System for evaluating a pathological stage of prostate cancer

• US 9,779,213 B2
• Filed: 08/28/2009
• Issued: 10/03/2017
• Est. Priority Date: 07/25/2008
• Status: Active Grant

• Current Assignee: Champalimaud Foundation
• Original Assignee: Fundação D. Anna De Sommer Champalimaud E Dr. Carlos Montez Champalimaud
• Inventors: Donovan, Michael, Khan, Faisal, Alter, Jason, Fernandez, Gerardo, Mesa-Tejada, Ricardo, Powell, Douglas, Zubek, Valentina Bayer, Hamann, Stefan, Cordon-Cardo, Carlos, Costa, Jose
• Primary Examiner(s): Whaley, Pablo S
• Application Number:
  US12/584,048
• Publication Number:
  US 20100184093A1
• Time in Patent Office: 2,958 Days
• Field of Search: None
• CPC Class Codes: G16B 25/00 ICT specially adapted for hybridisation; ICT specially adapted for gene or protein expression
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16H 50/50 for simulation or modelling of medical disorders
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients

Abstract

Clinical information, molecular information and/or computer-generated morphometric information is used in a predictive model for predicting the occurrence of a medical condition. In an embodiment, a model predicts whether a patient is likely to have a favorable pathological stage of prostate cancer, where the model is based on features including one or more (e.g., all) of preoperative PSA, Gleason Score, a measurement of expression of androgen receptor (AR) in epithelial and stromal nuclei and/or a measurement of expression of Ki67-positive epithelial nuclei, a morphometric measurement of a ratio of area of epithelial nuclei outside gland units to area of epithelial nuclei within gland units, and a morphometric measurement of area of epithelial nuclei distributed away from gland units. In some embodiments, quantitative measurements of protein expression in cell lines are utilized to objectively assess assay (e.g., multiplex immunofluorescence (IF)) performance and/or to normalize features for use within a predictive model.

Claims

1. A system for evaluating a pathological stage of a patient with respect to prostate cancer, the system comprising:

(1) a fluorescence imaging device configured to record at least one sample image of a tissue sample of a patient, the tissue sample treated with a plurality of fluorochrome labeled antibodies, wherein the antibodies are selected to bind with at least AR and Ki67;

(2) a database configured to store patient data including clinical feature data values for the patient comprising a value indicative of the biopsy Gleason score (bGS) and biopsy Gleason grade (bGG) of the patient and a value indicative of the level of Prostate Specific Antigen (PSA) in the blood of the patient;

(3) a processor configured by code executing therein to perform the following;

a. evaluate the at least one sample image recorded by the imaging device and generate;

(i) one or more molecular feature values indicative of a combined AR dynamic range where the measured bGG values is <

=3, and a total Ki67 where the measured bGG value is 4>

= by applying at least a segmentation analysis to the sample image using a quad-tree function to differentiate the sample image into background and non-background objects where the background objects have an average pixel intensity below a predetermined threshold; and

(ii) a plurality of morphometric measurements, including;

(a) the ratio of area of epithelial nuclei outside gland units to the area of epithelial nuclei within gland units, and (b) area of epithelial nuclei distributed away from gland units;

b. combine the molecular feature values, morphometric measurements and clinical feature data values into a patient dataset;

c. evaluate the patient dataset with a Support Vector Regression for Censored Data (SVRc) algorithm executed as code by the processor, where the SVRc algorithm is configured to output a value corresponding to a risk score for cancer occurrence based on the patient dataset,

wherein the SVRc algorithm is generated by performing regression, using code executed in the processor on a population dataset, where each member of the population has data corresponding to the clinical feature data values, molecular feature values and morphometric measurements of the patient dataset, and where the population includes members where a cancer indolence status is known (uncensored members) and members where a cancer indolence status is unknown (censored members), and the regression includes implementing, as code executed in the processor, a first loss function on the censored member data such that;

Loss(f(x),y,s=1)={C*_s(e−

ε

*_s) e>

ε

*_s

0 −

ε

_s≦

e≦

ε

*_s,

C_s(ε

_s−

e) e<

−

ε

_s,

where e=f(x)−

y; and

ε and

C values differentiate the penalties incurred for f(x)>

0 versus f(x)<

0; and

a second loss function on the uncensored member data such that;

Loss(f(x),y,s=0)={C*_n(e−

ε

*_n) e>

ε

*_n

0 −

ε

_n≦

e≦

ε

*_n,

C_n(ε

_n−

e) e<

−

ε

_n,

where e=f(x)−

y and ε

_n*≦

ε

_n and C_n*≧

C_n and where f(x) is the predicted time to event for sample x, s is the censorship value, and y is a target value corresponding to either the actual time to cancer reoccurrence for a member where a cancer occurrence status is known, or the last known observation time for a member where the cancer occurrence status is unknown;

d. assigns the patient to a high probability of cancer indolence where the output value is below 30 and assigns the patient to a low probability of cancer indolence where the output value is above 30; and

e. generates a report based on the updated patient dataset.

2. The system of claim 1, wherein the morphometric measurement of epithelial nuclei distributed away from gland units comprises a morphometric measurement of the area of epithelial nuclei with centers in a band located a predetermined distance away from lumen borders.

3. The system of claim 2, wherein the band located a predetermined distance away from the lumen borders comprises the region between three and five units away from a lumen border, where a unit is approximately equal to a diameter of one epithelial nucleus.

4. The system of claim 1, wherein the patient data further includes values corresponding to androgen receptor (AR) area in AMACR-positive tumor epithelial cells, dynamic range of AR intensity in tumor epithelial cells, and median lumen area.

5. The system of claim 1, wherein the patient dataset includes at least one additional clinical, molecular, or morphometric feature not provided in subparts (1), (2) and (3).

6. The system of claim 1, wherein the regression performed by the processor includes assigning a cancer indolence probability value of 1 to each member of the population that has:

(a) a pathological stage as assessed by prostatectomy data of T2 or less, (b) a total Gleason score as assessed at prostatectomy of six or less with each component of the Gleason score being three or less, and (c) a first PSA reading post-prostatectomy of approximately zero.

7. The system of claim 1, wherein the patient clinical features dataset of subpart (2) consists essentially of data available at the time of diagnosis of prostate cancer and from a biopsy of tissue of the patient and wherein the processor, is configured by code executing therein, to perform the following:

evaluate a modified patient dataset of a patient assigned a low probability of cancer indolence in subpart (3)(d) using the SVRC algorithm of subpart (3)(c), wherein the modified patient dataset includes a data value indicative that the patient has undergone a radical prostatectomy.