Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Jul 1.
Published in final edited form as: Health Phys. 2020 Jul;119(1):72–82. doi: 10.1097/HP.0000000000001244

SCIENTIFIC AND LOGISTICAL CONSIDERATIONS WHEN SCREENING FOR RADIATION RISKS BY USING BIODOSIMETRY BASED ON BIOLOGICAL EFFECTS OF RADIATION RATHER THAN DOSE: THE NEED FOR PRIOR MEASUREMENTS OF HOMOGENEITY AND DISTRIBUTION OF DOSE

Harold M Swartz 1,2,*, Ann Barry Flood 1, Vijay K Singh 3,4, Steven G Swarts 5
PMCID: PMC7269859  NIHMSID: NIHMS1549005  PMID: 32175928

Abstract

An effective medical response to a large-scale radiation event requires prompt and effective initial triage so that appropriate care can be provided to individuals with significant risk for severe acute radiation injury. Arguably, it would be advantageous to use injury rather than radiation dose for the initial assessment, i.e., use bioassays of biological damage. Such assays would be based on changes in intrinsic biological response elements, e.g., up- or down-regulation of genes, proteins, metabolites, blood cell counts, chromosomal aberrations, micronuclei, micro-RNA, cytokines, or transcriptomes. Using a framework to evaluate the feasibility of biodosimetry for triaging up to a million people in less than a week following a major radiation event, Part 1 analyzes the logistical feasibility and clinical needs for ensuring that biomarkers of organ-specific injury could be effectively utilized in this context. We conclude that the decision to use biomarkers of organ-specific injury would greatly benefit by first having independent knowledge of whether the person’s exposure was heterogeneous and, if so, what was the dose distribution (to determine which organs were exposed to high doses). In Part 2 we describe how these two essential needs for prior information (heterogeneity and dose distribution) could be obtained by using in vivo nail dosimetry. This novel physical biodosimetry method can also meet the needs for initial triage, providing non-invasive, point-of-care measurements made by non-experts with immediate dose estimates for 4 separate anatomical sites. Additionally, it uniquely provides immediate information as to whether the exposure was homogeneous and, if not, can estimate the dose distribution. We conclude that combining the capability of methods such as in vivo EPR nail dosimetry with bioassays to predict organ-specific damage would together allow effective use of medical resources to save lives.

Keywords: Bioassay, dosimetry, personnel, emergencies, radiological, emergency planning

INTRODUCTION:

After a radiation/nuclear event in which there is a potential for significant risk to large numbers of people being exposed, the first need is to rapidly determine who is at a significant risk of having severe acute effects from the radiation exposure. This step (to triage people whether to enter the medical care system or not), is particularly important when the number of potentially exposed individuals exceeds the capacity of the medical response efforts to closely follow all exposed individuals. The traditional approach has been to carry out the initial decisions based on estimates of dose received by the individual, using biodosimetry, if available, to decide who should move to the next stage of triage. Usually this is based on estimates of the radiation dose to which the individual is exposed and the associated biological injury (Sullivan et al. 2013; Coleman and Koerner 2016).

However, some experienced scientists and clinicians in the field of radiation biology are in favor of using indicators of biological effects rather than radiation dose per se. Their rationale is based on evidence suggesting that people receiving the same dose of radiation can manifest different response to the radiation injury, sustaining acute injury at different doses. Therefore, it is argued, that the initial decisions for triage should be based on a patient’s biological response, not on the radiation dose received. (Dainiak 2018; Singh et al. 2018).

While this has been accomplished by using the conventional approach for small scale incidents in which all exposed individuals at risk can be placed into an advanced medical care system and their biological responses closely monitored, this would not be feasible for a large scale scenario involving considerable number of exposed victims. New biodosimetry methods, as well as ways to improve traditional biodosimetry methods, e.g., via high throughput automation and computerized analyses, have been proposed to handle an emergency radiation scenario involving thousands of victims (Brenner et al. 2015; Flood et al. 2016; Paul and Amundson 2008; Rothkamm et al. 2013a and 2013b; Swartz et al. 2012; Wang et al. 2019; Xu et al. 2013; Rogan et al. 2016).

Therefore, considerable recent effort has been directed into developing screening approaches that are relevant to large-scale scenarios which can provide information that is more specific/accurate in predicting the possibility of developing serious injury to radiosensitive organs, e.g., hematopoietic system (lymphocyte, neutrophil, and platelet depletion in blood), gastrointestinal system (low citrulline levels in small intestine), lung, and kidney. Such assays would be based on changes in intrinsic biological response elements, e.g., up- or down-regulation of genes, proteins, metabolites, blood cell counts, chromosomal aberrations, micronuclei, micro-RNA, cytokines, or transcriptomes. This focus on organ injury has not only been advocated especially in the context of screening for acute risk, but also for assessing risk for subacute and chronic effects to specific organs, with the expectation that there would be the possibility of initiating radiation mitigating strategies to decrease the probability of long-term effects (NIH 2019).

The homogeneity of the exposure is another important factor that goes beyond dose in predicating the response, because the predominant mode of life-threatening acute effects due to moderate radiation dose is usually injury to the hematopoietic system. Therefore, the sparing of even a modest amount of bone marrow is known to greatly improve the probability of survival after receiving an otherwise potentially lethal dose of radiation (Prasanna et al. 2010a). Consequently, if it can be determined that the exposure is very heterogeneous, then it might be especially useful to have indicators of organ-specific damage, i.e., such information is useful because the individual is less likely to have succumbed to acute suppression of the hematopoietic system.

We discuss three aspects crucial to the success of organ-specific biodosimetry:

  1. What characteristics are important for a bioassay method to have, in order for it to be an effective tool to screen for biological damage in specific organs?

  2. How should such a method be integrated into an effective early triage system?

  3. What additional information could make the use of organ-specific bioassays of organ damage more effective for the initial medical response in the context of a large-scale radiation scenario?

PART 1: REQUIREMENTS FOR THE EFFECTIVE USE OF BIOMARKERS OF ORGAN-SPECIFIC BIOLOGICAL DAMAGE FOR INITIAL TRIAGE

The mechanism of most or all of the biomarker induction assays whose purpose is to predict organ-specific injury is based on dynamic changes in the levels of naturally occurring biological response elements, such as genes and/or products produced through metabolism after radiation damage occurs. Fig. 1 illustrates the general dynamic flow of these changes over time following radiation exposure (shown for upregulated responses in this case). (Fig. 1a illustrates biologically based biodosimeters; Fig. 1b illustrates physically based biodosimeters.) The dynamics in Fig. 1a apply to all biologically-based biomarkers, including those intended to estimate dose/system-based injury as well as those intended to predict organ-specific injury.

Figure 1:

Figure 1:

Open in a new tab

Typical Time Course following Exposure to Ionizing Radiation of: (a) biologically-based and (b) physically-based biodosimetry markers: Schema shows upregulated response for both types

AXES: x = level of biomarker at baseline and in response to exposure; y = time following initial perturbation.

TIME POINTS: 1 = initial state prior to injury (baseline level); 2 = time when response becomes detectable; 3 = time when change becomes relatively stable; 4 = time when response starts to decline; 5 = time when changes becomes stabilized ~ at baseline value. Notes (a) This illustration shows an initial increase in the biomarkers; analogous dynamics in the opposite direction would occur with a radiation-induced decrease. (b) Not portrayed here is individual variation in the baseline level and degree of response to the same biomarker at the same exposure. (adapted from Swarts et al. 2018)

Because these dynamics are common to all biologically-based biomarkers, we can build upon our previous framework (Flood et al., 2011, 2014 and 2016) that evaluated the practical feasibility of methods of biodosimetry used for estimating radiation dose or the likelihood of having life-threatening systemic injuries. This framework can also analyze the logistical needs and feasibility of using organ-specific biomarkers in the context of planning the initial medical response to large radiation scenarios.

As noted in our previous analyses, in order to effectively use a biodosimeter or biomarker in a large radiation scenario, it must be suitable for use under the circumstances that are likely to be present, and the assays must be completed within the specified window of time. For our purposes, we use the standard planning scenarios in the USA, which stipulates that each method intended for triage during a large scale radiation scenario must be capable of acquiring the data for at least 1 million people within ~6 days and make such data available to triage decision-makers (Sullivan et al. 2013; Coleman and Koerner 2016).

In order to evaluate the feasibility and practicality of achieving this goal for a large scale radiation scenario, it is important to consider the full timeframe needed, i.e., starting with collecting a sample from a potential victim in a disaster setting, to delivering the results to the decision-makers responsible for deciding on the next steps for the individual, with an emphasis on whether the individual should be brought into the health care system for further evaluation or not. The length of the various time intervals can vary widely depending on the method of biodosimetry being used and on the specifics of the emergency (e.g., type of radiation, numbers of people potentially exposed, timeliness of the response team). These varying times and logistical challenges in turn impact the number of potential victims who can be screened in a timely manner to save lives—our ultimate goal.

In our previous publications, we identified a number of principal time intervals (briefly listed below) and provided estimates for most of the major biodosimetry methods being considered (Flood et al. 2011, 2014 and 2016). Using the best available estimates of these time intervals for each method, we then predicted how many people could be evaluated in the 6 days following a large scale event. Note that this list of principal time intervals applies to all types of biodosimetry/biomarker methods, including the organ-specific biomarkers proposed. Below (in italics) we also describe attributes that are especially important for biologically-based biomarkers such as organ-specific assays.

Principal Time Intervals for Evaluation Biomarkers

  1. The initial time from the event before a valid sample can be collected. This is indicated in Fig. 1a by both the Latent Period, i.e., before changes are detectable plus the Evolving Changes period, i.e., when the level of the biomarker is changing rapidly. Sampling during these times would underestimate the magnitude of the response and may result in a false negative reading. Both the baseline level and the magnitude of the evolving period may vary among individuals, due to intrinsic variability, the impact of prior events, or concomitant physical injury or stress. In the case of organ specific markers, individual variation in the biomarker may also be impacted by radiation damage to other organs.

  2. The time interval during which a valid sample can be obtained. For virtually all biologically based biomarkers, the time during which the biomarker is stable and most accurately reflective of the magnitude of the response to the exposure/injury will end, and the level will start to return toward the baseline. Unfortunately, for many biologically-based biodosimeters, the end of the stable period for sampling occurs before or during the time when acute medical treatment is being decided. (In contrast, for some, such as the dicentric chromosome analysis (DCA) assay, and for most physically based biodosimeters, the diagnostic period can extend for months or years, making the sample valid to obtain for long term surveillance of health effects [Simon et al. 2019].) The interval during which an accurate quantitative response can be obtained is indicated as the Diagnostic Period in Fig. 1. As noted above, the length of time during which the sample is valid to collect may vary among individuals due to intrinsic physiological variability, the impact of prior events, or concomitant physical injury or stress.

  3. The time required to locate the victims where the sampling and/or measurements will be made and to deliver the equipment, provide facilities and personnel needed. If assays are to be completed in the field, this time period includes the time required to place the equipment needed to carry out the assay and trained personnel at the site.

  4. The time and resources needed to collect a valid sample from each victim. This time includes preparing it for transportation, storage or processing, obtaining identifying information sufficient to link results uniquely to an individual, and recording key demographics or medical information needed to interpret results for a given victim. As needed, it will also include any time needed to determine that the sample is valid, as described above in 2. It is likely that most organ-specific biomarkers will be based on fairly readily obtained body fluids so that the logistics will be similar for most types of biologically-based assays. Obtaining a sufficient, accurate, and uniquely identified sample from a large number of individuals will be challenging for any type of biodosimetry.

  5. The preparation time needed for each sample to be ready for conducting the actual assays. Some assays do not require additional preparation before a measurement can be run; others require hours or days before measurements can be made.

  6. The time it takes to transport the sample to the appropriate facilities needed to perform the assay. If such transport is needed, this could be challenging both because of the intrinsic logistics of dealing with a large number of samples to transport out of an area and the severely compromised infrastructure in the affected area.

  7. The number of samples that can be processed in a given time. These challenges include both the availability of sufficient amounts of the equipment needed for carrying out the assays and the appropriate manpower, e.g., having sufficient personnel is a known bottleneck for the DCA (Maznyk et al. 2012; Wilkins et al. 2011).

  8. The time it takes to measure the prepared sample to obtain the desired result in an appropriate format suitable for use for the persons making the decision on the next steps for the victim. Depending on the specific assay, this can vary from almost no time to a considerable additional time.

  9. The time to deliver the information to the appropriate medical response decision maker. This includes any time needed to locate the victim in order to carry out any actions. This may be challenging for all assays that need to be done remotely, because usual forms for communication systems such as cell phones and internet are likely to be compromised and the victims may not be locatable at their usual contact sites at home or work.

The above time windows focus on entire time period needed for analysis from the onset of the event until the results are made available to the decision maker and victim. However, there is another factor that is critical to the success of a given biomarker. Namely, the results must be made available during the “window of opportunity” in which medical interventions can be effectively administered to the victim. The duration for this time window depends on the particular radiation mitigator or therapeutic that is being considered. The time of administration for the medicinal for the optimal effect might be shorter than the 6 days allocated in the planning scenario to process samples for up to one million people. For example, in the case of Neupogen and Neulasta (Amgen 2019a and 2019b), it is only 24 h, and for Leukine, it is 48 h post-radiation exposure (Sanofi-Aventis U.S. LLC, 2019). Currently, only these three countermeasures are FDA-approved as radio-mitigators for the indication of hematopoietic acute radiation syndrome (H-ARS) (Farese and MacVittie 2015; Singh and Seed 2018).

There are several additional basic science/clinical issues that need to be taken into consideration before organ specific biomarkers can be optimally utilized. Arguably, several of the biodosimetry methods/assays that have been developed for initial triage in large scale scenarios are in fact ‘organ-specific’, markers that assess damage to the hematopoietic system, e.g., DCA. However, the organ-specific biomarkers of interest in this article are those being developed with the specific intent to predict the magnitude of a given organ’s damage, with the intent to inform treatment directed toward specific organ injury:

  1. Organ-specific biomarkers intended to predict organ damage will be most useful when the dose is heterogeneous and when the victim is likely to survive any acute assault on the hematopoietic system. That is because the first line of treatment for homogeneous and life-threatening doses will focus on the hematopoietic system; thereafter, organ-specific damage assessments may be useful for the next stage of triage and treatment planning. However, when the dose is heterogeneous and thought to be life-threatening to specific radiation-sensitive organs, targeted treatment or mitigators can help save lives in the short and long-term and would greatly benefit from identifying the magnitude of the damage to the specific organ so that resources can be optimally deployed to those who can benefit most. Additionally, in order to identify which organs to assess, it also is crucial to know the distribution of the dose.

  2. It will be important to know whether and how the biomarker of radiation injury itself is affected by:
    1. the individual’s overall health and lifestyle habits, e.g., smoking;
    2. concomitant perturbations from the event experienced by the individual, including physical trauma and stress;
    3. whether the technical validity of the assay is impacted by damage to other organ systems, such as suppression of the immune response.

In summary, after a radiation scenario that puts large numbers of individuals at risk for clinically significant acute or long term health consequences, in order to derive full benefit from the use of organ-specific biomarkers as well as to provide essential information so that informed decision-making can be implemented, several essential types of information that are needed. The information is needed to determine who is at risk, what is the type and severity of the risk, and to guide appropriate medical intervention with treatment and/or the use of mitigators. In particular, prior to deciding which organ-specific biomarker of damage should be used, the following information is important to know:

  • Whether the individual has a credible risk of a having had a clinically significant exposure to radiation;

  • Whether the exposure was homogeneous (total-body) or heterogeneous (partial body);

  • If the exposure is homogeneous, whether the individual is expected to survive the initial injury to the hematopoietic system;

  • If the exposure was heterogeneous, the distribution of dose throughout the body. Knowing the distribution of dose would be a key to deciding which organs to assess for damage, i.e., which are unlikely to have received a high dose vs those at risk, in order to target injured organs for effective and efficient treatment. The dose distribution could also influence the decision to prioritize the patient for treatment. Knowledge of the dose to the organ would additionally provide a very useful check on potential false positive or false negative results from the organ-based biomarkers.

PART 2: POTENTIAL ROLE OF IN VIVO NAIL DOSIMETRY TO ENHANCE THE USEFULNESS OF ORGAN-SPECIFIC BIOMARKERS IN PLANNING MEDICAL RESPONSE FOR A LARGE-SCALE SCENARIO

In Part 1 of this paper, we concluded there are two important factors for the effective utilization of biomarkers of organ-specific injury, i.e., determine first whether the exposure is homogeneous (total-body exposure) and, if it is not, to determine the distribution of the dose. If the radiation exposure is mostly homogeneous, with no areas of bone marrow receiving substantially lower doses, then biodosimetry based on physical dose should be adequate for effective screening for risk of ARS. That is because, for mostly homogeneous exposure, the physical dose can be assumed to be the same for all organ systems and, for exposures in the treatable range, the predominant short-term risk for morbidity and mortality will be acute suppression of the bone marrow. In addition, the information on dose obtained using in vivo nail EPR, combined with the information from the organ-specific biomarkers, will provide potentially important data about the heterogeneity of biological responses to irradiation.

On the other hand, if the exposure is significantly heterogeneous, then it would be desirable to have detailed information on the distribution of the dose to various organs. That would enable the most effective use of the biomarkers of organ specific damage, because they could be utilized in those subjects with the highest possibility of significant damage to particular organs. This could contribute significantly to understand and treat acute and long-term effects.

In Part 2 we consider how a newly emerging physical biodosimetry method based on the use of Electron Paramagnetic Resonance (EPR) to measure in vivo the level of radiation induced free radicals in the nails of hands and feet can provide the types of information needed to meet these objectives.

EPR is a magnetic resonance technique which, for this application, uses a moderate magnetic field (340 mT) and an electromagnetic frequency that is similar to the frequency used in microwave ovens. EPR selectively responds to unpaired electrons such as those occurring in free radicals. EPR nail dosimetry is based on the measurement of the relatively stable (lasting at least several weeks) radiation induced free radicals generated in the keratin of the human nail, both fingernails and toenails. These radicals are produced by the direct interactions of the reactive intermediates generated by the ionizing radiation with keratin, and the intensity of the EPR signals of these radicals is directly proportional to the dose received.

Characteristics of in vivo EPR nail dosimetry that make them effective and complementary to the use of biomarkers of organ damage

Data on the feasibility of in vivo EPR nail dosimetry have been previously reported (Swarts et al 2018a). Swarts et al. also reported on the dose resolution (about 10 Gy) that had been achieved with the preliminary work and discussed areas of planned improvements that are expected to achieve a resolution of at least 1 Gy. In the following discussion we assume the successful implementation of the improvements and therefore assume a resolution of ~1 Gy. In vivo EPR nail dosimetry (Swartz et al 2014b) has several characteristics that are favorable for their original intended use (initial triage in large scale radiation events) and which in aggregate indicate that this technique can provide information, not available from any other biodosimetry technique, that is necessary in order for biomarkers of organ-specific damage to be effectively deployed. These characteristics are described below and, in italics, their relationship to the characteristics of biologically-based bioassays.

  1. In vivo EPR nail dosimetry is based on a physical process (generation of stable free radicals, proportional to dose in the dose ranges of interest) that is not confounded by the types of trauma and stress that are likely to occur in a disaster (Coleman and Koerner 2016). This is in contrast to the biological assays that can be impacted by concomitant stress or injury. Therefore, EPR measurements can be used to evaluate whether biomarkers of organ-specific damage could be misleadingly high if these conditions were present.

  2. The measurable effect of radiation on nails occurs instantaneously upon radiation exposure, is independent of the dose rate of exposure, and reflects the cumulative dose at the site of the nails. Although there is a need to confirm the observation, preliminary data indicate that in vivo the radiation-induced signal remains stable from the time of exposure to several weeks or months afterwards. (If this is not the situation, then some modifications to the protocol will be needed, but these should not impact the utility of the methodology.) This is in contrast to biologically based biomarkers which typically have a time-dependent pattern of changes (as illustrated in Fig. 1a). (See Fig. 1b for the contrasting dynamics for in vivo nails which are essentially unchanging during the pertinent periods for triage.) Consequently, as discussed in Part 1 of this paper, almost all of the “Principal Time Intervals for Evaluation Biomarkers” delineated in Part 1 (i.e., those numbered 1,2,4-6, and 8,9) are minimal or non-existent for measurements of in vivo nails.

  3. The measurements can be made immediately after the event and at any time throughout the relevant time periods (Fig. 1b). In contrast, biologically based biomarkers often have a limited time during which a valid sample can be collected (the diagnostic period in Fig. 1a).

  4. Measurements can be repeated as desired because they are non-invasive and non-destructive of the ‘sample’ and, as noted above, the response remains stable during the relevant time period when measurements would be appropriate. This provides opportunities for confirmation and validation of initial results, including modifying the measurement techniques to enhance accuracy. Such a capability might also be available for some biologically-based bioassays.

  5. EPR measurements can be carried out in the designated temporary facilities that are set-up near the event site as part of the response to the scenario. The instrument is easily deployable and can be operated with minimal needs for supplies and by operators with limited training, i.e., training can be completed via a few-minute video embedded into the software. In contrast, at least some of the biomarkers are likely to require that analyses be carried out using specialized equipment and experts located remotely.

  6. The results are immediately available. Data processing of the measurements by the software will produce results instantaneously following completion of data acquisition that are appropriate for the use by the medical decision makers. This capability is likely to vary with the type of biologically based biomarker assays.

  7. Measurements are completely noninvasive; there is no sample to be collected, as the measurement is done in vivo. Most biomarker assays will require collection of a sample, (often a small amount of blood, but sometimes a small amount of tissue, feces, or urine), which is then brought to the analytic instrument in the field or at a distant site.

  8. Measurements of nails from multiple limbs can be used to compute dose distribution immediately, providing unambiguous evidence on probability of the exposure being homogeneous. It should also be feasible to compute the distribution of dose throughout the body if the exposure was heterogeneous. The simulation models will be particularly focused on providing the dose to any and all radiation-sensitive organs. The software can produce this information at the same time as providing individual dose estimates for each limb. This type of information is unlikely to be available from biomarker assays. Although some have been argued to distinguish partial body vs whole body exposures, the biomarkers cannot be used to establish dose distribution (Vaurijoux et al. 2012).

  9. The EPR dosimetric measurements can be made with throughput times of less than 5 minutes of measurement time per subject from ‘sampling’ the nails on each limb to providing all of the results. In contrast, throughput for biologically-based biomarkers varies from several minutes to several days, not including any limitations on when samples can be taken after exposure and any transportation times to send samples to distant laboratories and the time to send results back to the medical response team.

  10. Because the method is based on physical changes that are unlikely to be perturbed by disease or stress (Swartz 2016), patients undergoing therapeutic whole-body irradiation (or partial body irradiation that include exposure of the nails) are suitable test subjects, providing a means to test the effectiveness of measurements made directly in human subjects who were exposed to radiation in vivo. Biologically-based biomarkers typically are potentially confounded by the underlying diseases and their medical treatments, such as chemotherapy. Therefore, they need to be developed by a preclinical model irradiating and measuring in vivo in animals.

Method for making measurements of radiation dose with EPR in vivo nail dosimetry

In vivo nail biodosimetry is applicable to the entire population as long as the individual can cooperate. (Separate versions would need to be developed for unconscious subjects and infants/very young children, because the existing versions assume that the subject can actively cooperate with being measured, especially in regard to restricting motion of the limbs.) Details of the prototype for the resonator and methods for measuring nails in vivo are reported elsewhere (Swarts et al. 2018; Sidabras et al. 2014). Briefly, each digit that is to be measured is first placed into a digit/resonator holder (see Fig 2). This support system facilitates accurate and rapid placement of the nail plate under a novel Surface Resonator Array (SRA) (Sidabras et al. 2014) that is held in place inside one of four magnets, each of which is placed for ideal use for a specific limb (Fig. 3). The magnetic field of the permanent magnet provides a uniform magnetic field in the region of the nail. Up to five digits per limb can be measured simultaneously within each magnet.

Figure 2:

Figure 2:

Open in a new tab

Support for finger and SRA resonator (shown on one digit only): Support is placed on digits outside of spectrometer and is used for proper placement of the digit into the resonator held inside the magnet

Figure 3:

Figure 3:

Open in a new tab

Person being measured by in vivo nail EPR spectroscopy on all four limbs simultaneously

The region of the nail that is measured is deliberately located away from the margins of the nails; therefore, there will be no need to be concerned with dirt under the nails or the impact of recent clipping, etc. (He et al. 2014; Wilcox et al. 2010, Wang et al. 2015). Variation in curvature and size of the nailbed (Murdan 2011) has been investigated and appears not to be a problem in preliminary studies. bThere is a remaining need to determine if the in vivo measurements of nails are perturbed by the presence of nail polish or other cosmetic treatments (Trompier et al. 2015). If perturbations are determined to impact in vivo measurements, there will be a need for a step to remove the polish/cosmetic treatment prior to the measurements.

It is feasible to make simultaneous measurements on each digit. In addition, measurements can be carried out simultaneously on all four limbs (Fig 3). The total measurement time can be less than five minutes, because all measurements of the digits can be done simultaneously. All aspects of the technique can readily be automated so that the measurements can be carried out by operators with no previous experience in the technique. The results, expressed as dose in each measured digit and aggregated by limb, will be immediately available at the conclusion of the measurements.

The data output will also include an automated calculation of the probability that the exposure was inhomogeneous and the extent (dose distribution) of the heterogeneity. Although estimates of dose distribution are not yet fully developed, it should be quite feasible, based on techniques already in use in radiation therapy, to go from doses at each of the four limbs to a map of the dose distribution throughout the whole body, including estimates of doses to each radiation-sensitive organ system (Petroccia et al. 2017; Herve et al. 2007; Borrego et al. 2017; Sands et al. 2017; Wayson and Bolch 2018; Khailov et al. 2015). This calculation should be able to be made automatically by the software, with the results available immediately after the measurements are completed.

Use of in vivo EPR nail dosimetry to determine homogeneity of the exposure

While the need to have information about the homogeneity of the exposure to radiation has long been recognized as essential in order to determine the risk for ARS (Coleman and Koerner 2016), there is currently no other method available for determining the homogeneity of an exposure to radiation in the context of a large-scale radiation incident. In small incidents such as accidents involving ionizing radiation, determining homogeneity of exposure usually requires physical dosimetry, but, in large-scale scenarios, there seems to be no practical method to do this, e.g., by having pre-placed physical dosimeters throughout the general population and then using computational reconstruction of the exposure based on where each person was located at the time of the event.

Biologically-based biodosimetry does not currently have the capability to provide rapid indications of the homogeneity of the exposure, and none can estimate dose distribution. While in principle the DCA assay can provide some indication of the presence of heterogeneous exposure, this technique requires considerable time because of the need to transport the samples to laboratories outside of the affected area and the intrinsic times for sampling and completing the assay (Prasanna et al. 2010b). As already noted, there also are significant logistical problems with available capacity to complete large numbers of assays within the required timeframe of a large-scale radiation scenario.

On the other hand, EPR in vivo nail dosimetry has the capability of making the determination of heterogeneity of exposure, in a few minutes, directly in the individual in the field, eliminating the need to send samples to a distant site. It can provide rapid and robust indications of homogeneity by measurements at four widely separated anatomical sites, the two hands and the two feet.

If simplification is needed, the simplest situation would be to use estimates based on measuring only a single digit on two limbs, which would provide a limited but useful indicator of homogeneity of exposure. A more robust indication could be obtained by measuring a digit on at least 3 limbs. Because each digit is measured individually with its own resonator, the precision of the measurements on each limb can be enhanced by measuring additional digits on the same limb.

Use of EPR nail dosimetry to determine the full distribution of the radiation dose

In order to determine the physical distribution of exposure dose, geometrically defined information seems likely to be essential. This is unlikely to be achieved by assays that utilize changes in body fluids. While in principle information on distribution might be obtained by biopsy at multiple sites, that is very unlikely to be practical for a large scale scenario.

One of the very attractive features of in vivo EPR nail dosimetry is that it can obtain information on dose at four distinctly different sites. As noted above, with these data we should be able to provide a robust indication of whether the exposure was homogeneous unless there were markedly heterogeneous exposures to the head or to a very small volume elsewhere in the body. Based on existing utilization of point measurements to determine dose distribution and with the information on dose from four sites, it should be feasible to map the variation of the exposure over the entire body. The algorithm to carry out this calculation could include input on the quality of the radiation, the dose rate associated with the incident, and whether there were particles from fallout on any of the digits. (The method will not consider internal radiation, but that is not likely to be a major factor for acute effects [Swartz et al. 2014]).

While there is considerable experience in radiation therapy for making such calculations based on measurements at well characterized locations, in order to calculate dose distributions based on a few experimental dose measurements (Petroccia et al. 2015; Herve et al. 2007; Borrego et al. 2017; Sands et al. 2017; Wayson and Bolch 2018; Khailov et al. 2015), the situation gets more complex when there is some uncertainty in the position of the limbs, as would be likely for the hands. If the exposure was over a very short period of time, e.g., primarily from prompt radiation, it may be possible to have the subject describe the likely position of the hands. If the exposure occurs over a longer period of time, such as from fallout, then it will be reasonable to assume an averaged position.

The algorithm for calculating the whole body distribution of dose can include a consideration of the additional uncertainties, such as those due to the lack of precision of knowledge of the position of the limbs. While we will need to develop the specific algorithm based on the use of dose at the nails, we will be able to use well investigated techniques to do this as applied previously for diagnostic radiology, radiation therapy, and accident dosimetry (Petroccia et al. 2017; Herve et al. 2007; Borrego et al. 2017; Sands et al. 2017; Wayson and Bolch 2018; Khailov et al. 2015).

SUMMARY

In order to respond effectively to a large-scale radiation scenario in which there are many more individuals potentially at risk than can be brought immediately into the medical care system for direct medical observation for development of clinical manifestations, it is essential that effective triage be carried out as promptly as possible. There also is a need for the early identification of individuals who have received clinically significant exposures that could be ameliorated by available mitigators or early treatment.

Recently, there has been increased interest in using injury rather than dose to carry out the initial triage. The focus on estimating injury is based on the assumption that individuals with the same exposure can vary significantly in their organ-specific responses to radiation injury. More importantly, the assumption is that treatments will be more effective, if based on information that estimates an individual’s organ-specific damage. The suggested approach is the use of biomarkers that could provide organ-specific predictions of the radiation injury. While the biological base for such assumptions has not been fully demonstrated, work on the development of such bioassays is under development.

In Part 1, using prior analyses of the suitability of biodosimeters for the response to a large-scale radiation scenario, we examined the logistical and practical features of bioassays of organ-specific damage and whether there are important pieces of information that would make such biomarkers suitable for use in the early triage decisions. We concluded that there are two especially important pieces of information that are needed prior to being able to effectively utilize such biomarkers: 1) whether the exposure was homogeneous and 2) if there was significant heterogeneity, determination of the distribution of the dose.

The rationale for our conclusions rests on well understood clinical grounds: If radiation exposure is mostly homogeneous, with no area containing significant amounts of bone marrow receiving substantially lower doses, then biodosimetry based on physical dose should be adequate for effective triage for risk of ARS. On the other hand, if the exposure is significantly heterogeneous, then it would be desirable to have detailed information on the distribution of the dose. That would enable the most effective use of the biomarkers of organ specific injury, because they could be utilized in those subjects with the highest possibility of significant damage to particular organs to predict the need for acute and long term medical intervention, using criteria that presumably will be developed by the government in the plans for response.

In Part 2, we describe how a newly emerging physical biodosimetry method, i.e., the in vivo use of EPR to measure the amount of radiation induced radicals in the nails on all limbs, appears to be to provide the types of information needed to optimize the use of biomarkers of organ-specific injury. Because in vivo nail dosimetry obtains data from four separate, dispersed anatomical locations, it can be used to assess homogeneity, and, if there is significant heterogeneity, the measurements from the four limbs can be used to estimate the distribution of the dose to radiation-sensitive organs.

Also importantly, because in vivo nail dosimetry can rapidly provide quantitative data about the level of exposure and can be used at point-of-care settings, these measurements should clearly distinguish between the worried well and subjects who have received doses of radiation that could lead to significant morbidity and mortality, It also could help to identify subjects whose dose is likely to be greater than would be compatible with short term survival, so that they could be handled appropriately.

It is important to note that this combined approach is complementary, not competitive to the information obtained by bioassays based on organ specific damage. The combination of physical biodosimetry, biological biodosimetry, and clinical observation will optimally serve the goal of providing the most effective response to a large-scale radiation scenario.

Acknowledgments

Funding source: This study was funded in part by grant U19AI091173 from Centers for Medical Countermeasures Against Radiation (CMCR) in the National Institute of Allergy and Infectious Diseases (NIAID).

Footnotes

Conflicts of interests and sources of funding:

HMS and ABF: Co-owners of Clin-EPR, LLC, which makes and sells EPR instruments for clinical and preclinical investigational use only.

No other author has any other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript

VKS: The opinions or assertions contained herein are the private views of the authors and are not necessarily those of the Uniformed Services University of the Health Sciences or the Department of Defense, USA. Mention of trade names, commercial products, or organizations does not imply endorsement by the United States Government.

a.

Note: Some of these characteristics apply equally to EPR dosimetry based on nail clippings. However, this paper is restricted to in vivo nail dosimetry for two principal reasons. Measuring nail beds in vivo obviates the problems that occur because clipping of nails produces a confounding signal. To deal with these problems, clipped nails need to be sent to specialized laboratories for analysis and require complex handling of the clippings. In contrast, in vivo nail EPR can be accomplished at the point-of-care without special handling to eliminate the effects of clipping.

b.

Whether certain kinds of nail diseases and/or their treatment affect measurements has not yet been determined.

References

  1. Amgen Inc. Highlights of prescribing information for Neulasta (pegfilgrastim) injection for subcutaneous use [online]. Available at http://pi.amgen.com/united_states/neulasta/neulasta_pi_hcp_english.pdf. Last assessed 24 Aug 2019a.
  2. Amgen Inc. Highlights of prescribing information for Neupogen (filgrastim) injection for subcutaneous or intravenous use [online]. Available at http://pi.amgen.com/united_states/neupogen/neupogen_pi_hcp_english.pdf Last accessed 24 Aug 2019b.
  3. Borrego D, Siragusa DA, Balter S, Bolch WE. A hybrid phantom system for patient skin and organ dosimetry in fluoroscopically guided interventions. Med Phys Sep;44(9):4928–4942; 2017. doi: 10.1002/mp.12419. Epub 2017 Jul 28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Brenner DJ, Chao NJ, Greenberger JS, Guha C, McBride WH, Swartz HM, Williams JP. Are we ready for a radiological terrorist attack yet? Report from the CMCR Network. Int J Radiat Oncol Biol Phys 92(3):504–505; 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Coleman CN, Koerner JF. Biodosimetry: medicine, science, and systems to support the medical decision-maker following a large scale nuclear or radiation incident. Radiat Prot Dosim 172(1–3):38–46; 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Dainiak N Medical management of acute radiation syndrome and associated infections in a high-casualty incident. J Radiat Res 59(S2):ii54–ii64; 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Farese AM, MacVittie TJ. Filgrastim for the treatment of hematopoietic acute radiation syndrome. Drugs today. 51:537–48; 2015. [DOI] [PubMed] [Google Scholar]
  8. Flood AB, Nicolalde RJ, Demidenko E, Williams BB, Shapiro A, Wiley AL, Swartz HM. A framework for comparative evaluation of dosimetric methods to triage a large population following a radiological event. Radiat Measure 46(9): 916–922, 2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Flood AB, Boyle HK, Du G, Demidenko E, Nicolalde RJ, Williams BB, Swartz HM. Advances in a framework to compare bio-dosimetry methods for triage in large-scale radiation events. Radiat Prot Dosim 159(1–4):77–86; 2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Flood AB, Ali AN, Boyle HK, Du G, Satinsky VA, Swarts SG, Williams BB, Demidenko E, Schreiber W, Swartz HM. Evaluating the special needs of the military for radiation biodosimetry for tactical warfare against deployed troops: Comparing military to civilian needs for biodosimetry methods. Health Phys 111(2):169–82; 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. He X, Swarts SG, Demidenko E, Flood AB, Grinberg O, Gui J, Mariani M, Marsh SD, Ruuge AE, Sidabras JW, Tipikin D. Development and validation of an ex vivo electron paramagnetic resonance fingernail biodosimetric method. Radiat Prot Dosim 159(1–4):172–181; 2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Herve ML, Clairand I, Trompier F, Tikunov D, Bottollier-Depois JF. Relation between organ and whole body doses and local doses measured by ESR for standard and realistic neutron and photon external overexposures. Radiat Prot Dosim 125(1–4):355–360; 2007. [DOI] [PubMed] [Google Scholar]
  13. Khailov AM, Ivannikov AI, Skvortsov VG, Stepanenko VF, Orlenko SP, Flood AB, Williams BB, Swartz HM. Calculation of dose conversion factors for doses in the fingernails to organ doses at external gamma irradiation in air. Radiat Meas 82:1–7; 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Maznyk NA, Wilkins RC, Carr Z and Lloyd DC The capacity, capabilities and needs of the WHO BioDoseNet member laboratories. Radiat Prot Dosim 151, 611–620; 2012. [DOI] [PubMed] [Google Scholar]
  15. Murdan S Transverse fingernail curvature in adults: a quantitative evaluation and the influence of gender, age, and hand size and dominance. Int J Cosmet Sci 33(6):509–513; 2011. [DOI] [PubMed] [Google Scholar]
  16. National Institutes of Health (NIH), DHHS. RFA-AI-18–045: Radiation biodosimetry assays and devices [online]. Available at https://grants.nih.gov/grants/guide/rfa-files/RFA-Ai-18-045.html. Last accessed 21 Aug 2019.
  17. Paul S, Amundson SA. Development of gene expression signatures for practical radiation biodosimetry. Int J Radiat Oncol Biol Phys 71:1236–1244; 2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Petroccia H, Mendenhall N, Liu C, Hammer C, Culberson W, Thar T, Mitchell T, Li Z, Bolch W. A hybrid phantom Monte Carlo-based method for historical reconstruction of organ doses in patients treated with cobalt-60 for Hodgkin’s lymphoma. Phys Med Biol 62(15):6261–6289; 2017. doi: 10.1088/1361-6560/aa7c2f. [DOI] [PubMed] [Google Scholar]
  19. Prasanna PGS, Blakely WF, Bertho JM, Chute JP, Cohen EP, Goans RE, Grace MB, Lillis-Hearne PK, Lloyd DC, Lutgens L, Meineke V, Ossetrova NI, Romanyukha A, Saba JD, Weisdorf DJ, Wojcik A, Yukihara EG and Pellmar TC. Synopsis of partial-body radiation diagnostic biomarkers and medical management of radiation injury workshop. Radiat Res 173(2):245–53; 2010a. pmid:20095857. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Prasanna PGS, Moroni M, Pellmar TC. Triage dose assessment for partial-body exposure: dicentric analysis. Health Phys 98(2):244–51;2010b. pmid:20065689. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Rogan PK, Li Y, Wilkins RC, Flegal FN, Knoll JH. Radiation dose estimation by automated cytogenetic biodosimetry. Radiat Prot Dosim 172(1–3):207–217; 2016. [DOI] [PubMed] [Google Scholar]
  22. Rothkamm K, Barnard S, Ainsbury EA, Al-hafidh J, Barquinero JF, Lindholm C, Moquet J, Perala M, Roch-Lefevre S, Scherthan H, Thierens H, Vral A, Vandersickel V. Manual versus automated gamma-H2AX foci analysis across five European laboratories: can this assay be used for rapid biodosimetry in a large scale radiation accident? Mutat Res Genet Toxicol Environ Mutagen 756:170–173; 2013a. [DOI] [PubMed] [Google Scholar]
  23. Rothkamm K, Beinke C, Romm H, Badie C, Balagurunathan Y, Barnard S, Bernard N, Boulay-Greene H, Brengues M, De Amicis A, De Sanctis S, Greither R, Herodin F, Jones A, Kabacik S, Knie T, Kulka U, Lista F, Martigne P, Missel A, Moquet J, Oestreicher U, Peinnequin A, Poyot T, Roessler U, Scherthan H, Terbrueggen B, Thierens H, Valente M, Vral A, Zenhausern F, Meineke V, Braselmann H, Abend M. Comparison of established and emerging biodosimetry assays. Radiat Res 180:111–119; 2013b. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Sands MM, Borrego D, Maynard MR, Bahadori AA, Bolch WE. Comparison of methods for individualized astronaut organ dosimetry: Morphometry-based phantom library versus body contour autoscaling of a reference phantom. Life Sci Space Res 15:23–31, 2017. doi: 10.1016/j.lssr.2017.07.002. [DOI] [PubMed] [Google Scholar]
  25. Sanofi-Aventis U.S. LLC. Highlights of prescribing information for LEUKINE® (sargramostim) for injection, for subcutaneous or intravenous use [online]. Available at https://www.accessdata.fda.gov/drugsatfda_docs/label/2018/103362s5240lbl.pdf?utm_campaign=20180329%20MCMi&utm_medium=email&utm_source=Eloqua. Last accessed 24 Aug 2019.
  26. Sidabras JW, Varanasi SK, Mett RR, Swarts SG, Swartz HM, Hyde JS. A microwave resonator for limiting depth sensitivity for electron paramagnetic resonance spectroscopy of surfaces. Rev Sci Instrum 85(10):104707; 2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Simon SL, Bailey SM, Beck HL, Boice JD, Bouville A, Brill AB, Cornforth MN, Inskip PD, McKenna MJ, Mumma MT, Salazar SI, and Ukwuani A. Estimation of radiation doses to US military test participants from nuclear testing: a comparison of historical film-badge measurements, dose reconstruction and retrospective biodosimetry. Radiat Res 191(4):297–310; 2019. 10.1667/RR15247.1. [DOI] [PubMed] [Google Scholar]
  28. Singh VK, Santiago PT, Simas M, Garcia M, Fatanmi OO, Wise SY, Seed TM. Acute radiation syndrome: An update on biomarkers for radiation injury. J Radiat Cancer Res 9:132–46; 2018. [Google Scholar]
  29. Singh VK, Seed TM. An update on sargramostim for treatment of acute radiation syndrome. Drugs Today 54:679–93; 2018. [DOI] [PubMed] [Google Scholar]
  30. Sullivan JM, Prasanna PG, Grace MB, Wathen L, Wallace RL, Koerner JF, Coleman CN. Assessment of biodosimetry methods for a mass-casualty radiological incident: medical response and management considerations. Health Phys 105(6):540–554; 2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Swarts SG, Sidabras JW, Grinberg O, Tipikin DS, Kmiec M, Petryakov S, Schreiber W, Wood VA, Williams BB, Flood AB, Swartz HM. Developments in biodosimetry methods for triage, with a focus on x-band electron paramagnetic resonance in vivo fingernail dosimetry. Health Phys 115(1):140–150; 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Swartz HM. Using stable free radicals to obtain unique and clinically useful data in vivo in human subjects. Radiat Prot Dosim 172 (1–3): 3–15; 2016. doi: 10.1093/rpd/ncw323 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Swartz HM, Flood AB, Williams BB, Dong R, Swarts SG, He X, Grinberg O, Sidabras J, Demidenko E, Gui J, Gladstone DJ, LA Kmiec MMK, Lesniewski PN, Marsh SDP, Matthews TP, RJ, Pennington PM, Raynolds T, Salikhov I, Wilcox DE and Zaki BI. Electron paramagnetic resonance dosimetry for a largescale radiation incident. Health Phys 103(3):255–267; 2012. PMCID: PMC3649772 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Swartz HM, Flood AB, Williams BB, Meineke V, Dorr H. Comparison of the needs for biodosimetry for large-scale radiation events for military versus civilian populations. Health Phys 106(6):755–63; 2014. [DOI] [PubMed] [Google Scholar]
  35. Swartz HM, Williams BB, Flood AB. Overview of the principles and practice of biodosimetry. Radiat Environ Biophys 53:221–232; 2014b [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Trompier F, Romanyukha A, Swarts S, Reyes R, Gourier D. Influence of nail polish in EPR dosimetry with human nails. Radiat Meas 75:6–8, 2015. [Google Scholar]
  37. Vaurijoux A, Gregoire E, Roch-Lefevre S, Voisin P, Martin C, Voisin P, Roy L and Gruel G. Detection of partial-body exposure to ionizing radiation by the automatic detection of dicentrics. Radiat Res 178(4): 357–364, 2012. [DOI] [PubMed] [Google Scholar]
  38. Wang L, Wang X, Zhang W, Zhang H, Ruan S, Jiao L. Determining dosimetric properties and lowest detectable dose of fingernail clippings from their electron paramagnetic resonance signal. Health Phys 109(1):10–14; 2015. [DOI] [PubMed] [Google Scholar]
  39. Wang Q, Rodrigues MA, Repin M, Pampou S, Beaton-Green LA, Perrier J, Garty G, Brenner DJ, Turner HC, and Wilkins RC. Automated triage radiation biodosimetry: integrating imaging flow cytometry with high-throughput robotics to perform the cytokinesis- block micronucleus assay. Radiat Res 191(4):342–351; 2019. doi: 10.1667/RR15243.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Wayson MB, Bolch WE. Individualized adjustments to reference phantom internal organ dosimetry-scaling factors given knowledge of patient external anatomy Phys Med Biol. 63(8):085007; 2018. doi: 10.1088/1361-6560/aab72f. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Wilcox DE, He X, Gui J, Ruuge AE, Li H, Williams BB, Swartz HM. Dosimetry based on EPR spectral analysis of fingernail clippings. Health Phys 98(2):309; 2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Wilkins RC, Romm H, Oestreicher U, Marro L, Yoshida MA, Suto Y and Prasanna PGS. Biological dosimetry by the triage dicentric chromosome assay—further validation of international networking. Radiat Meas 46, 923–928; 2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Xu Y, Turner HC, Garty G, Brenner D. A rapid, quantitative method to characterize the human lymphocyte concentration for automated high-throughput radiation biodosimetry. Biomed Eng Res 2:16–19; 2013. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES