Context‐dependent accuracy of the cobas plasma separation card for HCV RNA viral load measurement

Collection and preservation of plasma are challenging in remote or under‐resourced settings. The cobas® Plasma Separation Card (PSC) is an alternative specimen type for blood‐borne pathogen nucleic acid quantitation. We assessed PSC as a specimen type for HCV RNA quantitation in Pakistan. Plasma from venous blood and PSC from finger prick blood were prepared at two sites: Site 1 (in Lahore, n = 199) consisted of laboratory‐based outpatient clinics. Specimens were prepared in the same facility and stored frozen. Site 2 was a catchment area within a resource‐limited, semi‐urban locality of Islamabad with limited access to healthcare services (n = 151). Community public health outreach staff collected blood and prepared the PSC in the participants' homes. Specimens were transported to the central hepatitis laboratory in Lahore to be stored frozen until tested. HCV RNA testing was performed using the cobas HCV RNA test in a central laboratory. Concordance with respect to RNA detectability was high at Site 1 (97.4%), but lower at Site 2 (82.4%). At Site 1, HCV viral load in plasma and PSC were well correlated across the linear range with a 0.21 log10 IU/mL mean bias toward higher concentrations in PSC. At Site 2, HCV viral load in plasma and PSC were poorly correlated. There was a 0.11 log10 IU/mL mean bias toward higher concentrations in PSC. PSC performance can be excellent in underserved settings where refrigerated transport of traditional specimens is difficult. In very challenging field settings, extra support must be provided to ensure correct specimen collection and handling.


| INTRODUC TI ON
Approximately 58 million individuals were infected with hepatitis C virus (HCV) worldwide in 2019, causing chronic liver disease and an estimated 290,000 deaths per year. 1 The World Health Organization has set an ambitious goal to eliminate viral hepatitis as a public health threat by 2030.However, in 2019, only 21% of people with chronic HCV infection had been diagnosed. 1Simplification of service delivery and care pathways to overcome HCV testing access barriers is needed to identify HCV-infected people and engage them in treatment and care, 1 especially in rural, remote areas of the world.
HCV infection screening is usually performed with anti-HCV antibody tests, followed by qualitative tests for HCV RNA.The preferred specimen for HCV RNA testing is plasma, which requires phlebotomy, centrifugation, and often storage and shipping of frozen plasma.These requirements may not be feasible in some contexts due to a shortage of trained phlebotomists, laboratory infrastructure, or refrigeration equipment.[6] Approximately 60 million people living in Pakistan are infected with HCV, almost 10 million of which are viremic and in need of diagnosis and treatment.Every year there are approximately 130,000 HCV-related deaths, 500,000 cases of liver cirrhosis, and 100,000 cases of liver cancer. 7 explored the utility, acceptability, and ease of use of the PSC as a specimen type to support HCV RNA testing in Pakistan and compared viral load (VL) results from PSC to those from paired plasma specimens using the cobas HCV RNA assay.Blood was collected by venipuncture using plasma preparation tubes (PPT).From the other hand, 140 μL of whole blood was collected by finger stick in each of three capillary tubes and transferred onto the spots marked on the PSC.Drying time was about 10-15 min.Plasma was separated and refrigerated for 2-3 days until transfer to the Punjab Hepatitis Central Laboratory in Lahore where it was frozen until testing (after about 8-12 days).It was fall during the specimen collection in Lahore (daytime ambient temperatures 20-30°C).

| Specimen collection and study design
Site 2 is a catchment area within an underserved, resourcelimited, semi-urban part of Islamabad with limited access to healthcare services and lacking basic infrastructure (i.e., no electricity, water, and limited furniture).HCV antibody testing was performed by community public health outreach staff in individual residences; blood specimens were collected; and PSC was prepared as in Site 1 but on site from those found to be anti-HCV antibody positive.Two technicians were involved, both trained phlebotomists with 10 years of experience.PPT were kept cool and transported on the same day to the federal government dispensary's lab in Islamabad where plasma was separated and frozen for 2-3 days.PSC were transported in the same way and stored at ambient temperature.It was summer during the specimen collection in Islamabad (daytime ambient temperatures 35-40°C).Plasma and PSC were then transported to the central hepatitis laboratory in Lahore where they were stored (plasma at −20°C, PSC at ambient temperature) until testing (after about 12-15 days).

| Laboratory testing
HCV RNA testing was performed using the cobas HCV RNA test (Roche Molecular Systems, Inc., Pleasanton, CA) at the Punjab Hepatitis Central Laboratory in Lahore, on a single cobas 6800 instrument.The two VL tests were performed using 0.5 mL of plasma and one of the three spots from the PSC simultaneously, according to the manufacturer's instructions.
Statistical analysis was performed using SAS v9.4 (The SAS Institute, Cary, NC).A correction factor of 44-fold was applied to PSC results before comparative analysis. 4

| Study participants
Approximately 5500 individuals were tested for anti-HCV antibodies: 2350 at Site 1 and 3150 at Site 2. A total of 350 participants meeting the eligibility criteria were identified, 199 at Site 1 and 151 at Site 2. The proportions of participants that were female, aged over 50 years, or treated were lower in Site 1 compared to Site 2 (Table S1).The majority of participants at both sites were illiterate.

| HCV VL in plasma and PSC
We used three categories for concordance analysis of HCV RNA results between plasma and PSC at each site: undetectable (target not detected, or TND), detectable but below the lower limit of quantitation for PSC (866 copies/mL), 4 and detectable and within the linear range.Overall, concordance was high at Site 1 (97.4%), with only four participants being mis-classified; three of the four had lower HCV RNA levels in PSC versus plasma (Table S2).Overall concordance was lower at Site 2 (82.4%), with 35 participants being mis-classified; 26 had higher HCV RNA levels in PSC versus plasma (Table S2).
At Site 1, HCV VL in plasma and PSC were well correlated across the linear range (Pearson correlation coefficient: 0.85; Figure 1A).
At Site 2, VL in plasma and PSC were poorly correlated across the linear range (Pearson correlation coefficient: 0.44; Figure 1C).Overall, there was a small bias toward higher concentrations in PSC (mean bias 0.11 log 10 IU/mL; 95% CI −1.7 to 1.9 log 10 IU/mL; Figure 1D).
At the lower end of the linear range (below approximately 4.4 log 10 IU/mL), VL measurements from PSC were higher than from plasma, while the opposite pattern was evident at the higher end of the linear range (Figure 1C).

| DISCUSS ION
Refrigerated transport of biological specimens in underserved or remote areas faces numerous hurdles including inadequate cold storage and limited transportation options.Point-of-care devices can partially address these challenges but struggle to match performance and volume capabilities of central laboratories.
The PSC is an alternative specimen collection method that eliminates the need for a trained phlebotomist and reliable cold chain.

F I G U R E 1 Deming regression (A and C
) and Bland-Altman (B and D) for Lahore (A and B) and Islamabad (C and D) sites.Points colored in red are below the LLOQ but above the LOD for both assays; these points are excluded from the statistical analysis but shown here for illustrative purposes.
We collected specimens from consenting individuals over 18 years of age with detectable anti-HCV antibodies (Abbott SD Bioline rapid test) in two settings (sites).Pregnant and lactating mothers, patients with chronic renal failure on dialysis, and migrants who had temporarily settled in the areas were excluded.Ethical clearance was obtained from Pakistan National Bioethics Committee Ref: No.4-87/ NBC-678/21/452.Site 1 consisted of three community health center outpatient clinics near Lahore (Shahkot, Sangla Hills and Nankana Sahib) with electricity, running water, appropriate furniture, and other basic infrastructure including on-site laboratory facilities.These clinics serve patients in various categories including untreated, anti-HCV antibody-positive individuals and patients undergoing anti-HCV treatment being monitored with HCV RNA testing.One technician was involved in specimen collection at each clinic.