OR WAIT null SECS
Retrospective analysis of clinical trial enrollment data evaluates the effectiveness of point-of-care testing in reducing later-screen failures.
In a global chronic kidney disease trial, a large pharmaceutical sponsor used point-of-care diagnostic testing (POCT) at investigator sites to pre-screen prospective subjects prior to sending samples for central lab (CL) screening in order to reduce CL screen failures. A screen failure is defined as a subject who does not get enrolled into the study for various reasons, documented during the screening process. Prior to initiating the pre-screening step in this Phase III study, the sponsor had experienced a 70% screen failure rate in the Phase II trials for the same pharmaceutical agent. The purpose of this review was to compare results from POCT and CL to determine whether POCT would have an impact on ruling out subjects who would not successfully pass the higher-cost CL screening process.
Design of review
A retrospective observational analysis of available POCT and CL results for estimated glomerular filtration rate (eGFR) and urine albumin-to-creatinine ratio (UACR) was completed after eight months of the Phase III study. Since the study protocol did not require sites to report pre-screening test results from POCT, sites were contacted directly to obtain POCT values from subject-coded records. The subject codes were then used to look up the corresponding CL values. The POCT and CL results were compared based on the screening criteria illustrated in Figure 1.
Method of pre-screening application with POCT
Pre-screening at investigator sites with POCT was introduced as part of the protocol before the CL screening at the beginning of Phase III, as shown in Figure 2 (see below). During the “pre-screening period,” each subject was asked to sign a pre-screening informed consent form. Pre-screened subjects with an eGFR < 60 mL/min/1.73 m² and a UACR ≥ 300 mg/g would then be enrolled into the “screening period.” This period consisted of two visits, including informed consent and collection of laboratory specimens to be evaluated by a commercial CL. This ongoing study currently involves 850 sites and approximately 4,000 subjects in 35 countries.
Each site participating in pre-screen testing was equipped by the sponsor with one Siemens CLINITEK® urinalysis analyzer and one Nova Biomedical StatSensor® CRE meter along with test reagents and quality control (QC) materials. For UACR testing with the Clinitek, first void samples were collected from the prospective subject. The Clinitek Microalbumin 2 Strips provide albumin, creatinine, and albumin-to-creatinine ratio results in one minute. Results were printed out from the device at the site. The eGFR was evaluated by collecting a 2-microliter whole blood sample from the prospective subject using a finger-stick method. The StatSensor CRE measures the creatinine within 30 seconds. Site staff entered subject parameters into the device to calculate the eGFR. Total test time for UACR and eGFR, with preparation, was less than 10 minutes.
Investigators and site staff were provided with didactic and hands-on workshop training on both POCT devices during planned investigator meetings and follow-up webinars. Each site that received the POCT devices was instructed to run QC upon completion of the device setups. The QC test values were entered by site along with the site number, device serial numbers, date of testing, and site demographics. All QC test values were confirmed to be within the acceptable range, according to the manufacturer’s package insert prior to subject testing. Sites were issued training certificates at the completion of the process.
Results of pre-screening application with POCT
Fifty-two records were obtained for comparison. There was 78.8% agreement in results from POCT and CL from this subject dataset. This suggests that the on-site POCT, performed at a significantly lower cost-per-test on the day of the prospective subject visit, could effectively filter out subjects who would represent a screen failure using the CL process.
After 12 months, the sponsor reported that screen failures were below 50% in the study using the pre-screening step with POCT.
Application of POCT in interventional pharmaceutical drug development clinical trials started around 1995. One of the early publications involving an encrypted POCT device used in a global pharmaceutical clinical trial documented measurement of activated clotting time during dosing of a platelet inhibitor for unstable angina.1 In this study, the heparin administration was systematically blinded and the activated partial thromboplastin time (aPTT) measurements were run from a whole blood sample at bedside using a POC microcoagulation device. The device generated a code in place of an actual aPTT value. The investigator called an IVRS to convert the subject’s coded aPTT into a heparin dose adjustment using a computerized standard nonogram to a laboratory equivalent. To reach the target therapeutic range, the POC microcoagulation device was used to check the aPTT six hours after the drug initiation and then six to 12 hours later. A POC diagnostic device that provided immediate, encrypted bedside test results was the tool essential to enabling clinicians to execute the protocol requirements of dose management within a therapeutic range.
The use of POCT in clinical trials has continued to increase due to technological advancements, as well as to the availability of a growing number of POC tests on these disposable, hand-held, or tabletop platforms. A substantial number of randomized controlled trials have shown that POCT correlates closely with CL.2 While glucose meters have been employed widely in diabetic drug trials to manage dose and monitor glycemic index, sponsors have recognized the utility of POC devices to accurately measure hemoglobin, prothrombin time, cholesterol, LDL, HDL, triglycerides, troponin, HbA1c, and brain natriuretic peptide from a small whole blood sample during a study subject visit. An increasing number of studies use mobile health devices or POCT applications in therapeutic areas ranging from asthma and cancer to schizophrenia and diabetes clinical trials. Results from a comparative mobile diabetes intervention study of 163 subjects found that adding a mobile coaching application with a POC glucose test, together with feedback on personalized analysis of blood glucose data and lifestyle behaviors via smartphones, substantially lowered glycated hemoglobin levels for more than a year.3
With the advent of continuous glucose monitoring (CGM), the potential to change the way studies are designed and data collected in real time has arrived. As test sensor technology improves, the test menu will likely expand on continuous monitoring platforms.
This report provides the first evidence from a large global study that pre-screens prospective subjects with POCT technology. While single and multi-analyte POCT devices have been used to make clinical decisions and collect data for primary outcomes previously, the use of two POCT devices to generate two diagnostic measures in an exclusion algorithm has not been found in the literature prior to this review. For infectious disease studies, screening tests during the subject visit at the investigator site can confirm presence of an infectious agent, antigen, or antibody from samples such as whole blood, serum, urine, fecal, genital, nasal, sputum, or oral swab. A confirmatory test can be run when samples arrive at a CL. The screening in this case can also be used to enroll a subject at the first visit and start treatment without delay.
Subjects can use POC devices like glucose meters at home; however, there are a limited number of tests that can be used in the US due to regulatory approvals. Since most global studies include US sites, the US regulations can thus be a default standard for the rest of the locations. Sponsors need to be aware of laboratory regulations involving POC device and In Vitro Diagnostic Directive (IVDD) designations such as home use, Clinical Laboratory Improvement Amendments (CLIA)-waived, moderately complex, and highly complex lab test devices.
In this application, the pre-screening with a discrete decision algorithm served a unique purpose: to reduce screen failures. The relative cost of performing the two pre-screen tests (UACR and eGFR) at the investigator site was lower than the sample shipment and testing by CL. And, since the time-to-results is less than 30 minutes, clinicians were empowered to complete the enrollment process on that first visit, provided the criteria were met. While the POC tests do not replace the CL tests, POC measures serve to raise the probability that the results from a subject’s CL samples will meet the inclusion criteria. Modeling the impact of pre-screening with POCT is also possible using documented sensitivity and specificity of tests along with cost comparison to the standard.
The devices used in this study are robust and easy to operate. Clinitek automatically reads a urinalysis strip and calculates the creatinine-to-microalbumin ratio. This simplified processing takes one minute to complete. Urinalysis strips are routinely read manually by comparing to a color chart; however, having the Clinitek to read, compute, display, store, and print the UACR minimizes operator interpretation and transcription error. StatSensor also made computation of eGFR simple with a built-in algorithm that used the whole blood creatinine result obtained from a small finger stick sample, along with the input of the subject’s sex, age, and race. Results are stored on the meter and can be printed or transmitted to a computer.
Training site staff on the proper use, maintenance, quality control, and troubleshooting of these devices is accomplished by hands-on demonstrations and videos. Support to the sites is available by email and phone so that questions and resupply can be addressed quickly. While there can be country-specific importation requirements, because these two devices were both 510K-cleared and CE-marked, completing customs documents could be handled in advance of the study to minimize shipment delays. The devices and reagents in this study were shipped to sites on five continents and in 35 countries, along with multilingual instruction manuals and user interfaces. Sites needing translation support were accommodated in order to resolve any training or instrument issues. Replacement instruments were held in regional depots to minimize time of shipment to within 48 hours to a site. Resupply of reagents was forecasted and planned according to the protocol, but sites could reorder by email or phone.
The takeaways from this approach are the following:
Possible improvements that could be made to this approach include the following:
For sponsors considering ways to take this approach and apply it in other types of trials, the following should be considered as means of increasing efficiency of the process:
With regards to cost savings, the total cost of pre-screening for UACR and CRE was less than the cost of the one-way shipment alone of 24-hour urine and blood samples to the CL. Depending on the number of subjects screened, the cost savings both in total cost and subject time could be significant. With the typical cost of a CL test running between $230-$675, the calculation of savings would include those subjects who were pre-screened at cost of the POC, then tested by CL screening, as well as those subjects who were pre-screened out of the study.
An example of cost savings that could be obtained by including a single pre-screen POCT in subject evaluation is shown in Table 1.
The implications of on-site testing using POC technology, pre-screening for enrollment, or monitoring for dose adjustment for future clinical trials include:
In this Phase III study, by using a novel two-test criterion as a pre-screen with POCT, a significant number of prospective subjects were ruled out of the higher-cost screening period during the first enrollment visit. The novel use of POCT for pre-screening at investigator sites provides a practical step toward shortening enrollment time and reducing the cost of more expensive CL screening tests with sample shipment by overnight couriers. Since this study did not require sites to record the pre-screen results, a statistical analysis of all the sample results was not possible. However, the potential exists to apply this methodology to address an anticipated high-screen failure rate for a study. Use of a model could predict the cost and time savings of incorporating pre-screening steps into a protocol. Further prospective studies are planned to evaluate the impact of using POCT in new drug studies to statistically determine improved efficiencies and return on investment.
Paul Savuto is President and Chief Financial Officer, Blinded Diagnostics, LLC; Jeffrey DuBois, MD, is Vice President of Medical and Scientific Affairs, Nova Biomedical Corp.; Steve Karuppan is CEO, Blinded Diagnostics
* Blinded Diagnostics wishes to thank Covance Central Labs Division, Nova Biomedical, and Dr. P. Jane Gale for their support during this clinical trial and with this review.
1. PARAGON Investigators. “International, Randomized, Controlled Trial of Lamifiban (a Platelet Glycoprotein IIb/IIIa Inhibitor), Heparin, or Both in Unstable Angina.” Circulation 97 (1998): 2386-95.
2. Andrew St John. “The Evidence to Support Point-of-care Testing.” Clin Biochem Rev 31 (2010): 111-119.
3. Niklas Morton and David Blackman. “The Growing Availability of Wearable Devices: A Perspective on Current Applications in Clinical Trials.” Appl Clin Trials, May 27, 2016.
1. “Global Point-of-care Diagnostics Market Worth $32.7 Billion by 2022.” medGadget, February 23, 2016.
2. Bienkowski, Robert and Norman M. Goldfarb. “Screen Failures in Clinical Trials: Financial Roulette or Cost of Doing Business?” Journal of Clinical Research Best Practices 4 (7), July 2008.
3. Lalkhen, Abdul Ghaaliq and Anthony McCluskey. “Clinical tests: Sensitivity and Specificity, Continuing Education in Anaesthesia.” Critical Care & Pain 8 (2008): 221-223.
4. Kost, Gerald J. Principles and Practices of Point-of-care Testing (Philadelphia: Lippincott, Williams, and Wilkins, 2002), 270-271.
5. Hertzberg, Vicki, Marc Chimowitz, Michael Lynn, Cristen Chester, William Asbury, and George Cotsonis. “Use of Dose Modification Schedules is Effective for Blinding Trials of Warfarin: Evidence from the WASID Study.” Clin Trials 5 (2008): 23-30.
6. Albers, Gregory W. and Hans-Christoph Diener, et al. “Ximelagatran vs Warfarin for Stroke Prevention in Subjects with Nonvalvular Atrial Fibrillation: a randomized trial.” JAMA 293 (2005): 690-98.
7. Rao, Sunil V., Chiara Melloni, et al. “Evaluation of a New Heparin Agent in Percutaneous Coronary Intervention: Results of the Phase 2 Evaluation of M118 in Percutaneous Coronary Intervention (EMINENCE) Trial.” Circulation 121 (2010): 1713-21.
8. Lincoff, A. Michael, John A. Bittl, et al. “Bivalirudin and Provisional Glycoprotein IIb/IIIa Blockade Compared with Heparin and Planned Glycoprotein IIb/IIIa Blockade during Percutaneous Coronary Intervention REPLACE-2 Randomized Trial.” JAMA 289 (2003): 853-63.
9. Francis, Charles W., Scott D. Berkowitz, et al. “Comparison of Ximelagatran with Warfarin for the Prevention of Venous Thromboembolism after Total Knee Replacement.” New Eng J Med 349 (2003): 1703-12.
10. Schulman, Sam, Clive Kearon, et al. “Dabigatran versus Warfarin in the Treatment of Acute Venous Embolism.” New Eng J Med 361 (2009): 2342-52.
11. Granger, Christopher B., John H. Alexander, et al. “Apixaban versus Warfarin in Subjects with Atrial Fibrillation.” New Eng J Med 365 (2011): 981-92.
12. Solomon, Scott D., Hajime Uno, et al. “Erythropoietic Response and Outcomes in Kidney Disease and Type 2 Diabetes.” New Eng J Med 363 (2010): 1146-55.
13. Patel, Manesh R., Kenneth W. Mahaffey, et al. “Rivaroxaban versus Warfarin in Nonvalvular Atrial Fibrillation.” New Eng J Med 365 (2011): 883-91.
14. Bang, Heejung and Jongbae J. Park. “Blinding in Clinical Trials: A Practical Approach.” Journal of Alternative and Complementary Medicine 19 (2013): 367-369.
15. Bang, Heejung, Stephen P. Flaherty, Jafar Kolahi, and Jongbae Park. “Blinding Assessment in Clinical Trials: A Review of Statistical Methods and a Proposal of Blinding Assessment Protocol.” Clin Res Reg Affairs 27 (2010): 42-51.
16. Ridker, Paul M., Samuel Z. Goldhaber, et al. “Long Term, Low Intensity Warfarin Therapy for the Prevention of Recurrent Venous Thromboembolism.” New Eng J Med 348 (2003): 1425-34.
17. Lopes, Renato D., John H. Alexander, et al. “Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation (ARISTOTLE) trial: Design and Rationale.” Am Heart J 159 (2010): 331-339.
18. Patti, Giuseppe, Annunziata Nusca, Fabio Mangiacapra, Laura Gatto, Andrea D’Ambrosio, Germano Di Sciascio. “Point-of-care Measurement of Clopidogrel Responsiveness Predicts Clinical Outcome in Subjects Undergoing Percutaneous Coronary Intervention: Results of the ARMYDA-PRO (Antiplatelet Therapy for Reduction of Myocardial Damage during Angioplasty-Platelet Reactivity Predicts Outcome) Study.” J Am Coll Cardiol 52 (2008): 1128–33.
19. Wiviott, Stephen D., Dietmar Trenk, et al. “Prasugrel Compared with High Loading- and Maintenance-Dose Clopidogrel in Subjects with Planned Percutaneous Coronary Intervention: The Prasugrel in Comparison to Clopidogrel for Inhibition of Platelet Activation and Aggregation-Thrombolysis in Myocardial Infarction 44 Trial.” Circulation 116 (2007): 2923-32.
20. Becker, Richard C., Scott D. Berkowitz, et al. “Rivaroxaban-Once Daily, Oral, Direct Factor Xa Inhibition Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation: Rationale and Design of the ROCKET AF Study.” Am Heart J 159 (2010): 340-47.
21. Ruff, Christian T., Robert P. Giugliano, Elliott M. Antman, Sharon E. Crugnale, Tomas Bocanegra, Michele Mercuri, James Hanyok, Indravadan Patel, Minggao Shi, Dan Salazar, Carolyn H. McCabe, and Eugene Braunwald. “Evaluation of the Novel Factor Xa Inhibitor Edoxaban Compared with Warfarin in Subjects with Atrial Fibrillation: Design and Rationale for the Effective Anticoagulation with Factor Xa Next Generation in Atrial Fibrillation-Thrombolysis In Myocardial Infarction Study 48 (ENGAGE AF-TIMI 48).” Am Heart J 160 (2010): 635-41.
22. Grignolo, Alberto and Sy Pretorius. “Strategies to Reduce Phase III Failure Rates.” White Paper in Appl Clin Trials, August 6, 2015.