esp lung eqa schemes - olv z
TRANSCRIPT
obo Prof. Dr. Els Dequeker
Biomedical Quality Assurance Research Unit
Department of Public Health & Primary Care
KU Leuven, Leuven, Belgium
Pathology symposium
OLV Ziekenhuis Aalst
10 December 2019
ESP Lung EQA schemesPrograms for quality assurance of IHC
and molecular tests in non-small cell lung cancer
Cleo Keppens
Domain:
• DNA analysis
• RNA analysis
• IHC (+ digital cases)
• FISH (+ digital cases)
Genes: ALK, ROS1, PD-L1,
EGFR, KRAS, BRAF, c-MET
≈ 230 labs, 37 countries
- Steering comittee with medical and technical experts
- Validating labs, experts and assessors
- Recognized EQA coordination center
http://lung.eqascheme.org
http://www.esp-pathology.org
Accredited as EQA provider by BELAC in
accordance with standard NBN EN ISO/IEC
17043:2010 (PT-215)
ESP Lung EQA Scheme: organized since 2012
2
EMA, European Medicines Agency;
FDA, US Food and Drug Administration.
2014: New
biomarker ROS1
included
2017: New joint pilot
EQA scheme for
ctDNA EGFR/KRAS
2012:
pilot round
2015: Technical
assessment for
ALK IHC
2017: New scheme set-up:
3 rounds
New biomarkers:
PD-L1,
KRAS (mol), BRAF (mol)
Expansion of scheme according to current needs
2018: New
biomarker:
genomic cMET
20132012 2014 2015 2016 2017 2018 2019
2016: Technical
assessment for
ROS1 IHC
2020
2019: RNA
analysis of
cMET
2020: RNA analysis
ALK and ROS1 in
combination with MET
2018: full joint
cfDNA scheme for
EGFR
2012: EMA
approves
Xalkori
2015: FDA
approves
osimertinib
2016: EMA
approves
osimertinib
2013: EMA/FDA
approve afatinib
2013: EGFR
pilot included
2016: EMA/FDA
update label
crizotinib to ROS+
2014: FDA
approves
ceritinib
2015: EMA
approves
ceritinib/alectinib
2016: EMA
approves
alectinib
2017:
FDA/EMA
Dabrafenib
+ trametinib
(BRAF
V600E)
2012-2019: Participant’s feedback
2017:
participants
workshop cfDNA
2016: survey on
cfDNA
implementation
2018:
Osimertinib
(1st L EGFR)
2018:
FDA/EMA
approves
durvaluma/
nivolumab/
Pembrolizu
mab
2015: EMA
approves
nivolumab
2019 FDA
approves
pembrolizumab
3
PD-L1 ALK ROS1 EGFR
KRAS/BRAF
C-MET
IHC IHC IHC
FISHFISH
Genomic
variant
analysis
Genomic
variant
analysis
RNA analysis
2017 2012 2014
20142012
2018
2019
2013
2017
Currently assessed biomarkers & techniques
4
ALK IHC
ROS1 IHC • Genotyping results
• Technical assessment
• Reporting
• Interpretation of IHC
digital cases
PD-L1 IHC
ALK FISH
ROS1 FISH
• Genotyping
• Interpretation of FISH
digital cases
• Reporting
EGFR, KRAS, BRAF • Genotyping results
• Reporting
C-MET
N=8 N=4
N=10
N=5
N=5
N=5 N=5
ESP Lung EQA scheme: Set-up
FFPE resection specimens
Digitals
5
PD-L1 ALK ROS1 EGFR
KRAS/BRAF
C-MET
IHC IHC IHC
FISHFISH
Genomic
variant
analysis
Genomic
variant
analysis
RNA analysis
Currently assessed biomarkers & techniques
6
7
Overview of PD-L1 test results
Year Subscheme N° laboratories% successful
laboratories*
Average
score (%)
2017 (pilot)
IHC
11 100% 96%
2018 10 90% 94%
2019 24 (ongoing)
2017 (pilot)IHC technical
assessment
9 89% 89%
2018 11 100% 89%
2019 24 (ongoing)
2019: sponsored by * ≥90% analysis score or ≥3/5 technical score
Year Subscheme N° laboratories% successful
laboratories*
Average
score (%)
2017 (pilot)
IHC
78 71% 91%
2018 72 85% 93%
2019 107 (ongoing)
2017 (pilot)IHC technical
assessment
67 81% 76%
2018 74 96% 86%
2019 107 (ongoing)
PD-L1 analysis methods (2018 – R3)
58%22%
7%
6%4% 3%
EU (n=72)
22C3 (Dako)
SP263 (Ventana)
E1L3N (cell signaling)
28-8 (Abcam)
QR1 (Quartett)
CAL10 (Biocare Medical)
SP142 (Ventana)82%
9%
9%
BE (n=11)
8
9
Overview of PD-L1 test results
Year Subscheme N° laboratories% successful
laboratories*
Average
score (%)
2017 (pilot)
IHC
11 100% 96%
2018 10 90% 94%
2019 24 (ongoing)
2017 (pilot) IHC technical
assessment
9 89% 89%
2018 11 100% 89%
2019 24 (ongoing)
2019: sponsored by * ≥90% analysis score or ≥3/5 technical score
Year Subscheme N° laboratories% successful
laboratories*
Average
score (%)
2017 (pilot)
IHC
78 71% 91%
2018 72 85% 93%
2019 107 (ongoing)
2017 (pilot) IHC technical
assessment
67 81% 76%
2018 74 96% 86%
2019 107 (ongoing)
Keppens C et al. 2019. PD-L1 immunohistochemistry in non-small cell lung cancer:
unraveling differences in staining quality and interpretation. Under review.
Technical assessment of staining performance
(2017-2018)
▪ 1 TMA slide – 3 µm
▪ Team of min. 2 pathologists
▪ Microscopic review of staining quality
compared to optimal staning for specific antibody
▪ Indivual comments on staining quality
3 cases (2017)
4 cases (2018)
+ Tonsil control
Link of staining score (on 5 points) to
▪ Laboratory setting (type, accreditation)
▪ Experience (# samples, laboratory size,..)
▪ Applied protocols
5 Excellent staining
4 Pass + minor remark
3 Deficiency without clinical effect
2 Deficiency with clinical effect
1Failure to stain the slides, no
interpretation possible
10
Keppens C et al. 2019. PD-L1 immunohistochemistry in non-small cell lung cancer:
unraveling differences in staining quality and interpretation. Under review.
Primary antibody
Epitope retrieval kit
Detection system# times
used (%) (n=141)
Method code
OR (95% CI) relative to method
a b c d e f
22C3 (Dako)
Cc1 (Ventana) OptiView DAB IHC Detection Kit (Ventana)
35 (24.8%)
a /1.247
[0.547; 2.843]
4.075 [1.314;
12.632]*
0.129 [0.042;
0.396]***
1.769 [0.482; 6.494]
1.211 [0.422; 3.481]
EnVisionFLEX Target Retrieval Solution, low pH (Dako)
Envision flex (Dako)
32 (22.7%)
b0.802
[0.352; 1.828]
/3.269
[0.979; 10.919]
0.103 [0.029;
0.371]***
1.419 [0.361; 5.579]
0.972 [0.311; 3.037]
various various 1 (0.7%) c0.245
[0.079; 0.761]*
0.306[0.092; 1.022]
/0.032
[0.007; 0,147]****
0.434 [0.090; 2.102]
0.297[0.075; 1.178]
SP263 (Ventana)
Cc1 (Ventana) OptiView DAB IHC Detection Kit (Ventana)
24 (17.0%)
d
7.752 [2.525;
23.810]***
9.709 [2.695;
34.483]***
31.660 [6.809;
147.22]****
/13.746 [2.699;
70.007]**
9.412 [2.466;
35.916]**
E1L3N (cell signaling)
Bond Epitope Retrieval 2 (Leica)or homebrew EDTA
various 6 (4.3%) e0.565
[0.154; 2.075]
0.705[0.179; 2.770]
2.304[0.476; 11.111]
0.073[0.014;
0.371]**/
0.685[0.149; 3.155]
Other various various 1 (0.7%) f0.826
[0.287; 2.370]
1.029[0.329; 3.215]
3.364 [0.849; 13.321]
0.106[0.028;
0.406]**
1.461 [0.317; 6.719]
/
Comparison of PD-L1 IHC assays
OR=Odds Ratio [90% CI], *p<0.05, **p<0.01, ***p<0.001.
11
Keppens C et al. 2019. PD-L1 immunohistochemistry in non-small cell lung cancer:
unraveling differences in staining quality and interpretation. Under review.
Primary antibody
Epitope retrieval kit
Detection system# times
used (%) (n=141)
Method code
OR (95% CI) relative to method
a b c d e f
22C3 (Dako)
Cc1 (Ventana) OptiView DAB IHC Detection Kit (Ventana)
35 (24.8%)
a /1.247
[0.547; 2.843]
4.075 [1.314;
12.632]*
0.129 [0.042;
0.396]***
1.769 [0.482; 6.494]
1.211 [0.422; 3.481]
EnVisionFLEX Target Retrieval Solution, low pH (Dako)
Envision flex (Dako)
32 (22.7%)
b0.802
[0.352; 1.828]
/3.269
[0.979; 10.919]
0.103 [0.029;
0.371]***
1.419 [0.361; 5.579]
0.972 [0.311; 3.037]
various various 1 (0.7%) c0.245
[0.079; 0.761]*
0.306[0.092; 1.022]
/0.032
[0.007; 0,147]****
0.434 [0.090; 2.102]
0.297[0.075; 1.178]
SP263 (Ventana)
Cc1 (Ventana) OptiView DAB IHC Detection Kit (Ventana)
24 (17.0%)
d
7.752 [2.525;
23.810]***
9.709 [2.695;
34.483]***
31.660 [6.809;
147.22]****
/13.746 [2.699;
70.007]**
9.412 [2.466;
35.916]**
E1L3N (cell signaling)
Bond Epitope Retrieval 2 (Leica)or homebrew EDTA
various 6 (4.3%) e0.565
[0.154; 2.075]
0.705[0.179; 2.770]
2.304[0.476; 11.111]
0.073[0.014;
0.371]**/
0.685[0.149; 3.155]
Other various various 1 (0.7%) f0.826
[0.287; 2.370]
1.029[0.329; 3.215]
3.364 [0.849; 13.321]
0.106[0.028;
0.406]**
1.461 [0.317; 6.719]
/
Comparison of PD-L1 IHC assays
OR=Odds Ratio [90% CI], *p<0.05, **p<0.01, ***p<0.001.
12
Keppens C et al. 2019. PD-L1 immunohistochemistry in non-small cell lung cancer:
unraveling differences in staining quality and interpretation. Under review.
Primary antibody
Epitope retrieval kit
Detection system# times
used (%) (n=141)
Method code
OR (95% CI) relative to method
a b c d e f
22C3 (Dako)
Cc1 (Ventana) OptiView DAB IHC Detection Kit (Ventana)
35 (24.8%)
a /1.247
[0.547; 2.843]
4.075 [1.314;
12.632]*
0.129 [0.042;
0.396]***
1.769 [0.482; 6.494]
1.211 [0.422; 3.481]
EnVisionFLEX Target Retrieval Solution, low pH (Dako)
Envision flex (Dako)
32 (22.7%)
b0.802
[0.352; 1.828]
/3.269
[0.979; 10.919]
0.103 [0.029;
0.371]***
1.419 [0.361; 5.579]
0.972 [0.311; 3.037]
various various 1 (0.7%) c0.245
[0.079; 0.761]*
0.306[0.092; 1.022]
/0.032
[0.007; 0,147]****
0.434 [0.090; 2.102]
0.297[0.075; 1.178]
SP263 (Ventana)
Cc1 (Ventana) OptiView DAB IHC Detection Kit (Ventana)
24 (17.0%)
d
7.752 [2.525;
23.810]***
9.709 [2.695;
34.483]***
31.660 [6.809;
147.22]****
/13.746 [2.699;
70.007]**
9.412 [2.466;
35.916]**
E1L3N (cell signaling)
Bond Epitope Retrieval 2 (Leica)or homebrew EDTA
various 6 (4.3%) e0.565
[0.154; 2.075]
0.705[0.179; 2.770]
2.304[0.476; 11.111]
0.073[0.014;
0.371]**/
0.685[0.149; 3.155]
Other various various 1 (0.7%) f0.826
[0.287; 2.370]
1.029[0.329; 3.215]
3.364 [0.849; 13.321]
0.106[0.028;
0.406]**
1.461 [0.317; 6.719]
/
Comparison of PD-L1 IHC assays
OR=Odds Ratio [90% CI], *p<0.05, **p<0.01, ***p<0.001.
13
Optimal/weak PD-L1 stainings: an example
Keppens C et al. 2019. PD-L1 immunohistochemistry in non-small cell lung cancer:
unraveling differences in staining quality and interpretation. Under review.
Optim
al
Suboptim
al
Weak +
Cytoplasmic stainingExcessive background
staining
Background
stainingWeak +
Cytoplasmic staining
Weak
staining
14
Staining artefacts
Keppens C et al. 2019. PD-L1 immunohistochemistry in non-small cell lung cancer:
unraveling differences in staining quality and interpretation. Under review.
More than 1 artefact could be observed per laboratory
15
Keppens C et al. 2019. PD-L1 immunohistochemistry in non-small cell lung cancer:
unraveling differences in staining quality and interpretation. Under review.
Staining quality directly affects TPS
estimations & technical failures
Importance
of including
a technical
assessment
to improve
outcomes!
IRR=Incidence Rate Ratio [90% CI]
16
Keppens C et al. 2019. PD-L1 immunohistochemistry in non-small cell lung cancer:
unraveling differences in staining quality and interpretation. Under review.
Technical assessment of staining performance
(2017-2018)
▪ 32 different protocols
❑ 22C3 (Dako) 56.7%
❑ SP263 (Ventana) 19.1%
❑ E1L3N (Cell Signaling) 7.1%
▪ Staining artefacts
❑ Very weak or weak antigen demonstration 63.0%
❑ Excessive background staining 19.8%
▪ Positive influences
❑ EQA scheme year
❑ Use of CE-IVD kit without protocol manipulation LDT
❑ Antibody dilution: RTU
❑ Laboratory accreditation (overestimations only)
17
PD-L1 ALK ROS1 EGFR
KRAS/BRAF
C-MET
IHC IHC IHC
FISHFISH
Genomic
variant
analysis
Genomic
variant
analysis
RNA analysis
Currently assessed biomarkers & techniques
18
ALK ROS1
Year Subscheme # participantsSuccessful
participants*# participants
Successful
participants*
2012
IHC
analysis
29 (pilot) 52% / /
2013 58 64% / /
2014 96 70% 31 (pilot) 90%
2015 95 92% 31 58%
2016 102 88% 36 94%
2017 109 83% 53 94%
2018 99 90% 58 97%
2019 104 (ongoing) 79 (ongoing)
2015
IHC technical
assessment
73 (pilot) 90% / /
2016 90 95% 31 (pilot) 100%
2017 95 97% 52 90%
2018 93 96% 54 98%
2019 104 (ongoing) 79 (ongoing)
19
Tembuyser L, Tack V, Zwaenepoel K, et al. Plos One 2014;9(11);e112159
Keppens C et al. Oncotarget 2018; 9(29); pp: 20524-20538
Overview of ALK/ROS1 IHC test results
* ≥90% analysis score or ≥3/5 technical score
20
Tembuyser L, Tack V, Zwaenepoel K, et al. Plos One 2014;9(11);e112159
Keppens C et al. Oncotarget 2018; 9(29); pp: 20524-20538
Overview of ALK/ROS1 IHC test results
ALK ROS1
Year Subscheme # participantsSuccessful
participants*# participants
Successful
participants*
2012
IHC
analysis
(pilot) / /
2013 1 100% / /
2014 11 64% 5 (pilot) 100%
2015 12 100% 4 50%
2016 11 82% 7 100%
2017 11 73% 8 100%
2018 13 77% 9 89%
2019 9 (ongoing) 12 (ongoing)
2015
IHC technical
assessment
11 (pilot) 100% / /
2016 9 89% 5 (pilot) 100%
2017 8 100% 12 100%
2018 14 93% 8 100%
2019 9 (ongoing) 12 (ongoing)
* ≥90% analysis score or ≥3/5 technical score
ALK/ROS1 IHC antibodies (2018 – R3)
48%
23%
8%
7%
5%2%
7%
ALK EU (n=99) D5F3 (Ventana)
D5F3 (Cell SignallingTechnology)1A4 (Origene)
5A4 (Leica)
5A4 (Novocastra)
5A4 (Abcam)
Other
36%
36%
14%
7%7%
ALK BE (n=14)
ROS1 (n8):
• 100% D4D6 (Cell Signalling
Technology)
ROS1 (n=58):
• 91% D4D6 (Cell Signalling Technology)
• 4% EP282 (Epitomics)
• 2% D4D6 (Genemed)
• 2% D4D6 (Ozyme)
• 2% oti1a1 (ThermoFisher)
21
ALK ROS1
Year Subscheme # participantsSuccessful
participants*# participants
Successful
participants*
2012
IHC
analysis
29 (pilot) 52% / /
2013 58 64% / /
2014 96 70% 31 (pilot) 90%
2015 95 92% 31 58%
2016 102 88% 36 94%
2017 109 83% 53 94%
2018 99 90% 58 97%
2019 104 (ongoing) 79 (ongoing)
2015
IHC technical
assessment
73 (pilot) 90% / /
2016 90 95% 31 (pilot) 100%
2017 95 97% 52 90%
2018 93 96% 54 98%
2019 104 (ongoing) 79 (ongoing)
22
Tembuyser L, Tack V, Zwaenepoel K, et al. Plos One 2014;9(11);e112159
Keppens C et al. Oncotarget 2018; 9(29); pp: 20524-20538
Overview of ALK/ROS1 IHC test results
* ≥90% analysis score or ≥3/5 technical score
Keppens C et al. 2019. Staining performance of ALK and ROS1 immunohistochemistry and influence
on molecular interpretation in NSCLC. In preparation.
Technical assessment of staining performance
(2015-2018)
▪ Similar set-up as for PD-L1
Marker ALK ROS1
EQA scheme year 2015 2016 2017 2018 2016 2017 2018
# participants 73 91 96 92 31 52 54
Average staining
score
(on a total of 5
points)
3,7 4,2 4,3 4,1 4,0 3,9 4,5
# assessorsOne group of 2
One group of 32 3 3 3 3 2
Sample type All resectionsAll
resections
All
resections
All
resections
All
resections
-2 resections
-1 cell-line with
a positive and
negative core
All
resections
# cases provided 5 5 3 3 5 4 3
# positive cases 2 2 1 2 2 2 1
# negative cases 3 3 2 1 3 2 2
Abbreviations: #, number; ALK, anaplastic lymphoma kinase;
EQA, External quality assessment; ROS1, ROS proto-oncogene 1.
23
Keppens C et al. 2019. Staining performance of ALK and ROS1 immunohistochemistry and influence
on molecular interpretation in NSCLC. In preparation.
Influencers of staining performance
Marker ALK ROS1
More samples tested in last 12 months
# staff involved in testing
More successive EQA participations
Later EQA scheme year
Laboratory setting
Laboratory accreditation
Use of a CE-IVD kit N/A
Switched protocol between schemes
Antibody dilution
Incubation time > 60 min.
Incubation temperature > 40 °C
Incubation at room temperature
24
Staining quality directly affects false-positives,
false-negatives and technical failures
Similar to PD-L1 -> EQA providers should assess both aspects!
+ Fewer analysis failures when more samples are tested annually,
more failures when a new method is introduced
Keppens C et al. 2019. Staining performance of ALK and ROS1 immunohistochemistry and influence
on molecular interpretation in NSCLC. In preparation.
IRR=Incidence Rate Ratio [90% CI]
25
PD-L1 ALK ROS1 EGFR
KRAS/BRAF
C-MET
IHC IHC IHC
FISHFISH
Genomic
variant
analysis
Genomic
variant
analysis
RNA analysis
Currently assessed biomarkers & techniques
26
27
Year Subscheme # laboratories % of labs successful**
2012 (pilot)ALK FISH 54 72%
ALK FISH digital 67 82%
2013ALK FISH 104 68%
ALK FISH digital 106 74%
2014ALK FISH 116 69%
ALK FISH digital 81 educational
2015
ALK FISH + digital*
111 79%
2016 113 82%
2017 116 79%
2018 103 98%
2019 (ongoing)
2014 (pilot)
ROS1 FISH + digital*
56 64%
2015 68 78%
2016 71 70%
2017 85 82%
2018 86 94%
2019 (ongoing)
Overview of ALK/ROS1 FISH test results
*Combination of wet and digital cases
** >90% analysis score
Tembuyser L, Tack V, Zwaenepoel K, et al. Plos One 2014;9(11);e112159
Keppens C et al. Oncotarget 2018; 9(29); pp: 20524-20538 27
28
Year Subscheme # laboratories % of labs successful**
2012 (pilot)ALK FISH 7 ND
ALK FISH digital 7 ND
2013ALK FISH 9 56%
ALK FISH digital 8 75%
2014ALK FISH 9 67%
ALK FISH digital 9 educational
2015
ALK FISH + digital*
6 100%
2016 5 100%
2017 7 57%
2018 6 50%
2019 5 (ongoing)
2014 (pilot)
ROS1 FISH + digital*
4 ND
2015 3 100%
2016 4 75%
2017 6 83%
2018 6 83%
2019 5 (ongoing)
Overview of ALK/ROS1 FISH test results
*Combination of wet and digital cases
** >90% analysis score
Tembuyser L, Tack V, Zwaenepoel K, et al. Plos One 2014;9(11);e112159
Keppens C et al. Oncotarget 2018; 9(29); pp: 20524-20538 28
ALK FISH methods EU (n=103) % BE (n=6) %
Vysis ALK break apart FISH Probe (Abbott) 48 (46.6%) 4 (66.7%)
ZytoLight SPEC ALK Dual Color Break Apart Probe (ZytoVision) 21 (20.4%)
ZytoLight SPEC ALK/EML4 TriCheck Probe (ZytoVision) 8 (7.8%)
ALK FISH DNA Probe, Split Signal (Dako) 5 (4.9%
Non-commercial method 5 (4.9%)
Repeat-Free Poseidon ALK (2p23) Break Probe (Kreatech) 4 (3.9%)
ALK IQFISH Break-Apart Probe for Dako Omnis (Dako, Agilent) 3 (2.9%) 2 (33.3%)
ROS FISH methods EU (n=86) % BE (n=6) %
ZytoLight SPEC ROS1 Dual Color Break Apart Probe
(ZytoVision)49 57,0%
2 33,3%
Repeat-Free Poseidon ROS1 (6q22) Break FISH Probe
(Kreatech Diagnostics)8 9,3%
6q22 ROS1 Break Apart FISH Probe RUO Kit (Abbott Molecular) 7 8,1% 1 16,7%
ROS1 break apart Probe (Cytocell) 4 4,7%
Vysis LSI ROS1 (Cen) SpectrumGreen Probe (Abbott) 3 3,5% 1 16,7%
Homebrew 3 3,5%
ROS1 IQFISH Break-Apart Probe for Dako Omnis (Dako,
Agilent)3 3,5%
2 33,3%
ALK/ROS1 FISH probes (2018 – R3)(most frequently used)
29
↑ errors for ROS1
↓ errors for ALK
Keppens C, et al. Oncotarget 2018; 9(29); pp: 20524-20538
Longitudinal analysis:
Errors IHC > errors FISH
Errors FISH > errors FISH digital
30
Error rates in earliest schemes
Error rates are calculated taking into account the number of false positive and false negative results as
well as the total number of samples for which a result was submitted, with the exclusion of educational
cases for which more than or equal to 25% of the participants were not able to obtain a result.
30
PD-L1 ALK ROS1 EGFR
KRAS/BRAFC-MET
IHC IHC IHC
FISHFISH
Genomic
variant
analysis
Genomic
variant
analysis
RNA analysis
Currently assessed biomarkers & techniques
31
2018 MET pilot EQA scheme
Sample
numberOutcome MET NM_000245.3 VAF Errors (%)
Technical
failures (%)
L18.MET1WT
SNP c.2975C>T, p.(T992I)
/
19%1/37 (3) 0/37 (0)
L18.MET2 c.3023_3028+9del 24% 9/37 (24) 0/37 (0)
L18.MET3c.3028+3A>G
SNP c.2975C>T, p.(T992I)
62%
40%8/37 (22) 1/37 (3)
L18.MET4 WT / 0/37 (0) 0/37 (0)
L18.MET5 WT / 0/37 (0) 1/37 (3)
▪ For genomic DNA analysis (mutations)
▪ 49 laboratories
▪ 12 laboratories used RNA tests (ex 14 skipping)
Not included in assessment
Separate RNA evaluation in 2019 EQA
▪ NGS: 31/37, non-commercial sequencing: 6/37
SNPs not considered to calculate scores
Successful state not determined (pilot)
▪ Average score: 90.4% (n=37)
32
PD-L1 ALK ROS1 EGFR
KRAS/BRAFC-MET
IHC IHC IHC
FISHFISH
Genomic
variant
analysis
Genomic
variant
analysis
RNA analysis
Currently assessed biomarkers & techniques
33
EU BE
Year Gene(s) # participantsSuccessful
particpants**# participants
Successful
particpants**
2013 (pilot)EGFR 107 educational 10 educational
KRAS 92 educational 5 educational
2014 EGFR 144 61% 8 75%
2015 EGFR 114 52% 10 70%
2016 EGFR 97 71% 7 100%
2017*
EGFR 101 71% 9 89%
KRAS 51 98% 9 100%
BRAF 47 98% 7 100%
2018*
EGFR 98 82% 8 100%
KRAS 56 91% 6 100%
BRAF 55 96% 6 100%
2019*
EGFR 95 (ongoing) 7 (ongoing)
KRAS 70 (ongoing) 7 (ongoing)
BRAF 73 (ongoing) 7 (ongoing)
34
Keppens C et al. Oncotarget 2018; 9(29); pp: 20524-20538
Overview of molecular test results
*EGFR = mandatory, KRAS/BRAF=optional
**>90% analysis score
0
10
20
30
40
50
60
70
101 144 114 97 101 98 95
2013 2014 2015 2016 2017 2018 2019
% p
arti
cip
ants
EGFR methods
Prescreening
Commercial test kit
NGS
LDT
N=
Technological expansion: NGS in favour of LDTs
NGS=next-generation sequencing, LDT=Laboratory developed test35
https://varnomen.hgvs.org/
Keppens C, et al. Hum Mutat. 2019 Sep 25. doi: 10.1002/humu.23926.
HGVS nomenclature during molecular schemes
Scheme year 2013 2014 2015 2016 2017 2018
Gene Variant % HGVS compliant nomenclature (N)
EGFR
(NM_005228.5)
c.2155G>A
p.(Gly719Ser)/
3.8
(131)/
6.3
(32)/ /
c.2155G>T
p.(Gly719Cys)
0.0
(37)/
9.1
(44)/ / /
c.2235_2249del
p.(Glu746_Ala750del) / /
4.7
(121)***
25.9
(27)
47.2
(36)***/
c.2236_2250del
p.(Glu746_Ala750del)/
5.3
(113)/ / /
20.0
(95)**
c.2303G>T
p.(Ser768Ile)
0.0
(17)/
13.8
(29)
8.7
(23)/ /
c.2369C>T
p.(Thr790Met)
7.0
(71)
4.3
(93)**
5.2
(97)*
18.1
(83)
21.7
(92)**
16.5
(194)
c.2573T>G
p.(Leu858Arg)
3.0
(168)***
0.0
(85)***
6.6
(150)**
18.0
(130)
38.6
(133)***
27.1
(96)***
Recurring variants are shown only. In case a variant was distributed in both ESP and Gen&Tiss EQA schemes within a specific year, average
percentages for both schemes were presented. /: Variant not distributed during this scheme year. Asterisks represent a statistical difference
compared to other scheme years. Chi-squared test or Fisher’s Exact test for cell counts below 5. *p<0.05, **p<0.01, ***p<0.001.
Abbreviations: EGFR, Epidermal Growth Factor Receptor; HGVS, Human genome Variation Society;
36
Keppens C, et al. Hum Mutat. 2019 Sep 25. doi: 10.1002/humu.23926.
HGVS compliance is related to used analysis
techniques
*p<0.05, **p<0.01, ***p<0.001.
NGS = next-generation sequencing
37
Quality Management System
External Quality Assessment
Tools to assure quality of care and
laboratory testing
- Requirements in some European countries
- Hospital and laboratory accreditation
- More and more evidence that obtaining / holding an accreditation is a step forward
38
The outer line indicates the significant results (p < 0.05). The inner line shows the significance level of 0.05.
All markers in the centre of the figure showed no significant result.
Tack V. et al, Br J Cancer; 2018; Vol. 119; iss. 5; pp. 605 - 614
ESP EQA 2013-2016:
Accreditation leads to:
-successful EQA scores (p =0.018) (score
≥ 90%),
-fewer analysis errors (p = 0.002).
Laboratory setting:
-less analysis errors were made in a
university and research background,
compared to (private) laboratories and
industry laboratories (p = 0.016 and p =
0.012, respectively).
A higher nr of samples tested/year:
-increasing probability to have a successful
EQA
score (p = 0.009)
Effect on EQA performance
39
Longitudinal follow-up of problems reported
by EQA participants (2015-2018)
Deviating EQA results:
o 325 individual surveys
o 184 unique laboratories ~34 countries
o 514 analyzed cases (NSCLC + mCRC)
0 10 20 30 40 50 60 70 80 90 100
FISH (wet cases)
FISH (digital cases)
IHC (wet cases)
IHC (digital cases)
IHC (technical assessment)
Variant analysis
% cases (n=514)
Pre-analytical Analytical Post-analytical Unknown
40Keppens C. et al, Managing deviating EQA results: strategies of clinical laboratories testing for
oncological biomarkers. 2019. Submitted.
Longitudinal follow-up of problems reported
by EQA participants (2015-2018)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
99 102 99 108 66 10
ALK ROS1 PD-L1 EGFR KRAS WT
% c
ase
s
Cause
Unknown/other
Technical problem/equipment
Clerical error/personnel error
Problem with EQA material
Methodological/reagent issue
Interpretation error
41Keppens C. et al, Managing deviating EQA results: strategies of clinical laboratories testing for
oncological biomarkers. 2019. Submitted.
Importance of quality assessment during
all phases in the total test process
For laboratories:
ISO15189:2012a:
“Interlaboratory comparison programme(s) chosen by the laboratory shall, as far as possible, provide clinically
relevant challenges that mimic patient samples and have the effect of checking the entire examination
process, including pre-examination procedures, and post-examination procedures, where possible.”
For EQA providers:
Guideline on the requirements for EQA
providersb:
“Three aspects of molecular testing should be
covered by an EQA program: Pathology review,
the molecular test itself, and reporting.”
a. ISO 15189:2012 Medical laboratories -
Particular requirements for quality and
competence.
b. van Krieken JH et al. (2013) Virchows Arch.
42
ISO 15189:2012 Medical laboratories - Particular requirements for quality and competence (International Organization for Standardization)
van Krieken JH et al. Guideline on the requirements of external quality assessment programs in molecular pathology. Virchows Arch. 462:27-37 (2013)
CAP - Laboratory Accreditation Program - Molecular Pathology Checklist (2014)
Gulley M.L. et al. Clinical laboratory reports in molecular pathology. Arch Pathol Lab Med. 131:852-863 (2007)
ESP EQA SCHEMES FOR NSCLC:
Standards and guidelines on reporting
43
ESP Lung EQA: scoring of the post-analytical phase
Per sample evaluation:
• Test results (genotype given)
• Interpretation given and correct
General items:
• Patient name
• Date of birth
Method information:
• Method used
• Sensitivity/treshold of method
• List of mutations tested
• Reference sequence
44
Average reporting scores over time
0
10
20
30
40
50
60
70
80
90
100
ALK ROS1 ALK ROS1 PD-L1 EGFR/KRAS/BRAF
MET
FISH IHC Variantanalysis
Ave
rag
e r
ep
ort
ing
sc
ore
(%
)
EQA scheme
2012 2013 2014 2015 2016 2017 2018
*2013-2016: EGFR only,
2017-2018: EGFR/KRAS/BRAF
*
45
BE
Technique-specific information (2018)
0102030405060708090
100
Me
tho
d u
se
d, in
cl.
aberr
ations teste
d
Me
tho
d t
resh
old
# n
eo
pla
stic c
ells
an
aly
se
d
# c
ells
with s
plit
/sin
gle
sig
nal
Me
tho
d u
se
d, in
cl. a
ntib
od
y
IHC
in
terp
reta
tio
n c
rite
ria
Ne
op
lastic c
ell
co
nte
nt
Me
tho
ds u
se
d
Me
tho
d s
en
sitiv
ity
Sp
ecific
atio
n o
f te
ste
dm
uta
tio
ns
Re
fere
nce
se
qu
en
ce
FISH IHC variantanalysis
Ite
m p
rese
nt a
nd
co
rre
ct (%
)
ALK ROS1 PD-L1 Molecular scheme MET
46
Upcoming ESP Lung EQA schemes
ESP Lung EQA scheme
2020
IQN Path cfDNA EQA scheme
2019-2020
Registrations open now until 18 JAN 2019 Registration period: TBD
3 rounds:
▪ March (wet samples)
▪ May (digital cases)
▪ September (wet samples)
Sample distribution:
▪ February 2020
▪ 5 plasma samples
▪ common EGFR variants
5 digital images for ALK/ROS1 FISH schemes
4 digitals for PD-L1 IHC
Technical evaluation for IHC schemes
5 pre-validated FFPE resections
10 for molecular DNA scheme
1 report/subscheme (mol. DNA scheme: 2 reports)
47
2020: Free participation to PD-L1 subscheme
for BE participants thanks to
The 2020 ESP Lung EQA scheme
New features
PD-L1 ALK ROS1 EGFR
KRAS/BRAF
C-MET
IHC IHC IHC
FISHFISH
Genomic variant analysis
RNA-based analysis
One combined scheme
EGFR = mandatory
New scheme for detection of
ALK/ROS1 fusions and
MET exon 14 skippng
48
New participants platform
www.eqascheme.org
▪ Modernized look
▪ Ease of use
▪ Inegrated datasheets
▪ Automated emailing
▪ One platform for all EQA schemes
Lung, cfDNA, Cystic Fibrosis,
Clonality, …
❑ CF 2020: ongoing
❑ Clonality 2020: January 2020
❑ cfDNA 2020: January 2020
❑ Lung 2021: November 2020
49
Scheme coordinator and assistant coordinators Prof. Dr. Dequeker E., Gentens R., Hombroeckx E, Tembuyser L., Tack
V., Keppens C., Dufraing K., Van Casteren K, Nauwelaers I
European Society of Pathology Sc. Director Aldieri R., Byrhanga S, Short M,
The reference laboratories University Medical Center Groningen, Groningen, The Netherlands
Erasmus MC, Rotterdam, The Netherlands,
Radboud UMC, Nijmegen, The Netherlands
University Hospital Antwerp, Edegem, Belgium
Charles University, Hradec Králové, Czech Republic
Center of Excellence on Aging, Chieti, Italy
Hospital University Vall d´Hebron, Barcelona, Spain
Institut for Pathology, Cologne, Germany
Institute for Pathology, Basel, Switzerland
Laboratorio de Dianas Terapéuticas, Madrid, Spain
UCL advanced diagnostics, London, UK
VU University Medical Centre, Amsterdam, The Netherlands
UZ Leuven, Leuven, Belgium
The medical/technical experts Prof. Dr. Schuuring E., Dr. ‘t Hart N.,
Dr. von der Thüsen J, Prof. Dr. Ryska A,
Prof. Dr. Pauwels P., Dr. Zwaenepoel K.,
Prof. Dr. Ales Ryska, prof. Dr. Erik Thunnissen
The assessors Bubendorf L., Cabillic F., Delen S., Marchetti A., Miller K.,
Pauwels P., Rouleau E, Ryska A., Schuuring E., Stenzinger A,
t Hart N., Thunnissen E., Tornillo L.., Warth A., Weichert W.,
Zwaenepoel K
Scheme sponsors AstraZeneca Belgium, Pfizer Oncology, Bristol-Myers Squibb, Roche
Turkey, Pfizer (Poland, Croatia, Serbia and Central Balkan Countries),
Pfizer (Czech Republic, Hungary, Slovakia)
The participating laboratories
Acknowledgements