|Year : 2012 | Volume
| Issue : 3 | Page : 105-110
Evaluation of internal hazards in medical cyclotrons
Ex-BARC & Ex-IAEA; Consultant, Atomic Energy Regulatory Board, Mumbai, Maharashtra, India
|Date of Web Publication||5-Sep-2013|
M R Iyer
Bungalow D4, Raj Kunj Society, Chembur, Mumbai 400074
Source of Support: None, Conflict of Interest: None
A large number of medical cyclotrons of various types and makes are now in operation in India and their number is ever increasing. A careful analysis of the various safety issues particularly in case of possible accidental conditions is required for a realistic evaluation of their impact. Apart from the external dose involved, internal dose issues under normal operation and in case of abnormal operational conditions such as target rupture, accidents, spills, etc., which are generally neglected also need evaluation. There is a need for carrying out worst scenario analysis and the possible dose consequences to the operating staff as well as the public due to releases through the stack. The paper carries out an analysis for a typical release of activity into the vault environment and the dose implications. Along with any measured air activity measurements in the vault of operating cyclotrons this would resolve the issue one way or other. During radio pharmaceutical processing a substantial fraction of the volatile positron emitting radiopharmaceuticals are released into the atmosphere. In some cyclotrons a provision of an air trap for holding the air is mentioned. Analysis of possible dose to a member of the public using typical release rates is also presented and shown to be not negligible. A short review of such analysis in literature is carried out to show that the possible internal dose consequences cannot be ruled out and need to be addressed to in the safety analysis of these facilities for regulatory controls. Methods for proper calibration of stack monitors are indicated. In case of location of medical cyclotrons in crowded areas replete with high rise buildings, it may be necessary "to insure engineered safety features to ensure zero discharges from the machines."
Keywords: Internal dose, medical cyclotron, radiation safety
|How to cite this article:|
Iyer M R. Evaluation of internal hazards in medical cyclotrons. Radiat Prot Environ 2012;35:105-10
| Introduction|| |
Medical Accelerators are used for production of positron emitting positron emission tomography (PET) isotopes. Most machines are used to produce 18 F, though 11 C, 13 N, 15 O, etc., are also produced often and rarely 124 I, 64 Cu, 99m Tc etc., In this paper, we deal with mainly fluoro-deoxy-glucose (FDG) production since this is more common . Risk analysis in machines producing longer lived isotopes need to be addressed to separately.
| Types of Medical Accelerators and Features|| |
Medical cyclotrons of are of two types:
Self-shielded machines are generally of 11-16 MeV, 60 uAmp. These are with retractable interlocked shield located in a vault with concrete walls of about 50 cm and entrance door interlocked with machine operation.
Bunker type machines of 16-30 MeV are located in a vault and the irradiation ports may be located in caves, housed in a vault with concrete shield walls of generally 2 m thick and an entry maze to meet the regulatory stipulation of design dose rate of 1 uSv/h in all accessible areas.
Medical cyclotrons are equipped with chemical processing modules in separate areas. The operations, mostly automatic are carried out remotely behind shielded cells equipped with lead glass windows. The sample transfer is carried out pneumatically through tubes embedded under floor with sufficient shielding. The synthesized radio pharmaceuticals are dispensed automatically into vials and loaded into shielded casks for transport and patient administration.
In general, four curies of activity is produced in a single lot. Small amount of activities are transferred to quality control laboratory for quality control purposes. These involve steps of manual handling and have potential for contamination.
The vault air and the chemistry module air are provided with ventilation. The air containing the radioactive materials escaping into the vault and chemistry module environment are discharged through a stack after passing through filter banks. In some brands of synthesis modules the air is passed through a nitrogen trap to remove the airborne activity in the cell and then released through the stack.  As an additional precaution some facilities have pressurized hold up tanks through which air is passed. Regulatory agencies in some countries seem to insist on this.
Apart from the activity produced in the target, gaseous activation products such as 41 Ar are also discharged through the stack from the cryotron vault air. The stacks are generally equipped with absolute filters and charcoal filters.
A stack of about 2.5 times taller than the surrounding structures is generally recommended. Medical cyclotrons are generally located in the vicinity of public locations. No definite regulatory guidelines exist on locating these machines.
A careful analysis of the various safety issues particularly in case of possible accidental conditions is required for a realistic analysis of their impact. Apart from the external dose involved, internal dose issues under normal operation and in case of abnormal operational conditions such as target rupture, accidents, spills etc., also need evaluation. There is a need for carrying out worst scenario analysis and the possible dose consequences to the operating staff as well as the public due to releases through the stack.
The target is in the form of water enriched in 18 O, which forms 18 F by (p, n) reaction. Target rupture can happen due to improper cooling or other reasons. A target rupture is considered to be an event, which calls for close follow-up to ensure that the operating personnel do not receive undue dose. External dose is controlled by controlling entry into the vault after a target rupture until the radio isotope has decayed for around seven half-lives. Only the external dose is checked before entry. Possible internal dose due to inhalation is not considered generally, but needs evaluation.
During radio pharmaceutical processing a substantial fraction of the volatile positron emitting radiopharmaceuticals are released into the atmosphere.  Rigorous quantification of the activity discharged through the stack is not always available and monitors are often not properly calibrated. In general, a ventilation rate of 10 air changes is designed in most cyclotrons, but no results of evaluation of efficacy of the ventilation are available.
| Biokinetic Aspects of 18 F and Dose Conversion Factors|| |
Internal dose results by inhalation, ingestion, injection or absorption through skin of radioactive materials. Unlike in the case of external radiation the radiation dose cannot be estimated directly in case of internal doses. There can be inhomogeneous uptake in various organs of the radionuclide depending on the bio-kinetic aspects of different organs. Further even in the case of the same radionuclide the absorption and excretion depends on the chemical nature of the radionuclide. This give rise to the need of modeling of the body compartments with its bio-kinetic parameters. The dose depends on the mass of the organ, which again differs from individual to individual and hence the dose calculation uses average organ weight for the population. In general, the internal dose calculations need inputs from body excretion measurement of the radionuclide known as bio assay or by means of measurements of the radioactive emissions from the whole body using highly sensitive whole body counting and spectrometry. Various organs have different radio sensitivity and the whole body dose is obtained by weighting the organ dose with the weighting factors. The radiations have also different weighting factors depending on the specific ionization and are different for alpha beta and gamma radiations doses. Equivalent dose is
HT = ΣWRDT,R (1)
W R is radiation weighting factors for radiation R and for organ T.
Effective dose = ΣWTHT (2)
W T is tissue weighting factor for organ T.
The radionuclide resident in the body continues to irradiate the tissues until it is completely excreted or decayed and integrated dose for 50 years of active life span of an individual is considered. The calculation of dose is depending on physiological parameters of source and target organs and is arrived at using models. The resident radioactivity is governed by its physical half-life and also the biological half-life, a combination of these two is known as effective half-life. Usually, the biological half-life is much lower than physical half-life. For inhalation dose special lung models have been developed by International Commission on Radiological Protection (ICRP). Another input to this is the particle size of the inhaled particles. The gastrointestinal tract model uses various compartments for the different internal organs such liver stomach is used. Metabolic models make use of the kinetic of the substance after it has entered the blood stream.
The bio kinetics of 18 F is well-studied and available in the literature. Though, it is relatively short lived and merits little attention the studies have been carried out in detail since it is very widely used in PET studies. The studies are mostly concentrated on the ingestion route. FDG is a glucose analog used in the characterization of glucose metabolism for myocardium investigations and cerebral metabolism and in the diagnosis of cancer diseases. A fraction of 0.3 of the activity in other organs and tissues is considered to be excreted in urine with a biological half-life of 12 min (25%) and 1.5 h (75%) following the ICRP kidney bladder model. There is a significant variation in dose coefficient for a child, being a factor of 10 higher and is important in considering the public dose due to inhalation. As can be seen the inhalation dose conversion coefficient is higher than that through the ingestion route.
18 F has a physical half-life 1.83 h. Following ingestion/inhalation, there is an initial uptake of FDG in heart (0.04%), brain (0.08%), liver (0.05%), lungs (0.03%) and all other tissues (0.8%).  The retention in the specified source organs is considered to be infinite. A fraction of 0.3 of the activity in other organs and tissues is considered to be excreted in urine with biological half-times of 12 min (25%) and 1.5 h (75%), according to kidney bladder model.
The dose conversion coefficients for inhalation and ingestion (in Sv/Bq) are given below: 
Thus if 1 mCi is inhaled by a person it can lead to a dose of 1 mSv and in case of infants it is around 10 mSv.
Assessment of internal dose due to inhalation (occupational workers) in medical cyclotrons
The dose is estimated on the basis of the following assumptions:
Assume 1 mCi (3.7E7Bq) of 18 F is released into the vault atmosphere. This, in most cases represent only about 0.025% of the built up activity in the target, which seems to be reasonable.
- Cyclotron vault volume - 50 m 3
- Ventilation - 10 air changes per hour.
Concentration assuming uniform mixing would be 7.4 E5 Bq/m 3 .
So the total intake for an entry into the accelerator vault for 10 min would be 148000 Bq (breathing rate 1.2 m 3 /h).
Using the dose conversion factor this would lead to a dose of 4 uSv.
On the other hand if the entry was after a short time after the incident the localized concentration would be much higher due to improper mixing. If it is assumed that only 1 m 3 of air was initially available for mixing this would lead to a dose of 200 uSv. Thus, the dose is not trivial and need to be investigated in detail. The actual dose can be evaluated only if some test measurements of the air activity is carried out.
| Inhalation Dose to the Public Due to Routine Release of 18 F from Medical Cyclotrons|| |
The exhaust air from the vault as well as from the chemistry modules are mixed and let out through the stack after passing through high efficiency particulate air filter and charcoal filters. The exhaust of the pharmaceuticals from the chemistry module would be chronic as a significant fraction of the materials is expected to get released as airborne during the routine synthesis.
Assumptions and formulae
About 20 mCi (740 MBq) of activity is assumed to be discharged through the stack per month, which is just 0.01% of the total typical production of 200Ci (7400 GBq)/month. Measurements in some medical cyclotrons also indicate that this much of activity is discharged through the stack per month.  Assuming 2 h of operation per day and 20 days/month release rate per second is 5500 Bq/s. The dose assessments based on this are given below:
D (Sv/s) = R (Bq/s) × X/Q (s/m 3 ) × B (m 3 /S) × DCF (Sv/Bq) ,, (3)
D is the inhalation dose in Sv/s
R is the release rate 5500 Bq/s.
X/Q is the dilution factor at a distance of 100 m, 3.0E-3 s/m 3 (for a stack height of 5 meters; calculated based on Gaussian plume dispersion model).
B is the inhalation rate - 3.3E-4 m 3 /s. 
DCF is the Dose Conversion Factor 2.8E-11 Sv/Bq. 
However, this factor for an infant is 2.6E-10, a factor of 10 higher.
Public dose estimates
Based on these assumptions and formulation the committed dose works out to 5.5E-4 uSv/h. For 520 h of operation in a year this is equivalent to 0.29 uSv/year, which is trivial.
Considering the location of medical cyclotrons the dilution factor taken above may not be achievable.  Often taller structures overlooking the stack appear subsequently! The most pessimistic estimate would be the concentration at the stack outlet not taking into account any dilution factor and then the dose would be 90 uSv (10% of limit for public of 1 mSv/year). It is not for suggesting that one should not take into account any dilution factor but it would be prudent to realize that the dose would be in the range of 0.3-90 uSv. Another factor to be considered is that for an infant this dose could be 10 times higher.
The estimates show that it is necessary to quantify the emission realistically by actual stack monitoring measurements and this should be enforced, since the estimated emission of 740 MBq would depend on machine operation parameters and one need to be cautious about any abnormal releases, which cannot be ruled out.
Compared to this our Nuclear Power Plants (NPP) do not contribute to more than 2% (20 uSv/year) of the limit for public dose and keeps meticulous record of yearly doses to the public from all our NPPs. And in a few years the number of cyclotrons in India may be 10 times more than NPPs.
Analysis of the impact is required even to neglect these!
| Internal Dose from Spills|| |
The internal dose implications in case of a spill or personal contamination also need to be analyzed to understand the implications of these. These are possible through inhalation or absorption through skin. These issues are familiar to safety professionals in the nuclear fuel cycle. These are not found to be considered in the Safety Assessment Reports (SAR), though this type of incidents cannot be ruled out either due to personal negligence or system failure.
| Review of Reported Analyses and Unusual Occurrences in Medical Cyclotrons|| |
It is not as though there are no cases of accidents in cyclotrons. Few reported cases are available in publications as also analysis of risks from inhalation of radioactivity in Medical Cyclotrons is also reported in the literature.
One such was an incident in the cyclotron at Harvard medical school, USA in 2009 which resulted in the operator receiving an extremity dose of 100 Rad (1Gy), twice the limit.  This was due to an operator error resulting in accidental delivery of the irradiated target material to the wrong process cell. The Radiochemist was exposed to a vial containing 1.6 Ci of F-18 in contact with his upper arm. Preliminary dose estimate was 50-100 rem to upper arm. Added to that he thought the alarming personal dosimeter was giving false alarm! However, promptness of reporting to National Research Council was impressive. Such things do happen and regulatory norms need to be tightened to take care of these.
Calandrino et al. report in Health Physics, 2009 about the evaluation of the risks and doses for the internal contamination of the radiochemistry staff in a high workload medical cyclotron facility.  The dose from internal contamination from inhalation of radioactive gas leakage from the cells by personnel involved in the synthesis processes are estimated from urine sample and whole body measurements. Dose was in the range of 100 and 500 uSv/year. They recommend continuous and accurate air monitoring in the lab must be installed and calibrated in view of the impracticability of assessing internal dose by urine analysis. During 18 months they had four instances of contaminated air leakage with personnel present giving a probability of 1 in 100. They stress the importance of leaving the lab during the synthesis process whenever possible in order to limit low chronic intake and related additional committed dose to the staff.
The implications of target rupture are not discussed generally in SARs except qualitatively stating that only "negligible" activity gets leaked. However, in a medical cyclotron in Iran the dose rates after a target rupture were actually measured by 7 International Commission of Radiation Units spherical body phantoms placed inside the liquid target room.  The results showed that the repair can be started immediately after stopping of the proton bombardment only if the target has been ejected from the target room and the duration of bombardment has not taken more than 10 min.
Clark et al.  carried out atmospheric tracer test and assessment of potential accident at the national medical cyclotron in Australia using atmospheric tracer release of SF 6 and noticed strong influence of cyclotron building and surround structures on the downward movement of tracer. Atmospheric dilution factors were applied to release of 123 I and 123 Xe from the cyclotron using the measured dilution factor of 1.9 E-3, which they obtained from their tracer studies. These studies show that atmospheric modeling may underestimate the dose and that the public dose from medical cyclotron is considered in many countries.
Shallu et al. analyzed target rupture scenario.  They consider target rupture scenario as one of the major emergency situations in medical cyclotrons since there is potential of overexposure to working personnel. They make recommendations to reduce personal exposure during such events. However, they have not dealt with the aspect of internal exposure in such scenarios.
Safety analysis report of a RDS Eclipse medical cyclotron arrives at a stack exhaust dose of 0,0175 mR (0.175 uSv) for an assumed vented puff of 1 Ci (37 GBq). However, they assumed a stack height of 25 m height, which is unusual. For the occupational internal exposure they arrive at a dose of 98 mR (0.98 mSv) for a worst case of full load of gaseous activity vented into the cyclotron room. The maximum external exposure is stated to be 546 mR (5.46 mSv) for infinite occupancy after the incident. These are however extreme cases. However, it is worth noting that this SAR includes such analysis. 
It is imperative that the stack monitors are calibrated in terms of becquerel per second emission and a record of such emissions are kept and submitted to the regulatory agency. A limit for stack emission based on the impacted dose to a member of the public need to be arrived at by the regulatory agency to be adhered to by all medical cyclotrons. In nuclear reactors generally the limit is kept at a few percent of the overall regulatory limit of 1 mSv/year to a member of the public. Cyclotron operators are generally unfamiliar with calibration procedures. However, there are many published papers on the method used for calibration and for rechecking this periodically. Monitoring services are available in some countries to provide the necessary support for this. Many manufacturers do give a calibration factor for stack monitors to convert the cps to bequerals per unit time. However, a protocol for checking of calibration is required to be issued to the facilities along with the regulatory license. The background when the facility is not operating need to be established and stored in the software so that this is always subtracted before applying the calibration factor to arrive at the activity released. The calibration methods vary from using an aliquot of 18 F, computational procedures using volume source calculation, calibration using point sources of positron emitters at different points in the measuring chamber etc. 
| Conclusions|| |
The estimates on internal exposure assuming certain extreme conditions and the quoted reports of such analyses and incidences show the need for regulatory steps to take care of these.
Many accelerators incorporate liquid nitrogen trap for the air borne radioactive products during synthesis and make the air from the reaction vessel inside the synthesis module to pass through this trap, which condenses the moisture containing air borne activity and then put the air back into the glove box atmosphere. In some cyclotrons, the air exhaust from the chemistry modules is trapped in a balloon and allowed to be released only after the decay. In yet some other cyclotrons there are a further provision to hold up the discharges in pressurized tanks and released through the stack after decay. Some self-shielded machines have separate ventilation ducts from inside the shield to promptly release any air activity that maybe produced inside.  Thus varying methods are adopted for containing released radioactivity. However, efficacy of these steps needs to be established in controlled experiments.
The ventilation is generally taken for granted and never measured to ensure that there are no locked up packets in the areas. It stands to reason that the whole volume of vault air would not be available for dilution soon after the release and any attempted evaluation on the basis of complete dilution would understate the internal hazard.
Though the glove boxes are supposed to under negative pressure and the leakage is not supposed to take place into the working environment of the synthesis modules some reports  do report airborne activity in the working areas and even assessment of the internal dose using bio dosimetry. They even recommend the operator to be away from the outside of the synthesis module as an added precaution. All these show that the internal dose in medical cyclotrons cannot be neglected and its absence taken for granted.
Appropriate regulatory steps to minimize the internal hazards have to be backed up by investigations on the measurement of radionuclide air concentrations in the vault atmosphere during target rupture incidences and at the stack outlet and appropriate calibration for the stack monitor. In the safety review of most facilities this is stated to be negligible yet no quantification exists on the leakages into the vault atmosphere, which then gets discharged through the stack. Data need to be acquired by simulating such incidents to understand the implications which might or might not be significant. Measurements and safety analysis are required even if it is to neglect these consequences. Air sampling methods suitable are very familiar to radiation protection professional in the nuclear fuel cycle and can easily be adopted. These devices can also be operated remotely in no-entry areas. Air sampling kits using bubbler traps/charcoal filters can be used for assessment of air activity. Stack monitors need to be calibrated, the methodology for this is available.  Often medical cyclotron operators feel helpless in carrying out these calibrations and may need help from professional radiation protection groups.
In view of the location of medical cyclotrons in public locations and their large numbers the concept followed in some cyclotrons of hold up tanks of the stack releases might be worth looking into. In case of location of medical cyclotrons in crowded areas replete with high rise buildings etc., it is necessary "to insure engineered safety features to ensure zero discharges from the machines." Such features as liquid nitrogen trap, pressurized hold up tanks etc., are used by some cyclotrons.
| References|| |
|1.||Rajan MG, Radiation Medicine Centre, BARC. Personal Communication. |
|2.||Mukherjee B. A real-time positron monitor for the estimation of stack effluent releases from PET medical cyclotron facilities. Appl Radiation & Isotopes 2002;57:899-905. |
|3.||ICRP Publication 106, Radiation dose to patients from radiopharmaceuticals - Addendum 3 to ICRP Publication 53. Ann. ICRP 2008;38(1-2). |
|4.||Siemens self shielded machine (private communication). |
|5.||IAEA Safety Series Report 119, Generic models for use in assessing the impact of discharges, IAEA Vienna, 2001. |
|6.||IAEA International Basic Safety Standards, Safety Series 115. International Atomic Energy Agency Vienna. 2011. |
|7.||Clark H. Atmospheric tracer test and assessment of a potential accident at the National Medical Cyclotron, Camperdown, Australia, Report ANSTO E 710; 1994. |
|8.||Martel C, Tarzia JP. Response to an event in a medical cyclotron, Health Physics Society Meeting; 2009. |
|9.||Caladrino R. Intake risk and dose evaluation methods for workers in radiochemistry labs of a medical cyclotron facility. Health Phys. 2009;97:315-21. |
|10.||Seyed M. Assessment of the staff absorbed dose related to cyclotron operation and service in the production of F-18 radiopharmaceuticals, NUKLEONIKA 2012;57(3):407-10. |
|11.||Shallu VS, Sharma SD, Kumar R, Sarin B. Target foil rupture scenario. J. Med. Phys 2009;34:161-6. |
|12.||Mukherjee B. A real-time positron monitor for the estimation of stack effluent releases from PET medical cyclotron facilities. Appl Radiat Isot 2002;57:899-905. |