Getting smart with coagulation

1 INTRODUCTION

“Coagulation” is a word often used in this journal. One dictionary definition identifies coagulation as “the action or process of a liquid, especially blood, changing to a solid or semi-solid state.” Coagulation is an essential element of hemostasis and fundamentally reflects the activation of secondary hemostasis, with thrombin generation eventually triggering the conversion of the major hemostasis protein fibrinogen (soluble) to fibrin (insoluble), which in vivo is ultimately effective to stabilize the primary (prevalently platelet) clot after activation of primary hemostasis. Inadequate coagulation can arise in a variety of disorders.¹ In vitro laboratory tests of coagulation tend to separate some components of primary hemostasis, viz the platelets, and assess secondary hemostasis in a cell-free system. The main routine coagulation assays are the activated partial thromboplastin time (APTT) and the prothrombin time (PT).^{2, 3} In a laboratory setting, these tests are performed using platelet poor plasma, are sensitive to presence or absence of various drugs and exogenous inhibitors of coagulation factors, and are thus used to screen for the presence of hemostasis dysfunction, or to monitor various types of anticoagulant therapy. For example, APTT is often used for monitoring heparin therapy,^{2, 4} and PT, as typically converted to an INR (international normalized ratio), used for monitoring vitamin K antagonist (VKA) therapy (e.g., warfarin, acenocumarol).³ In particular, patients under VKA therapy require regular monitoring of INR to ensure this remains within a recommended therapeutic range (e.g., 2.0–3.0 for most indications); INR values below and above such ranges increase the risk of adverse events (e.g., thrombotic recurrence or bleeding, respectively).

There are two main test avenues for such regular INR monitoring: automated coagulation analyzers or point of care (POC) instruments. Each approach has advantages and disadvantages. Coagulation analyzers are now almost ubiquitous in our hospitals, performing hundreds of coagulation assays daily. These instruments use either optical or mechanical monitoring of the coagulation process to record a clotting time at reactions’ end. Automated coagulation analyzers are now rather easy to manage, very accurate and have high throughput, and INR testing can be accomplished using small volumes of citrate-anticoagulated plasma (in the range of µl), with blood collected from in-patients or out-patients, or via general practitioner (GP) clinics. However, this requires patient attendance at a collection facility, and a venous blood sample to be collected (typically 2–5 ml in volume). Also, the procedure needs time for processing (including centrifugation steps), and for tests to be performed and reported, and then for a clinician or health-care provider to relay results to the patient, with advice about potential VKA dose adjustments. The modern generation of POC devices encompasses accurate instrumentation, and although they have slower output and can typically only measure one sample at a time, they are well suited for INR testing, as they can be performed on capillary blood (i.e., needle prick), with results available in minutes, even outside the laboratory environment. Indeed, such devices can also be used directly by patients, who can then self-monitor and undertake dose adjustments more or less in real time. However, such devices, and their consumables/running costs, may be cost prohibitive in developing countries or for low-income populations, and also require definition and implementation of criteria of quality assessment. Moreover, overall costs for patient self-testing/self-management using standard POC devices seems to be higher than standard routine clinic-based care for patients receiving long-term anticoagulation.⁵

2 GETTING SMART WITH THE PT/INR

Wouldn’t it be great if we could offer PT/INR testing to those who currently have limited access to POC tests or even classical analyzer testing? This unmet need may finally be around the corner. In a recent paper published in Nature Communications, Chan et al. described a novel approach that enabled micro-mechanical blood clot testing using smartphones.⁶ The authors noted that smartphones were widely utilized, even in poor or developing countries, as well as by those with low relative incomes. The authors utilized the smartphones’ vibrator and camera to respectively mix blood (or plasma) with test reagents and start a coagulation reaction and then to monitor and detect clot formation. The ultra-low-cost method employed a small cup containing a small copper particle, to which a small volume (10 µl) of blood (or plasma) could be added, followed by addition of commercial PT reagent to activate coagulation. The clot formation causes cessation of movement in the vibrating copper particle, which is then captured by the smartphone’s camera, representing the clot time. Purpose-developed algorithms can convert this information into an INR. The device itself is small, can be 3D-printed, and the copper particles reused. The INRs generated by the device compared well to corresponding INRs from both automated analyzers and POC instruments. The authors undertook several additional evaluations looking at temperature, blood volume, use of blood versus plasma, vibration strength, camera optics, and so on, thereby demonstrating optimal conditions for use. Several makes of smartphones were also assessed. However, the study appeared to be laboratory based, and samples assessed derived from citrate anticoagulated venous blood. Thus, an interesting omission from the study was comparative assessment of capillary blood, which is more typically applied for POC testing, and presumably the target sample for the eventual application in the field. Notably, capillary blood collection is also more vulnerable to preanalytical issues such as type and profundity of the cut, as well as external conditions (disinfection, environmental temperature). As often the case with studies that self-report on costing, and as recently reviewed by Geoffrey Wool,⁷ this was probably underestimated. For example, the authors utilized commercial thromboplastin reagent, and stated “the 20 μl of activator needed for our system would cost 0.87 cents.”⁶ However, the cost was based on a price of “$52.16 for 120 ml of solvent and lyophilized thromboplastin.” Once reconstituted, the reagent would have a finite life. Thus, although wastage in the study was probably moderate, it is unlikely that the entire content of a commercial vial of reagent would be utilized by patients using the system, so accurate cost estimation should also include activator eventually discarded post-expiry. Nevertheless, the study did indeed provide a proof of concept to the idea of smartphone INR monitoring.⁶

3 DO WE EVEN NEED THE INR ANYMORE?

Some of you may be thinking, Do we even still need the INR anymore? Haven’t the DOACs (direct oral anticoagulants), without need for monitoring, supplanted the need to perform INRs?” Of interest, Chan et al. identified that “warfarin remains the most commonly prescribed outpatient blood thinner.”⁶ We suspect that this depends on geographic location. Certainly, within Australia and Italy (our geographic localities), the DOACs, in particular apixaban and rivaroxaban, are now more often prescribed than VKAs, especially for indications such as atrial fibrillation.^8-10 However, DOACs are not recommended for all indications, nor may they be the best choices for certain patients (e.g., those with impaired renal function, extremes of body weight, triple positive for antiphospholipid antibodies), and their urgent reversal (e.g., in trauma patients or in those needing urgent surgery) is more challenging than with traditional VKAs. Furthermore, DOACs are more expensive drugs than VKAs, as are the tests used for their measurement (when this becomes necessary), so the cost of DOAC implementation may be prohibitive in some geographies, particularly those locations in which the smartphone INR may be most useful. Naturally, regulatory approvals, which may differ according to geography, will also drive the differential use of DOACs versus VKAs. So, in summary, VKAs, being classical anticoagulants, are likely to be available everywhere, but DOAC availability, especially to low-income patients, may be less so.

As another potential caveat to continued use of the INR for monitoring VKA use is data that show that alternate approaches may have better utility. For example, one assay called the PT-FiiX, a new modified PT that is only sensitive to reductions in factor II or factor X, is reported to provide a better alternative to INR for monitoring of VKA therapy.^{11, 12} However, that assay is not yet widely available. Another alternative to INR monitoring is monitoring by specific factor levels, in particular factor X.^13-18 But again, that approach is not widely used.

4 WHERE TO NEXT?

Of course, having a proof of principle test capable of utilizing existing smartphone technology is just the first step in a long road to providing such technology to the masses. First, studies should be completed to evaluate the use of the technology in the field, with capillary blood, perhaps alongside the more standard POC devices. Second, perhaps some clinical studies should also be performed to associate the differentially derived INRs with outcomes, even perhaps “time in therapeutic range.” The real-word efficacy of this approach should also be tested in an appropriate population, that is, the patients themselves, who are supposed to self-collect the sample before testing without supervision of skilled personnel. The device itself will also need to be produced, perhaps in mass, or individually via 3D printing. Further validation will be required, as it is not fully known how many current smartphones are really fit for this purpose. The algorithms, or perhaps a smart-app, need to be made available in a user-friendly way. “Training” will also need to be provided for end users—perhaps YouTube or social media videos would best align to targeted users? Perhaps instead of the INR, a form of PT-FiiX can be developed for use? It also remains unknown who will be responsible for verifying the device on a particular smartphone, or whether a given ISI (international sensitivity index) and MNPT (mean normalized PT) for a given thromboplastin time reagent will be valid across a range of smartphones, or will multiple specific smartphone ISIs and MNPTs be required? We identify, for example, that the authors⁶ utilized a correction factor in their INR calculation. Whereas the reagent ISI was marked as 1.23, the ISI had to be adjusted upward with an “adjustment factor” of 0.48 to enable comparability with established lab and POC procedures; thus, the effective reagent ISI for smartphone use was 1.23 + 0.48 (or 1.71). So, “off-the-shelf” ISIs and MNPTs, generally suitable for lab-based INR testing, may not be suitable for smartphone-based INR testing. If mass produced, there will obviously be several mark-ups applied to make the process profitable for private investors, and so costs will obviously rise.

5 CONCLUSION

Contrary to the inherent meaning of the term, smartphones are no longer devices used for only making phone calls, as their usage is now widespread for hundreds (thousands?) of other reasons. Smartphones are probably the devices that we use most frequently in our day-to-day lives. Smartphones are always with us, and they hence represent the most practical tool that can be used for health monitoring, with a kaleidoscope of vital parameters that can be captured, analyzed, or even remotely transmitted, including lab tests, as predicted by one of us a decade ago.¹⁹ Their possible usage for monitoring anticoagulant therapy should therefore be seen as another important opportunity, which thus extends their usage for chronic disease monitoring (e.g., for measuring hemoglobin in patients with anemia, glycosylated hemoglobin in those with diabetes, and so forth). ^20-24 Nonetheless, some important aspects still need to be defined before they will be ready for prime time in VKAs monitoring (e.g., is self-adjustment of VKA therapy reliable using the device? How does one check the proper function of the device? What is an acceptable quality control procedure on the device? Should there be intermittent correlation with a well maintained commercial POC or laboratory-based INR to ensure ongoing reliability? What about regulatory approvals or oversight? etc.).

CONFLICTS OF INTEREST

The authors have no conflicts of interest.