Compositions, methods and kits for diagnosis of lung cancer
Inventors
Kearney, Paul Edward • Fang, Kenneth Charles • Li, Xiao-Jun • Hayward, Clive
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-11193935-B2
Publication Date
2021-12-07
Expiration Date
Abstract
The present invention provides methods for identifying biomarker proteins that exhibit differential expression in subjects with a first lung condition versus healthy subjects or subjects with a second lung condition. The present invention also provides compositions comprising these biomarker proteins and methods of using these biomarker proteins or panels thereof to diagnose, classify, and monitor various lung conditions. The methods and compositions provided herein may be used to diagnose or classify a subject as having lung cancer or a non-cancerous condition, and to distinguish between different types of cancer (e.g., malignant versus benign, SCLC versus NSCLC).
Core Innovation
The invention provides a method of scoring a pulmonary nodule in a subject by assaying expression of a plurality of proteins comprising ALDOA_HUMAN, FRIL_HUMAN, LG3BP_HUMAN, TSP1_HUMAN and COIA1_HUMAN from a biological sample. Protein expression is determined from proteolytically digested fragments using selected reaction monitoring mass spectrometry by contacting the fragments with labeled synthetic peptide fragments and detecting peptide transitions comprising at least ALQASALK (SEQ ID NO:25) (401.25, 617.4), AVGLAGTFR (SEQ ID NO:26) (446.26, 721.4), GFLLLASLR (SEQ ID NO:27) (495.31, 559.4), LGGPEAGLGEYLFER (SEQ ID NO:28) (804.4, 1083.6), and VEIFYR (SEQ ID NO:29) (413.73, 598.3).
A score is calculated from the protein expression determined from the biological sample, wherein the score is determined as P_s = 1/[1+exp(−α−Σ_i=1^5 β_i*{hacek over (I)}_i,s −γ*{hacek over (I)}_COIA1*{hacek over (I)}_FRIL)]. The {hacek over (I)}_i,s values are Box-Cox transformed and normalized intensity of transition i in the sample. The logistic regression formulation includes a panel-specific constant α, logistic regression coefficients β_i, and a coefficient γ for an interaction term comprising FRIL and COIA1.
The invention further provides a method of determining that a pulmonary nodule in a subject is not lung cancer by producing peptide fragments from a panel of proteins present in a biological sample using a proteolytic enzyme. The produced peptide fragments are combined with labeled, synthetic peptide fragments corresponding to the produced peptide fragments, and selected reaction monitoring mass spectrometry is performed to measure abundance by detecting the specified peptide transitions. A score is calculated using the same form of interaction-term logistic regression, and lung cancer is ruled out when the score is closer to 0.0 than to 1.0.
Claims Coverage
The document includes two independent claims. Across these claims, the inventive coverage centers on a protein-panel SRM measurement of specified peptide transitions and a logistic-regression score that includes a FRIL–COIA1 interaction term, with a decision rule for ruling out lung cancer using the score proximity to 0.0.
Protein-panel SRM scoring for pulmonary nodules
Assaying protein expression of a plurality of proteins comprising at least ALDOA_HUMAN, FRIL_HUMAN, LG3BP_HUMAN, TSP1_HUMAN and COIA1_HUMAN from a biological sample obtained from the subject by selected reaction monitoring mass spectrometry of proteolytically digested fragments, by contacting the fragments to labeled synthetic peptide fragments and detecting peptide transitions comprising at least ALQASALK (SEQ ID NO:25) (401.25, 617.4), AVGLAGTFR (SEQ ID NO:26) (446.26, 721.4), GFLLLASLR (SEQ ID NO:27) (495.31, 559.4), LGGPEAGLGEYLFER (SEQ ID NO:28) (804.4, 1083.6), and VEIFYR (SEQ ID NO:29) (413.73, 598.3); calculating a score from the protein expression using P_s = 1/[1+exp(−α−Σ_i=1^5 β_i*{hacek over (I)}_i,s −γ*{hacek over (I)}_COIA1*{hacek over (I)}_FRIL)], where {hacek over (I)}_i,s is Box-Cox transformed and normalized intensity, and γ is a coefficient for the interaction term.
Ruling out lung cancer using an interaction-term score
Determining that a pulmonary nodule in a subject is not lung cancer by contacting a biological sample with a proteolytic enzyme to produce peptide fragments from a panel comprising ALDOA_HUMAN, FRIL_HUMAN, LG3BP_HUMAN, TSP1_HUMAN and COIA1_HUMAN; combining the produced peptide fragments with labeled, synthetic peptide fragments corresponding to the produced peptide fragments; performing selected reaction monitoring mass spectrometry to measure abundance by detecting peptide transitions comprising at least ALQASALK (SEQ ID NO:25) (401.25, 617.4), AVGLAGTFR (SEQ ID NO:26) (446.26, 721.4), GFLLLASLR (SEQ ID NO:27) (495.31, 559.4), LGGPEAGLGEYLFER (SEQ ID NO:28) (804.4, 1083.6), and VEIFYR (SEQ ID NO:29) (413.73, 598.3); calculating a score using P_s = 1/[1+exp(−α−Σ_i=1^5 β_i*{hacek over (I)}_i,s −γ*{hacek over (I)}_COIA1*{hacek over (I)}_FRIL)], where γ is a coefficient for the interaction term comprising peptide fragments from FRIL_HUMAN and COIA1_HUMAN; ruling out lung cancer if the score is closer to 0.0 than to 1.0.
The claims jointly cover SRM-based detection of specified peptide transitions from a five-protein panel combined with a logistic-regression scoring model that includes a coefficient γ for an interaction term between FRIL and COIA1, with a decision rule for ruling out lung cancer based on whether the score is closer to 0.0 than to 1.0.
Stated Advantages
Documented Applications
Determining that a pulmonary nodule in a subject is not lung cancer using the disclosed scoring approach.
Interested in licensing this patent?