- A list of facilities that produce molybdenum-99 for medical use, including an indication of whether these facilities utilize highly enriched uranium.
- A review of international production of molybdenum-99 over the previous 5 years, including whether any new production was brought online; whether any facilities halted production unexpectedly; and whether any facilities used for production were decommissioned or otherwise permanently removed from service.
This chapter provides information on the molybdenum-99 (Mo-99) supply chain as of June 2016 and Mo-99 production for the period January 20091 to June 2016. Although not explicitly requested by Congress, the chapter also describes plans by current global Mo-99 suppliers to expand their capacities to supply Mo-99 to the market. Plans to develop domestic (U.S.) supplies of Mo-99 are described in Chapter 4.
This chapter also provides current and potential future estimates of
1 The previous Academies report on Mo-99 production was completed in late 2008 and published in 2009 (NRC, 2009). The present report examines Mo-99 production trends from January 2009 to the present and Mo-99 supply disruptions from 2007 to the present.
Mo-99 production and supply. Sidebar 3.1 defines the quantities used to characterize production and supply.
Companies in Australia, Canada, Europe, South Africa, and the United States currently participate in the global supply chain for Mo-99/technetium-99m (Tc-99m) (see Figure 3.1). These include
TABLE 3.1 Target Suppliers for Global Mo-99 Production as of June 2016
|Target Supplier||Mo-99 Supplier||Target Type|
|CERCA (France)||IRE (Belgium)||HEU, LEU|
|CERCA||Mallinckrodt (Netherlands)||HEU, LEU|
|CERCA||NTP (South Africa)||LEU|
|CNL (Canada)||Nordion (Canada)a||HEU|
NOTES: CERCA = Compagnie pour l’Etude et la Réalisation de Combustibles Atomiques; HEU = highly enriched uranium; LEU = low enriched uranium; NTP = Nuclear Technology Products Radioisotopes.
a Nordion purifies, packages, and ships Mo-99 produced at CNL. It does not process targets.
- Five Mo-99 suppliers (see Table 3.3),
- Six technetium generator suppliers (see Table 3.4), and
- Several Tc-99m-labeled radiopharmaceutical suppliers.
The portion of the supply chain that provides Mo-99 and associated medical isotopes to the United States is also illustrated in Figure 3.1.
3.1.1 Target Suppliers
Between approximately 9,000 and 10,000 targets are used annually to produce Mo-99 for medical use; about 80 percent of these targets contain highly enriched uranium (HEU) and 20 percent contain low enriched uranium (LEU).2 Four companies currently supply these targets (see Table 3.1):
- Compagnie pour l’Etude et la Réalisation de Combustibles Atomiques3 (CERCA), France;
- Comision Nacional de Ernergia Atomica4 (CNEA), Argentina;
- Canadian Nuclear Laboratories (CNL), Canada; and
- Nuclear Technology Products (NTP) Radioisotopes (“NTP”), South Africa.
2 ANSTO and NTP collectively consume about 2,000 LEU targets per year.
3 The English translation is “Company for the Study and Production of Atomic Fuels.”
4 The English translation is “Atomic Energy Commission of Argentina.”
Brief descriptions of these suppliers are provided in the following subsections.
BWX Technologies, Inc., in Lynchburg, Virginia, previously supplied HEU targets for the MAPLE reactors at CNL. These reactors were constructed to produce Mo-99 for Nordion but were never put into commercial operation (see NRC, 2009, pp. 115-120, and Section 18.104.22.168 in this chapter).
CERCA is part of AREVA-NP (Nuclear Parts Center), a subsidiary of AREVA, a private company that is 80 percent owned by the French government.5 Its target manufacturing facilities are located in Romans, France. The company currently produces 93 percent uranium-235 (U-235) HEU and 19.75 percent U-235 LEU targets. Based on the company’s customer projections the company expects to discontinue HEU target production by the end of 2016 (see Chapter 5).
CERCA produces HEU and LEU targets for Institut National des Radioéléments (IRE) and Mallinckrodt Pharmaceuticals (“Mallinckrodt”) (these companies are described in Section 3.1.3); the United States supplies the HEU used in these targets;6 the LEU used in these targets is provided by various suppliers. CERCA also produces LEU targets for the Australian Nuclear Science and Technology Organisation (ANSTO) and NTP Radioisotopes. The LEU used in these targets is supplied by the United States (for ANSTO’s targets) and the Russian Federation (for NTP’s targets).
CERCA estimates that it currently supplies the majority of the targets used to produce Mo-99 for medical use. CERCA also estimates that it has the manufacturing capacity to supply all of the targets used globally for Mo-99 production.7
The committee estimates that CERCA currently supplies at least 60 percent of the targets used by global Mo-99 suppliers; the company’s proportion of target supply probably varies from year to year depending on the timing of target purchases by Mo-99 suppliers. The committee also estimates that CERCA’s share of the global market for target supply could increase to almost 100 percent once Nordion stops producing Mo-99 (after October 2016) and global Mo-99 suppliers convert to LEU targets (2019). Some global Mo-99 suppliers are seeking alternate sources of target supplies, which could reduce CERCA’s market share in the future.
Other target suppliers (described below) provide the remainder global
5 In 2017 the majority of AREVA-NP’s reactor business will be sold to Électricité de France; see http://www.world-nuclear-news.org/C-Areva-outlines-restructuring-plan-1506164.html.
6 CERCA receives HEU through a 1960 U.S./Euratom supply agreement.
7 Berndt Stepnik, CERCA, verbal communication, October 19, 2015.
TABLE 3.2 Reactors That Irradiate Targets for Global Mo-99 Suppliers as of June 2016
|Reactor||Country||Power (MWt)||Fuel Type||Target Type||Start of Operation (year)|
NOTES: BR-2 = Belgian Reactor 2; HEU = highly enriched uranium; LEU = low enriched uranium; HFR = High Flux Reactor; NRU = National Research Universal; OPAL = Open Pool Australian Lightwater; SAFARI-1 = South African Fundamental Atomic Research Installation 1.
a Most reactors do not have the capability to produce Mo-99 all of the days they operate.
b Production weeks per year is typically derived by dividing normal operating days per year by 7. However, this is not always the case. For example Maria runs short irradiation cycles that typically last less than a week, after which the reactor is stopped to remove the HEU targets for Mo-99 production. Therefore, the number of anticipated Mo-99 production weeks for Maria is 36 and not 29 (or else 200/7) (Kevin Charlton, OECD-NEA, written communication, September 15, 2015).
supply of targets used for Mo-99 production. The committee was unable to obtain reliable estimates of the percentages of targets supplied by each of these companies because this information is proprietary.
CNEA supplies 19.75 percent LEU targets to Argentina, Australia, and Egypt from its Constituyentes Atomic Centre in Buenos Aires. The LEU used in the targets is supplied by the United States.
|Reactor Operation License Expiration (year)||Normal Operating Days (days/year)a||Mo-99 Production (weeks/year)b||Available Production Capacity per week (6-day Ci Mo-99/week)c||Available Production Capacity per year (6-day Ci Mo-99/year)||Percent Global Available Production Capacityd|
c Available production capacity is the theoretical maximum capacity that can be normally produced. It is a measure of actual capacity and target usage levels.
d A reactor’s percent global available production capacity is calculated by dividing the available production capacity per year for that reactor by the total available capacity per year for the seven reactors (i.e., 998,750 6-day Ci of Mo-99/year) and multiplying the result by 100.
f For OPAL this is the estimated end of operation and not the reactor’s operation license expiration date.
SOURCE: Modified from OECD-NEA (2016).
22.214.171.124 Canadian Nuclear Laboratories
CNL is the private-sector entity that manages and operates Atomic Energy of Canada Limited’s (AECL’s) Chalk River Laboratories8 under a government-owned, contractor-operated arrangement. It produces 93 percent HEU targets for production of medical isotopes in the National Research Universal (NRU) reactor; these targets are processed by CNL and Nordion. The HEU used in these targets is supplied by the United States.
8 CNL is located near Chalk River, Ontario, Canada.
TABLE 3.3 Global Mo-99 Suppliers as of June 2016
|Supplier||Country||Target Type||Expected Conversion to Using LEU Targetsa (year)||Mo-99 Supply (weeks/year||Available Supply Capacity (6-day Ci Mo-99/week)||Available Supply Capacity (6-day Ci Mo-99/year)||Global Supply Capacity (%)|
NOTES: ANSTO = Australian Nuclear Science and Technology Organisation; HEU = highly enriched uranium; LEU = low enriched uranium; IRE= Institut National des Radioéléments; NTP= Nuclear Technology Products Radioisotopes.
a The reactor used to irradiate these targets may be fueled with LEU or HEU.
c Information gathered by the committee during site visit to ANSTO.
e Information gathered by the committee during site visit to IRE.
SOURCE: Modified from OECD-NEA (2016).
TABLE 3.4 Major Technetium Generator Suppliers as of June 2016
|Country||Generator Supplier||Generator Name||Target Type for Mo-99 Contained in Generator||Global Generator Market in 2015 (%)|
|United Kingdom||GE Healthcare||Drytec™||HEU||10|
|United States||Lantheus Medical Imaging||TechneLite®||LEU and HEU||20|
|Netherlands and United States||Mallinckrodt Pharmaceuticals||Ultra-Technekow™ DTE Generator||HEU||60|
|South Africa||NTP Radioisotopes||NovaTec-PTM||LEU||<2|
NOTES: ANSTO = Australian Nuclear Science and Technology Organisation; HEU = highly enriched uranium; LEU = low enriched uranium; IBA = Ion Beam Applications.
SOURCE: Global generator market in 2015 (right-most column) from Brown (2015).
NTP is located at the South African nuclear complex at Pelindaba. It supplies its own HEU targets (45 percent enriched) and purchases LEU targets (19.75 percent) from CERCA. The HEU is of South African origin and the LEU is currently of Russian origin.
3.1.2 Irradiation Services Suppliers
Almost all of the Mo-99 for medical use is currently produced by irradiating uranium targets in seven research reactors:
- Belgian Reactor-2 (BR-2), Belgium
- High Flux Reactor (HFR), the Netherlands
- LVR-15, Czech Republic
- Maria, Poland
- NRU, Canada
- Open Pool Australian Lightwater (OPAL), Australia
- South Africa Fundamental Atomic Research Installation 1 (SAFARI-1), South Africa
An additional research reactor, OSIRIS in France, produced Mo-99 for medical use until it was shut down in December 2015. The seven currently operating reactors listed above produce over 95 percent of Mo-99 used globally for medical use, and they produce all of the Mo-99 for medical use in the United States. The remaining (~5 percent) supply of Mo-99 for medical use is produced primarily for regional use in research reactors in other countries, for example Argentina, Egypt, Indonesia, and Russia.
Table 3.2 provides information about the seven currently operating reactors and their Mo-99 production capacities. Note particularly:
- Six reactors (BR-2, HFR, LVR-15, Maria, NRU, and SAFARI-1) irradiate HEU targets to produce Mo-99 for medical use; one of these reactors (BR-2) is fueled with HEU. About 75 percent of the global supply of Mo-99 is produced with HEU targets.
- Two reactors (OPAL and SAFARI-1) irradiate LEU targets to produce Mo-99 for medical use. About 25 percent of the global supply of Mo-99 is produced with LEU targets.
- All but one of these reactors (OPAL) were constructed in the 1970s or earlier. Their average age is 53 years. The OPAL reactor became operational in 2006.
- On average, each of these seven reactors operates about 240 days (~67 percent) per year. These reactors do not produce Mo-99 on all of the days they operate.
- The combined available production capacity (see Sidebar 3.1) for these seven reactors is almost 28,000 6-day curies (Ci) of Mo-99 per week when they are operating. This is about three times the current weekly demand for Mo-99, which is estimated to be about 9,000 6-day Ci per week (OECD-NEA, 2016; see Chapter 6).
Additional information about these seven reactors is provided in the following sections.
BR-2 is a 100 megawatt-thermal9 (MWt), tank-type,10 HEU-fueled, light-water-cooled research reactor located at Mol, Belgium. It is operated by the Belgian Nuclear Research Center (SCK•CEN)11 and is used primarily
9 MWt is the thermal (heat) output of a reactor. Research reactors are not used to generate electricity.
10 The reactor core is contained in a tank of water, and water is actively circulated through the core to remove heat.
11 Dutch: Studiecentrum voor Kernenergie; French: Centre d’Étude de l’Energie Nucléair; English: Belgian Nuclear Research Centre.
for the testing of reactor fuels and materials. BR-2 currently accounts for about 21 percent of global available production capacity.12 Mo-99 produced in this reactor is purified and distributed through IRE and Mallinckrodt. BR-2 also produces I-131 and Xe-133 for distribution by IRE. (These Mo-99 suppliers are described in Section 3.1.3 in this chapter.)
BR-2 started operation in 1961 and its current operating license expires in 2026. It was shut down in March 2015 for an 18-month refurbishment; the refurbishment work included replacement of the beryllium reflector, a critical component of the reactor.13 The work was completed in July 2016 and the reactor has resumed irradiating targets for Mo-99 production.14
BR-2 produces Mo-99 about 140 days per year. SCK•CEN is considering expanding the reactor’s capacity for producing Mo-99 by increasing its irradiation schedule to 190 days per year (Ponsard, 2015). This expansion, which could increase BR-2’s available production capacity from 7,800 to 10,530 6-day Ci per week, is expected to occur gradually.15
SCK•CEN is developing MYRRHA (Multi-purpose hYbrid Research Reactor for High-tech Applications), a multifunctional experimental irradiation facility, to replace BR-2. MYRRHA would be the world’s first liquid-metal-cooled nuclear reactor driven by a particle accelerator. Construction of MYRRHA is planned for the period 2017-2021 and full commissioning of the facility is scheduled for 2022-2024.16 MYRRHA would start producing Mo-99 after 2026.17
HFR is a 45 MWt, tank-type, LEU-fueled, light-water-cooled research reactor located in Petten, the Netherlands. It is owned by the Institute for Energy of the Joint Research Centre of the European Commission and is operated by the Nuclear Research and Consultancy Group (NRG). NRG also holds the reactor’s operating license. The reactor supports nuclear research and development, radioisotope production, and industry irradiation services.
HFR started operation in 1961 and was converted from HEU to LEU fuel in 2006. The reactor currently accounts for about 23 percent of global
12 Production capacity is the maximum capacity of the reactor to produce Mo-99 on a routine basis. Most reactors produce Mo-99 below their production capacity. The percentage of global available production capacity for each reactor describes the reactor’s share of total annual available production capacity (see Table 3.2).
14 Bernard Ponsard, SCK•CEN, written communication, August 2, 2016.
15 Bernard Ponsard, SCK•CEN, written communication, August 2, 2016.
17 Bernard Ponsard, SCK•CEN, written communication, August 2, 2016.
available production capacity, but its capacity will increase from 5,400 to 6,200 6-day Ci per week starting in 2017 (OECD-NEA, 2016). Mo-99 produced in HFR is distributed through IRE and Mallinckrodt. HFR also produces I-131 and Xe-133 for distrubtion by IRE.
HFR will probably be shut down when its current operating license expires in 2024. A privately funded replacement reactor, PALLAS, is planned to be built at Petten. PALLAS will likely be a 55 MWt, tank-type reactor. A licensable design is expected to be completed by 2017 and construction by 2023 (WNN, 2014). The funding sources for PALLAS have not yet been identified.
LVR-15 is a 10 MWt, tank-type, LEU-fueled, light-water-cooled research reactor situated at Research Centre Řež near Prague, Czech Republic. It is owned and operated by the Nuclear Research Institute Řež, plc. Its main missions are in materials testing, industry irradiation services, and radioisotope production. LVR-15 currently accounts for about 7 percent of global available production capacity. LVR-15 produces Mo-99, I-131, and Xe-133 for distribution by IRE. Mallinckrodt has an agreement with IRE to share access to this reactor in exchange for IRE’s access to the Maria reactor.
LVR-15 started operation in 1957 and its current operating license expires in 2028. The reactor converted from HEU to LEU fuel in 2010.
The Maria reactor is a 30 MWt, pool-type,18 LEU-fueled, light-water-cooled research reactor located at Świerk-Otwock, near Warsaw, Poland. It is operated by Poland’s National Center for Nuclear Research (NCBJ) (POLATOM) Radioisotope Center. Maria began irradiating Mo-99 targets in 2010 to help ease the isotope shortages due to unplanned shutdowns of the NRU and HFR reactors (WNN, 2010; see Section 3.3.1 in this chapter). It currently accounts for about 9 percent of global available production capacity. Mo-99 produced in Maria is distributed by Mallinckrodt. IRE plans to irradiate targets for Mo-99 production in Maria in the near future under the aforementioned access agreement with Mallinckrodt.
Maria started operation in 1974 and its current operating license expires in 2030. It was taken offline in 1985 for a complete redesign and resumed normal operations in 1993. It completed conversion to LEU fuel in 2012.
18 The reactor core is contained in an open pool of water and water is actively circulated through the core to remove heat.
The NRU reactor is a 135 MWt, tank-type, LEU-fueled, heavy-water-cooled and moderated research reactor located at CNL near Chalk River, Ontario, Canada. It is owned by the Canadian government and operated by CNL. The reactor is used for industrial and medical radioisotope production, neutron beam research, and materials research and development for CANDU power reactors.19
The NRU reactor started operation in 1957 and was converted from HEU to LEU fuel in 1991. It produces Mo-99 and Xe-133 from HEU targets and I-131 from tellurium targets. All of these isotopes are distributed through Nordion. NRU accounted for 40 to 60 percent of global available production capacity for several decades. It currently accounts for about 19 percent of available annual production capacity (Brady and Pruneau, 2015).
Two dedicated isotope production reactors, Multipurpose Applied Physics Lattice Experiment (MAPLE) 1 and 2, and a new dedicated target processing facility were constructed to replace the NRU reactor. These new facilities were scheduled to produce Mo-99 starting in 2000. The Mo-99 production capacity of these new facilities was to exceed the then-current global demand for Mo-99 (see OECD-NEA, 2010). However, the MAPLE reactors were never used to produce Mo-99 because of technical and regulatory problems that were deemed too expensive to address (IAEA, 2009). In 2008, AECL terminated the MAPLE project. In 2011, the operating license for the Chalk River Site, including the NRU, was renewed to October 31, 2016.
The Canadian government announced in 2015 that it would extend NRU’s operations through March 2018 “to help support global medical isotope demand between 2016 and 2018 in the unexpected circumstances of shortages” (NRCan, 2015). The Canadian Nuclear Safety Commission (CNSC) has approved the extension of the license to March 31, 2018. NRU will continue to operate between the end of October 2016 and the end of March 2018 but will not be used for routine production of Mo-99. CNL’s facility for processing irradiated targets will be kept in hot standby during this period as well.
The Canadian government is currently investing in four projects to produce Mo-99/Tc-99m for Canadian domestic consumption:
- Canadian Isotope Innovations (CII), a for-profit spinoff from Canadian Light Source, to produce Mo-99 via the 100Mo(γ,n)99Mo reaction (see Chapter 2) using linear accelerators. The company plans to start Mo-99 production in 2019 (De Jong, 2015).
19 CANDU (CANada Deuterium Uranium) power reactors are pressurized heavy water reactors developed by Canada and used commercially to generate electricity.
- Prairie Isotope Production Enterprise, a not-for-profit organization formed in 2009 to develop Tc-99m supply for the Canadian health care sector, and also to produce Mo-99 via the 100Mo(γ,n)99Mo reaction using linear accelerators. The company plans to start Mo-99 production in 2016 (Saunders, 2015).
- A consortium between Advanced Cyclotron Systems, Inc., CRCHUS (the research center of the Centre Hospitalier Universitaire de Sherbrooke), and the University of Alberta to produce Tc-99m via the 100Mo(p,2n)99mTc reaction (see Chapter 2) using cyclotrons. The consortium partners plan to start Tc-99m production in 2017-2018 (Guérin, 2015).
- TRIUMF, Canada’s national laboratory for particle and nuclear physics and accelerator-based science, also to produce Tc-99m via the 100Mo(p,2n)99mTc reaction using cyclotrons. TRIUMF plans to start production in 2017 (Buckley, 2015).
Tc-99m production by these companies would have to take place close to end users because of the short (~6 hour) half-life for this radionuclide. At least two of these potential Canadian suppliers (Canadian Isotope Innovations and Advanced Cyclotron Systems) intend to provide Tc-99m to the U.S. market provided they can meet U.S. regulatory requirements.
OPAL is a 20 MWt, pool-type, LEU-fueled, light-water-cooled research reactor located in Lucas Heights, a suburb of Sydney, Australia. OPAL is owned and operated by ANSTO. The reactor is used for isotope production, silicon doping, neutron activation analysis, and neutron beam research. All of the reactor’s activities are scheduled around medical isotope production. ANSTO increased its available production capacity in 2016 from 1,000 to 1,750 6-day Ci per week and currently accounts for about 8 percent of global available production capacity. It also produces I-131 using tellurium targets and other isotopes used in nuclear medicine. These isotopes are produced and distributed by ANSTO.
OPAL replaced the High Flux Australian Reactor (HIFAR), which produced medical isotopes from 1958 to 2007. OPAL started operations in 2007 and production of Mo-99 in 2009. The reactor’s estimated end of operation is 2055. OPAL is the first research reactor in the world to be built to use LEU fuel and LEU targets for production of Mo-99. It is also the youngest reactor in the world that is used to produce Mo-99 on a commercial scale.
SAFARI-1 is a 20 MWt, pool-type, LEU-fueled, light-water-cooled research reactor located at Pelindaba, South Africa. The reactor is owned and operated by South African Nuclear Energy Corporation (NECSA). It was initially used for nuclear physics research programs, but its primary purpose today is production of radioisotopes, mostly Mo-99. It currently accounts for about 13 percent of global available production capacity. Targets are irradiated in SAFARI-1 and Mo-99 is produced and distributed by NTP Radioisotopes, a subsidiary of NECSA.
SAFARI-1 was commissioned in 1965 and was built in cooperation with the Atoms for Peace program run by the U.S. Atomic Energy Commission.20 The reactor was initially fueled with HEU supplied by the United States. South Africa developed its own 45 percent HEU fuel for the reactor after the United States cut off HEU fuel exports in 1975. SAFARI-1 converted to LEU fuel in 2009 (NECSA, 2009) and started using LEU targets for medical isotope production in 2010. The first ever supply of commercial-scale, LEU-derived Mo-99 to the United States was made that same year from NTP (WWN, 2011). The reactor currently irradiates Mo-99 targets with 45 percent HEU and 19.75 percent LEU. About 48 percent of Mo-99 production was from LEU targets in 2015; 80 percent of Mo-99 production was from LEU targets in the first quarter of 2016.
OSIRIS is a 70 MWt, pool-type, LEU-fueled, light-water-cooled research reactor located at Saclay Centre, France. It started operation in 1966 and was shut down in December 2015 after a safety and performance assessment showed that the reactor would not be a reliable irradiation facility even after a major refurbishment (ASN, 2014). OSIRIS was operated by France’s Atomic Energy Commission (CEA) and was used for irradiation tests of nuclear reactor fuels and structural materials and irradiation of targets for Mo-99 production. OSIRIS accounted for about 5 percent of available production capacity (2,400 6-day Ci per week) at the time it was shut down. Mo-99 produced in OSIRIS was distributed by IRE.
The Jules Horowitz Reactor (JHR), currently under construction at the CEA Cadarache Centre in France, will replace OSIRIS. The reactor is expected to start operation in 2020. It will have a weekly Mo-99 production capacity of 4,800 6-day Ci and produce Mo-99 32 weeks per year (OECD-NEA, 2016). JHR will provide irradiation capacity for materials,
20 The U.S. Atomic Energy Commission was the predecessor of the U.S. Department of Energy and the U.S. Nuclear Regulatory Commission.
fuel, and other engineering testing to support nuclear reactor development and for Mo-99 and other radioisotope production.
3.1.3 Mo-99 Suppliers
There are five global suppliers of Mo-99 at present (see Figure 3.1):
- ANSTO, Australia
- IRE, Belgium
- Mallinckrodt, Netherlands
- Nordion, Canada
- NTP, South Africa
All of these suppliers provide Mo-99 to the United States (see Figure 3.1). Their combined available supply capacity (see Sidebar 3.1) is 15,880 6-day Ci per week (see Table 3.3) when all of the suppliers are operating, somewhat less than twice the current global demand for Mo-99, estimated to be about 9,000 6-day Ci per week (OECD-NEA, 2016; see Chapter 6). Their annual available supply capacity is about 773,000 6-day Ci per year. This quantity accounts for the number of operating weeks per year for each supplier.
Mo-99 is also produced for regional use by other suppliers. Some of these suppliers are described in Section 3.2.2 in this chapter.
ANSTO’s target processing facility is located at the company’s Lucas Heights facility, which is also the location of the OPAL reactor. It currently accounts for about 6 percent of available annual supply capacity (1,100 6-day Ci per week) from processing LEU targets irradiated in OPAL. ANSTO currently exports Mo-99 to several countries, including the United States, China, Japan, and South Korea.
IRE’s target processing facility is located in Fleurus, Belgium. It currently accounts for about 24 percent of available supply capacity (3,600 6-day Ci per week) from processing HEU targets irradiated in BR-2, HFR, and LVR-15.21 It plans to obtain irradiated targets from Maria, JHR, and Forschungsreaktors München II (FRM-II) reactor (Germany) in the future. FRM-II has been operating since 2005 and is currently undergoing modifi-
21 IRE also irradiated targets in OSIRIS until 2015.
cations to allow for the irradiation of LEU targets starting in 2018. IRE also supplies I-131 to Mallinckrodt and unprocessed bulk radiochemical Xe-133 to Lantheus Medical Imaging (“Lantheus”) for processing and sale.22
Mallinckrodt23 processes irradiated targets at the Petten site in the Netherlands in a joint venture with NRG, the operator of HFR. It currently accounts for about 24 percent of available annual supply capacity (3,500 6-day Ci per week) by processing targets irradiated in BR-2, HFR, LVR-15, and Maria. It may obtain irradiated targets from FRM-II and JHR in the future. Mallinckrodt purchases I-131 from IRE and Nordion for sale in the United States and other countries. The company plans to purchase Xe-133 from Nordion once its new Mo-99 production process is implemented (see Chapter 4).
As the committee was finalizing this report for publication, Mallinckrodt announced that it had entered into an agreement with Ion Beam Applications (IBA) Molecular, a technetium generator supplier in Europe (see Section 3.1.4 in this chapter), to sell its nuclear imaging business.24
Nordion has the largest Mo-99 supply capacity of the current global suppliers. It obtains bulk Mo-99 from CNL and purifies and distributes it at its Kanata, Ontario, Canada, facility. It currently accounts for about 29 percent of available annual supply capacity (4,680 6-day Ci per week), down from 40 to 60 percent prior to 2010.
Nordion will stop supplying Mo-99 once the NRU reactor stops production after October 31, 2016. The company is developing a new Mo-99 production process in cooperation with General Atomics and the University of Missouri Research Reactor Center. This process is described in Chapter 4.
NTP, a subsidiary of NECSA, produces Mo-99 from HEU (45 percent enriched) and LEU (19.75 percent enriched) targets, which are irradiated in the SAFARI-1 reactor and processed in an adjacent facility. It currently
22 Lantheus received Food and Drug Administration approval for Xe-133 sourced from IRE on June 10, 2016, and made the first commercial shipment on June 30, 2016.
23 Mallinckrodt was spun off from Covidien in 2013.
accounts for about 17 percent of available annual supply capacity (3,000 6-day Ci per week), about half of which was produced with LEU targets. The company also produces and sells I-131 to the global market.
3.1.4 Technetium Generator Suppliers
There are six major suppliers of technetium generators at present (see Table 3.4):
- ANSTO, Australia
- General Electric (GE) Healthcare, United Kingdom (UK)
- IBA Molecular, France
- Lantheus Medical Imaging, United States
- Mallinckrodt, the Netherlands and United States
- NTP, South Africa
These are not the only technetium generator suppliers in the world. For example, there are also suppliers in India, Japan, Poland, Russia, and Turkey.
Mallinckrodt and Lantheus supply about 80 percent of the technetium generators used globally (see Table 3.4) and most of the generators used in the United States. (GE Healthcare also provides technetium generators manufactured in the UK to their commercial nuclear pharmacies in the United States [Business Wire, 2014]). Both companies have generator manufacturing facilities in the United States; Mallinckrodt also manufactures generators in Europe. As noted earlier in this chapter, Mallinckrodt announced that the company has entered into an agreement with IBA Molecular to sell its nuclear imaging business. If this agreement is carried out, the number of major suppliers of technetium generators will be reduced by one.
Mallinckrodt supplies most of the Mo-99 used in its technetium generators but has backup supply arrangements with other global Mo-99 suppliers. Lantheus purchases Mo-99 in roughly equal proportions from ANSTO, IRE, Nordion, and NTP.25
The other technetium generator suppliers shown in Table 3.4 supply generators locally and regionally. For example, ANSTO supplies generators to China, Hong Kong, Indonesia, Myanmar, New Zealand, the Philippines, Singapore, Taiwan, Thailand, and Vietnam. GE Healthcare primarily distributes generators in Europe.
25 Ira Goldman, Lantheus Medical Imaging, written communication, June 3, 2016.
3.1.5 Tc-99m Suppliers
Tc-99m-labeled radiopharmaceuticals are supplied to hospitals and clinics from a large number of nuclear pharmacies. Supply is a local business. In the United States, there are four commercial nuclear pharmacies with national chain store operations: Cardinal Health™, GE Healthcare, Triad Isotopes, and United Pharmacy Partners (UPPI, LLC) (see Figure 3.1). The latter is a network of locally owned nuclear pharmacies and independent nuclear pharmacies.
Several Mo-99 suppliers have initiated or announced plans to increase their Mo-99 available supply capacities in the near future. Plans by existing suppliers are described in Section 3.2.1; plans by potential new suppliers are described in Section 3.2.2.
3.2.1 Existing Global Suppliers
Three existing global Mo-99 suppliers (ANSTO, Mallinckrodt, and NTP) have announced plans to expand their available supply capacities. If all of these plans are realized, available supply capacities would be increased by about 4,400 6-day Ci per week by the end of 2017 (see Figure 3.2). This added capacity would almost offset the loss of available supply capacity (4,680 6-day Ci per week) when Canada (NRU/Nordion) stops producing and supplying Mo-99 after October 31, 2016. Additional details are provided in the following subsections.
ANSTO plans to increase its available supply capacity from 1,100 to 3,500 6-day Ci per week by mid-2017. This increase in capacity will be accomplished by making additional irradiations and processing runs. Most of the increased supply is targeted for the United States. The Australian government has loaned the necessary funding to ANSTO to construct a new target processing facility and a new radioactive waste treatment plant to enable this increased supply capacity. Construction of these facilities is under way.
Mallinckrodt plans to increase its available supply capacity from 3,500 to 5,000 6-day Ci per week by the end of 2017. This will be accomplished
by increasing target processing runs from four to six times per week. The company is also considering irradiating additional targets. These could be irradiated in the FRM II reactor in Germany and/or the reactor currently under construction (JHR) in France.
NTP plans to increase its available supply capacity from 3,000 to 3,500 6-day curies per week by September 2017.26 The company may use a number of strategies to increase capacity, including irradiating additional targets and improving target processing efficiencies.
3.2.2 Potential New Global and Regional Suppliers
Additional new global or regional supplies of Mo-99 may become available from other countries in the future. Plans by Argentina, Brazil,
26 Tina Eboka, NTP, written communication, May 31, 2016.
South Korea, and Russia are described in the following subsections. Other countries may also have plans to enter the Mo-99 supply market, but these plans have not been publicized and are therefore not well enough known to be discussed here.
Argentina currently supplies about 400 6-day Ci per week by irradiating LEU targets at the RA-3 reactor, which is operated by CNEA. This reactor is scheduled to be shut down in 2027; a 30 MWt replacement reactor, RA-10, is expected to be operational in 2020 (OECD-NEA, 2016). This new reactor will support increased radioisotope production to cover future demands; provide fuel and materials testing irradiation facilities to support national technology development; and offer modern neutron applications in science and technology (Sánchez et al., 2014).
The RA-10 reactor is designed by INVAP (Investigación Aplicada), the Argentina nuclear technology company that designed and built the OPAL reactor (WNN, 2013). RA-10 is expected to start commercial production of Mo-99 with an available supply capacity at 2,500 6-day Ci per week to cover domestic and regional needs (OECD-NEA, 2015). This is an increase of 1,900 6-day Ci per week compared to existing capacity levels.
The Brazilian government has formalized the decision to build RMB, a 30 MWt, open-pool reactor that will be part of a new nuclear research center near the city of São Paulo (Perrotta and Soare, 2015). The project is part of a joint declaration with Argentina to develop multipurpose reactors and to demonstrate their common interest in promoting peaceful use of nuclear energy (Merco Press, 2013).
RMB will serve three purposes: neutron activation analysis; materials and fuels testing; and radioisotope production, including production of Mo-99. (Brazil currently imports all of the Mo-99 needed for medical use.) The reactor is expected to start commercial production of Mo-99 with an available supply capacity of 1,000 6-day Ci per week (OECD-NEA, 2016).
The RMB project is still at the design phase and further progress depends on funding. It is not expected to produce any Mo-99 before at least 2021.
126.96.36.199 South Korea
South Korea has been producing I-131 and Ir-192 on a commercial scale at HANARO (High-flux Advanced Neutron Application ReactOr)
since 1995 but has been importing Mo-99 from other countries (Wu et al., 2013). The country produces technetium generators for domestic use and export, and it also imports generators for domestic use.27
The Korean government began an effort in 2012 to build the Kijang Research Reactor (KJRR) to supply Mo-99, I-131, I-125, and Ir-192 (Wu et al., 2013). This 15 MWt, pool-type, LEU-fueled, light-water-cooled reactor will be constructed near Busan City by the Korea Atomic Energy Research Institute. Several facilities are being constructed: the reactor, a radioisotope production facility, an LEU Mo-99 target production facility, a neutron transmutation doping facility, and a radioactive waste treatment facility (Lim et al., 2011).
A construction permit for the reactor was in review by the regulatory body in December 2015. Production of Mo-99 is expected to start in 2020 (OECD-NEA, 2016). The available supply capacity for Mo-99 is expected to be 2,000 6-day Ci per week, with initial supply of 400 6-day Ci per week. The Mo-99 produced at KJRR will be supplied domestically and regionally.28
Russia has been producing Mo-99 for decades in two reactors: the WWR-TS reactor at the Karpov Institute in Obninsk and, starting from 2013, the RBT-10/2 and RBT-6 reactors at the Research Institute of Atomic Reactors (RIAR) in Dimitrovgrad. These reactors and associated Mo-99 processing facilities are owned and operated by the Russian government agency ROSATOM. All three reactors use HEU fuel and targets. Currently, the three reactors supply hundreds 6-day Ci per week of Mo-99 for domestic consumption (demand in Russia is about 100 6-day Ci per week; Zhuikov, 2014); regular export to Iran; and export on a trial basis to Canada, India, the Philippines, Poland, and Saudi Arabia (Khlopkov et al., 2014). Recently, Russia signed a contract with the National Nuclear Energy Commission in Brazil to supply Mo-99.29
The Russian government plans to become a global supplier of Mo-99 and to capture about a 20 percent share of the global market but has not announced a schedule for doing so. The government plans to increase Mo-99 supplies by irradiating targets at the RBT-10/2 and RBT-6 reactors at RIAR in Dimitrovgrad. There are two facilities at RIAR for processing
27 Jun S. Lee, Director, RI Research Division, KAERI, written communication, March 8, 2016.
28 Jun S. Lee, Director, RI Research Division, KAERI, written communication, August 20, 2015.
irradiated targets from RBT-10/2 and RBT-6 with a combined available supply capacity of 900 to 1,000 6-day Ci per week. In addition to the RIAR reactors, other reactors in Russia are being assessed for feasibility to supply Mo-99 on a commercial basis.
188.8.131.52 Other Potential Suppliers
Several other countries produce Mo-99 for domestic use and could potentially expand production in the future30:
Consumption of Mo-99 in these countries typically ranges from a few tens to about 200 6-day Ci per week. This demand is met by a combination of local production and imports. Other countries may produce Mo-99/Tc-99m exclusively for local use.
The global Mo-99 supply chain is inherently fragile. The fragility stems from three factors:
- Mo-99 and its daughter isotope Tc-99m have short half-lives (66 and 6 hours, respectively) and therefore cannot be stockpiled. These radioisotopes need to be produced and delivered to the supply chain on a weekly or more frequent basis.
- Global supply of Mo-99 relies on a small number of reactors (seven currently; Table 3.2) and a small number of suppliers (five currently; Table 3.3), as noted previously.
- With the exception of the OPAL reactor, which is only 10 years old, the remaining six reactors that are used to irradiate targets for Mo-99 production are on average 53 years old.
Mo-99 production has been interrupted unexpectedly on numerous occasions since 2009 because of unplanned shutdowns of these aging reactors. These interruptions have caused Mo-99 supply shortages and in some cases, severe shortages. Table 3.5 lists some of the notable interruptions
in Mo-99 supply due to planned and unplanned shutdowns of reactors and Mo-99 suppliers from 2009 to present. Some of interruptions were discussed in the previous Academies’ report on medical isotope production (NRC, 2009); see especially Chapter 3 in that report. Additional discussion is provided in the following subsections.
The committee is not aware of any interruptions of Mo-99 supply due to target supplier shutdowns (see, e.g., European Observatory on the Supply of Medical Radioisotopes ). However, there was a several-month interruption of technetium generator supplies from Covidien (now Mallinckrodt) in 2005 due to routine sterility assurance process revalidation. There was another interruption of Mo-99 supply in 2007 due to the shutdown of the Covidien technetium generator manufacturing facility for 1 month. The cause of the shutdown was a Mo-99 breakthrough in the company’s technetium generators and subsequent recall of these generators.
3.3.1 Reactor Shutdowns
Supplies of Mo-99 were repeatedly and severely affected in 2007 because of planned and unplanned shutdowns of two major reactors: NRU in Canada and HFR in the Netherlands. NRU shut down in November 2007 for a planned 5-day maintenance outage, which was then voluntarily extended by the reactor owner/operator (AECL) to install seismically qualified emergency power systems for two of the reactor’s cooling pumps. These upgrades were required for compliance with AECL’s (now CNL’s) 2006 operating license issued by the CNSC. The shutdown, which lasted for about 2 months, occurred without any coordination with other reactors and interrupted supplies of Mo-99 to North America. NRU resumed operations after the Canadian parliament passed emergency legislation authorizing restart before the seismic updates were complete, countermanding the CNSC (Ljunggren et al., 2007).
The NRU shut down again for 14 months starting in May 2009 because of a vessel leak and subsequent repair. HFR shut down at about that same time because of cooling system leaks. The HFR shutdowns occurred from August 2008 to February 2009, an additional month in 2009, and 6 more months in 2010. NRU and HFR provided about 40 percent and 30 percent, respectively, of global Mo-99 supplies at the time of these shutdowns. Their overlapping shutdowns caused major disruptions in global supplies in 2009-2010 and led to the cancellation or postponement of diagnostic imaging procedures in some countries, including the United States. Increased Mo-99 supplies from Europe could not offset these supply losses because there was insufficient target processing capacity.
The United States was among the countries most seriously impacted by the 2009-2010 supply shortages, likely due to the country’s heavy reliance
on NRU and Nordion for Mo-99 supplies (see Collier, 2008). Canada, Japan, and Korea were also severely affected.
Unplanned Mo-99 supply interruptions occurred on at least five occasions since 2009-2010 (see Table 3.5). In 2013-2014, for example, interruptions occurred because of the simultaneous shutdowns of two major reactors (HFR and SAFARI-1).
In March 2015, BR-2 started an 18-month refurbishment that was completed in July 2016. The refurbishment work included the replacement of the beryllium reflector (NEI, 2014). The shutdown of the BR-2 reactor was planned and did not cause any major interruptions in Mo-99 supplies. However, an unplanned outage in HFR in October 2015 resulted in some minor supply disruptions because BR-2 was not operating.
3.3.2 Target Processing Facility Shutdowns
Target processing facilities shutdowns are potentially more serious than reactor shutdowns because they can temporarily remove a larger share of global Mo-99 supply. This is particularly true for IRE and Mallinckrodt, the two global Mo-99 suppliers that obtain target irradiation services from multiple reactors in Europe.
There have been two major shutdowns of target processing facilities since 2009:
- In 2013, positive pressure in a hot cell caused a leak of radioactive noble gases at NTP’s target processing facility. No workers were injured, but the facility was shut down for 2 months while the problem was diagnosed and fixed.
- Mallinckrodt was shut down in 2013-2014 due to an unplanned outage at HFR.
IRE was also shut down for about 3 months in 2008 because of an unplanned release of I-131 to the environment. The leak went undetected for several days and resulted in cumulative release of about 45 gigabecquerels (GBq) of I-131, which exceeded the authorized annual release of I-131 from this facility (Federaal Agentschap voor Nucleaire Controle, 2008).
3.3.3 Transportation Denials and Delays
Mo-99 supply is interrupted frequently because of transportation denials and delays. These interruptions are typically resolved within hours or a few days and do not lead to severe interruptions in Mo-99 supply.
According to an IAEA analysis (Esarey, 2015), common reasons for transportation denials or delays include transit logistics problems involving
customs declarations; airlines being concerned about radiation or lacking information about how to handle radioactive substances; and confusing regulations on transportation of radioactive material. Additional reasons for refusing air transport of Mo-99 include the pilots denying the added freight weight.
Two other transportation-related events have caused Mo-99 supply shortages since 2009:
- The eruption of the Eyjafjallajökull volcano in Iceland in April 2010 grounded transatlantic air travel and air travel within Europe for a week. Mo-99 suppliers explored alternative transportation options to deliver Mo-99 (Triad Isotopes, 2010).
- The terrorist attacks in Belgium in March 2016 led the Belgian government to close all public transport in the country’s capital, Brussels, including the airport. The Belgian government prohibited transport of all radioactive material in the country for a couple of days, causing IRE to halt production for one day.
Since the 2009-2010 Mo-99 supply shortages, organizations participating in the global supply chain for Mo-99/Tc-99m have worked together to improve the stability of supply through the following four initiatives:
- Development of outage reserve capacity (see Sidebar 3.1) at several levels in the Mo-99/Tc-99m supply chain.
- Coordination of reactor and processing facility outages.
- Enhanced communications among supply chain participants.
- Creation of Mo-99 supplier alliances.
These initiatives are described in the following subsections.
3.4.1 Development of Outage Reserve Capacity
On average, research reactors operate 67 percent of the time annually (see Table 3.2). Under normal circumstances (i.e., when there are no unexpected or prolonged reactor shutdowns), these reactors have the capacity to irradiate sufficient targets to supply the 9,000 6-day Ci per week global demand for Mo-99.
Reactors can utilize their outage reserve capacity (ORC; see Sidebar 3.1) to produce additional Mo-99 to fill supply gaps created by outages in other reactors. This ORC has become especially important since 2009-2010 because extended reactor shutdowns have become more frequent (see
Table 3.5). ORC needs to be available on short notice—typically within about 48 hours—to fill supply gaps (OECD-NEA, 2013). There are a number of ways to create reserve capacity: for example, by dedicating additional days in a reactor’s schedule for irradiation of targets, or increasing the number of available positions in the reactor to irradiate targets.
Provision for reserve capacity is a key principle for ensuring a long-term secure supply of Mo-99/Tc-99m (OECD-NEA, 2011a). The OECD-NEA has proposed that Mo-99 suppliers should hold sufficient paid target irradiation reserve capacity at reactors to replace the largest irradiation services supplier in their supply chain. Organizations further down the supply chain should also hold a similar level of reserve capacity. This is described in the OECD-NEA report as the (n–1) criterion, which sets the desired capacity to meet Mo-99 demand at
Current demand + 35% Outage reserve capacity = Desired capacity
The desired capacity for Mo-99 suppliers is
9,000 6-day Ci per week + 0.35 (9,000) = 12,150 6-day Ci per week
A 35 percent ORC would not have been sufficient to prevent Mo-99 shortages in 2013, when two irradiation services suppliers (HFR and SAFARI-1) were in outage at the same time. The reserve capacity necessary to cover an (n–2) scenario (i.e., replacing the two largest irradiation services suppliers) has not been determined by OECD-NEA.31 According to supply chain participants surveyed by OECD-NEA, however, a 50 percent ORC would give an acceptable probability of a reliable supply of irradiated targets to the processor.
Reserve capacity levels based on the (n–1) or (n–2) criteria change with time as reactors are removed from service, new reactors offer irradiation services, and existing reactors expand target irradiation capacities. Reserve capacity levels can also differ between irradiation services suppliers and Mo-99 suppliers. In 2015, for example, technetium generator suppliers and Mo-99 suppliers had access to reserve capacities that were estimated to be on average about 14,800 6-day Ci per week with a large week-to-week variation (see Figure 3.3).
3.4.2 Coordination of Facility Outages
The Association of Imaging Producers and Equipment Suppliers (AIPES)32 works with irradiation services suppliers and Mo-99 suppliers to minimize Mo-99 supply disruptions by coordinating research reactor outage schedules. Schedule coordination began during the 2009-2010 reactor outages (see Section 3.3 in this chapter) and reduced the impacts of Mo-99 shortages in Europe. For example, the OSIRIS reactor in France was originally scheduled for a 5-month maintenance shutdown starting April 2010. The shutdown was postponed until mid-June 2010, however, because two other reactors (NRU and HFR) were shut down in early 2010 (see Table 3.5).
Following the 2009-2010 Mo-99 supply shortages, the coordination of reactor scheduling through AIPES has become more future looking (see Figure 3.4). AIPES periodically reviews reactor operating schedules and identifies periods with high potentials for multiple outages. Irradiation services suppliers and AIPES work together to minimize supply disruptions during these periods, either by modifying reactor outage schedules or utilizing reactor reserve capacities to fill Mo-99 production gaps. Communication of operating schedules is restricted to relevant stakeholders only.
AIPES also developed the VERSAILLES Mo-99 MODEL, an analysis tool to improve reactor operation scheduling and to assess the impacts of schedules on Mo-99 production. The model was validated using 2013 and 2014 data and is now being used to assess maximum global Mo-99 production capacity on a week-by-week basis (see Figure 3.3). The model helps to identify periods of increased potential for Mo-99 supply shortages by taking into account current and planned future reactor outage schedules. The model has been especially useful in planning reactor operating schedules to accommodate the extended scheduled shutdown of the BR-2 reactor for beryllium reflector replacement (February 2015-July 2016); the permanent shutdown of the OSIRIS reactor (December 2015); the planned cessation of Mo-99 production in the NRU reactor (after October 2016); and the expected 2016-2017 transition period to production of Mo-99 from LEU targets in Europe.33
3.4.3 Enhanced Communications
The 2009-2010 Mo-99 supply shortages demonstrated that frequent communications among supply chain participants can help to mitigate
32 AIPES provides a forum for addressing specific radiopharmaceutical issues, similar to CORAR (Council on Radionuclides and Radiopharmaceuticals) in the United States. It also lobbies the European Commission on issues related to diagnostic imaging.
33 Bernard Ponsard, SCK•CEN, written communication, February 11, 2016.
TABLE 3.5 Interruptions of Mo-99 Supply (2009-2016)
|Year(s)||Reactor or Processing Facility||Duration of Facility Shutdown||Planned or Unplanned Shutdown||Reason for Shutdown|
|2008-2009||HFR (Netherlands)||6 months||unplanned||Gas bubbles detected in the main cooling system|
|2009-2010||NRU (Canada)||14 months||unplanned||Reactor vessel welding repairs|
|2010||HFR||6 months||unplanned||Repair of a primary cooling pipework|
|2012-2013||HFR||8 months||unplanned||Repair of a primary cooling pipework|
|2013-2014||HFR||6 months||unplanned||Issue with control rod|
|2013-2014||NTP (South Africa)||2 months||unplanned||Positive pressure in a hot cell caused a leak of noble gases|
|2013-2014||SAFARI-1 (South Africa)||2 months||unplanned||NTP processing facilities shutdown|
|2013-2014||Mallinckrodt (Netherlands)||6 months||unplanned||HFR unplanned outage|
|2015||HFR||1 reactor cycle||unplanned||Maintenance|
|2015-June 2016||BR-2 (Belgium)||16 months||planned||Major refurbishment|
NOTES: BR-2 = Belgian Reactor 2; HFR= High Flux Reactor; NRU = National Research Universal; NTP = Nuclear Technology Products; SAFARI-1 = South African Fundamental Atomic Research Installation 1.
the consequences of supply disruptions. Technetium generator suppliers, for example, issued regular press releases during the 2009-2010 shortages to keep their customers (Tc-99m suppliers and the medical community) informed about actions being taken to address the Mo-99 supply shortages. The actions communicated included the following (Puthenedam, 2010; Triad Isotopes, 2009a,b):
- Entering into new Mo-99 supply agreements to maximize access to Mo-99.
- Providing projection calendars for technetium generator deliveries to assist with planning of medical procedures.
- Adjusting generator production and distribution schedules to meet customer needs.
- Providing advice to nuclear pharmacy operators and the medical community on how to maximize the availability of Tc-99m.
- Increasing production of alternate radiopharmaceuticals that could be used in place of Tc-99m radiopharmaceuticals.34
Nuclear pharmacy operators have created dedicated web pages explaining the events that led to the shortages as well as actions taken to address these issues. Actions included the following (Cardinal Health, 2009):
- Diversifying purchases of technetium generators across suppliers.
- Improving efficiencies of technetium generator elutions.
Professional societies such as the Society of Nuclear Medicine and Molecular Imaging35 provided recommendations for dealing with the Mo-99 shortages (SNMMI, 2010). The recommendations focused on maximizing existing Tc-99m supplies by scheduling patients around Tc-99m availability, use of low-dose imaging protocols, and utilization of alternate radiopharmaceuticals or procedures to replace Tc-99m-based procedures.
The communication enhancements that were initiated during the 2009-2010 Mo-99 shortages continue today.
3.4.4 Creation of Mo-99 Supplier Alliances
At present, ANSTO, IRE, and NTP support each other through backup-supply arrangements. Additionally, ANSTO and NTP are forming the Southern Radioisotopes Alliance Inc. (SRA) as a marketing and sales alliance for the supply of Mo-99.36 The alliance aims to help optimize supply routes and reduce decay losses. It also allows ANSTO and NTP to coordinate their reactor schedules and provide mutual backup capability. ANSTO and NTP will sell Mo-99 to the alliance, which in turn will supply it to customers. ANSTO and NTP also have an agreement to sell Mo-99 to
34 For example, Lantheus increased thallium production by almost 300 percent during the times of significant Mo-99 shortages to meet cardiac imaging needs.
35 Formerly the Society of Nuclear Medicine.
each other to meet supply needs during facility outages. NTP and IRE also have a separate Mo-99 supply arrangement, but IRE is not part of SRA.
FINDING 1A: As of June 2016, most (~95 percent) of the global supply of molybdenum-99 for medical use is produced in seven research reactors located in Australia, Canada, Europe, and South Africa and supplied from five target processing facilities in those same locations. The remainder (~5 percent) of the global supply is produced in other locations for regional use.
Information about the seven reactors that produce most of the Mo-99 for medical use, including their average Mo-99 production levels in 6-day Ci per week, is provided in Table 3.2. The global supply of Mo-99 is schematically illustrated in Figure 3.1.
FINDING 1B: As of June 2016, about 75 percent of the global supply of molybdenum-99 for medical use is produced by irradiating highly enriched uranium targets in six research reactors; one of these reactors is also fueled with highly enriched uranium. The remaining 25 percent of the global supply is produced by irradiating low enriched uranium targets in two research reactors.37
Table 3.2 provides information on HEU and LEU use in targets used to produce Mo-99 and in the fuel used in the reactors that irradiate these targets. Five reactors produce Mo-99 exclusively with HEU targets; one reactor produces Mo-99 exclusively with LEU targets; and one reactor produces Mo-99 using both HEU and LEU targets. One of the reactors used to produce Mo-99 with HEU targets is also fueled with HEU.
FINDING 2A: New molybdenum-99 supplies have become available since 2009, and expansions in available supply capacity are planned by current and new suppliers: A supplier in Australia (Australian Nuclear Science and Technology Organisation) has entered the global supply market and plans to expand its available supply capacity; existing global suppliers in Europe (Mallinckrodt) and South Africa (NTP Radioiso
37 One research reactor irradiates both HEU and LEU targets.
topes) have initiated plans to expand their available supply capacities; and the Russian Federation plans to become a global supplier.
FINDING 2B: Reactors in France (OSIRIS) and Canada (NRU) have halted or announced plans to halt molybdenum-99 production since 2009. These shutdowns have reduced/will reduce available production capacity and reserve production capacity that could be used to cover supply shortages if they occur.
There have been several changes in global supply of Mo-99 since the 2009 Academies report on medical isotope production (NRC, 2009). ANSTO became a global Mo-99 supplier in 2009 and now accounts for about 6 percent of available supply capacity (1,100 6-day Ci per week). The company plans to further expand Mo-99 supply in 2017, which would increase its available supply capacity by over a factor of three (to 3,500 6-day Ci per week). Mallinckrodt and NTP also plan to expand their available supply capacities by 1,500 6-day Ci per week and 500 6-day Ci per week, respectively, in 2017. If all of these plans are realized, available supply capacity would be increased by 4,400 6-day Ci per week by the end of 2017 (see Figure 3.2). This added capacity would almost offset the loss of available production capacity from NRU/Nordion (4,680 6-day Ci per week). However, if expansions are not realized on time, available supply capacity would fall below the desired capacity.
Other organizations have plans to expand or initiate supply of Mo-99. Argentina, Brazil, and South Korea are building new reactors to provide regional supplies of Mo-99. Russia plans to become a global supplier of Mo-99 and capture about a 20 percent share of the global market but has not published a schedule for doing so.
The OSIRIS reactor, which was permanently shut down in December 2015, and the NRU reactor, which will stop producing Mo-99 at the end of October 2016, have a combined available production capacity of over 7,000 6-day Ci per week. This is a little less than half of the current available production capacity for Mo-99 globally. The Canadian reactor also produces Xe-133 and I-131.
FINDING 2C: Molybdenum-99 production and supply were disrupted unexpectedly in 2009-2010 because of prolonged unplanned reactor and target processing facility shutdowns. These shutdowns caused protracted and severe molybdenum-99 supply shortages in the United States and some other countries. Shorter supply interruptions have also occurred as a result of shorter planned and unplanned reactor and target processing facility shutdowns and transport disruptions.
FINDING 2D: Coordinated actions taken by governments, molybdenum-99 suppliers, technetium generator suppliers, technetium-99m suppliers, and others since the 2009-2010 supply shortages have improved the resilience of the global supply chain, minimized supply disruptions during unplanned reactor and processing facility shutdowns, and increased molybdenum-99/technetium-99m utilization efficiencies. Supply vulnerabilities remain, however, owing to the small number of participating organizations at some steps in the supply chain.
Mo-99 production has been disrupted unexpectedly on numerous occasions since 2009. Some of these disruptions resulted in severe supply shortages in the United States and other countries. Disruptions in 2009-2010 occurred when Canada’s NRU and Europe’s HFR reactors were simultaneously shut down for extended periods. These shutdowns caused major disruptions in Mo-99 supplies and in diagnostic imaging procedures in some countries, including Canada and the United States. Supply interruptions have also occurred as a result of transportation denials and delays. However, these are typically resolved within hours or a few days.
Several actions have been taken since the 2009-2010 supply shortages to improve the resilience of the Mo-99 supply chain. These actions, which are described in Section 3.4, involve the development of ORC, coordination of reactor and target processing facility outages, enhanced communications among supply chain participants, and the creation of Mo-99 supplier alliances.
In spite of these actions, however, vulnerabilities remain in some parts of the supply chain owing to the small number of participating organizations. This is particularly true for the front end of the supply chain, where one company (CERCA) provides the majority of the targets used to produce Mo-99. This large market share also gives this manufacturer strong pricing power. The relatively small number of global Mo-99 suppliers is another potential point of vulnerability, particularly after one of them (Nordion) ceases supplying Mo-99 after October 2016. See Chapter 7 for additional discussion.