U.S. patent application number 14/212567 was filed with the patent office on 2014-07-17 for method for predicting whether a cancer patient will not benefit from platinum-based chemotherapy agents.
The applicant listed for this patent is Biodesix, Inc.. Invention is credited to Julia Grigorieva, Heinrich Roder, Joanna Roder.
Application Number | 20140200825 14/212567 |
Document ID | / |
Family ID | 48045747 |
Filed Date | 2014-07-17 |
United States Patent
Application |
20140200825 |
Kind Code |
A1 |
Roder; Joanna ; et
al. |
July 17, 2014 |
Method for predicting whether a cancer patient will not benefit
from platinum-based chemotherapy agents
Abstract
A testing method for identification whether a cancer patient is
a member of a group or class of cancer patients that are not likely
to benefit from administration of a platinum-based chemotherapy
agent, e.g., cisplatin, carboplatin or analogs thereof, either
alone or in combination with other non-platinum chemotherapy
agents, e.g., gemcitabine and paclitaxel. This identification can
be made in advance of treatment. The method uses a mass
spectrometer obtaining a mass spectrum of a blood-based sample from
the patient, and a computer operating as a classifier and using a
stored training set comprising class-labeled spectra from other
cancer patients.
Inventors: |
Roder; Joanna; (Steamboat
Springs, CO) ; Roder; Heinrich; (Steamboat Springs,
CO) ; Grigorieva; Julia; (Steamboat Springs,
CO) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Biodesix, Inc. |
Boulder |
CO |
US |
|
|
Family ID: |
48045747 |
Appl. No.: |
14/212567 |
Filed: |
March 14, 2014 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
13836064 |
Mar 15, 2013 |
8718996 |
|
|
14212567 |
|
|
|
|
61668077 |
Jul 5, 2012 |
|
|
|
Current U.S.
Class: |
702/19 |
Current CPC
Class: |
G16H 50/50 20180101;
G01N 2800/52 20130101; G16C 20/70 20190201; G01N 33/6848 20130101;
H01J 49/26 20130101; G01N 33/57488 20130101; H01J 49/0027 20130101;
Y02A 90/10 20180101; G16H 50/20 20180101 |
Class at
Publication: |
702/19 |
International
Class: |
G06F 19/00 20060101
G06F019/00; H01J 49/26 20060101 H01J049/26 |
Claims
1. A method for guiding treatment of a cancer patient, comprising
the steps of: a) obtaining a blood-based sample from the patient;
b) obtaining a mass spectrum of the blood-based sample with the aid
of a mass spectrometer; c) in a programmed computer, (1) performing
predefined pre-processing steps on the mass spectrum, (2) obtaining
integrated intensity values of selected features at one or more
predefined m/z ranges in the mass spectrum after the pre-processing
steps are performed and (3) comparing the integrated intensity
values with a training set comprising class-labeled mass spectra
from blood-based samples of other cancer patients to thereby
classifying the mass spectrum with a class label, and d) if the
class label is Poor or the equivalent, the patient is predicted to
not benefit from treatment in the form of administration of a
platinum-based chemotherapy agent and is thereby guided towards a
treatment regimen not containing platinum agents.
2. The method of claim 1, wherein in step d) if the class label is
Poor or the equivalent, the patient is predicted to not benefit
from treatment in the form of administration of a combination of a
non-platinum chemotherapy agent and a platinum-based chemotherapy
agent and the patient is guided towards treatment regimen not
containing platinum agents.
3. The method of claim 1, wherein the training set comprises
class-labeled spectra from cancer patients, the class labels
indicating whether the patients obtained benefit from treatment by
administration of an epidermal growth factor receptor inhibitor,
the class label Poor or the equivalent assigned to members in the
training set for those cancer patients that did not benefit from
the administration of an epidermal growth factor receptor
inhibitor, the class label Good or the equivalent assigned to
members in the training set for those cancer patients that did
benefit from the administration of an epidermal growth factor
receptor inhibitor.
4. The method of claim 3, wherein the class-labeled spectra in the
training set are obtained from non-small-cell lung cancer (NSCLC)
patients.
5. The method of claim 1, wherein the platinum-based chemotherapy
agent comprises cisplatin, carboplatin or analogs thereof.
6. The method of any of claim 1, wherein the treatment regimen not
containing platinum agents comprises gemcitabine or the analog or
docetaxel or the analog, and wherein the platinum-based
chemotherapy agent comprises cisplatin or analogs thereof.
7. The method of claim 1, wherein the cancer patient comprises a
patient having either ovarian cancer or non-small cell lung
cancer.
8. The method of claim 7, wherein the platinum-based chemotherapy
agent comprises cisplatin, carboplatin or analogs thereof.
9. The method of claim 7, wherein the treatment regimen not
containing platinum agents comprises gemcitabine or the analog or
docetaxel or the analog, and wherein the platinum-based
chemotherapy agent comprises cisplatin or analogs thereof.
10. The method of claim 7, wherein the training set comprises
class-labeled spectra from non-small-cell lung cancer (NSCLC)
patients, the class labels indicating whether the patients obtained
benefit from treatment by administration of an epidermal growth
factor receptor inhibitor, the class label Poor assigned to members
in the training set for those NSCLC patients that did not benefit
from the administration of an epidermal growth factor receptor
inhibitor.
11. The method of claim 1, wherein the cancer patient has either
colorectal cancer or head and neck cancer.
12. A method for guiding treatment of a patient having either
non-small-cell lung cancer, ovarian cancer, head and neck cancer,
breast cancer or colorectal cancer, comprising the steps of: a)
obtaining a blood-based sample from the patient; b) obtaining a
mass spectrum of the blood-based sample with the aid of a mass
spectrometer; c) in a programmed computer, (1) performing
predefined pre-processing steps on the mass spectrum, (2) obtaining
integrated intensity values of selected features at one or more
predefined m/z ranges in the mass spectrum after the pre-processing
steps are performed and (3) comparing the integrated intensity
values with a training set comprising class-labeled mass spectra
from blood-based samples of other cancer patients to thereby
classifying the mass spectrum with a class label, wherein the
wherein the training set comprises class-labeled spectra from
cancer patients, the class labels indicating whether the patients
obtained benefit from treatment by administration of an epidermal
growth factor receptor inhibitor, the class label Poor or the
equivalent assigned to members in the training set for those cancer
patients that did not benefit from the administration of an
epidermal growth factor receptor inhibitor, the class label Good or
the equivalent assigned to members in the training set for those
cancer patients that did benefit from the administration of an
epidermal growth factor receptor inhibitor, and d) if the class
label for the blood-based sample of the patient classified by the
programmed computer in step c) 3) is Poor or the equivalent, the
patient is predicted to not benefit from treatment in the form of
administration of a platinum-based chemotherapy agent and is
thereby guided towards a treatment regimen not containing platinum
agents.
13. The method of claim 12, wherein the treatment regimen not
containing platinum agents comprises gemcitabine or the analog or
docetaxel or the analog, and wherein the platinum-based
chemotherapy agent comprises cisplatin or analogs thereof.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application claims priority as a continuation of U.S.
application Ser. No. 13/836,064 filed Mar. 15, 2013, now allowed.
Ser. No. 13/836,064 filed Mar. 15, 2013 claims priority benefits
under 35 U.S.C. .sctn.119(e) to U.S. provisional application Ser.
No. 61/668,077 filed Jul. 5, 2012, the entire content of which is
incorporated by reference herein.
BACKGROUND
[0002] This invention relates generally to the field of methods for
guiding the treatment of cancer patients with chemotherapy agents.
More particularly, this invention relates to a method of
predicting, in advance of initiating treatment, whether a patient
is a member of a class of patients that are unlikely to benefit
from administration of platinum-based chemotherapy agents, such as
cisplatin, carboplatin, and analogs thereof. The methods of this
disclosure use mass spectral data obtained from a blood-based
sample of the patient, a computer configured as a classifier
operating on the mass spectral data, and a training set comprising
class-labeled spectra from other cancer patients.
[0003] The assignee of the present invention, Biodesix, Inc., has
developed a test known as VeriStrat. One of the uses of VeriStrat
is that it predicts whether Non-Small Cell Lung Cancer (NSCLC)
patients are likely or not likely to benefit from treatment with
drugs targeting the Epidermal Growth Factor Receptor (EGFR)
pathway, e.g. EGFR inhibitors such as erlotinib. The test is
described in U.S. Pat. No. 7,736,905, the content of which is
incorporated by reference herein. Additional applications of the
test are also described in U.S. Pat. Nos. 7,858,390; 7,858,389 and
7,867,775, the contents of which are incorporated by reference
herein.
[0004] In brief, the VeriStrat test is based on serum and/or plasma
samples of cancer patients. Through a combination of MALDI-TOF mass
spectrometry and data analysis algorithms implemented in a
computer, the commercial version of the test compares a set of
eight integrated peak intensities at predefined m/z ranges in the
mass spectrum of the patient sample (after pre-processing steps are
performed) with those from a training cohort, and generates a class
label for the patient sample: either VeriStrat Good, VeriStrat
Poor, or VeriStrat "indeterminate." In multiple clinical validation
studies it has been shown that patients, whose pre-treatment
serum/plasma is classified as VeriStrat Good, have significantly
better outcome when treated with epidermal growth factor receptor
inhibitor drugs than those patients whose sample is classified as
VeriStrat Poor. In a few cases (less than 2%) no determination can
be made, resulting in a VeriStrat indeterminate label.
[0005] The applicants have further discovered that the VeriStrat
test is also predictive for whether head and neck squamous cell
carcinoma and colorectal cancer patients are likely to have better
or worse outcomes from treatment with certain anti-cancer drugs, as
described in U.S. Pat. Nos. 8,024,282; 7,906,342; 7,879,620;
7,867,775; 7,858,390; 7,858,389 and 7,736,905.
[0006] In pending U.S. patent application Ser. No. 13/356,730 filed
Jan. 24, 2012, we have described how the VeriStrat test also
predicts that breast cancer patients having the VeriStrat Poor
signature are unlikely to obtain good clinical outcomes from
endocrine therapy alone, including for example an aromatase
inhibitor such as letrozole. In that document, we have also
described how hormone receptor positive breast cancer patients,
regardless of their HER2 status, which have the VeriStrat Poor
signature are likely to benefit from administration of a
combination treatment comprising administration of a targeted
anti-cancer drug in addition to an endocrine therapy drug.
[0007] Platinum-based chemotherapy drugs, including cisplatin
(cis-PtCl.sub.2(NH.sub.3).sub.2) and analogs thereof, are used to
treat various kinds of cancers, including sarcomas, lymphomas, and
carcinomas. The drug reacts in vivo, binding to and causing
crosslinking of DNA, which interferes with cell division by
mitosis, ultimately triggering apoptosis (programmed cell death).
Cisplatin combination therapy is a cornerstone of treatment of many
cancers. However, while initial platinum responsiveness is high,
some patients do not respond to treatment, and the majority of
cancer patients will eventually relapse with cisplatin resistant
disease. There is a need in the art to be able to predict in
advance of treatment whether a patient will not obtain benefit from
cisplatin (alone or in combination treatment) as if this
determination can be made at an early stage in treatment the
patient can be guided towards other non-platinum therapies that are
more likely to provide clinical benefit. This invention meets that
need.
SUMMARY
[0008] We have discovered that our mass spectral test described in
the above-referenced documents can be used to identify whether a
cancer patient is a member of a class of cancer patients that are
not likely to benefit from platinum-based chemotherapy agents such
as cisplatin, carboplatin, or analogs thereof, either alone or in
combination with other non-platinum-based anticancer agents, e.g.,
anti-mitotic agents such as docetaxel or nucleoside analogs such as
gemcitabine. Such patients are identified when the classifier
assigns the Poor class label to the sample's mass spectrum.
[0009] These discoveries are implemented in the form of practical,
concrete and useful tests. One example is a testing method which
identifies, in advance of treatment, whether a particular cancer
patient (e.g., NSCLC, ovarian cancer patient) is a member of a
group or class of cancer patients that are not likely to benefit
from administration of a platinum-based chemotherapy agent, e.g.,
cisplatin, carboplatin or analogs thereof, either alone or in
combination with other non-platinum chemotherapy agents. If the
patient is identified as being a member of this group (by virtue of
having the Poor class label), the patient is directed towards an
anticancer treatment regimen that does not contain platinum-based
chemotherapy agents. Accordingly, the patient is steered at the
outset away from treatments that are not likely to benefit the
patient and towards other treatment options that are more likely to
provide some clinical benefit.
[0010] One embodiment of the test is in the form of a method for
guiding treatment of a NSCLC patient. The method includes the steps
of: a) obtaining a blood-based sample from the patient; b)
obtaining a mass-spectrum of the blood-based sample with the aid of
a mass spectrometer; and c) in a programmed computer, performing 1)
predefined pre-processing steps on the mass spectrum (e.g.,
background subtraction, normalization and spectral alignment), 2)
obtaining integrated intensity values of selected features at one
or more predefined m/z ranges in the spectrum after the
pre-processing steps are performed, and 3) comparing the integrated
intensity values with a training set comprising class-labeled
spectra from other cancer patients and thereby classifying the mass
spectrum with a class label. This step 3) is a classification step
and can be implemented in various ways, including with the aid of a
K-nearest neighbor classification algorithm implemented in the
computer as described below. The method further includes step d),
wherein if the class label is Poor or the equivalent, the patient
is predicted to not benefit from treatment in the form of
administration of a platinum-based chemotherapy agent and is
thereby guided towards a treatment regimen not containing platinum
agents.
[0011] In a variation of the method, at step d) if the class label
is Poor or the equivalent, the patient is predicted to not benefit
from treatment in the form of administration of a combination of a
non-platinum chemotherapy agent and a platinum-based chemotherapy
agent and the patient is guided towards treatment regimen not
containing platinum agents.
[0012] While we describe specific data supporting this invention in
the context of non-small cell lung cancer and ovarian cancer, our
previous work in this field across many different types of cancer
(see the above-cited patent documents and the application of Julia
Grigorieva et al., U.S. Ser. No. 12/932,295 filed Feb. 22, 2011)
leads us to conclude that the VeriStrat test is also predictive for
cancer patient non-benefit from platinum-based chemotherapy in
general; i.e., that those cancer patients having the VeriStrat Poor
signature generally are not likely to benefit from platinum-based
chemotherapy agents. That is, the method recited above can be
performed on blood-based samples of cancer patients and if the
class label is Poor or the equivalent the patient is predicted to
not benefit from treatment in the form of administration of a
platinum chemotherapy agent, a combination of platinum chemotherapy
agents, or combinations of a non-platinum chemotherapy agent and a
platinum-based chemotherapy agent, and the patient is thereby
guided towards a treatment regimen not containing platinum
agents.
BRIEF DESCRIPTION OF THE DRAWINGS
[0013] FIG. 1 is a Kaplan-Meier plot of progression free survival
(PFS) by VeriStrat group for NSCLC patients treated with
gemcitabine alone.
[0014] FIG. 2 is a Kaplan-Meier plot of overall survival (OS) by
VeriStrat group for NSCLC patients treated with gemcitabine
alone.
[0015] FIG. 3 is a Kaplan-Meier plot of time to progression (TTP)
by VeriStrat group for advanced NSCLC patients treated with
cisplatin as first line treatment. Most of the patients in the data
shown in FIG. 3 were administered gemcitabine in addition to
cisplatin.
[0016] FIG. 4 is a Kaplan-Meier plot of OS by VeriStrat group for
advanced NSCLC patients treated with cisplatin as first line
treatment. Most of the patients in the data shown in FIG. 4 were
administered gemcitabine in addition to cisplatin.
[0017] FIG. 5 is a flow-chart of a process for conducting mass
spectrometry on a blood-based sample and classifying the spectrum
as Good or Poor using a class-labeled training set. If the patient
sample is classified under the method as Poor or the equivalent,
the patient is predicted to not benefit from platinum-based
chemotherapy agents and is guided towards a treatment regimen not
containing platinum-based chemotherapy agents.
[0018] FIG. 6 is a Kaplan-Meier plot of overall survival for a
subgroup analysis of one of the cohorts studied in Taguchi et al.,
Mass spectrometry to classify non-small-cell lung cancer patients
for clinical outcome after treatment with epidermal growth factor
receptor tyrosine kinase inhibitors: a multicohort
cross-institutional study. J Natl Cancer Inst 2007; 99:838-46,
showing that NSCLC patients treated with docetaxel exhibit similar
overall survival rates regardless of VeriStrat label for such
patients.
[0019] FIGS. 7A and 7B are Kaplan-Meier plots of disease free
survival (DFS) and overall survival (OS), respectively, for ovarian
cancer patients who, after surgery, were treated with carboplatin
and paclitaxel. The results show that patients classified as
VeriStrat Poor have shorter disease free survival and overall
survival than VeriStrat Good patients when treated with the
platinum-based chemotherapy.
DETAILED DESCRIPTION
I. Predictive Tests for Platinum-Based Chemotherapy Agents, Related
Classifiers and Systems
[0020] The discovery that the VeriStrat test is predictive of
cancer patients not benefitting from platinum based chemotherapy
agents resulted from our analysis of VeriStrat separation of
survival curves in patients enrolled in an ovarian cancer study,
and in two separate lung cancer studies, one of which involved a
study of elderly NSCLC patients treated with gemcitabine and a
second study of advanced NSCLC patients receiving cisplatin-based
chemotherapy as first line treatment. In this second study, most of
the patients were known to have received cisplatin in combination
with gemcitabine. The first lung cancer study will be referred to
as "the LCCC0512 study" and the second lung cancer study will be
referred to as "the Italian study." Our results from analysis of
the ovarian cancer study and both lung cancer studies are described
below.
[0021] In both the ovarian cancer study and the lung cancer
studies, our work involved obtaining plasma or serum samples from
patients enrolled in the studies, obtaining mass spectra of such
samples, performing certain pre-processing steps on the spectra,
and then subjecting the spectra to a classifier we have developed
and described in our U.S. Pat. No. 7,736,905. The classifier
assigned a class label to the samples, either Good or Poor or in a
few instances "indeterminate." The class labels for the patient
samples were assigned using a K-nearest neighbor (K-NN)
classification algorithm based on a comparison of the spectra,
after preprocessing and calculation of integrated intensity values
of selected features at one or more predefined m/z ranges in the
spectra, with a training set of class-labeled spectra from other
cancer patients.
[0022] In our work, the training set used by the classification
algorithm consisted of class-labeled spectra from a population of
non-small cell lung cancer patients, with the class-label for a
spectrum in the training set being Good if the associated patient
benefitted from administration of an epidermal growth factor
receptor inhibitor (EGFR-I) in the treatment of NSCLC, whereas the
class label Poor was assigned to spectra for patients who did not
benefit from such drugs. This training set and the classifier were
the subject of extensive validation studies. The method of
conducting our mass-spectral testing and classification of
blood-based samples is explained in further detail below.
[0023] The LCCC0512 study was a randomized phase II trial of
first-line treatment with gemcitabine, erlotinib or the combination
in elderly (over 70 years old) patients with advanced non-small
cell lung cancer (NSCLC). Patients had stage IIIB or IV NSCLC and
ECOG performance status of 0-2. Patients were randomized to receive
either a maximum of 4 cycles of gemcitabine (1200 mg/m.sup.2 days
1,8 every 21 days), 150 mg erlotinib daily, or a maximum of 4
cycles of gemcitabine (1000 mg/m.sup.2 days 1,8 every 21 days) with
concurrent 100 mg erlotinib daily. Primary endpoint was
progression-free survival at 6 months. Patients randomized to
receive gemcitabine monotherapy were encouraged to continue on
study to receive erlotinib after progression. A total of 146
patients were enrolled. No significant differences in PFS or OS
were found between treatment arms.
[0024] Pretreatment serum or plasma samples from patients in the
LCCC0512 trial were available for VeriStrat testing. 124 samples
were received. Sixty four samples were classified as VeriStrat
Good, 39 as VeriStrat Poor and 7 as Indeterminate
(undefined/equivocal classification). Of the remaining samples 13
could not be processed due to hemolysis and data could not be
collected from one sample. For 5 samples where VeriStrat testing
was performed, the samples could not be matched to clinical data
from the trial. This left 98 samples with VeriStrat classification
of Good or Poor and clinical data for statistical analysis. The
majority of the samples tested were plasma, but where plasma was
not available for a few patients, serum samples were used. Of the
samples on which VeriStrat testing yielded a result of Good or Poor
only 5 were identified as serum.
[0025] Survival analysis was carried out using SAS Enterprise Guide
(SAS 9.2). Difference between groups was assessed using log-rank p
values. Both univariate and multivariate hazard ratios were
evaluated using Cox proportional hazard models. Kaplan-Meier plots
and p values for contingency tables were generated using PRISM
(GraphPad). For purposes of this disclosure, the data of interest
pertain to the "gemcitabine" arm.
[0026] FIG. 1 is a Kaplan-Meier plot of progression free survival
(PFS) by VeriStrat ("VS") group for NSCLC patients in the LCCC0512
study treated with gemcitabine alone. Both VeriStrat groups have
similar PFS (log-rank p=0.67; hazard ratio (HR)=1.21, 95%
Confidence Interval (CI): 0.51-2.88). The median PFS is 137 days
(95% CI: 13-210 days) in the VeriStrat Poor group and 133 days (95%
CI: 37-164 days) in the VeriStrat Good group.
[0027] FIG. 2 is a Kaplan-Meier plot of overall survival (OS) by
VeriStrat group for NSCLC patients treated with gemcitabine alone.
Both VeriStrat groups have similar OS (log-rank p=0.64; HR=0.82,
95% CI: 0.35-1.90). Median OS is 197 days (95% CI: 34-348 days) in
the VeriStrat Poor group and 201 days (95% CI: 119-326 days) in the
VeriStrat Good group.
[0028] The "Italian study" consisted of the VeriStrat analysis of
pretreatment plasma samples from advanced non-small cell lung
cancer (NSCLC) patients. The cohort consisted of 33 patients
treated at San Raffaele Hospital in Milan, Italy and 112 patients
treated in Perugia, Italy. All patients received cisplatin-based
chemotherapy as first-line treatment for advanced NSCLC. Patients
treated at Perugia were known to have received the specific
combination of cisplatin and gemcitabine, while the specific
therapy received by patients treated in Milan is not known, only
that it is was cisplatin-based. Although the study involved
patients with ECOG performance status (PS) 0-3, most patients had
PS 0 (43%) or 1 (48%). Most patients (66%) presented with stage IV,
metastatic disease, with the remaining patients having stage IIIA
or IIIB disease. Eighty three percent of patients were current or
former smokers and 79% were male.
[0029] VeriStrat testing was performed on plasma samples collected
prior to commencement of cisplatin treatment. Of the 145 available
samples, 83 (57%) were classified as VeriStrat Good, 58 (40%) as
VeriStrat Poor and the remaining 4 samples received the equivocal
indeterminate classification (see discussion of FIG. 5 below).
[0030] Time to progression (TTP) in the Italian study was analyzed
by VeriStrat classification, see FIG. 3. The VeriStrat Good group
had significantly better TTP than the VeriStrat Poor group, hazard
ratio (HR)=0.64 (95% CI: 0.45-0.91), log-rank p=0.015. The median
TTP in the Good group was 177 days, compared with 103 days in the
Poor group.
[0031] When overall survival (OS) in the Italian study was analyzed
by VeriStrat classification, the VeriStrat Good group was found to
have significantly longer OS than the VeriStrat Poor group, see
FIG. 4. In FIG. 4, the HR=0.48 (95% CI: 0.33-0.71), log-rank
p=0.0002. The median survival time was 368 days in the VeriStrat
Good group and only 205.5 days in the Poor group.
[0032] Analysis of the response to the cisplatin-based chemotherapy
showed that response was significantly correlated with VeriStrat
classification (p value for chi-squared test=0.049), see Table 1
below. Objective response was significantly correlated with
VeriStrat classification (Fisher's exact test p=0.025) with 38% vs.
20% in the evaluable patients classified as VeriStrat Good and Poor
respectively. Disease control showed a trend to a significant
correlation with VeriStrat classification in patients evaluable for
response (67% vs 49% in Good vs Poor).
TABLE-US-00001 TABLE 1 Response to cisplatin-based chemotherapy by
VeriStrat classification VeriStrat Good VeriStrat Poor P value PD
27 28 0.049 PR 31 11 SD 23 16 Objective Response - Yes 31 11 0.025
Objective Response - No 50 44 Disease Control - Yes 54 27 0.056
Disease Control - No 27 28
[0033] From the Italian study, we conclude that, when treated with
cisplatin-based chemotherapy, VeriStrat Good patients had
significantly better TTP and OS than VeriStrat Poor patients. The
differences in median TTP and median OS between Good and Poor
groups were clinically meaningful, being more than 70 days for TTP
and more than 160 days in OS. In addition, objective response rate
was significantly greater in the Good group than in the Poor group
and there was a strong trend to a similar significantly larger
disease control rate.
[0034] Furthermore, because most of the patients in the Italian
study received the combination of gemcitabine and cisplatin,
comparison of the Kaplan-Meier plots of FIGS. 3 and 4 with the
Kaplan-Meier plots of the LCCC0512 study (FIGS. 1 and 2) is
insightful. When one studies FIGS. 1-4 together, we see that in the
treatment arm of gemcitabine +cisplatin the Kaplan-Meier plots show
clear separation by VS groups (FIGS. 3 and 4), and in particular in
FIG. 4 the VS Poor patients do much worse than the VS Good
patients, whereas in the treatment arm of gemcitabine alone the
Kaplan-Meier plots show no separation by VS groups at all (FIGS. 1
and 2). Hence, one can conclude that there is likely a differential
treatment effect between the two treatments for the VS groups.
[0035] Another example of separation of VeriStrat groups treated
with platinum-containing therapy was observed in ovarian cancer
patients receiving carboplatin and paclitaxel after surgery. See
FIGS. 7A and 7B. The data shown in FIGS. 7A and 7B reflect
VeriStrat testing on a subset of 96 patients from a larger study of
about 160 patients. In this study, serum samples were taken from
ovarian cancer patients before surgery and retrospectively analyzed
with the VeriStrat test. After surgery, the patients were treated
with carboplatin and paclitaxel. The results indicate that patients
classified as VeriStrat Poor have shorter disease free survival
(FIG. 7A) and shorter overall survival (FIG. 7B) than VeriStrat
Good patients when treated with the platinum-based
chemotherapy.
[0036] For FIG. 7A, the log-rank (Mantel-Cox) Test P value=0.0512.
Median disease free survival times for VeriStrat Good patients was
20 months whereas it was 11 months for VeriStrat Poor patients. The
data have a Hazard Ratio of 0.4906 and a 95% Confidence Index (CI)
of ratio 0.2398 to 1.004. In FIG. 7B the log-rank Test P
value=0.0036, the median overall survival time for VeriStrat Good
patients is undefined and for VeriStrat Poor patients is 21 months.
The data have a Hazard Ratio of 0.3061 and a 95% Confidence Index
(CI) of ratio 0.1379 to 0.6795.
[0037] Given our previous experience with VeriStrat separation in
NSCLC and other cancers, we believe that the difference shown in
FIGS. 1-4 and 7A-7B lies principally in the Poor group--i.e., that
they probably benefit very little from the combination of
gemcitabine+cisplatin (or possibly other combinations of
chemotherapy based on a platinum agent), but can get outcomes
similar to Good patients if they receive gemcitabine alone, or
another non-platinum agent, for example, docetaxel. Our results in
a subgroup of NSCLC patients treated with docetaxel show no
difference in overall survival between VeriStrat Good and VeriStrat
Poor patients (hazard ratio 1.03, p=0.95, see FIG. 6 for a
Kaplan-Meier plot)
[0038] We further believe that VS Poor NSCLC patients are likely to
do worse, or certainly no better, on gemcitabine+cisplatin than
they would on gemcitabine alone, probably because they do badly on
a platinum agent and addition of the gemcitabine does not correct
this. On gemcitabine alone VS Poor NSCLC patients do as well as the
VS Good patients (FIGS. 1 and 2). Therefore, it can be concluded
that the VS Poor class label predicts that the NSCLC patients are
not likely to benefit from platinum-based chemotherapy agent (e.g.,
cisplatin or analogs thereof) either singly or in combination
treatments with non-platinum chemotherapy agents.
[0039] We further observe that the addition of platinum-based
agents to other anticancer therapies that might otherwise help
those patients classified as VeriStrat Poor negates the beneficial
effects of such non-platinum-based anticancer therapies.
[0040] While we cannot exclude the possibility that the VeriStrat
Good patients may do better on gemcitabine+cisplatin than
gemcitabine alone, we believe the above considerations (in addition
to the ovarian cancer data presented in FIGS. 7A and 7B) support
the hypothesis that the difference in separation between VeriStrat
Good and VeriStrat Poor groups treated with gemcitabine+cisplatin
is due to the VeriStrat Poor patients not obtaining benefit from
anticancer agents combined with platinum chemotherapy, rather than
only the VeriStrat Good patients obtaining benefit from adding
cisplatin to gemcitabine.
[0041] The VeriStrat Test and Guiding Patient Treatment
[0042] As noted above, practical useful tests follow from the
discoveries of this disclosure. One aspect is that our testing
method identifies whether a particular cancer patient is a member
of a group of cancer patients that are not likely to benefit from
administration of a platinum-based chemotherapy agent, e.g.,
cisplatin, carboplatin or analogs thereof, either alone or in
combination with other non-platinum chemotherapy agents, e.g.,
gemcitabine. This identification can be made in advance of
treatment.
[0043] In one example, a method for guiding treatment of a NSCLC
patient is described, comprising the steps of: a) obtaining a
blood-based sample from the patient; b) obtaining a mass-spectrum
of the blood-based sample with the aid of a mass spectrometer; c)
in a programmed computer, performing predefined pre-processing
steps on the mass spectrum, obtaining integrated intensity values
of selected features in the spectrum over predefined m/z ranges
after the pre-processing steps are performed, and comparing the
integrated intensity values with a training set comprising
class-labeled spectra from other cancer patients and classifying
the mass spectrum with a class label, and d) if the class label is
Poor or the equivalent, the patient is predicted to not benefit
from treatment in the form of administration of a platinum-based
chemotherapy agent and is thereby guided towards a treatment
regimen not containing platinum agents.
[0044] The test is illustrated in flow chart form in FIG. 5 as a
process 300.
[0045] At step 302, a serum or plasma sample is obtained from the
patient. In one embodiment, the serum samples are separated into
three aliquots and the mass spectrometry and subsequent steps 304,
306 (including sub-steps 308, 310 and 312), 314, 316 and 318 are
performed independently on each of the aliquots. The number of
aliquots can vary, for example there may be 4, 5 or 10 aliquots,
and each aliquot is subject to the subsequent processing steps.
[0046] At step 304, the sample (aliquot) is subject to mass
spectrometry. A preferred method of mass spectrometry is matrix
assisted laser desorption ionization (MALDI) time of flight (TOF)
mass spectrometry, but other methods are possible. Mass
spectrometry produces mass spectra consisting of data points that
represent intensity values at a multitude of mass/charge (m/z)
values, as is conventional in the art. In one example embodiment,
the samples are thawed and centrifuged at 1500 rpm for five minutes
at four degrees Celsius. Further, the samples may be diluted 1:10,
or 1:5, in MilliQ water. Diluted samples may be spotted in randomly
allocated positions on a MALDI plate in triplicate (i.e., on three
different MALDI targets). After 0.75 ul of diluted sample is
spotted on a MALDI plate, 0.75 ul of 35 mg/ml sinapinic acid (in
50% acetonitrile and 0.1% trifluoroacetic acid (TFA)) may be added
and mixed by pipetting up and down five times. Plates may be
allowed to dry at room temperature. It should be understood that
other techniques and procedures may be utilized for preparing and
processing samples in accordance with the principles of the present
invention.
[0047] Mass spectra may be acquired for positive ions in linear
mode using a Voyager DE-PRO or DE-STR MALDI TOF mass spectrometer
with automated or manual collection of the spectra. (Of course,
other MALDI TOF instruments could be used, e.g., instruments of
Bruker Corporation). Seventy five or one hundred spectra are
collected from seven or five positions within each MALDI spot in
order to generate an average of 525 or 500 spectra for each sample
specimen. Spectra are externally calibrated using a mixture of
protein standards (Insulin (bovine), thioredoxin (E. coli), and
Apomyglobin (equine)).
[0048] At step 306, the spectra obtained in step 304 are subject to
one or more pre-defined pre-processing steps. The pre-processing
steps 306 are implemented in a general purpose computer using
software instructions that operate on the mass spectral data
obtained in step 304. The pre-processing steps 306 include
background subtraction (step 308), normalization (step 310) and
alignment (step 312). The step of background subtraction preferably
involves generating a robust, asymmetrical estimate of background
in the spectrum and subtracts the background from the spectrum.
Step 308 uses the background subtraction techniques described in
U.S. Pat. No. 7,736,905, which is incorporated by reference herein.
The normalization step 310 involves a normalization of the
background subtracted spectrum. The normalization can take the form
of a partial ion current normalization, or a total ion current
normalization, as described in U.S. Pat. No. 7,736,905. Step 312 as
described in U.S. Pat. No. 7,736,905 aligns the normalized,
background subtracted spectrum to a predefined mass scale, which
can be obtained from investigation of the training set used by the
classifier.
[0049] Once the pre-processing steps 306 are performed, the process
300 proceeds to step 314 of obtaining integrated intensities in the
spectrum over predefined m/z ranges. The normalized and background
subtracted intensity values may be integrated over these m/z
ranges. This integrated value (i.e., the sum of intensities within
the corresponding predefined m/z range is assigned to a feature.
Predefined m/z ranges may be defined as the interval around the
average m/z position of the corresponding feature with a width
corresponding to the peak width at this m/z position. This step is
also disclosed in further detail in U.S. Pat. No. 7,736,905.
[0050] At step 314, the integrated values of intensities in the
spectrum are obtained at one or more of the following m/z ranges:
[0051] 5732 to 5795 [0052] 5811 to 5875 [0053] 6398 to 6469 [0054]
11376 to 11515 [0055] 11459 to 11599 [0056] 11614 to 11756 [0057]
11687 to 11831 [0058] 11830 to 11976 [0059] 12375 to 12529 [0060]
12502 to 12656 [0061] 23183 to 23525 [0062] 23279 to 23622 and
[0063] 65902 to 67502.
[0064] In a preferred embodiment, values are obtained at eight of
these m/z ranges shown in Table 2 below. The significance, and
methods of discovery of these peaks, is explained in the U.S. Pat.
No. 7,736,905. In practice the above widths (ranges) may vary.
[0065] At step 316, the values obtained at step 314 are supplied to
a classifier, which in the illustrated embodiment is a K-nearest
neighbor (KNN) classifier. The classifier makes use of a training
set of class labeled spectra from a multitude of other patients
(which may be NSCLC cancer patients, or other solid epithelial
cancer patients, e.g., HNSCC, breast cancer). The application of
the KNN classification algorithm to the values at 314 and the
training set is explained in U.S. Pat. NO. 7,736,905. Other
classifiers can be used, including a probabilistic KNN classifier
or other classifier. In the illustrated embodiment, the training
set is in the form of class-labeled spectra from NSCLC patients
that either did or did not benefit from administration of EGFR
inhibitors, those that did benefit being labeled "Good" and those
that did not labeled "Poor."
[0066] At step 318, the classifier produces a label for the
spectrum, either Good, or Poor. As mentioned above, steps 304-318
are performed in parallel on the three separate aliquots from a
given patient sample (or whatever number of aliquots are used). At
step 320, a check is made to determine whether all the aliquots
produce the same class label. If not, an undefined (or
Indeterminate) result is returned as indicated at step 322. If all
aliquots produce the same label, the label is reported as indicated
at step 324.
[0067] As described in this document, new and unexpected uses of
the class label reported at step 324 are disclosed. For example,
those NSCLC cancer patients labeled Poor in accordance with the
VeriStrat test are unlikely to benefit from treatment in the form
of a platinum-based chemotherapy agent, such as cisplatin or
analogs thereof. As another example, if the NSCLC patient is
identified as Poor in accordance with the test, then the patient is
not likely to benefit from administration of a combination of a
platinum-based chemotherapy agent and a non-platinum-based
chemotherapy agent, such as the combination of gemcitabine and
cisplatin.
[0068] It will be understood that steps 306, 314, 316 and 318 are
typically performed in a programmed general purpose computer using
software coding the pre-processing steps 306, the obtaining of
integrated intensity values in step 314, the application of the KNN
classification algorithm in step 316 and the generation of the
class label in step 318. The training set of class labeled spectra
used in step 116 is stored in memory in the computer or in a memory
accessible to the computer.
[0069] The method and programmed computer may be advantageously
implemented at a laboratory test processing center as described in
our prior patent application publication U.S. Pat. No.
7,736,905.
TABLE-US-00002 TABLE 2 Peaks used in VeriStrat. Peak number m/z 1
5843 2 11445 3 11529 4 11685 5 11759 6 11903 7 12452 8 12579
[0070] It will further be noted that step 304, conducting MALDI-TOF
mass spectrometry on the sample, can be performed in accordance
with the so-called "deep-MALDI" methods of conducting mass
spectrometry as described in US provisional application Ser. No.
61/652,394 filed May 29, 2012, the entire content of which is
incorporated by reference herein. Additionally, the features used
for classification may be obtained by investigation of spectra
obtained from a multitude of samples subject to the "deep-MALDI"
method as well.
[0071] Generalization of the VeriStrat Test as Predictive of
Patient Non-Benefit from Platinum-Based Chemotherapy to Other
Cancers, including HNSCC (Head and Neck Cancer), Ovarian, CRC,
Breast and Others
[0072] The VeriStrat Poor signature has been found in many
different epithelial cancer types, including NSCLC, squamous cell
carcinoma of the head and neck (SSCHN, or HNSCC), colorectal
cancer, renal cell carcinoma, melanoma, pancreatic cancer, breast
cancer and ovarian cancer, but has not been identified in patients
without cancer. Hence, while not specific to lung cancer, it is
related to the patient's cancer. In the absence of systemic
treatment, patients classified as VeriStrat Poor have inferior
outcomes (in terms of OS and PFS) as compared to patients
classified at VeriStrat Good (Carbone D. et al., 2nd European Lung
Cancer Conference, April 2010. J Thorac Oncol 2010; 5(5) S80 abstr.
203O) and hence VeriStrat Poor classification is measuring an
innate property of the disease, or specific disease state.
[0073] In addition to its prognostic capability, the VeriStrat test
has demonstrated a significant predictive power, the ability to
predict differential treatment benefit between VeriStrat groups, in
multiple trials across multiple tumor types, including NSCLC and
breast cancer, including the studies mentioned in this application
and in EGF30008, see U.S. patent application Ser. No. 13/356,730
filed Jan. 24, 2012.
[0074] Previous studies have shown that separation in outcomes
between VeriStrat groups depends on the treatment regimen, but is
not restricted to the type of cancer. For example, VeriStrat Good
patients have been shown to have better progression-free survival
and overall survival than VeriStrat Poor patients when treated with
EGFR inhibitors not only when they have NSCLC, but also when they
have SSCHN or colorectal cancer (see, e.g., the previous patent
literature of Biodesix, Inc. cited above, and Cancer Epidemiol
Biomarkers Prey. 2010 February; 19(2):358-65; Taguchi F. et al., J.
Nat. Cancer Institute, 2007 v. 99 (11), 838-846); Chung C, Seeley
E, Roder H, et al. Detection of tumor epidermal growth factor
receptor pathway dependence by serum mass spectrometry in cancer
patients. Cancer Epidemiol Biomarkers Prev. 2010 February;
19(2):358-65). Similarly, we have data in other treatment regimens
across tumor types where there is no separation between VeriStrat
groups, e.g. non-platinum based chemotherapy in HNSCC (Chung et al.
article, supra) and NSCLC (Stinchcombe, T. E. et al. A randomized
phase II trial of first-line treatment with gemcitabine, erlotinib,
or gemcitabine and erlotinib in elderly patients (age >/=70
years) with stage IIIB/IV non-small cell lung cancer. J Thorac
Oncol 6, 1569-77 (2011).
[0075] On the basis of this body of evidence over many tumor types,
we hypothesize that pretreatment VeriStrat Poor status is
identifying a specific disease state across tumor types and that
there are classes of therapies where we observe inferior outcomes
for VS Poor patients across tumor types (e.g. EGFRI monotherapy)
and others where we observe similar outcomes for both VS groups
(e.g. non-platinum chemotherapy). This leads us to conclude that it
is likely that the inferior outcomes observed in NSCLC patients
treated with a chemotherapy regimen containing a platinum-based
agent, will not be unique to NSCLC, but rather should be expected
in all cancer types where we observe VeriStrat Poor patients and
platinum-based chemotherapy is used.
* * * * *