Rachel B. Keller
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Rachel B. Keller has a total of 213 co-authors. Here are the top co-authors by number of shared publications:
- Jason Laurence Hornick
Department of Pathology, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA
130 shared publications - Jeffrey William Clark
Harvard Medical School, Boston, Massachusetts
126 shared publications - Geoffrey I. Shapiro
Dana-Farber Cancer Institute 1 , Boston, Massachusetts
122 shared publications - Douglas A. Rubinson
Dana-Farber Cancer Institute 1 , Boston, Massachusetts
121 shared publications - Hersh Gupta
Dana-Farber Brigham and Women's Cancer Center, Department of Medical Oncology, Dana-Farber Cancer Institute and Harvard Medical School, Boston, Massachusetts
121 shared publications - Srivatsan Raghavan
Dana-Farber Cancer Institute 1 , Boston, Massachusetts
121 shared publications
Latest Publications
- Open Access
- 5 October 2026
- Published by Elsevier in SSRN Electronic Journal
NTRK fusion testing: Comparison across different assays in a series of 12 cases and pan-TRK immunohistochemistry screening in a prospective cohort
NTRK fusions are rare in cancers but are important to identify as they predict a strong response to targeted therapy. Many testing modalities are available to test for these fusions with varying performance. Pan-TRK immunohistochemistry (IHC) has been suggested as a potential screening tool due to its low cost. We retrieved 12 cases with NTRK fusions from our department archive and performed the following tests to compare their performance in detecting NTRK fusions: pan-TRK IHC, real-time polymerase chain reaction (PCR), amplicon-based RNA next generation sequencing (NGS), anchored multiplex PCR, hybrid capture RNA NGS and NTRK1 fluorescent in situ hybridization (FISH) (for cases with NTRK1 fusions). All the molecular assays showed largely concordant results except for NTRK1 FISH which showed poor sensitivity (only 2 out of 5 NTRK1 fusion cases were FISH positive). Pan-TRK IHC was positive in all cases except for a thyroid carcinoma with ETV6::NTRK3 fusion. We also performed pan-TRK IHC testing on a prospective pan-cancer cohort of 300 cases. Thirty-two cases (10.7%) were positive for pan-TRK IHC but only 1 tested positive for an NTRK fusion (ETV6::NTRK3) on molecular testing. We also found that using a cut-off value of ≥50% tumour cell staining for cytoplasmic/membranous pattern of staining and nuclear staining in any percentage of tumour cells compared to the conventional 1% threshold improved the sensitivity and specificity of pan-TRK IHC. In summary, an unselected pan-cancer cohort yielded only 0.3% (1/300) of NTRK fusions. RNA based molecular assays showed similar performance while NTRK1 FISH has poorer sensitivity.
- 5 October 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessNTRK fusion testing: Comparison across different assays in a series of 12 cases and pan-TRK immunohistochemistry screening in a prospective cohort
NTRK fusions are rare in cancers but are important to identify as they predict a strong response to targeted therapy. Many testing modalities are available to test for these fusions with varying performance. Pan-TRK immunohistochemistry (IHC) has been suggested as a potential screening tool due to its low cost. We retrieved 12 cases with NTRK fusions from our department archive and performed the following tests to compare their performance in detecting NTRK fusions: pan-TRK IHC, real-time polymerase chain reaction (PCR), amplicon-based RNA next generation sequencing (NGS), anchored multiplex PCR, hybrid capture RNA NGS and NTRK1 fluorescent in situ hybridization (FISH) (for cases with NTRK1 fusions). All the molecular assays showed largely concordant results except for NTRK1 FISH which showed poor sensitivity (only 2 out of 5 NTRK1 fusion cases were FISH positive). Pan-TRK IHC was positive in all cases except for a thyroid carcinoma with ETV6::NTRK3 fusion. We also performed pan-TRK IHC testing on a prospective pan-cancer cohort of 300 cases. Thirty-two cases (10.7%) were positive for pan-TRK IHC but only 1 tested positive for an NTRK fusion (ETV6::NTRK3) on molecular testing. We also found that using a cut-off value of ≥50% tumour cell staining for cytoplasmic/membranous pattern of staining and nuclear staining in any percentage of tumour cells compared to the conventional 1% threshold improved the sensitivity and specificity of pan-TRK IHC. In summary, an unselected pan-cancer cohort yielded only 0.3% (1/300) of NTRK fusions. RNA based molecular assays showed similar performance while NTRK1 FISH has poorer sensitivity.
- Open Access
- 18 June 2026
- Published by American Society of Clinical Oncology (ASCO) in JCO Oncology Practice
- p. OP2501191
Turnaround Time of Comprehensive Genomic Profiling in Lung Cancer and Other Solid Tumors
PURPOSE: Given the importance of timely comprehensive genomic profiling (CGP) for the treatment of cancer, this study aimed to provide generalizable estimates of turnaround time (TAT) for tissue-based CGP. Secondary aims were to identify TAT trends over time and explain variation based on clinical, demographic, and administrative factors. METHODS: Data for solid tumor samples profiled from 2018 through 2024 from a commercial laboratory were analyzed. Descriptive statistics were used to summarize and compare TAT between sites. Linear regression was used to assess for yearly trends and to measure associations between clinical, demographic, and administrative variables and TAT. RESULTS: A total of 271,574 solid tumor malignancies originating from 5,497 clinical sites were included in the analysis. The overall TAT (specimen collection to CGP results reporting) decreased from a median of 43 days (IQR, 31-72) to 32 days (IQR, 22-51) from 2018 to 2024. When comparing overall TAT for 2024 by clinical site, the shortest and longest overall TAT quintiles were 23.1 days (IQR, 20.6-24.6) and 46.5 days (IQR, 42.7-50.8), respectively ( P < .001). There was variability in timeliness of ordering between sites with a median time from biopsy to CGP ordering of 10.0 days (IQR, 8.0-12.0) versus 30.8 days (IQR, 26.6-35.0) for the shortest and longest quintiles, respectively. Factors associated with shorter ordering and overall TAT included, but were not limited to, higher clinical site ordering volumes, cancer type, and ordering method. CONCLUSION: Delay in the time from biopsy to CGP ordering was the largest contributor to protracted TAT and varied between clinical sites. Increasingly, the appropriate treatment of cancer is dependent on biomarker testing and delays in testing likely contribute to suboptimal patient care.
- 18 June 2026
- Published by American Society of Clinical Oncology (ASCO) in JCO Oncology Practice
- p. OP2501191
Open AccessTurnaround Time of Comprehensive Genomic Profiling in Lung Cancer and Other Solid Tumors
PURPOSE: Given the importance of timely comprehensive genomic profiling (CGP) for the treatment of cancer, this study aimed to provide generalizable estimates of turnaround time (TAT) for tissue-based CGP. Secondary aims were to identify TAT trends over time and explain variation based on clinical, demographic, and administrative factors. METHODS: Data for solid tumor samples profiled from 2018 through 2024 from a commercial laboratory were analyzed. Descriptive statistics were used to summarize and compare TAT between sites. Linear regression was used to assess for yearly trends and to measure associations between clinical, demographic, and administrative variables and TAT. RESULTS: A total of 271,574 solid tumor malignancies originating from 5,497 clinical sites were included in the analysis. The overall TAT (specimen collection to CGP results reporting) decreased from a median of 43 days (IQR, 31-72) to 32 days (IQR, 22-51) from 2018 to 2024. When comparing overall TAT for 2024 by clinical site, the shortest and longest overall TAT quintiles were 23.1 days (IQR, 20.6-24.6) and 46.5 days (IQR, 42.7-50.8), respectively ( P < .001). There was variability in timeliness of ordering between sites with a median time from biopsy to CGP ordering of 10.0 days (IQR, 8.0-12.0) versus 30.8 days (IQR, 26.6-35.0) for the shortest and longest quintiles, respectively. Factors associated with shorter ordering and overall TAT included, but were not limited to, higher clinical site ordering volumes, cancer type, and ordering method. CONCLUSION: Delay in the time from biopsy to CGP ordering was the largest contributor to protracted TAT and varied between clinical sites. Increasingly, the appropriate treatment of cancer is dependent on biomarker testing and delays in testing likely contribute to suboptimal patient care.
- Open Access
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Supplementary Data from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
SUPPLEMENTARY MATERIALS & METHODS Foundation Medicine Comprehensive Genomic Profiling
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Data from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
SUPPLEMENTARY MATERIALS & METHODS Foundation Medicine Comprehensive Genomic Profiling
- Open Access
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Supplementary Figure S10 from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
Supplementary Figure S10. HER2-Directed Therapy Response Modifying Alterations In Select Cancers (Related To Figure 4) Co-occurrence of alterations that have been reported to impact response to HER2-directed therapies (mAB, TKI, ADC) in tumors with ERBB2 amplification or ERBB2 activating mutations including TP53 MUT, PIK3CA MUT, CCNE1 AMP, RB1 MUT/DEL, PTEN MUT/DEL, ERBB3 AMP/MUT, EGFR MUT, MET AMP/Ex14 MUT, and IGF1R AMP. Patients with multiple ERBB2 alteration types (e.g., AMP + MUT) were excluded from the analysis. ADC, Antibody-Drug Conjugate; AMP, Amplification; mAB, Monoclonal Antibodies; MUT, Mutation (SNV, Indel); TKI, Tyrosine Kinase Inhibitor.
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure S10 from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
Supplementary Figure S10. HER2-Directed Therapy Response Modifying Alterations In Select Cancers (Related To Figure 4) Co-occurrence of alterations that have been reported to impact response to HER2-directed therapies (mAB, TKI, ADC) in tumors with ERBB2 amplification or ERBB2 activating mutations including TP53 MUT, PIK3CA MUT, CCNE1 AMP, RB1 MUT/DEL, PTEN MUT/DEL, ERBB3 AMP/MUT, EGFR MUT, MET AMP/Ex14 MUT, and IGF1R AMP. Patients with multiple ERBB2 alteration types (e.g., AMP + MUT) were excluded from the analysis. ADC, Antibody-Drug Conjugate; AMP, Amplification; mAB, Monoclonal Antibodies; MUT, Mutation (SNV, Indel); TKI, Tyrosine Kinase Inhibitor.
- Open Access
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Supplementary Table S1 from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
Supplementary Table S1. Histology Breakdown Of Top 5 ERBB2 MUT Cancer Types
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Table S1 from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
Supplementary Table S1. Histology Breakdown Of Top 5 ERBB2 MUT Cancer Types
- Open Access
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Supplementary Figure S6 from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
Supplementary Figure S6. ERBB2 Mutation Clonality Distribution of clonal fraction (VAF/TP) for ERBB2 mutation classes across select cancers. A dashed line at clonal fraction of 0.25 (25%) is shown as an estimated threshold for clonality1,2. ECD, Extracellular Domain; KD, Kinase Domain; TMD, Transmembrane Domain; TP, Computational Tumor Purity; VAF, Variant Allele Frequency. 1Sottoriva A, Graham TA. A pan-cancer signature of neutral tumor evolution. bioRxiv. 2015;014894. 2Bozic I, Gerold JM, Nowak MA. Quantifying Clonal and Subclonal Passenger Mutations in Cancer Evolution. PLoS Comput Biol. 2016;12:e1004731.
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure S6 from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
Supplementary Figure S6. ERBB2 Mutation Clonality Distribution of clonal fraction (VAF/TP) for ERBB2 mutation classes across select cancers. A dashed line at clonal fraction of 0.25 (25%) is shown as an estimated threshold for clonality1,2. ECD, Extracellular Domain; KD, Kinase Domain; TMD, Transmembrane Domain; TP, Computational Tumor Purity; VAF, Variant Allele Frequency. 1Sottoriva A, Graham TA. A pan-cancer signature of neutral tumor evolution. bioRxiv. 2015;014894. 2Bozic I, Gerold JM, Nowak MA. Quantifying Clonal and Subclonal Passenger Mutations in Cancer Evolution. PLoS Comput Biol. 2016;12:e1004731.
- Open Access
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Supplementary Figure S4 from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
Supplementary Figure S4. ERBB3 Mutation Landscape Of Select Cancers Shown for each cancer type are (Left) the distribution of mutations across key HER3 protein domains, (Middle) lollipop plot showing the location and incidence of mutations across the HER3 protein (NM_001982) with the 5 most frequently mutated codons labeled, and (Right) plot showing distribution of mutations among the 10 most frequently mutated codons. The number of mutations is indicated. AA, Amino Acid; ECD, Extracellular Domain; KD, Kinase Domain; TMD, Transmembrane Domain.
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure S4 from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
Supplementary Figure S4. ERBB3 Mutation Landscape Of Select Cancers Shown for each cancer type are (Left) the distribution of mutations across key HER3 protein domains, (Middle) lollipop plot showing the location and incidence of mutations across the HER3 protein (NM_001982) with the 5 most frequently mutated codons labeled, and (Right) plot showing distribution of mutations among the 10 most frequently mutated codons. The number of mutations is indicated. AA, Amino Acid; ECD, Extracellular Domain; KD, Kinase Domain; TMD, Transmembrane Domain.
- Open Access
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Supplementary Figure S3 from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
Supplementary Figure S3. ERBB2 Mutation Landscape Of Select Cancers Shown for each cancer type are (Left) the distribution of mutations across key HER2 protein domains, (Middle) lollipop plot showing the location and incidence of mutations across the HER2 protein (NM_004448) with the 5 most frequently mutated codons labeled, and (Right) plot showing distribution of mutations among the 10 most frequently mutated codons. The number of mutations is indicated. AA, Amino Acid; ECD, Extracellular Domain; KD, Kinase Domain; TMD, Transmembrane Domain.
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure S3 from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
Supplementary Figure S3. ERBB2 Mutation Landscape Of Select Cancers Shown for each cancer type are (Left) the distribution of mutations across key HER2 protein domains, (Middle) lollipop plot showing the location and incidence of mutations across the HER2 protein (NM_004448) with the 5 most frequently mutated codons labeled, and (Right) plot showing distribution of mutations among the 10 most frequently mutated codons. The number of mutations is indicated. AA, Amino Acid; ECD, Extracellular Domain; KD, Kinase Domain; TMD, Transmembrane Domain.
- Open Access
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Supplementary Table S2 from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
Supplementary Table S2. Histology Breakdown Of Top 5 ERBB3 MUT Cancer Types
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Table S2 from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
Supplementary Table S2. Histology Breakdown Of Top 5 ERBB3 MUT Cancer Types
- Open Access
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Supplementary Figure S11 from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
Supplementary Figure S11. Pan-Tumor Landscape Of ERBB2 Activating Alterations Detected In The MSKCC Clinicogenomic Cohort a) Prevalence of ERBB2 activating alterations (known or likely functional significance) across solid tumors in the MSKCC clinicogenomic cohort. ‘Multiple’ includes patients with AMP + MUT, AMP + RE, MUT + RE, AMP + MUT + RE and patients with >1 MUT. b) Cancer type distribution of ERBB2 MUT tumors. AMP, Amplification (CN ≥ 6); MUT, Mutation (SNV, Indel); RE, Rearrangement.
- 17 June 2026
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure S11 from Diverse ERBB2/ERBB3 Activating Alterations and Coalterations Have Implications for HER2/3-Targeted Therapies across Solid Tumors
Supplementary Figure S11. Pan-Tumor Landscape Of ERBB2 Activating Alterations Detected In The MSKCC Clinicogenomic Cohort a) Prevalence of ERBB2 activating alterations (known or likely functional significance) across solid tumors in the MSKCC clinicogenomic cohort. ‘Multiple’ includes patients with AMP + MUT, AMP + RE, MUT + RE, AMP + MUT + RE and patients with >1 MUT. b) Cancer type distribution of ERBB2 MUT tumors. AMP, Amplification (CN ≥ 6); MUT, Mutation (SNV, Indel); RE, Rearrangement.