS, South Korea .2025 .500
226 225
D: [ Ra(d,3n) Ac] (linear Nusano, United States #160,000 Under evaluation
accelerator)
E: [226Ra(g,n)225Ra!225Ac] NorthStar, United States 2023 3,700–15,000 .37,000 [.1,000] Rhodotron: nca 225
Ac
(photoconverter)
Pantera, Belgium 2027 3,700–5,000
TerraPower, United 2029 3,700–5,000
States
225
Niowave, United States 2023 #18,000 Linac: nca Ac
Hitachi, Japan .2024 .3,700
226 225 225
F: [ Ra(n,2n) Ra! Ac] Nusano, United States #44,000 Under evaluation
(n from d on beryllium
target)

ORNL 5 Oak Ridge National Laboratory; nca 5 no carrier added; IPPE 5 I.I. Leypunsky Institute of Physics and Power Engineering,
Obninsk; est. 5 estimated; JRC-ITE 5 Joint Research Centre–Institute for Transuranium Elements; TRIUMF 5 TRI University Meson
Facility; BNL 5 Brookhaven National Laboratory; LANL 5 Los Alamos National Laboratory; EOB 5 end of bombardment; CNL 5 Canadian
Nuclear Laboratories; BWXT 5 BWX Technologies Inc.; ITM 5 Isotope Technologies Munich SE; INR 5 Institute for Nuclear Research of the
Russian Academy of Science; CIAE 5 China Institute of Atomic Energy; KIRAMS 5 Korea Institute of Radiological and Medical Sciences.
Data in brackets are curies.

2 THE JOURNAL OF NUCLEAR MEDICINE
Vol. 64
No. 10
October 2023

should allow large-capacity production. Facilities are under con- number of treated patients remains low. The situation becomes dif-
struction in the United States and Europe. ferent if millions of doses are used yearly and a fraction is dissemi-
nated in nature. We yet have a few years for the authorities at the
ISSUES TO BE SOLVED national level to think seriously about this issue, which is not speci-
Even if high capacity and reliable access to 225Ac is confirmed fic to 225Ac and will—over the long term—also affect other long-
from 2025 on, production of actinium still raises a series of other half-life radionuclides, including 131I, 177Lu, and 161Tb (8). This
issues, some of which are on their way to being solved. issue creates an opening for a next generation of radiotherapeutic
agents with shorter half-lives such as 67Cu, 211At, and 212Pb, but
Access to Larger Amounts of 226Ra this is another story.
Although access and handling of 232Th does not seem to be an
issue, access to larger amounts of 226Ra remains questionable.
CONCLUSION
Most of the companies are on their way to finding a solution,
either by getting access to domestic (waste) stocks or by extracting Altogether, the worldwide 225Ac production capacity by 2032,
radium from older devices (e.g., paintings, radiotherapy tools, and estimated largely above 25 TBq (670 Ci), should be sufficient to
older brachytherapy material). In the worst case, this issue should produce at least 2 million patient doses a year. With an additional
also be solved by 2025. investment remaining below $100 million (e.g., accelerator),
another 300,000 doses could be produced per new site. The situa-
Need for Additional Safety Investments
tion is much more comfortable than for 177Lu and 161Tb, for which
Handling of 226Ra is a more complex issue because this radio-
access to reactors will soon become the bottleneck. In the mean-
nuclide generates the gas 222Rn, which is difficult to store and
time, 225Ac-labeled radiopharmaceuticals are just the beginning of
trap. The increase in patients will proportionally lead to a higher
a success story spanning 15-plus years.
need for radium and higher production of radon, leading to the
need for additional safety investments.
DISCLOSURE
Risk of Explosion and Contamination
Cyclotron and photoconversion technologies may face the risk No potential conflict of interest relevant to this article was
of explosion of the radioactive radium target as cooling will remain reported.
difficult and the limiting factor, leading to potential contamina-
tion with long-half-life radionuclides. The increase in capacity
REFERENCES
will therefore be performed stepwise, and the upper limit remains
theoretic. 1. Goethals PE, Zimmermann R. Nuclear Medicine Report and Directory. 9th ed.
MEDraysintell; 2022.
Potential Toxicity and Limited Therapeutic Efficacy 2. Study of RYZ101 compared with SOC in pts w inoperable SSTR1 well-differentiated
225
Ac is an a-emitter that decays through a 6-step cascade to GEP-NET that has progressed following 177Lu-SSA therapy (ACTION-1).
ClinicalTrials.gov website. https://clinicaltrials.gov/ct2/show/NCT05477576. Published
stable 209Bi, releasing consecutively 4 a- and 2 b-particles. Recoil July 28, 2022. Updated July 12, 2023. Accessed August 2, 2023.
effects and the restricted capacity of chelators to trap decay pro- 3. Morgenstern A, Apostolidis C, Bruchertseifer F. Supply and clinical application of
ducts limit therapeutic efficacy to the first emitted a-particle. The actinium-225 and bismuth-213. Semin Nucl Med. 2020;50:119–123.
additional a- and b-particles released by the daughter radionu- 4. Radchenko V, Morgenstern A, Jalilian AJ, et al. Production and supply of a-particle-
clides decay elsewhere in the body. 223Ra shows a similar profile, emitting radionuclides for targeted a-therapy. J Nucl Med. 2021;62:1495–1503.
5. Jiang Z, Revskaya E, Fisher DR, et al. In vivo evaluation of free and chelated
and the associated drug (Xofigo) was approved by the authorities accelerator-produced actinium-225: radiation dosimetry and toxicity results. Curr
without a request for additional studies. Potential toxic effects sim- Radiopharm. 2018;11:215–222.
ply define the level of the maximum tolerated dose. 6. Sgouros G, He B, Ray N, et al. Dosimetric impact of Ac-227 in accelerator-
produced Ac-225 for alpha-emitter radiopharmaceutical therapy of patients with
Contamination Issues hematological malignancies: a pharmacokinetic modeling analysis. EJNMMI Phys.
It is still not clear whether patients eventually will be handled on 2021;8:60.
7. Robertson AK, Ramogida CF, Schaffer P, Radchenko V. Development of 225Ac
an outpatient basis or will be required to stay in shielded rooms,
radiopharmaceuticals: TRIUMF perspectives and experiences. Curr Radiopharm.
but patients treated with 225Ac (half-life, 9.92 d) cannot be kept suf- 2018;11:156–172.
ficiently long at therapeutic centers to collect all their waste until 8. Zimmermann R. Alpha-emitters from an industrial perspective: 2023–2032. Pre-
full decay. Releasing this waste in nature is not a problem if the sented at: TAT12 Conference; March 1, 2023; Cape Town, South Africa.

IS ACTINIUM REALLY HAPPENING?
Zimmermann 3