ML25017A059

Evaluation of Maccs Ki Modeling Capability

ML25017A059
Person / Time
Issue date:	03/11/2025
From:	Etter N NRC/RES/DSA/AAB
To:
References
Download: ML25017A059 (1)

Text

3 Evaluation of MACCS KI Modeling Capability Noah Etter Accident Analysis Branch Division of Systems Analysis

Contents 1 INTRODUCTION...........................................................................................................3 2

BACKGROUND.........................................................................................................3 2.1 Radioiodine and KI.............................................................................................3 2.2 Selection of phenomena for investigation..........................................................6 3 METHODOLOGY..........................................................................................................6 3.1 Assumptions.......................................................................................................6 3.2 Summary of modeling cases..............................................................................7

4. RESULTS.....................................................................................................................8 4.1 Base Case Description.......................................................................................8 4.1.1 Definition of Sample Problem......................................................................8 4.2 KI-Enabled Test Case...........................................................................................11 4.3 Removal of Iodine and Tellurium Isotopes, No KI Test Case...............................12 4.4 Removal of Iodine and Tellurium Isotopes, KI-Enabled Test Case......................13 4.5 Sensitivity Analysis of KI Efficacy on Health Effects.............................................13 4.6 Sensitivity Analysis of Population Fraction on Health Effects...............................14 4.7 Note on Cohort Relocation....................................................................................15 4.8 Note on Parameters KI_EFF and KINAME...........................................................16 5

SUMMARY

AND CONCLUSIONS..........................................................................16 6

Future Work.............................................................................................................19 7

References..............................................................................................................20

3 1 INTRODUCTION Potassium iodide (KI) is effective for protection of the thyroid against the potential harmful effects of radioiodine. Federal KI policy (67 FR 1355, dated January 10, 2002) recommends the use of KI for emergency workers, institutionalized persons and the general public. NRC regulations in 10 CFR 50.47(b)(10) require the consideration of KI as a supplement to evacuation and sheltering. Yet there are limited quantitative analyses on the efficacy of various KI implementation strategies (e.g., predistribution, stored at central location, available at shelters, sensitivity of ingestion timing).

Quantitative analyses and additional data are needed to inform, in consultation with other Federal agencies, the need for updating implementing guidance and strategies in support of Federal KI policy.

KI is a supplemental protective measure to reduce radioactive iodine uptake to the thyroid. However, the ideal strategy for KI ingestion and distribution is unclear.

Consequence analysis models in MACCS can simulate KI ingestion in protective action modeling. The efficacy of KI ingestion depends on many factors including the ability of cohorts to find or obtain KI during an emergency, ingestion timing, pre-existing stable iodine saturation of the thyroid gland, and relative importance of KI compared to evacuation and sheltering. Uncertainty surrounding KI distribution and ingestion may lead to less than optimal protective action recommendations. This uncertainty warrants a review of recent literature and a sensitivity analyses conducted in MACCS to better understand the efficacy of KI ingestion and distribution strategies.

As part of Research Assistance Request (RAR) NSIR-2024-5 (ADAMS Accession No. ML24027A003), the Office of Nuclear Safety and Incident Response (NSIR) requested that research be undertaken to review recent literature and perform MACCS sensitivity analyses to understand the efficacy of KI ingestion and distribution strategies. A full understanding of the effects of the MACCS KI model is a crucial first step in designing such sensitivity analyses. This work supports RAR-NSIR-2024-5 by improving staff understanding of the workings of the MACCS KI model.

4 2

BACKGROUND This section provides a concise background appropriate to the subsequent work, including information on radioiodine, thyroid uptake, KI distribution strategies and a brief description of iodine biokinetics.

2.1 Radioiodine and KI Radioiodine is produced through the fission of uranium atoms in a nuclear reactor.

During a severe reactor accident, this volatile radionuclide may potentially be released in quantities significant enough to merit specific protective actions to limit uptake into the body.

The primary organ at risk from radioiodine is the thyroid, a gland located near the base of the neck. The purpose of this organ is to produce thyroid hormones (e.g., thyroxine, tetraiodothyronine, triiodothyronine) which are distributed throughout the body through the bloodstream. The thyroid requires iodine to produce these hormones, which is typically received through iodine present in the everyday diet (National Institutes of Health, 2010). The thyroid has no mechanism for distinguishing between radioactive and non-radioactive iodine, thus all isotopes of iodine are absorbed by the thyroid. If Iodine-131 is released in large concentrations, a relatively large radiological dose to the thyroid is possible (American Thyroid Association, 1990).

One supplemental protective action for an exposed population is the ingestion of stable iodine in the form of KI. It is important to note that KI is primarily used to prevent dose in newborns, children, or those with low amounts of iodine present in the thyroid, and has a limited utility, only protecting from inhaled radioiodine (National Library of Medicine, 2023). In the U.S., the distribution of KI for use by the public in the area surrounding a commercial nuclear power plant is a decision made by the states, and many different distribution strategies and guidelines exist. Two major strategies are pre-distribution and supplying individuals with KI directly prior to or after a radiological release. Another strategy is stockpiling, or supplying KI to various storehouses such that it may be distributed when and where needed. The effectiveness of KI is dependent on many factors such as when the KI is ingested (relative to when the individual is exposed to radioiodine) and the age of the individual ingesting the KI. Potassium Iodide has a shelf life of around five to seven years (U.S. Food and Drug Administration, 2020), which further complicates KI distribution strategies as supplies of KI must be restocked or reissued after expiration.

The effects of radioiodine in the body can be examined using a biokinetic model1 (ICRP, 2015). Inhaled iodine travels through the respiratory tract and is absorbed into the 1 From ICRP: A reference biokinetic model describes the intake, uptake, distribution, and retention of a radionuclide in various organs or tissues of the body, and the subsequent excretion from the body by various pathways.

5 blood. It can then be absorbed by the thyroid or be transferred through other organs. It is ultimately removed from the blood by the kidneys and excreted through urine. A more detailed model was developed by the ICRP as shown in Figure 1.

Figure 1: New Biokinetic Model of Iodine (ICRP, 2017)

Stable iodine in the form of KI can saturate the thyroid, reducing the uptake of radioiodine. MACCS simplifies the complex biokinetics by using a model that simply reduces the thyroid dose by a user-defined fraction. The MACCS model for the effect of KI ingestion is as follows:

,= (1 ),

The MACCS user manual (Sandia, 2023) describes the model as such:

, is the thyroid dose (Sv) from radioiodine through the inhalation pathways that an individual would receive with no KI ingestion using standard dose coefficients.

6 is the efficacy of the KI tablets in reducing thyroid doses from radioiodine where a value of one is complete protection and a value of 0 is no protection specified by the parameter EFFACY In addition, KI protection is only applicable to the fraction of the cohort (modeled populations) that ingest KI tablets. When KI protection is modeled, cohorts are further subdivided between those that ingest KI tablets and those that do not according to the parameter POPFRAC.

2.2 Selection of phenomena for investigation Several parameters were varied to validate that the KI model works as expected.

Bounding cases were defined where the KI model was activated but had 0 EFFACY and 1 POPFRAC;and 1 EFFACY and 0 POPFRAC. Another case was included in which isotopes of tellurium (which may decay into iodine) and iodine were removed.

Radioiodine is the only isotope affected by KI, thus by disabling all isotopes of iodine and tellurium, the bounds of the KI model can be tested. A detailed summary of these test cases is included in Section 3.2.

3 METHODOLOGY 3.1 Assumptions Primary modeling assumptions:

There is one non-evacuating cohort representing the entirety of the population.

This cohort is subject to the KI model when applicable and does not engage in evacuation or sheltering. The threshold for relocation is set to a high value to disable the MACCS dose-based relocation model. The effects of a radiological release and all early health effects are applied to the entirety of the affected population.

Only early phase exposures are modeled. That is, the MACCS EARLY module, but not the MACCS CHRONC module, was activated.

The dose response model follows the Linear No Threshold model as defined in the Point Estimates LNT sample problem model.2 2 The Point Estimates LNT base case is one of the sample problems distributed with the WINMACCS code suite.

7 3.2 Summary of modeling cases Table 1: Test Case Summary Case Number Description 1

Sample Problem with No KI 2

Sample Problem with KI 3

Sample Problem with No KI and No Te or I 4

Sample Problem with KI and No Te or I 5

Sensitivity Analysis on Linearity of Health Effects and KI Efficacy 6

Sensitivity Analysis on Linearity of Health Effects and Population Fraction

1. Case 1 is defined as the base case and was developed by modifying the Point Estimates LNT sample problem to represent a single evacuation cohort that does not evacuate, shelter or relocate. This serves as the base case for subsequent trials and does not include the KI model. It is important to note that the MACCS sample problem has an unusually high source term, leading to relatively high doses in order to demonstrate early health effects.

2. Case 2 is defined as a variation of the base case with the KI model activated with full efficacy and population fraction. The expected results are a decrease in thyroidal health effects and a decreased peak dose, but no change in remaining health effects.

3. Case 3 is defined as a variation of the base case with all isotopes of iodine and tellurium removed from the inventory. The expected results are that implementing the KI model will cause no changes in health effects or peak dose, as all isotopes of iodine, as well as tellurium that can decay into iodine, are removed.

4. Case 4 is defined as a variation of the base case with all isotopes of iodine and tellurium removed from the inventory with the KI model in effect. The expected results are that all health effects and dose is identical to case 3, as KI should have no effect on isotopes outside of iodine.

5. Case 5 is defined as a variation of the base case in which KI is activated and the parameter EFFACY is adjusted over a range from 0-1. The expected results are that as the parameter EFFACY increases, health effects and peak dose should decrease linearly.

6. Case 6 is defined as a variation of the base case in which KI is activated and the parameter POPFRAC is adjusted over a range from 0-1. The expected results are that as the parameter POPFRAC increases, health effects and peak dose should decrease linearly. These results should be identical to Case 5, as both EFFACY and POPFRAC are simple multipliers on the inhaled dose from radioiodine.

8 A detailed look at each case is included in the results section.

4. RESULTS 4.1 Base Case Description The base case used was the NRC Point Estimates LNT sample problem. This was chosen for two reasons: 1) it is available to any user through the WINMACCS code distribution website;2) probabilities in this sample are point estimates and the actual values do not change between trials. This ensures that all users who run the NRC Point Estimates LNT sample problem will use the same values, and thus all test cases can be replicated with identical output results. Results include total health effect cases computed by the MACCS Type 1 output integrated over the full radius of the sample problem (0-1609 km) as well as the peak total effective dose (L-ICRP60ED) on the spatial grid over the full radius of the sample problem. The health effects tabulated include both deterministic effects (total cases of thyroiditis and hypothyroidism) and stochastic effects (total fatal cancer cases and total fatal thyroid cancer cases). The exact input files are included in the Point Estimate LNT Sample Dataset available on the MACCS download site.

4.1.1 Definition of Sample Problem Table 2: Base Case Notable General Properties Property Value Early Consequences

Late Consequences N/A MACCS Cyclical File Set

Dose Response Model Linear No Threshold Activate KI Model N/A Evacuation Problem Model None (LASMOV=0)

Number of Cohorts 1

Early Fatality Effects

Early Injury Effects

Latent Cancer Effects from Early Exposure

The notable general properties of the sample problem are included in Table 2. A brief description of each is included below. Properties were considered notable if they are changes from the Point Estimates LNT sample problem, or if they offer insight to the evaluation of the MACCS KI model.

9 Early Consequences: The focus of the evaluation is on early health effects, as opposed to any latent effects.

Late Consequences: Late consequences were disabled in all cases.

MACCS Cyclical File Set: A cyclical file set was used for the sensitivity analyses to decrease the chance of user error when varying parameters EFFACY and POPFRC.

Dose Response Model: All cases use the linear no-threshold model, allowing for the full range of potential dose effects to be quantified.

Activate KI Model: This parameter allows for the use of the KI model and was activated or deactivated depending on the case.

Evacuation Problem Model: LASMOV=0 removes the outer evacuation boundary, beyond which evacuees are assumed to disappear from the EARLY health effects model and assume no additional dose. With LASMOV=0, no protective actions are credited to reduce exposure to offsite populations.

Number of Cohorts: Use of one cohort constrains the KI model to affect the entire population.

Early Fatality Effects/Early Injury Effects/Latent Cancer Effects: Activating these parameters allow for all health effects to be quantified.

Table 3: Base Case Notable EARLY Properties Property Value KI_EFF L-ICRP60ED KINAME A-THYROID L_THYROID Duration of EARLY phase 40 Days Normal Relocation Action Time 3.456E+06 s Normal Relocation Dose Threshold 1E+10 Sv Hot Spot Relocation Time 3.456E+06 s Hot Spot Relocation Dose Threshold 1E+10 Sv EVAKEY None Cohort Fraction 1

KI Population Fraction N/A KI Efficacy N/A KI_EFF: This parameter varies what organ is affected by KI. Selecting L-ICRP60ED allows for the modeling of the dose savings from KI ingestion.

KINAME: This parameter functions in the same way as KI_EFF. Choosing the thyroidal organs ensures that the dose savings from KI are modeled accurately.

10 Duration of EARLY Phase: The maximum allowable time was chosen to ensure that the full scope of the EARLY phase was captured.

Normal Relocation Action Time: The maximum allowable time was chosen to ensure relocation would not occur.

Normal Relocation Dose Threshold: The maximum allowable dose threshold was chosen to ensure that relocation does not occur.

Hot Spot Relocation Time: The maximum allowable time was chosen to ensure that relocation does not occur.

Hot Spot Relocation Dose Threshold: The maximum allowable dose threshold was chosen to ensure that relocation does not occur. These parameters are redundancies to ensure that no protective actions skew the results.

EVAKEY: None ensures that no evacuation is modeled.

Cohort Fraction: This determines how much of the population is modeled within a given cohort. An input of 1 indicates that there is only 1 cohort that encompasses the entirety of the population.

KI Population Fraction: This parameter (POPFRAC) is varied depending on the case.

KI Efficacy: This parameter (EFFACY) is varied depending on the case.

Using these parameters as a base case, the following health effects were obtained.

These health effects are the only effects that changed when implementing the KI model.

These health effects represent the frequency of events over the entire cohort. Because cohorts can vary greatly in size depending on location, these results are representative as a relative frequency of events.

Table 5: LNT Point Estimate Health Effects and Peak Dose Health Effect Mean Value (0-1609 km)

Thyroiditis 5.11E+00 Hypothyroidism 1.12E+02 Total Cancer Fatalities 7.28E+02 Thyroid Cancer Fatalities 4.60E+01 Peak Total Effective Dose on Spatial Grid (Sv) 2.47E+02

• Peak total effective dose was reported at the first sector (0-0.2 km)

Table 5 includes the average number of health effects as well as the peak dose on the spatial grid. These results are used as the basis for comparison for all subsequent cases. The peak total effective dose on spatial grid refers to the dose at the first sector.

As stated earlier, the peak dose is quite large because the source term is unusually high. These values are only intended to serve as a base case for comparison in testing the KI model.

11 4.2 KI-Enabled Test Case The next case represents the Point Estimates LNT sample problem with the KI model enabled, and the parameters POPFRAC and EFFACY, representing the fraction of population that ingests KI and the efficacy factor for the KI model respectively, both set to 1. This differs from the default values of POPFRAC = 0.5 and EFFACY = 0.7 and were changed to represent the bounding case for KI implementation. The expected results are that all exposure of the thyroid from iodine is eliminated, but that thyroid-related adverse health effects are not fully eliminated because other isotopes may also result in exposure of the thyroid. A summary of all changes is included below.

Table 6: KI Case Notable Changes Property Value Activate KI Model

KI Population Fraction 1

KI Efficacy 1

Table 7 contains the notable health effects and peak dose with the KI model activated.

A physical interpretation for this case is that KI is 100% effective to 100% of the affected population that took KI. This does not fully negate thyroidal health effects due to the presence of other dose pathways but does drastically reduce them. In reality KI efficacy is a function of time taken prior to or after exposure.

Table 7: Health Effects with KI Enabled Health Effect Mean Value (0-1609 km)

Thyroiditis 7.26E-02 Hypothyroidism 1.37E+01 Total Cancer Fatalities 6.85E+02 Thyroid Cancer Fatalities 3.23E+00 Peak Total Effective Dose on Spatial Grid (Sv) 1.79E+02

• Peak total effective dose was reported at the first sector (0-0.2 km)

As expected, thyroid-related adverse health effects decreased considerably. These include thyroiditis, hypothyroidism, and cancer fatalities due to thyroid cancer. The early thyroid injury frequencies decreased by around 95% and thyroidal cancer fatalities decreased by 93%. The total cancer fatalities decreased by around 6% and the peak dose on the spatial grid decreased by around 26%.

12 4.3 Removal of Iodine and Tellurium Isotopes, No KI Test Case The next test case was the removal of isotopes of tellurium and iodine. The purpose was to validate that the KI model only affected isotopes of iodine. Tellurium isotopes were also removed as they have the potential to decay into iodine and would result in an increase to thyroidal dose. A summary of changes is included below.

Table 8: No KI Case with No I or Te Isotopic Changes Isotope Inventory (Bq)

I-131 0

I-132 0

I-133 0

I-134 0

Te-127 0

Te-127m 0

Te-129 0

Te-129m 0

Te-131 0

Te-131m 0

Te-132 0

The following table contains the health effects and peak dose when removing all isotopes of iodine and tellurium. The physical basis for this would be if a population was exposed to a radiological release that contained no iodine or tellurium.

Table 9: Health Effects with No KI and No I or Te Isotopes Health Effect Mean Value (0-1609 km)

Thyroiditis 2.21E-02 Hypothyroidism 8.48E+00 Total Cancer Fatalities 5.15E+02 Thyroid Cancer Fatalities 2.19E+00 Peak Total Effective Dose on Spatial Grid (Sv) 1.45E+02

• Peak total effective dose was reported at the first sector (0-0.2 km)

As expected, all health effects decreased in frequency due to the decrease in amount of released radionuclides. These results can be directly compared to a test case in which the KI model was reintroduced, with the expectation that there should be no difference in health effects between the two cases.

13 4.4 Removal of Iodine and Tellurium Isotopes, KI-Enabled Test Case The next test case is to test whether the health effects change when KI is enabled, and with no isotopes of iodine or tellurium. The physical interpretation for this would be if KI was distributed to a population who was exposed to a radiological release that contained no iodine or tellurium. The isotopic changes are identical to those contained in Table 8. A summary of additional model changes is included below.

Table 10: KI Case with No I or Te Notable Changes Property Value Activate KI Model

KI Population Fraction 1

KI Efficacy 1

Table 11 contains the mean frequency of health effects and the peak dose.

Table 11: Health Effects with No I or Te Isotopes, with KI Enabled Health Effect Mean Value (0-1609 km)

Thyroiditis 2.21E-02 Hypothyroidism 8.48E+00 Total Cancer Fatalities 5.15E+02 Thyroid Cancer Fatalities 2.19E+00 Peak Total Effective Dose on Spatial Grid (Sv) 1.45E+02

• Peak total effective dose was reported at the first sector (0-0.2 km)

The mean frequency of health effects did not change whether the KI model was in effect or not. This indicates that the KI model only affects the dose received from iodine but retains all other doses and dose pathways.

4.5 Sensitivity Analysis of KI Efficacy on Health Effects With the test cases verified, sensitivity analyses on the effectiveness of a basic KI distribution strategy could be tested. This essentially encapsulates a pre-distribution strategy in which an affected population would not have to travel to receive KI, and the fraction of the population who ingests KI and the efficacy of this KI can be directly adjusted. The first sensitivity analysis was to test the effects of the KI efficacy parameter EFFACY. A table summarizing these effects is included below.

14 Table 12: Sensitivity Analysis Results for Parameter EFFACY Health Effect (0km-1609 km)

Thyroiditis Hypothyroid ism TOTAL Cancer Fatalities Thyroidal Cancer Fatalities Peak Dose

[Sv]

0 EFFACY 5.11E+00 1.12E+02 7.28E+02 4.60E+01 2.47E+02 0.1 EFFACY 4.32E+00 1.02E+02 7.23E+02 4.15E+01 2.40E+02 0.2 EFFACY 3.58E+00 9.15E+01 7.19E+02 3.71E+01 2.33E+02 0.3 EFFACY 2.89E+00 8.13E+01 7.15E+02 3.27E+01 2.26E+02 0.4 EFFACY 2.29E+00 7.11E+01 7.10E+02 2.83E+01 2.20E+02 0.5 EFFACY 1.74E+00 6.10E+01 7.06E+02 2.39E+01 2.13E+02 0.6 EFFACY 1.25E+00 5.10E+01 7.01E+02 1.96E+01 2.06E+02 0.7 EFFACY 8.37E-01 4.11E+01 6.97E+02 1.53E+01 1.99E+02 0.8 EFFACY 4.93E-01 3.15E+01 6.93E+02 1.11E+01 1.93E+02 0.9 EFFACY 2.39E-01 2.22E+01 6.89E+02 7.11E+00 1.86E+02 1 EFFACY 7.26E-02 1.37E+01 6.85E+02 3.23E+00 1.79E+02 4.6 Sensitivity Analysis of Population Fraction on Health Effects The input parameters EFFACY and POPFRAC can best be thought of as static multipliers on the inhalation dose to the thyroid due to iodine. With that in mind, modifying EFFACY and POPFRAC should have equivalent results. To validate this, the sensitivity analysis was repeated with the parameter POPFRAC, and the results are included in Table 13.

15 Table 13: Sensitivity Analysis Results for Parameter POPFRAC Health Effect (0km-1609km)

Thyroiditis Hypothyroid ism TOTAL Cancer Fatalities Thyroidal Cancer Fatalities Peak Dose

[Sv]

0 POPFRAC 5.11E+00 1.12E+02 7.28E+02 4.60E+01 2.47E+02 0.1 POPFRAC 4.32E+00 1.02E+02 7.23E+02 4.15E+01 2.40E+02 0.2 POPFRAC 3.58E+00 9.15E+01 7.19E+02 3.71E+01 2.33E+02 0.3 POPFRAC 2.89E+00 8.13E+01 7.15E+02 3.27E+01 2.26E+02 0.4 POPFRAC 2.29E+00 7.11E+01 7.10E+02 2.83E+01 2.20E+02 0.5 POPFRAC 1.74E+00 6.10E+01 7.06E+02 2.39E+01 2.13E+02 0.6 POPFRAC 1.25E+00 5.10E+01 7.01E+02 1.96E+01 2.06E+02 0.7 POPFRAC 8.37E-01 4.11E+01 6.97E+02 1.53E+01 1.99E+02 0.8 POPFRAC 4.93E-01 3.15E+01 6.93E+02 1.11E+01 1.93E+02 0.9 POPFRAC 2.39E-01 2.22E+01 6.89E+02 7.11E+00 1.86E+02 1.0 POPFRAC 7.26E-02 1.37E+01 6.85E+02 3.23E+00 1.79E+02 As expected, the results from varying POPFRAC and EFFACY are identical, verifying that the two inputs can be considered equivalent in their weighting. To summarize, POPFRAC represents the fraction of the population of a given cohort that ingests KI.

EFFACY represents the percent effectiveness of the KI ingested. Both act as a multiplier on the total population dose. As an example, with a POPFRAC of 0.5 and an EFFACY of 0.3, 50% of the total population would ingest KI that decreases the thyroid dose due to radioiodine inhalation by 30%.

4.7 Note on Cohort Relocation During initial testing, one unexpected result was an increase in health effects unrelated to the thyroid, such as breast cancer, when implementing the KI model. At full efficacy and population fraction, the non-thyroidal cancers increased by less than 1%, despite having disabled evacuation and sheltering. This issue was identified to be a dose threshold for relocation no longer being met with the application of KI, resulting in slightly higher doses. Once the parameters for relocation were increased to ensure that no protective actions would be taken, the increased frequency of non-thyroidal cancers ceased. This relocation threshold can be illustrated by the figure below. Note that the purpose of this figure is to illustrate a particular MACCS behavior and does not necessarily indicate that this behavior is significant in assessing dose consequences.

16 Figure 2: Erroneous Increase in Non-Thyroidal Health Effects Following Failure to Fully Negate Relocation 4.8 Note on Parameters KI_EFF and KINAME MACCS allows for the adjustment of the organ that is affected by KI, specifically with the parameter EFFACY. There are two pertinent parameters that are defined in a similar way, creating some ambiguity as to the exact purposes of parameters KINAME and KI_EFF. The current definitions according to the MACCS user manual (Sandia, 2023) are included below:

KI_EFF defines the effective organ affected by the ingestion of KI.

KINAME provides a drop-down menu to select the organ names affected by the KI model.

Testing indicated that these parameters function identically. As an example, having an organ as KI_EFF with a KI efficacy of 1 resulted in the same dose reduction as the organ used as an input to parameter KINAME with a KI efficacy of 1. Further testing is recommended to validate the functionality of these parameters.

5

SUMMARY

AND CONCLUSIONS The results from this analysis are summarized in Figures 3 and 4 below. Using the sample problem, the Total Cancer Fatalities as a function of KI efficacy appears to be a linear function. Using the Point Estimates LNT sample problem with the previously listed

17 parameters, there is a decrease of around 6% in total cancers over a 1000-mile radius with full efficacy.

Figure 3: Total Cancer Fatalities as a Function of KI Efficacy The following figure illustrates the peak dose at 100 meters as a function of KI efficacy.

It is important to note that this decrease is only due to dose received to the thyroid from radioiodine inhalation. At full efficacy, the peak dose on the spatial grid of the sample problem decreased by around 68 Sv with a percent decrease of around 28%. This seems to indicate that in the case of the sample problem, around 28% of the peak early-phase effective dose (L-ICRP60ED) at the first spatial interval (0-0.2 km, or 0-200 m) could be attributed to radioiodine inhalation. Note that this assumes all protective actions are removed.

6.60E+02 6.70E+02 6.80E+02 6.90E+02 7.00E+02 7.10E+02 7.20E+02 7.30E+02 7.40E+02 0

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

y = -4.3091x + 731.85 R2 = 0.9993 Total Cancer Fatalities vs. KI Efficacy CAN FAT/TOTAL EFFACY Fraction 0.00E+00 5.00E+01 1.00E+02 1.50E+02 2.00E+02 2.50E+02 3.00E+02 0

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

y = -6.7636x + 253.49 R2 = 0.9998 Peak Dose [Sv] vs. KI Efficacy Peak Dose [Sv]

EFFACY Fraction

18 Figure 4: Peak Total Dose on Spatial Grid as a Function of KI Efficacy The KI distribution model available in MACCS can be conceptualized as a linear function of dose from the inhalation pathway to the thyroid due to iodine. KI primarily affects the dose to one organ from the inhalation of radioiodine. However, as observed from recent results, there is a noticeable non-linearity between the peak dose reduction and the corresponding impact on latent cancer fatalities. For instance, while the peak dose (presumably at close range) drops by 28% with full KI efficacy, the total health latent fatality cases in the 0-1000 mile region decrease by only 6%. This disparity suggests that other factors such as the MACCS high dose versus low dose cancer fatality model, variations in the dose-response relationship at different distances, or a shift in the significance of iodine inhalation with radial and off-centerline distances may influence the results. Modifying the dose coefficient factor (DCF) file can better represent at-risk populations such as children or individuals with low iodine intake, and the timing of iodine ingestion can be adjusted using EFFACY.

19 6

Future Work There were several areas of this study that may warrant additional research. The MACCS KI model is limited to two scaling factors on thyroid dose due to radioiodine.

The biokinetic transport of KI ingestion cannot presently be modeled by MACCS, so a quantitative comparison between the current KI model and a model that can allow for biokinetic transport, such as the compartment model illustrated by Figure 1, may inform future code changes or policy regarding KI distribution. In addition, given the observed non-linearity between peak dose reduction and latent cancer fatalities, further investigation into the MACCS cancer fatality model may be beneficial. For example, the current KI distribution model in MACCS treats the dose to the thyroid from inhaled iodine as a linear function, but recent findings suggest that this linearity does not translate into proportional reductions in health outcomes. Peak dose reductions of 28%

only lead to a 6% decrease in total latent cancer fatalities over a 0-1000 mile radius, which highlights complexities in dose-response relationships. Other factors such as the radial distribution of iodine inhalation risk, variations in dose-response at different distances through differences in radioisotope distribution, or the effects of high versus low doses on cancer outcomes, should also be explored. Additionally, modifications to the dose coefficient factor (DCF) file, particularly to account for vulnerable populations like children or individuals with low iodine intake, and adjustments to iodine ingestion timing using EFFACY, could provide a more accurate representation of at-risk populations and improve the effectiveness of KI distribution policies.

20 7

References National Research Council (U.S.). Distribution and Administration of Potassium Iodide in the Event of a Nuclear Incident. National Academies Press, 2004. Prepared by the Committee to Assess Distribution and Administration of Potassium Iodide in the Event of a Nuclear Incident.

National Institutes of Health, Office of Dietary Supplements. Iodide Factsheet for Consumers. National Institutes of Health, 2010.

American Thyroid Association. "Potassium Iodide Stockpile for Nuclear Accidents."

JAMA, vol. 264, no. 6, 1990, pp. 730-731.

Torti, J., and Correa, R. "Potassium Iodide." National Libraries of Medicine, 2023.

U.S. Food and Drug Administration. "Potassium Iodide Tablets - Shelf Life Extension."

FDA Guidance Documents, Apr. 7, 2020.

ICRP. Occupational Intakes of Radionuclides: Part 1. ICRP Publication 130. Annals of the ICRP, vol. 44, no. 2, 2015.

ICRP. Occupational Intakes of Radionuclides: Part 3. ICRP Publication 137. Annals of the ICRP, vol. 46, no. 3/4, 2017.

ICRP. 1990 Recommendations of the International Commission on Radiological Protection. ICRP Publication 60. Annals of the ICRP, vol. 21, no. 1-3, 1991.

Sandia National Laboratories. MACCS User Guide - Version 4.2. Prepared for the U.S.

Nuclear Regulatory Commission, Office of Nuclear Regulatory Research, Washington, DC, March 2023 (SAND2023-01315). Albuquerque, NM: Accident Consequence Modeling and Analysis Department.