To date no biodosimetric system or assay optimally meets all of the requirements discussed above. We will use the RABiT system (Garty et al. 2010
; Garty et al. 2011
) with which we are intimately familiar as a case study for all such systems. We expect the RABiT to be typical of any automated, centralized biodosimetry system using blood-based assays, particularly when considering sample collection logistics. The RABiT approach is to use well established biodosimetry assays which are currently performed manually, and fully automate them using robotic technology, a multi-well plate platform, and advanced imaging approaches. Using ‘mature’ assays has considerable advantages in that 1) the assays are already well characterized, and 2) there is a more direct regulatory route to deployment, as the assays are already in use.
The RABiT analyzes fingerstick-derived blood samples (~30 μl,), either to estimate past radiation dose, or to identify individuals exposed above/below a cutoff dose. The RABiT fully automates two mature, but formerly manual, biodosimetry assays (Cytokinesis Block Micronucleus assay (CBMN) (International Atomic Energy Agency 2001
; Fenech et al. 2003
) and phosphorylation of the histone H2AX (γ-H2AX) (Nakamura et al. 2006
; Turner et al. 2011
)), thus converting them to ultra high throughput. Recent reliability and performance testing (Garty et al. 2011
) indicated a maximum throughput of 30,000 samples per RABiT machine per day is achievable. We are currently partnered with Northrop Grumman Security Systems (Linthicum, MD, USA), to develop a field-deployable system based on the RABiT prototype. As part of this effort we will be conducting large scale testing to demonstrate that the high-throughput system can achieve equivalent dose estimates to those obtained by manual processing.
The RABiT was designed as a flexible robotically-based system, to potentially allow routine multi-use applications in a hospital or clinic setting. Examples are cytogenetic assays such as amniocentesis, or potentially for multiplex immunoassays, such as screening for multiple cytokines. A simple adaptation of the RABiT technology would allow rapid screening for individual radiosensitivity, with potential applications both for radiation oncology and radiology. These alternate uses are currently under preliminary study.
The two current RABiT assays were chosen both for their maturity and for their potential to be fully automated. Both have advantages and disadvantages (Amundson et al. 2001
). Specifically, the γ-H2AX assay is rapid (<2 h) but the signal lasts only a few days post exposure (Redon et al. 2010
). By contrast, the micronucleus signal is stable for many months, but the current micronucleus assay, though now high-throughput, takes ~70 hours to generate the dose estimate.
To achieve very high throughputs, the RABiT contains the following key technological innovations:
- Use of small volumes of blood (~30 μl) from a standard lancet fingerstick; this is a minimally invasive, and thus potentially high throughput, approach – conventional venipuncture is not compatible with ultra high throughput.
- Complete robotically-based automation of the biology, with biological processing and imaging performed in situ in multi-well plates. This allows rapid processing of multiple simultaneous samples. The use of filter bottomed multiwell plates prevents loss of lymphocytes during fluid removal steps.
- Innovations in high-speed imaging allow rapid analysis following biological processing.
The RABiT prototype, under testing at Columbia University is shown in . The RABiT consists of 7 stations arranged around a SCARA (Selective Compliant Articulated Robot Arm) that transfers samples from station to station:
Breadboard prototype of the RABiT system.
Blood samples arriving from the field are placed into centrifuge buckets on the input stage. A centrifuge is used for separating lymphocytes from red blood cells. At each centrifugation cycle 384 capillaries are spun simultaneously. The lymphocyte harvest station (Garty et al. 2011
) transfers the lymphocyte band from the capillaries to a 96-well plate for further processing. The biological assays are performed in an automated liquid handling system where reagents can be added or removed as needed. A robotically controlled incubator is used for lymphocyte culturing in the CBMN assay.
After the lymphocytes have been fixed and stained the plates are moved to a transfer to substrate system (Chen et al. 2010
) where the filter bottoms are removed from the multiwell plates and sealed between two layers of transparent tape. Finally the lymphocytes are imaged using a custom built imaging system.
By modifying the number of cells scored per sample, throughput and sensitivity can be adjusted. During the initial triage, only a small number of cells may be scored to obtain a crude dose estimate (e.g. above/below 2 Gy). At a later stage, after all samples were triaged. They can be re-imaged and analyzed at higher statistics (and therefore lower throughput) to achieve a more precise dose, as would be required for long-term follow up. As described in (Garty et al. 2011
; Turner et al. 2011
), the γ-H2AX assay, as implemented in the RABiT, is sensitive between 1 Gy and 8 Gy, with higher sensitivity achievable through a collection of higher statistics. The micronucleus assay has similar accuracy (McNamee et al. 2009
). This is well matched to the dose range required for triage following a radiological event, see below.
The RABiT system imposes several requirements on the sample collection process:
- Samples will be collected in the field and will need to be transported to the RABiT with no spillage and no cross contamination.
- The RABiT is designed to isolate lymphocytes, by centrifugation, from small volumes of whole blood, in heparin-coated capillaries. To ensure separation of lymphocytes out of whole blood samples, the blood needs to be layered above separation medium with no mixing.
- The lymphocytes in the collected blood need to be kept viable as the micronucleus assay requires them to be cultured to division.
- During transport, the blood may need to be kept chilled to prevent γ-H2AX foci repair (see Figure 4a in (Moroni et al. 2008)).
These requirements are not specific to the RABiT and would need to be maintained for almost any automated biodosimetry system, implementing a blood-based biodosimetry assay.
RABiT sample collection
Sample collection for the RABiT, as described below addresses the concerns noted above. As it does not require highly trained personnel, it can be easily merged into the emergency response scenario detailed above. Sample collectors, at the CRC or elsewhere, will draw the blood, by fingerstick and verify the contact information. Individuals with other injuries (e.g. trauma) will be triaged by a medical professional and can be evacuated to a hospital. Those who appear healthy, other than the possible radiation exposure, will be sent home after the sample collection. Samples will then be packed and transported to the RABiT (which may be across the hall or in a different state, but most likely, at the nearest large medical center).
Sample collection kit
In order to facilitate blood collection, by minimally trained individuals, we have developed a sample collection kit (), consisting of lancets, bar-coded, heparin coated, capillary tubes with matched personal data cards and patient tracking wristbands, alcohol wipes and sample holders for filled capillaries. The kit is designed to match the 32 samples that can be collected over a 2–3 hour collection period by one sampler. We envision a few hundred such collection kits would be kept at local emergency response stores, as part of the CRC Go-kit (May et al. 2007
) and would be ready to be used immediately. A much larger number of kits can then be stored at the Strategic National Stockpile (SNS), as part of the “12-hour push package” (Esbitt 2003
), and will arrive at the CRC within 12 hours of a request by local authorities.
(a) Sample collection kit, (b) data collection card with capillary, (c) close-up of bar-coded capillary, and (d) wristband.
Data collection card
On entering the CRC, Individuals will be handed a data collection card (), where they are to enter personal and contact information. In addition to the contact details, processing in the RABiT may require knowing the age, gender and smoking status so that this information is also included. The card has a printed barcode which is matched to the barcode etched on a heparinized PVC (Poly Vinyl Chloride) capillary (), attached to the card, and a detachable human readable version of the same code, with instructions on how to obtain the results of the blood test, is also provided. Alternatively, the card can contain an integrated self laminating wristband () which is detached and applied to the individual. The wristband contains information allowing the individual, or their medical caregiver to obtain the results of the blood test 1–3 days following the sample collection.
Since the RABiT requires 30 μl of blood and since multiple fingersticks would reduce the processing throughput, the reliability of the lancet in producing large blood volumes is critical. Although standard “diabetic” lancets are not sufficient for this purpose, as they are required to provide less than 5 μl of blood (Yum & Roe 1999
), other, commercially available, lancets have larger blades which penetrate deeper into the skin and typically result in 50μl of blood or more (Fruhstorfer 2000
; Garty et al. 2010
). Care should be made, to select this class of lancet for the sample collection kit.
Fingerstick sampling procedure
After loading about 30 μl of blood into the capillary, the sample collector then seals the top of the capillary with their (gloved) thumb, begins inserting it into the holder, which is preloaded with separation medium and sealing putty (), while releasing their thumb to allow trapped air to escape from the capillary.
Figure 4 Scheme of the sample collection. (a) Sample holder with sealing putty (P) and separation medium (M); (b) blood in a capillary (B); (c) capillary loaded into sample holder layering blood above the separation medium without mixing; and (d) photograph of (more ...)
As the blood in the capillary () does not reach its edge, when the capillary is inserted into the holder, an air bubble is trapped between the blood and separation medium, preventing their mixing during shipping (up to 24 hours).
The sealing putty is compressed into and around the capillary ensuring a seal (), requiring a small force to extract the capillary from the holder. This prevents the capillary from falling out even if the holder is inverted and vigorously shaken, but still allows the RABiT robotics to extract the capillary from the holder (Garty et al. 2011
). As the bottom of the capillary is sealed, the blood and separation medium cannot leak out. This procedure allows the sample to be collected by an individual with minimal training, while maintaining the required layering of the blood and separation medium and preventing contaminations. We have seen that the technique for this can be learned in a few minutes.
Transporting samples to the RABiT
After the capillary holder is filled with 32 capillaries, the top of the capillaries is sealed with a foam rubber mat, to prevent cross-contamination of the samples, and the capillary holder can be wrapped and shipped to the RABiT.
As the γ-H2AX assay, which does not require culturing the lymphocytes, provides a much faster processing (a few hours compared to 3 days for the micronucleus assay), it is the assay of choice for rapid triage. To reduce γ-H2AX signal decay during shipping, the samples need to be chilled to 4–10 °C. This can be done by adding ice packs in with the samples for shipping (see for example (Kendal et al. 1997
No such cooling needs to be done for the micronucleus assay. Indeed, we have verified that capillaries stored at room temperature for 24 to 48 hours, still contain a sufficient quantity of viable lymphocytes, which undergo mitosis when stimulated in the RABiT.