CLINICAL PERFORMANCE OF A RAPID SEPSIS TEST ON A NEAR-PATIENT MOLECULAR TESTING … · CLINICAL...
2
CLINICAL PERFORMANCE OF A RAPID SEPSIS TEST ON A NEAR-PATIENT MOLECULAR TESTING PLATFORM James Kirk 1 , Thomas Yager 1 , Silvia Cermelli 1 , Dayle Sampson 1 , Richard Brandon 1 , Peter Sillekens 2 , Inge Keuleers 2 , Thierry Vanhoey 2 , Russ R. Miller III 3 , Bert K. Lopansri 4 , Olaf Cremer 5 , Marc Bonten 5 Immunexpress, Seattle, WA, USA 1 / Biocartis NV., Belgium 2 / FirstHealth of the Carolinas, Pinehurst, NC, USA 3 Intermountain Medical Center, Murray, UT, USA 4 / University Medical Center Utrecht, Utrecht, Netherlands 5 The purpose of this study was to clinically validate a new, rapid version of the SeptiCyte ® assay on a near-patient testing platform (Biocartis’ Idylla™ platform*) using samples collected as part of previously established retrospective studies [1,2]. SeptiCyte ® LAB can differentiate between sepsis and systemic inflammatory response syndrome (SIRS) in critically ill patients. SeptiCyte ® LAB is the first- in-class sepsis diagnostic to gain FDA-clearance but currently has a complex workflow and turnaround time (TaT) of ~ 6 hours [3]. The assay in Idylla™ cartridge format is called SeptiCyte ® RAPID with a TaT of ~ 1 hour. INTRODUCTION SeptiCyte ® LAB was translated to the Biocartis Idylla™ near-patient testing platform and analytically validated. Patients (N=195) were recruited from previously established cohorts in two Intensive Care Units (ICU) located in the USA and Europe. For this validation study, 0.9mL of peripheral blood (collected in PAXgene TM tubes) was pipetted directly into the cartridge and inserted into the Idylla™ instrument. SeptiCyte ® RAPID results were reported as a SeptiScore ® between 0 and 15 with higher scores representing higher probability of sepsis. Assay performance metrics included technician hands-on-time (HoT), assay TaT, repeatability, reproducibility and Area Under ROC Curve (AUC). The comparator was retrospective physician diagnosis (RPD) using a panel of three independent physicians. METHODS For 195 cartridges run, the average technician HoT was 2 minutes, average TaT was 65 minutes, and coefficient of variation for a positive extraction control was 1.19% demonstrating repeatability. Correlation of SeptiScore ® values between SeptiCyte ® LAB and SeptiCyte ® RAPID, based upon a subset of N= 100 samples run on both platforms, was very high (r2>0.94) (Figure 1). Figure 2A shows the SeptiCyte ® RAPID ROC curve for differentiating sepsis and SIRS in the sample set (N=195) and a high AUC of 0.904. Figure 2B shows significant differentiation of sepsis and SIRS patients in box and whisker plot format with SeptiScore ® on the Y-axis and RPD on the X-axis. Figure 3 shows test TaT comparison between SeptiCyte ® LAB (FDA cleared, kit format) and SeptiCyte ® RAPID cartridges on the Idylla™ platform. Preliminary precision studies conducted with three sample types (Low, Medium, High SeptiScore ® pools), two lots of cartridges, three instruments and and across six days confirmed high reproducibility of the test (Figure 4). SeptiCyte ® RAPID cartridges (Figure 5) are fully integrated and include sample nucleic acid extraction, RT-qPCR and reporting in ~ 1 hour on the Idylla™ platform. RESULTS Figure 1: SeptiCyte ® LAB vs RAPID comparison SeptiCyte ® LAB SeptiCyte ® RAPID r 2 = 0.94 SeptiScore ® Figure 2 - SeptiCyte ® RAPID performance B A False positive fraction True positive fraction
Transcript of CLINICAL PERFORMANCE OF A RAPID SEPSIS TEST ON A NEAR-PATIENT MOLECULAR TESTING … · CLINICAL...
CLINICAL PERFORMANCE OF A RAPID SEPSIS TEST ON A NEAR-PATIENT MOLECULAR TESTING PLATFORM
James Kirk1, Thomas Yager1, Silvia Cermelli1, Dayle Sampson1, Richard Brandon1, Peter Sillekens2, Inge Keuleers2, Thierry Vanhoey2, Russ R. Miller III3, Bert K. Lopansri4, Olaf Cremer5, Marc Bonten5
Immunexpress, Seattle, WA, USA1 / Biocartis NV., Belgium2 / FirstHealth of the Carolinas, Pinehurst, NC, USA3
Intermountain Medical Center, Murray, UT, USA4 / University Medical Center Utrecht, Utrecht, Netherlands5
The purpose of this study was to clinically validate a new, rapid version of the SeptiCyte® assay on a near-patient testing platform (Biocartis’ Idylla™ platform*) using samples collected as part of previously established retrospective studies [1,2]. SeptiCyte® LAB can differentiate between sepsis and systemic inflammatory response syndrome (SIRS) in critically ill patients. SeptiCyte® LAB is the first-in-class sepsis diagnostic to gain FDA-clearance but currently has a complex workflow and turnaround time (TaT) of ~ 6 hours [3]. The assay in Idylla™ cartridge format is called SeptiCyte® RAPID with a TaT of ~ 1 hour.
INTRODUCTION
SeptiCyte® LAB was translated to the Biocartis Idylla™ near-patient testing platform and analytically validated. Patients (N=195) were recruited from previously established cohorts in two Intensive Care Units (ICU) located in the USA and Europe. For this validation study, 0.9mL of peripheral blood (collected in PAXgeneTM tubes) was pipetted directly into the cartridge and inserted into the Idylla™ instrument. SeptiCyte® RAPID results were reported as a SeptiScore® between 0 and 15 with higher scores representing higher probability of sepsis. Assay performance metrics included technician hands-on-time (HoT), assay TaT, repeatability, reproducibility and Area Under ROC Curve (AUC). The comparator was retrospective physician diagnosis (RPD) using a panel of three independent physicians.
METHODS
For 195 cartridges run, the average technician HoT was 2 minutes, average TaT was 65 minutes, and coefficient of variation for a positive extraction control was 1.19% demonstrating repeatability. Correlation of SeptiScore® values between SeptiCyte® LAB and SeptiCyte® RAPID, based upon a subset of N= 100 samples run on both platforms, was very high (r2>0.94) (Figure 1). Figure 2A shows the SeptiCyte® RAPID ROC curve for differentiating sepsis and SIRS in the sample set (N=195) and a high AUC of 0.904. Figure 2B shows significant differentiation of sepsis and SIRS patients in box and whisker plot format with SeptiScore® on the Y-axis and RPD on the X-axis. Figure 3 shows test TaT comparison between SeptiCyte® LAB (FDA cleared, kit format) and SeptiCyte® RAPID cartridges on the Idylla™ platform. Preliminary precision studies conducted with three sample types (Low, Medium, High SeptiScore® pools), two lots of cartridges, three instruments and and across six days confirmed high reproducibility of the test (Figure 4). SeptiCyte® RAPID cartridges (Figure 5) are fully integrated and include sample nucleic acid extraction, RT-qPCR and reporting in ~ 1 hour on the Idylla™ platform.
RESULTS
Figure 1: SeptiCyte® LAB vs RAPID comparison
8
r2 = 0.94
SeptiCyte® LAB
Sep
tiC
yte®
RA
PID
SeptiScore®
Figure 1: SeptiCyte® LAB vs RAPID comparisonSeptiCyte® RAPID demonstrates high correlation (r2 = 0.94) to the SeptiCyte® LAB (manual kit) for a cohort of clinical ICU patients (N = 100). In the case of SeptiCyte®
LAB, all clinical PAXgene™ blood samples were extracted using the PAXgene™ Blood RNA Kit, IVD version (Qiagen Cat No./ID: 762164).
8
r2 = 0.94
SeptiCyte® LAB
Sep
tiC
yte®
RA
PID
SeptiScore®
Figure 1: SeptiCyte® LAB vs RAPID comparisonSeptiCyte® RAPID demonstrates high correlation (r2 = 0.94) to the SeptiCyte® LAB (manual kit) for a cohort of clinical ICU patients (N = 100). In the case of SeptiCyte®
LAB, all clinical PAXgene™ blood samples were extracted using the PAXgene™ Blood RNA Kit, IVD version (Qiagen Cat No./ID: 762164).
Figure 2 - SeptiCyte® RAPID performance
9
Figure 2: SeptiCyte® RAPID performance(A): Receiver operating characteristic curve for SeptiScore®, calculated for the clinical dataset (N = 195). (B): SeptiCyte® RAPID strongly discriminates sepsis (N=106) from SIRS (N=89) cases.
A B
9
Figure 2: SeptiCyte® RAPID performance(A): Receiver operating characteristic curve for SeptiScore®, calculated for the clinical dataset (N = 195). (B): SeptiCyte® RAPID strongly discriminates sepsis (N=106) from SIRS (N=89) cases.
A B
9
Figure 2: SeptiCyte® RAPID performance(A): Receiver operating characteristic curve for SeptiScore®, calculated for the clinical dataset (N = 195). (B): SeptiCyte® RAPID strongly discriminates sepsis (N=106) from SIRS (N=89) cases.
A B
9
Figure 2: SeptiCyte® RAPID performance(A): Receiver operating characteristic curve for SeptiScore®, calculated for the clinical dataset (N = 195). (B): SeptiCyte® RAPID strongly discriminates sepsis (N=106) from SIRS (N=89) cases.
A B
False positive fraction
Tru
e p
osi
tive
fra
ctio
n
Poste ID: XXXX
Figure 1SeptiCyte® RAPID demonstrates high correlation (r2 = 0.94) to
the SeptiCyte® LAB (manual kit) for a cohort of clinical ICU patients (N = 100). In the case of SeptiCyte® LAB, all clinical PAXgene™ blood samples were extracted using the PAXgene™ Blood RNA Kit, IVD version (Qiagen Cat No./ID: 762164).
Figure 2(A): Receiver operating characteristic curve for SeptiScore®,
calculated for the clinical dataset (N = 195). (B): SeptiCyte® RAPID strongly discriminates sepsis (N=106) from SIRS (N=89) cases.
Figure 4Test reproducibility was determined with three sample types
(Low [gray], Medium [white], High [red] score pools), two lots of cartridges, and three Idylla™ instruments (INS). The study (N = 54) was conducted over six days (D), with three samples per instrument, per sample type, per day. Operator-to-operator variability was expected to be minimal and was not evaluated.
The reproducibility of the SeptiCyte® RAPID test demonstrates minimal variation (SD < 0.3) for each sample pool when considering all sources of variation.
References
[1] MARS (Molecular Diagnosis and Risk Stratification of Sepsis) www.clinicaltrials.gov identifier NCT01905033
[2] VENUS (Validation of Septic Gene Expression Using SeptiCyte®) www.clinicaltrials.gov identifier NCT02127502
[3] Miller RR III, Lopansri BK, Burke JP, et al. “Validation of a Host Response Assay, SeptiCyte® LAB, for Discriminating Sepsis from Systemic Inflammatory Response Syndrome in the ICU.” Am J Respir Crit Care Med. 2018 Oct 1;198(7):903-913.
Figure 4 - SeptiCyte®
RAPID reproducibility
11
Lot 1
INS 1 INS 2 INS 3
D1D3D5
D1D3D5
D1D3D5
D2D4D6
D2D4D6
D2D4D6Lot 2
Lot 1 Lot 2
10.51 +/- 0.20
6.69 +/- 0.13
3.59 +/- 0.20
10.45 +/- 0.20
6.73 +/- 0.21
3.65 +/- 0.16
Sep
tiSc
ore
®
0
5
10
15
Sep
tiSc
ore
®
0
5
10
15
Figure 4: SeptiCyte® RAPID reproducibility Test reproducibility was determined with three sample types (Low [gray], Medium [white], High [red] score pools), two lots of cartridges, and three Idylla™ instruments (INS). The study (N = 54) was conducted over six days (D). Operator-to-operator variability was expected to be minimal and was not evaluated.
The reproducibility of the SeptiCyte® RAPID test demonstrates minimal variation (SD < 0.3) for each sample pool when considering all sources of variation.
Figure 5 - Biocartis Idylla™ console, instrument and a SeptiCyte® RAPID cartridge Figure 5: Biocartis Idylla™ console, instrument and a SeptiCyte® RAPID cartridge
12
instrument
console
SeptiCyte® RAPID cartridge
CONCLUSIONS
• This is the first demonstration of a validated, fully-integrated, rapid, repeatable, reproducible, near-patient, immune-response sepsis diagnostic, providing accurate results in ~ 1 hr to differentiate sepsis from SIRS.
• It is anticipated to improve patient management for critically ill patients to be admitted to ICU.
• Further validation of the performance of SeptiCyte® RAPID is underway in real-time studies across multiple US sites.
Immunexpress uses the Idylla trademark under license from Biocartis NV.
Figure 3 - SeptiCyte® LAB and RAPID test turnaround time comparison
