This article is more than 5 years old. Information may no longer be current.
Biomarker panel predicts poor prognosis in early-stage lung cancer
Click Here to Manage Email Alerts
A prognostic classifier panel of three genetic markers identified patients with stage I lung cancer at high risk for recurrence following surgery, according to the results of a study conducted at the NCI’s Center for Cancer Research.
This biomarker panel could be used to select those patients with early-stage lung cancer who are more likely to recur and might benefit from adjuvant therapy.
Curtis C. Harris
“From a mechanistic standpoint and from a statistical standpoint, our strategy has been to add together biomarkers to see what might work for this patient population,” Curtis C. Harris, MD, chief of Laboratory of Human Carcinogenesis at NCI’s Center for Cancer Research at the NIH, told HemOnc Today. “We wanted to see if combining two mechanistic and statistically independent biomarkers into a classifier was better than having just one, and if combining three was better than two.”
Approximately 30% of patients with early-stage lung cancer experience a recurrence within 5 years of surgery with curative intent, according to Harris. The researchers incorporated epigenetic biomarkers to improve existing protein-coding gene and microRNA expression prognostic classifiers, in the hopes of achieving stronger means for identifying patients at a greater recurrence risk.
“The clinically important question is: Can you determine — with some degree of accuracy — who, among people diagnosed with a disease, will have a recurrence, or a poor prognosis?” Harris said. “We knew that about a third of patients with stage I lung cancer who underwent surgery with a curative intent would have a recurrence. With the use of these biomarkers, we strived to improve the identification of patients at the highest risk for a recurrence — those with even up to an 80% recurrence risk.”
Ana I. Robles
Harris, Ana I. Robles, PhD, staff scientist at the NCI’s Laboratory of Human Carcinogenesis, and international colleagues first conducted genome-wide screening of differential DNA methylation in adjacent tumor and non-tumor tissues from three cohorts of patients from the NCI, Norway and Japan. Results of this analysis revealed HOXA9 as a candidate prognostic marker. HOXA9 hypermethylation — a specific epigenetic change to the DNA that is not an alteration of DNA sequencing — appeared associated with 5-year lung cancer mortality.
The researchers then further assessed the prognostic value of HOXA9 promoter methylation alone and in combination with mRNA and microRNA biomarkers. They analyzed HOXA9 promoter methylation by pyrosequencing a validation cohort of patients with stage I lung adenocarcinoma from the NCI (n = 87) and Norway (n =17).
Patients identified as having a poor prognosis had a mean methylation of 40%, whereas patients with a better prognosis had a mean methylation of 17%. Patients with high HOXA9 promoter methylation (n = 23) — defined as above 40% — experienced worse disease-specific survival (HR = 2.6; P = .02) regardless of stage (stage IA vs. IB) and smoking history.
HOXA9 promoter methylation also appeared to be associated with shorter cancer-specific survival in patients who had not received chemotherapy in multivariable analyses adjusted for adjuvant therapy (HR = 3.7; P = .006).
These data were confirmed in a cohort of 113 patients from Japan, results of which showed HOXA9 promoter methylation was associated with shorter time to recurrence independent of stage or smoking history (HR = 3; P = .01).
Combining data from these two cohorts, researchers found HOXA9 promoter methylation was associated with outcomes for patients with stage I (HR = 3.1; P = .001), stage IA (HR = 3.4; P = .009) and stage IB (HR = 3.4; P = .015) disease, independently of smoking history, race and cohort membership.
“We didn’t know how methylation as a biomarker would compare with gene or microRNA expression,” Robles told HemOnc Today. “Methylation is one of the regulatory mechanisms for gene expression, so there was one possibility that it would not really add anything independent of gene expression if it only underlined the expression of genes already in the classifier. What we found is that this particular marker (HOXA9) is associated with prognosis independently of gene expression, pointing to a different kind or category of high-risk patients. In a way, there was no reason for us to expect that methylation would be completely independent of gene expression or add anything to a biomarker that already had gene and microRNA expression in it. It was surprising that it led to a better classifier.”
Researchers had previously found that high microRNA-21 and a four protein-coding gene classifier — composed of XPO1, BRCA1, HIF1α and DLC1 — identified patients with a poor prognosis in the same NCI/Norway and Japan cohorts.
In the current analysis, researchers observed that the outcome of patients in these cohorts appeared independently associated with the four protein-coding genes (HR = 2.8; P = .002), microRNA-21 expression (HR = 2.3; P = .01) and HOXA9 promoter methylation (HR = 2.4; P = .005).
When combined, a high score based on these biomarkers predicted a greater risk for cancer-specific death among therapy-naive patients in the NCI/Norway cohort (HR = 43; P = .001) and shorter time to recurrence in stage I patients from the Japan cohort (HR = 6.2; P = .005). The association persisted when researchers evaluated data from therapy-naive patients in a combined cohort (HR = 10.2; P = 3 x 10-5).
“We already knew that having two mechanistically independent biomarkers improved the prognostic classifier,” Harris said. “When we added DNA methylation, we showed that it was independent of the other two prognostic classifiers and, therefore, the combination proved to be more predictive of poor prognosis.”
Adjuvant chemotherapy and/or immunotherapy are currently not recommended for patients with stage 1 disease, largely because previous trials have failed to show decisive benefit in early-stage patients with lung cancer, according to the researchers. However, Harris and Robles believe that the use of their three-marker panel may identify the patients who are most likely to benefit from adjuvant therapy.
“In general, further therapy does not occur at this point in time until recurrence,” Harris said. “But if we can identify the individuals who have a high probability of recurrence, a more regular use of adjuvant therapy [in appropriate patients] may decrease the frequency of recurrence and get a better prognosis for those individuals.”
An increase in early-stage lung cancer diagnoses — potentially related to evolving diagnostic tools — will necessitate new treatment options for this patient population, according to Robles.
“There are much better diagnostic tools for lung cancer at this point,” Robles said. “Low-dose computed tomography is becoming more commonly used, and, as a result, we are likely going to see a shift of the patient population towards earlier-stage lung cancer diagnoses. So, there needs to be a new focus on treatment and management of early-stage lung cancer, and opportunities to incorporate new strategies for disease management.”
Harris recommends that the biomarker panel be validated in further studies.
“I think that additional studies with confirmatory results in different populations will cause medical oncologists to consider whether randomized trials of adjuvant chemotherapy for high-risk patients might be advantageous,” Harris said. – by Cameron Kelsall
For more information:
Curtis C. Harris, MD, can be reached at Laboratory of Human Carcinogenesis, NCI-CCR, NIH, 37 Convent Drive, Room 3068A, Bethesda, MD 20892; email: curtis_harris@nih.gov.
Ana I. Robles, PhD, can be reached at Laboratory of Human Carcinogenesis, NCI-CCR, NIH, 37 Convent Drive, Room 3060D, Bethesda, MD 20892; email: roblesa@mail.nih.gov.
Disclosure: Harris and Robles report no relevant financial disclosures.
Click Here to Manage Email Alerts