GeneNews Scientific Papershttps://www.myinnovativelab.com/scientific-papers/ Development and Validation of the ColonSentry Test
The predictive ability of the ColonSentry® test was demonstrated in two key studies: a North
American 10,000 patient case-control study that identified and validated the 7-gene biomarker panel,
and a Malaysian case-control validation study that confirmed the findings.13,21 Briefly, RNA profiling
of whole blood samples from 642 North American cases and controls matched by sex, ethnicity, and
BMI identified 7 genes whose expression pattern was associated with CRC (6 overexpressed genes
and 1 underexpressed gene). A training set of 112 CRC cases and 120 controls was first used to
identify the combination of genes whose expression pattern most significantly differentiated CRC
cases from controls. Researchers then developed a risk scale to assess the likelihood of an
individual having CRC using a logistic regression-based algorithm derived from the expression levels
of these 7 genes. To validate the predictive performance of the 7-gene panel, Marshall et al. blinded
410 samples (202 CRC; 208 controls) and scored this “test set” on the basis of the logistic
regression model.13
An independent Malaysian study then validated the 7-gene panel using blood samples from 99 CRC
patients and 111 controls.21 Blood samples were collected from an ethnically diverse Asian
population of patients referred to a colonoscopy clinic between August 2007 and November 2009.
All 6 genes of interest (ANXA3, CLEC4D, LMNB1, PRRG4, TNFAIP6, and VNN1) were upregulated
and the reference gene (IL2RB) was downregulated in CRC as compared with control samples,
confirming the pattern of differential gene expression found in the North American sample.21 Results
were highly statistically significant for both the Malaysian and North American studies.
A Dual Approach for Lung Cancer Screening
While the performance metrics of EarlyCDT-Lung compares favorably with annual CT screening it is the
combination of both tests that yields the most benefit. EarlyCDT-Lung offers a complementary approach to
annual CT screening which is the standard in the USA for early lung cancer detection if the patient meets the
high-risk criteria set by USPSTF.As a convenient blood test, EarlyCDT-Lung can be used when an individual
is at increased risk, but does not meet the criteria for annual CT screening or, when individuals are unwilling
or unable to undergo lifelong annual CT screening. In either case, a Moderate or High Level EarlyCDT-Lung
result can be followed by suitable CT scans to confirm the presence of lung cancer.
EarlyCDT-lung, in the 12,000 patient National Health Service, Scotland study, has shown that it can detect
lung cancer as much as 2 or more years before it may be seen via LDCT. A raised risk score, followed up
aggressively for 2 years, demonstrated a sensitivity of 81%, while maintaining a specificity of 91%. This
contrasts with the validation studies where, with only a six month follow up, sensitivity was 40%. The
difference is in aggressive follow-up for a significant enough time to allow LDCT to see it.
LDCT, while the standard for screening for lung cancer, typically will not differentiate well below a range of
2-4 mm. Also, there is an issue with differentiating between benign and malignant growths, with a false
positive rate of 96%. A combination of EarlyCDT-lung and LDCT should deliver a superior result than LDCT
alone.
The National Health Service ("NHS") Scotland-sponsored Early Cancer Detection Test – Lung Cancer
Scotland ("ECLS") Study of 12,000 high-risk smokers demonstrated a cancer detection rate (sensitivity) of
81% for EarlyCDT®-Lung in these initial results. While the control arm in the study has not been formally
assessed, the positivity rate was as expected with a specificity of 91%. The final data on the control arm will
be collected at the end of the study.
The value of EarlyCDT-Lung lies in its ability to detect early stage disease (Stages I and II) and multiple
histological subtypes, with an overall specificity of 93% (specificity of a High Risk result improves to 98%),
positive predictive value of 1 in 10 (PPV improves to 1 in 5 for a High Risk result) and overall accuracy of 92%
(97% for a High Risk result), assuming a lung cancer prevalence of 1.2%.