The Economics of NGS:
The Economics of NGS:
A Cost Comparison of NGS implementations with legacy technologies
Presenter: Dr. Peter Meintjes
September, 2016
Potential Conflict of Interest Declaration §
I work for Omixon, the commercial provider of NGS-based HLA typing product, Holotype HLA.
Overview §
Introduction and Background
§
NGS Overview
§
Small/Large lab modeling (24/96 samples/week)
§
Results
§
Summary and Conclusions
§
Questions
Introduction to NGS for HLA “Recent technological advances, cumulatively called NGS, provide the first opportunity to fully characterize and fully phase the HLA genes” §
The Technique – Long Range PCR Amplification – Clonal sequencing of Amplicons – Assembly of whole gene, nucleotide resolution sequences
§
The Goal – Automated, high-throughput analysis – Unambiguous results – No reflexive testing
NGS workflow Sample Prep (Pre-PCR)
Library Prep (Post-PCR)
Sequencing
Analysis
Scientific & Technical Benefits of NGS §
Complete resolution – nucleotide-level, 3-field or higher
§
Increased characterization/coverage – whole gene consensus sequences of all exons and introns
§
Allele phasing – Unambiguous typing
§
High-throughput batching of samples
Small/Large (24/96) lab modeling §
Small Lab – 24 samples/week (~1,250 samples/year) – 2x 12 samples/run
§
Medium Lab – 48 samples/week (~2,500 samples/year) – 2x 24 samples/run
§
Large Lab – 96 samples/week (~5,000 samples/year) – 2x 48 samples/run
§
Lab + small registry – 192 samples/week (~10,000 samples/year)
Capital Equipment Investment Instrument
Cost*
Illumina MiSeq
$100,000
Illumina MiSeq Support (Years 2 & 3)
$34,000
Size selection equipment (Optional)
$10,000
Plate fluorometer (Optional)
$20,000
qPCR machine (Optional)
$30,000
5x PCR machine
$35,000
64-bit computer with 16+ GB RAM
$3,000
1-2x Liquid handler (Optional, pre & post PCR)
$50,000+ (each)
*Based on figures from four US Labs, rounded to $1,000
The Importance of Pooling §
Traditional cost analysis are expressed as “$/locus” – NGS (specifically per-sample indexing or locus “pooling”) challenges this approach
§
Why? – Cost of enzymes and labor used prior to pooling scales per locus – Cost of enzymes and labor used after locus pooling scales per sample
Step
Scale mechanism
Manufacturing/QC
Vendor Specific
Pre-pooling (e.g. Amp’ing, tagmentation)
Increases per locus
Post-pooling (e.g. library prep)
Increases per sample
Sequencing
Decreases per sample
Labor
Per technician
NGS Cost/sample (relative)
§
Sequencing costs become a smaller proportion of cost/sample as you increase the number of samples
§
Labor increases ‘per locus’ prior to pooling
Cost Savings for Reflexive Testing Technique
Cost of Initial Typing
Cost of Ambiguity Reflex
Cost of Post-transplant Reflex
SSO
Low
Medium
High
SBT
High
High
High
NGS
Medium
None
None
§
Almost zero ambiguities means no reflexive pre-transplant tests
§
High resolution typing of solid organ donors also eliminates reflexing post-transplant when antibodies are observed
Cost/locus for NGS (Holotype)
§
Labor costs associated with SBT are very high
§
96 samples/run have lower sequencing cost/sample than 24 samples/run
§
Diminishing extra costs of additional samples and loci (if single technician assumption is met)
Hands on time per locus (Holotype)
§
SSO and NGS both have batching capabilities that drastically alter the hands on time per locus and per sample
Results/Benefits §
Time savings – Hands on time is lower – No high-res reflexing when antibodies detected – Interpretation times are near zero
§
Convenience – Scheduling is easier, batching is useful
§
Others – Capital equipment is relatively easily financed, or financed through the vendor (reagent rental) – Empowers labs to keep everything in house
Summary and Conclusions §
NGS is cost competitive per sample with SSO at 24 samples for 11 loci, and costs are further reduced with increasing number of samples
§
Earlier ‘locus pooling’ drives economic efficiencies for cost of goods and labor costs
§
28 hour turnaround from DNA to unambiguous typing means NGS is suitable for most SOT and all BMT workflows – Current NGS methods are not suitable for deceased donor typing
§
HLA Labs of the future will only need to support two workflows, deceased donor by preferred methods and everything else by NGS – – – –
Reduction of workflows Reduced inventories Less training Less validation
Questions?
Peter Meintjes Chief Commercial Officer [email protected]
A Cost Comparison of NGS implementations with legacy technologies
Presenter: Dr. Peter Meintjes
September, 2016
Potential Conflict of Interest Declaration §
I work for Omixon, the commercial provider of NGS-based HLA typing product, Holotype HLA.
Overview §
Introduction and Background
§
NGS Overview
§
Small/Large lab modeling (24/96 samples/week)
§
Results
§
Summary and Conclusions
§
Questions
Introduction to NGS for HLA “Recent technological advances, cumulatively called NGS, provide the first opportunity to fully characterize and fully phase the HLA genes” §
The Technique – Long Range PCR Amplification – Clonal sequencing of Amplicons – Assembly of whole gene, nucleotide resolution sequences
§
The Goal – Automated, high-throughput analysis – Unambiguous results – No reflexive testing
NGS workflow Sample Prep (Pre-PCR)
Library Prep (Post-PCR)
Sequencing
Analysis
Scientific & Technical Benefits of NGS §
Complete resolution – nucleotide-level, 3-field or higher
§
Increased characterization/coverage – whole gene consensus sequences of all exons and introns
§
Allele phasing – Unambiguous typing
§
High-throughput batching of samples
Small/Large (24/96) lab modeling §
Small Lab – 24 samples/week (~1,250 samples/year) – 2x 12 samples/run
§
Medium Lab – 48 samples/week (~2,500 samples/year) – 2x 24 samples/run
§
Large Lab – 96 samples/week (~5,000 samples/year) – 2x 48 samples/run
§
Lab + small registry – 192 samples/week (~10,000 samples/year)
Capital Equipment Investment Instrument
Cost*
Illumina MiSeq
$100,000
Illumina MiSeq Support (Years 2 & 3)
$34,000
Size selection equipment (Optional)
$10,000
Plate fluorometer (Optional)
$20,000
qPCR machine (Optional)
$30,000
5x PCR machine
$35,000
64-bit computer with 16+ GB RAM
$3,000
1-2x Liquid handler (Optional, pre & post PCR)
$50,000+ (each)
*Based on figures from four US Labs, rounded to $1,000
The Importance of Pooling §
Traditional cost analysis are expressed as “$/locus” – NGS (specifically per-sample indexing or locus “pooling”) challenges this approach
§
Why? – Cost of enzymes and labor used prior to pooling scales per locus – Cost of enzymes and labor used after locus pooling scales per sample
Step
Scale mechanism
Manufacturing/QC
Vendor Specific
Pre-pooling (e.g. Amp’ing, tagmentation)
Increases per locus
Post-pooling (e.g. library prep)
Increases per sample
Sequencing
Decreases per sample
Labor
Per technician
NGS Cost/sample (relative)
§
Sequencing costs become a smaller proportion of cost/sample as you increase the number of samples
§
Labor increases ‘per locus’ prior to pooling
Cost Savings for Reflexive Testing Technique
Cost of Initial Typing
Cost of Ambiguity Reflex
Cost of Post-transplant Reflex
SSO
Low
Medium
High
SBT
High
High
High
NGS
Medium
None
None
§
Almost zero ambiguities means no reflexive pre-transplant tests
§
High resolution typing of solid organ donors also eliminates reflexing post-transplant when antibodies are observed
Cost/locus for NGS (Holotype)
§
Labor costs associated with SBT are very high
§
96 samples/run have lower sequencing cost/sample than 24 samples/run
§
Diminishing extra costs of additional samples and loci (if single technician assumption is met)
Hands on time per locus (Holotype)
§
SSO and NGS both have batching capabilities that drastically alter the hands on time per locus and per sample
Results/Benefits §
Time savings – Hands on time is lower – No high-res reflexing when antibodies detected – Interpretation times are near zero
§
Convenience – Scheduling is easier, batching is useful
§
Others – Capital equipment is relatively easily financed, or financed through the vendor (reagent rental) – Empowers labs to keep everything in house
Summary and Conclusions §
NGS is cost competitive per sample with SSO at 24 samples for 11 loci, and costs are further reduced with increasing number of samples
§
Earlier ‘locus pooling’ drives economic efficiencies for cost of goods and labor costs
§
28 hour turnaround from DNA to unambiguous typing means NGS is suitable for most SOT and all BMT workflows – Current NGS methods are not suitable for deceased donor typing
§
HLA Labs of the future will only need to support two workflows, deceased donor by preferred methods and everything else by NGS – – – –
Reduction of workflows Reduced inventories Less training Less validation
Questions?
Peter Meintjes Chief Commercial Officer [email protected]
