Method and system for determining whether a drug will be effective on a patient with a disease

• US 7,736,905 B2
• Filed: 03/31/2006
• Issued: 06/15/2010
• Est. Priority Date: 03/31/2006
• Status: Active Grant

• Current Assignee: Biodesix, Inc.
• Original Assignee: Biodesix, Inc.
• Inventors: Tsypin, Maxim, Grigorieva, Julia, Roder, Heinrich
• Primary Examiner(s): Gakh; Yelena G
• Assistant Examiner(s): XU, XIAOYUN
• Application Number:
  US11/396,328
• Publication Number:
  US 20070231921A1
• Time in Patent Office: 1,537 Days
• Field of Search: None
• US Class Current: 436/173
• CPC Class Codes: A61P 35/00 Antineoplastic agents
• CPC Class Codes: G06F 18/2115 by evaluating different subsets according to an optimisation criterion, e.g. class separability, forward selection or backward elimination
• CPC Class Codes: G06F 18/24147 Distances to closest patterns, e.g. nearest neighbour classification
• CPC Class Codes: G06V 20/69 Microscopic objects, e.g. biological cells or cellular parts
• CPC Class Codes: G06V 2201/03 Recognition of patterns in medical or anatomical images
• CPC Class Codes: G16B 20/00 ICT specially adapted for functional genomics or proteomics, e.g. genotype-phenotype associations
• CPC Class Codes: G16B 40/00 ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
• CPC Class Codes: G16B 40/20 Supervised data analysis
• CPC Class Codes: Y02A 90/10 Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
• CPC Class Codes: Y10T 436/24 Nuclear magnetic resonance, electron spin resonance or other spin effects or mass spectrometry

Abstract

A process of determining whether a patient with a disease or disorder will be responsive to a drug, used to treat the disease or disorder, including obtaining a test spectrum produced by a mass spectrometer from a serum produced from the patient. The test spectrum may be processed to determine a relation to a group of class labeled spectra produced from respective serum from other patients having the or similar clinical stage same disease or disorder and known to have responded or not responded to the drug. Based on the relation of the test spectrum to the group of class labeled spectra, a determination may be made as to whether the patient will be responsive to the drug.

Claims

1. A method of identifying a non-small cell lung cancer patient as being likely to benefit from treatment with gefitinib or erlotinib targeting an epidermal growth factor receptor pathway or not likely to benefit from treatment with gefitinib or erlotinib, comprising the steps of:

a) obtaining a mass spectrum from a blood-based sample from the non-small cell lung cancer patient;

b) performing one or more predefined pre-processing steps on the mass spectrum obtained in step a);

c) obtaining integrated intensity values of selected features in said spectrum at one or more predefined m/z ranges after the pre-processing steps on the mass spectrum in step b) have been performed;

d) using the values obtained in step c) in a classification algorithm using a training set comprising class-labeled spectra produced from blood-based samples from other on-small cell lung cancer patients to identify the patient as being either likely or not likely to benefit from gefitinib or erlotinib.

2. The method of claim 1, wherein said one or more predefined m/z ranges comprise one or more m/z ranges selected from the group of m/z ranges consisting of:

5732 to 5795

5811 to 5875

6398 to 6469

11376 to 11515

11459 to 11599

11614 to 11756

11687 to 11831

11830 to 11976

12375 to 12529

23183 to 23525

23279 to 23622

65902 to 67502.

3. The method of claim 2, wherein step c) obtains integrated intensity values from said spectrum at least 8 of said m/z ranges in the group.

4. The method of claim 2, wherein step c) obtains integrated intensity values from said spectrum at all of the m/z ranges in the group.

5. The method of claim 1, wherein the classification algorithm comprises a K-nearest neighbors classification algorithm.

6. The method of claim 5, wherein the K-nearest neighbors classification algorithm comprises a probabilistic classification algorithm.

7. The method of claim 1, wherein the one or more pre-processing steps includes a step of subtracting background contained in the spectrum.

8. The method of claim 7, wherein the step of subtracting background is performed using a robust asymmetric estimate for the background contained in the spectrum.

9. The method of claim 7, wherein the one or more pre-processing steps includes a step of normalizing the background subtracted spectrum.

10. The method of claim 9, wherein the step of normalizing the background subtracted spectrum comprises performing a partial ion current normalization.

11. The method of claim 9, wherein the step of normalizing the background subtracted spectrum comprises performing a total ion current normalization.

12. The method of claim 7, wherein the one or more pre-processing steps includes a step of aligning the background subtracted spectrum to a predefined mass scale.

13. The method of claim 1, wherein the mass spectrum is obtained from a MALDI mass spectrometer.

14. Apparatus configured to identify a non-small cell lung cancer patient as being likely to benefit from treatment with gefitinib or erlotinib targeting an epidermal growth factor receptor pathway or not likely to benefit from treatment with gefitinib or erlotinib, comprising:

a storage device storing a mass spectrum of a blood-based sample from the non-small cell lung cancer non-small cell lung cancer patient, and

a processor executing software, instructions configured to

a) obtain integrated intensity values of features in said mass spectrum at one or more m/z ranges, the m/z ranges selected from the group of m/z ranges consisting of;

5732 to 5795

5811 to 5875

6398 to 6469

11376 to 11515

11459 to 11599

11614 to 11756

11687 to 11831

11830 to 11976

12375 to 12529

23183 to 23525

23279 to 23622

65902 to 67502; and

b) apply a classification algorithm operating on the values of the features in the spectrum at the selected one or more m/z ranges and using a training set comprising class-labeled spectra produced from blood-based samples from other non-small cell lung cancer patients to identify the patient as being either likely or not likely to benefit from gefitinib or erlotinib targeting an epidermal growth factor receptor pathway.

15. The apparatus of claim 14, wherein the instructions obtains integrated intensity values at least eight of the m/z ranges in the group.

16. The apparatus of claim 15, wherein the instructions obtains integrated intensity values at all of the m/z ranges in the group.

17. The apparatus of claim 14, wherein the classification algorithm comprises a K-nearest neighbor classification algorithm.

18. The apparatus of claim 17, wherein the K-nearest neighbor classification algorithm comprises a probabilistic K-nearest neighbor classification algorithm.