User login
Comparison of Precision Oncology Annotation Services in the National Precision Oncology Program
BACKGROUND
The National Precision Oncology Program (NPOP) provides comprehensive genomic profiling (CGP) through external vendors to patients within the Veterans Affairs Healthcare System who meet testing guidelines. We sought to assess the concordance of cancer therapy recommendations between Foundation Medicine (FM), one of the NPOP vendors, and OncoKB, an FDA-recognized public precision oncology knowledge database, which annotates human genetic variants associated with therapies guidance at varying levels of evidence.
METHODS
We selected FM CGP test reports with at least one therapy recommendation regardless of FDA approval or level of evidence were selected to compare FM and OncoKB therapy annotations of different mutation types, including short variants (SVs), rearrangements, and copy number alterations (CNAs) between 02/01/2019-03/13/2023. Therapy recommendations of annotations for unique combinations of gene, variant, and cancer type from FM and OncoKB were compared. Comparisons were scored as an Exact Match (EM) if FM and OncoKB therapy annotation was the same or a Partial Match (PM) if the FM therapy annotation was a subset of OncoKB’s or vice versa.
RESULTS
For annotations involving FDA-approved therapies, a total of 10,435 cases were compared for SVs, 546 for rearrangements, and 732 for CNAs. Among SVs annotations, 7,029 (67.4%) were EM and 787 (7.5%) were PM. Of rearrangement annotations, 328 (60.1%) were EM and 95 (17.4%) were PM. Of CNA annotations, 469 (64.1%) were EM and 28 (3.8%) were PM. For off-label therapies, agreement between annotation sources was much lower in all above scenarios. Examples included 3022 (29%) cases were identified as EM plus PM for SVs, 324 (59.3%) for rearrangements, and 42 (5.7%) for CNAs.
CONCLUSIONS
Therapy recommendations were inconsistent between FM and OncoKB annotation services, with a substantial disagreement among both FDA-approved and off-label therapy annotations. The limitation of time difference of annotations performed between FM and OncoKB therapy annotations accounted for some disagreement. Establishing accuracy and improving concordance between different annotation services is needed to better match treatments to patients and improve provider trust and reliability of annotation service.
BACKGROUND
The National Precision Oncology Program (NPOP) provides comprehensive genomic profiling (CGP) through external vendors to patients within the Veterans Affairs Healthcare System who meet testing guidelines. We sought to assess the concordance of cancer therapy recommendations between Foundation Medicine (FM), one of the NPOP vendors, and OncoKB, an FDA-recognized public precision oncology knowledge database, which annotates human genetic variants associated with therapies guidance at varying levels of evidence.
METHODS
We selected FM CGP test reports with at least one therapy recommendation regardless of FDA approval or level of evidence were selected to compare FM and OncoKB therapy annotations of different mutation types, including short variants (SVs), rearrangements, and copy number alterations (CNAs) between 02/01/2019-03/13/2023. Therapy recommendations of annotations for unique combinations of gene, variant, and cancer type from FM and OncoKB were compared. Comparisons were scored as an Exact Match (EM) if FM and OncoKB therapy annotation was the same or a Partial Match (PM) if the FM therapy annotation was a subset of OncoKB’s or vice versa.
RESULTS
For annotations involving FDA-approved therapies, a total of 10,435 cases were compared for SVs, 546 for rearrangements, and 732 for CNAs. Among SVs annotations, 7,029 (67.4%) were EM and 787 (7.5%) were PM. Of rearrangement annotations, 328 (60.1%) were EM and 95 (17.4%) were PM. Of CNA annotations, 469 (64.1%) were EM and 28 (3.8%) were PM. For off-label therapies, agreement between annotation sources was much lower in all above scenarios. Examples included 3022 (29%) cases were identified as EM plus PM for SVs, 324 (59.3%) for rearrangements, and 42 (5.7%) for CNAs.
CONCLUSIONS
Therapy recommendations were inconsistent between FM and OncoKB annotation services, with a substantial disagreement among both FDA-approved and off-label therapy annotations. The limitation of time difference of annotations performed between FM and OncoKB therapy annotations accounted for some disagreement. Establishing accuracy and improving concordance between different annotation services is needed to better match treatments to patients and improve provider trust and reliability of annotation service.
BACKGROUND
The National Precision Oncology Program (NPOP) provides comprehensive genomic profiling (CGP) through external vendors to patients within the Veterans Affairs Healthcare System who meet testing guidelines. We sought to assess the concordance of cancer therapy recommendations between Foundation Medicine (FM), one of the NPOP vendors, and OncoKB, an FDA-recognized public precision oncology knowledge database, which annotates human genetic variants associated with therapies guidance at varying levels of evidence.
METHODS
We selected FM CGP test reports with at least one therapy recommendation regardless of FDA approval or level of evidence were selected to compare FM and OncoKB therapy annotations of different mutation types, including short variants (SVs), rearrangements, and copy number alterations (CNAs) between 02/01/2019-03/13/2023. Therapy recommendations of annotations for unique combinations of gene, variant, and cancer type from FM and OncoKB were compared. Comparisons were scored as an Exact Match (EM) if FM and OncoKB therapy annotation was the same or a Partial Match (PM) if the FM therapy annotation was a subset of OncoKB’s or vice versa.
RESULTS
For annotations involving FDA-approved therapies, a total of 10,435 cases were compared for SVs, 546 for rearrangements, and 732 for CNAs. Among SVs annotations, 7,029 (67.4%) were EM and 787 (7.5%) were PM. Of rearrangement annotations, 328 (60.1%) were EM and 95 (17.4%) were PM. Of CNA annotations, 469 (64.1%) were EM and 28 (3.8%) were PM. For off-label therapies, agreement between annotation sources was much lower in all above scenarios. Examples included 3022 (29%) cases were identified as EM plus PM for SVs, 324 (59.3%) for rearrangements, and 42 (5.7%) for CNAs.
CONCLUSIONS
Therapy recommendations were inconsistent between FM and OncoKB annotation services, with a substantial disagreement among both FDA-approved and off-label therapy annotations. The limitation of time difference of annotations performed between FM and OncoKB therapy annotations accounted for some disagreement. Establishing accuracy and improving concordance between different annotation services is needed to better match treatments to patients and improve provider trust and reliability of annotation service.
Utilization and Clinical Benefit of Immune Checkpoint Inhibitor in Veterans With Microsatellite Instability-High Prostate Cancer
Background
The utilization of immune checkpoint inhibitors (ICI) in prostate cancer (PC) can be very effective for patients with mismatch repair-deficiency (as identified by MSI-H by PCR/NGS or dMMR IHC). The use of ICI in this patient population has been associated with high rates of durable response. There is limited published data on factors that may influence patient response and outcomes. The aim of this study is to describe the utilization of and tumor response to ICI in this patient population.
Methods
This is a retrospective study of men with MSI-H PC reported by somatic genomic testing from April 1, 2015 to March 31, 2022 through the VA National Precision Oncology Program (NPOP), who received at least one dose of ICI. The primary objectives are to describe the incidence of MSI-H PC and the utilization of ICI. Descriptive statistics and Kaplan- Meier estimator were used for secondary objectives to determine the prostate-specific antigen decline of at least 50% (PSA50), clinical progression free survival (cPFS), time on ICI as a function of number of prior therapies, the extent of metastasis prior to initiation of ICI, and the correlation of MMR genetic alterations with treatment response.
Results
66 patients with MSI-H PC were identified (1.5% of a total of 4267 patients with PC tested through NPOP). 23 patients (35%) received at least one dose of ICI. 12 of 23 patients (52%) had PSA response. PSA50 responses occurred in 6 patients (50%) and 5 continued to have durable PSA50 at six months. Median cPFS was 280 days (95% CI: 105 days-not reached) and the estimated PFS at six months was 72.2% (95% CI: 35.7%-90.2%). 8 of 12 (67%) responders have received multiple lines of therapy for M1 PC. 8 of 12 patients (67%) had high-volume disease at ICI initiation. Of those patients with a MMR genetic alteration, patients with MLH1 (3/3) and MSH2 (6/8) alterations responded more frequently than those with MSH6 alterations (1/4).
Conclusions
MSI-H PC is rare but response rates to ICI are high and durable. Patients with MLH1 and MSH2 alterations appeared to respond more frequently than those with MSH6. Additional follow-up is ongoing.
Background
The utilization of immune checkpoint inhibitors (ICI) in prostate cancer (PC) can be very effective for patients with mismatch repair-deficiency (as identified by MSI-H by PCR/NGS or dMMR IHC). The use of ICI in this patient population has been associated with high rates of durable response. There is limited published data on factors that may influence patient response and outcomes. The aim of this study is to describe the utilization of and tumor response to ICI in this patient population.
Methods
This is a retrospective study of men with MSI-H PC reported by somatic genomic testing from April 1, 2015 to March 31, 2022 through the VA National Precision Oncology Program (NPOP), who received at least one dose of ICI. The primary objectives are to describe the incidence of MSI-H PC and the utilization of ICI. Descriptive statistics and Kaplan- Meier estimator were used for secondary objectives to determine the prostate-specific antigen decline of at least 50% (PSA50), clinical progression free survival (cPFS), time on ICI as a function of number of prior therapies, the extent of metastasis prior to initiation of ICI, and the correlation of MMR genetic alterations with treatment response.
Results
66 patients with MSI-H PC were identified (1.5% of a total of 4267 patients with PC tested through NPOP). 23 patients (35%) received at least one dose of ICI. 12 of 23 patients (52%) had PSA response. PSA50 responses occurred in 6 patients (50%) and 5 continued to have durable PSA50 at six months. Median cPFS was 280 days (95% CI: 105 days-not reached) and the estimated PFS at six months was 72.2% (95% CI: 35.7%-90.2%). 8 of 12 (67%) responders have received multiple lines of therapy for M1 PC. 8 of 12 patients (67%) had high-volume disease at ICI initiation. Of those patients with a MMR genetic alteration, patients with MLH1 (3/3) and MSH2 (6/8) alterations responded more frequently than those with MSH6 alterations (1/4).
Conclusions
MSI-H PC is rare but response rates to ICI are high and durable. Patients with MLH1 and MSH2 alterations appeared to respond more frequently than those with MSH6. Additional follow-up is ongoing.
Background
The utilization of immune checkpoint inhibitors (ICI) in prostate cancer (PC) can be very effective for patients with mismatch repair-deficiency (as identified by MSI-H by PCR/NGS or dMMR IHC). The use of ICI in this patient population has been associated with high rates of durable response. There is limited published data on factors that may influence patient response and outcomes. The aim of this study is to describe the utilization of and tumor response to ICI in this patient population.
Methods
This is a retrospective study of men with MSI-H PC reported by somatic genomic testing from April 1, 2015 to March 31, 2022 through the VA National Precision Oncology Program (NPOP), who received at least one dose of ICI. The primary objectives are to describe the incidence of MSI-H PC and the utilization of ICI. Descriptive statistics and Kaplan- Meier estimator were used for secondary objectives to determine the prostate-specific antigen decline of at least 50% (PSA50), clinical progression free survival (cPFS), time on ICI as a function of number of prior therapies, the extent of metastasis prior to initiation of ICI, and the correlation of MMR genetic alterations with treatment response.
Results
66 patients with MSI-H PC were identified (1.5% of a total of 4267 patients with PC tested through NPOP). 23 patients (35%) received at least one dose of ICI. 12 of 23 patients (52%) had PSA response. PSA50 responses occurred in 6 patients (50%) and 5 continued to have durable PSA50 at six months. Median cPFS was 280 days (95% CI: 105 days-not reached) and the estimated PFS at six months was 72.2% (95% CI: 35.7%-90.2%). 8 of 12 (67%) responders have received multiple lines of therapy for M1 PC. 8 of 12 patients (67%) had high-volume disease at ICI initiation. Of those patients with a MMR genetic alteration, patients with MLH1 (3/3) and MSH2 (6/8) alterations responded more frequently than those with MSH6 alterations (1/4).
Conclusions
MSI-H PC is rare but response rates to ICI are high and durable. Patients with MLH1 and MSH2 alterations appeared to respond more frequently than those with MSH6. Additional follow-up is ongoing.
Outcomes of Off-Label Use of Molecular Targeted Agent Therapy in a Large-Scale Precision Oncology Program
Background
Increasing utilization of comprehensive genomic profiling (CGP) and a growing number of targeted agents (TAs) has led to substantial improvements in outcomes among patients with cancer with actionable mutations. We sought to evaluate real-world use and outcomes of off-label TA among patients who underwent CGP.
Methods
The National Precision Oncology Program database and VA Corporate Data Warehouse were queried to identify patients who underwent CGP between Feb 2019 and Dec 2021 and were prescribed 1 of 73 TAs for malignancy. OncoKB (accessed March 2022) annotations were used to select patients who received offlabel TAs based upon CGP results. Chart abstraction was performed in April 2022 to review response to offlabel TAs, toxicities, and treatment duration.
Results
Of 18,686 patients who underwent CGP, 2,107 (11%) were prescribed a TA and 170 (0.9%) were prescribed a total of 185 off-label TA regimens. Median age was 68 years, 88% were male, and 82% had prior systemic therapy, with 28% receiving 3 or more prior lines. Frequency of off-label TA prescriptions was highest for patients with unknown primary (CUP) (9%), thyroid (8%), and breast (6%) cancers. Most frequently targeted alterations involved ERBB2 (22%), BRAF (22%), and BRCA1/BRCA2/ATM (20%). Among the 161 regimens prescribed > 4 weeks, 44 (27%) led to complete or partial response, and 63 (39%) were administered for 6 months or longer or are continuing. Median progression free (PFS) and overall survival (OS) were 5.3 (95% CI, 4.3–6.5) and 9.9 (95% CI, 8.3–12.4) months, respectively. Patients with OncoKb level 2/3A/3B annotations had improved PFS (HR 0.45; 95% CI, 0.24-0.82; P = .01) and OS (HR 0.27; 95% CI, 0.15-0.48; P < .005) compared to level 4 treatments. Across all 185 regimens prescribed, 30 (16.2%) were discontinued due to toxicity, and further systemic treatment was prescribed subsequently in 58 (31.3%).
Conclusions
While the overall use of off-label TAs is low, nearly 10% of CUP and thyroid cancer patients who undergo CGP are prescribed TAs for off-label indications. More than one-quarter of off-label TA regimens lead to treatment response. Treatments associated with level 4 annotations lead to worse outcomes than TAs bearing superior levels of evidence.
Background
Increasing utilization of comprehensive genomic profiling (CGP) and a growing number of targeted agents (TAs) has led to substantial improvements in outcomes among patients with cancer with actionable mutations. We sought to evaluate real-world use and outcomes of off-label TA among patients who underwent CGP.
Methods
The National Precision Oncology Program database and VA Corporate Data Warehouse were queried to identify patients who underwent CGP between Feb 2019 and Dec 2021 and were prescribed 1 of 73 TAs for malignancy. OncoKB (accessed March 2022) annotations were used to select patients who received offlabel TAs based upon CGP results. Chart abstraction was performed in April 2022 to review response to offlabel TAs, toxicities, and treatment duration.
Results
Of 18,686 patients who underwent CGP, 2,107 (11%) were prescribed a TA and 170 (0.9%) were prescribed a total of 185 off-label TA regimens. Median age was 68 years, 88% were male, and 82% had prior systemic therapy, with 28% receiving 3 or more prior lines. Frequency of off-label TA prescriptions was highest for patients with unknown primary (CUP) (9%), thyroid (8%), and breast (6%) cancers. Most frequently targeted alterations involved ERBB2 (22%), BRAF (22%), and BRCA1/BRCA2/ATM (20%). Among the 161 regimens prescribed > 4 weeks, 44 (27%) led to complete or partial response, and 63 (39%) were administered for 6 months or longer or are continuing. Median progression free (PFS) and overall survival (OS) were 5.3 (95% CI, 4.3–6.5) and 9.9 (95% CI, 8.3–12.4) months, respectively. Patients with OncoKb level 2/3A/3B annotations had improved PFS (HR 0.45; 95% CI, 0.24-0.82; P = .01) and OS (HR 0.27; 95% CI, 0.15-0.48; P < .005) compared to level 4 treatments. Across all 185 regimens prescribed, 30 (16.2%) were discontinued due to toxicity, and further systemic treatment was prescribed subsequently in 58 (31.3%).
Conclusions
While the overall use of off-label TAs is low, nearly 10% of CUP and thyroid cancer patients who undergo CGP are prescribed TAs for off-label indications. More than one-quarter of off-label TA regimens lead to treatment response. Treatments associated with level 4 annotations lead to worse outcomes than TAs bearing superior levels of evidence.
Background
Increasing utilization of comprehensive genomic profiling (CGP) and a growing number of targeted agents (TAs) has led to substantial improvements in outcomes among patients with cancer with actionable mutations. We sought to evaluate real-world use and outcomes of off-label TA among patients who underwent CGP.
Methods
The National Precision Oncology Program database and VA Corporate Data Warehouse were queried to identify patients who underwent CGP between Feb 2019 and Dec 2021 and were prescribed 1 of 73 TAs for malignancy. OncoKB (accessed March 2022) annotations were used to select patients who received offlabel TAs based upon CGP results. Chart abstraction was performed in April 2022 to review response to offlabel TAs, toxicities, and treatment duration.
Results
Of 18,686 patients who underwent CGP, 2,107 (11%) were prescribed a TA and 170 (0.9%) were prescribed a total of 185 off-label TA regimens. Median age was 68 years, 88% were male, and 82% had prior systemic therapy, with 28% receiving 3 or more prior lines. Frequency of off-label TA prescriptions was highest for patients with unknown primary (CUP) (9%), thyroid (8%), and breast (6%) cancers. Most frequently targeted alterations involved ERBB2 (22%), BRAF (22%), and BRCA1/BRCA2/ATM (20%). Among the 161 regimens prescribed > 4 weeks, 44 (27%) led to complete or partial response, and 63 (39%) were administered for 6 months or longer or are continuing. Median progression free (PFS) and overall survival (OS) were 5.3 (95% CI, 4.3–6.5) and 9.9 (95% CI, 8.3–12.4) months, respectively. Patients with OncoKb level 2/3A/3B annotations had improved PFS (HR 0.45; 95% CI, 0.24-0.82; P = .01) and OS (HR 0.27; 95% CI, 0.15-0.48; P < .005) compared to level 4 treatments. Across all 185 regimens prescribed, 30 (16.2%) were discontinued due to toxicity, and further systemic treatment was prescribed subsequently in 58 (31.3%).
Conclusions
While the overall use of off-label TAs is low, nearly 10% of CUP and thyroid cancer patients who undergo CGP are prescribed TAs for off-label indications. More than one-quarter of off-label TA regimens lead to treatment response. Treatments associated with level 4 annotations lead to worse outcomes than TAs bearing superior levels of evidence.