Genomic Profile Predicts Recurrence Risk in Colorectal Cancer

An investigational test based on an 18-gene profile predicts recurrence risk and may help inform decisions about the need for adjuvant chemotherapy in patients with localized colorectal cancer.

The test, called ColoPrint, was validated in a study using tumor samples from 135 patients who underwent curative surgery for stage II colorectal cancer.

Just 5% of patients with a low-risk test result had a distant metastatic recurrence within 5 years, compared with 20% of their counterparts with a high-risk result. The data were presented during a teleconference held Jan. 18 in advance of a meeting on gastrointestinal cancers sponsored by the American Society of Clinical Oncology (ASCO).

"Our results confirm previous validation studies of ColoPrint," lead investigator Dr. Robert Rosenberg said. "ColoPrint is able to predict the development of distant metastases of stage II patients and facilitates the identification of patients who might not need additional chemotherapy."

"This study constitutes the second independent validation of the profile ColoPrint, and that makes it unique in itself," he added. "No other prognostic genomic profile in colon cancer has been submitted yet to a second independent validation."

The investigators used the microarray-based test to analyze fresh frozen tumor tissue from 135 patients who underwent curative resection of stage II colorectal cancer at the University Hospital of the Technical University in Munich between 1987 and 2003.

The 18-gene profile includes some genes already known to be associated with carcinogenesis and metastasis, explained Dr. Rosenberg, an assistant professor and surgeon at the hospital, "but at least five genes have not been described yet."

Agendia, the test manufacturer, ran ColoPrint on samples while blinded to patient data. The investigators in Munich assessed the association of test results with clinical outcome.

Study results showed that the test classified the majority of patients (74%) as having a low risk of recurrence, and indeed, just 5% of these patients developed distant metastases within 5 years.

The remaining patients (26%) were classified as having a high recurrence risk, and 20% of this group developed distant metastases during this time period.

In univariate analyses, the ColoPrint result was the only factor significantly predicting the risk of distant metastases, with a 4.1-fold greater likelihood for patients with a high-risk, compared with a low-risk, result (P = .009).

In contrast, none of a variety of clinical and pathologic factors – age, sex, tumor location, grade, number of lymph nodes assessed, pathologic T stage, or the combination of factors in the ASCO risk score (T4 stage, perforation, fewer than 12 lymph nodes assessed, and high grade) – were significant predictors.

In multivariate analyses, a high-risk ColoPrint test result independently predicted a more than quadrupling of the likelihood of recurrence (hazard ratio, 4.3; P = .01). Again, the ASCO risk score was not significantly associated with this outcome.

When patients were stratified by both ColoPrint result and ASCO risk score, agreement was limited. "ASCO risk factors and ColoPrint identify different patients. There is overlap but also a clear discordance," Dr. Rosenberg said. "It will therefore be interesting to find out if and how clinical factors can add to the value of ColoPrint."

Those questions are being addressed in an ongoing international study, called PARSC (Prospective Analysis of Risk Stratification by ColoPrint), which is prospectively comparing the genomic profile with clinical risk factors in 600 patients with stage II colon cancer. "We expect first results from PARSC this year," he commented.

Other tests and markers are also being used to assess recurrence risk in patients with stage II colorectal cancer, acknowledged Dr. Rosenberg. One such marker is microsatellite instability (MSI), with previous research showing that an MSI-high status is associated with a good prognosis.

"We were able to see that the good prognosis of MSI-high patients was recognized by ColoPrint in most of the patients," he said: fully 86% of those having a low-risk test result had a high MSI status. "Only one of the MSI-high patients experienced a relapse, and this patient was correctly identified as ColoPrint high-risk, so the results are independent of the MSI status."

The Oncotype DX colon cancer assay (manufactured by Genomic Health) is also being used to estimate prognosis in these patients and, like ColoPrint, taps a tumor’s genetic information, "but the tests are completely different," Dr. Rosenberg asserted.

Key differences include the ways in which the genomic signatures were developed, the genes they assess, the technology used for the assay, and the nature of the test result, with ColoPrint giving a binary result and Oncotype Dx having an intermediate category, he said.

Offering a clinical perspective, Dr. Jennifer C. Obel, moderator of the teleconference and a medical oncologist with the NorthShore University HealthSystem in Evanston, Ill., noted that the two tests "are very similar in some regards in terms of the questions they answer to identify stage II patients at higher risk of recurrence." Hence, clinicians’ choice between them might be driven by logistic factors, such as the steps required for tissue preparation. At the same time, she cautioned that, thus far, both tests have been shown to provide only prognostic information.

"Those tests help to further identify patients at higher risk for recurrence, though neither test helps to ... predict benefit from chemotherapy," said Dr. Obel, who did not report any relevant conflicts of interest.

Dr. Rosenberg reported that he had no relevant conflicts of interest. Some of the other coinvestigators have employment or leadership positions with or own stock in Agendia.

An investigational test based on an 18-gene profile predicts recurrence risk and may help inform decisions about the need for adjuvant chemotherapy in patients with localized colorectal cancer.

The test, called ColoPrint, was validated in a study using tumor samples from 135 patients who underwent curative surgery for stage II colorectal cancer.

Just 5% of patients with a low-risk test result had a distant metastatic recurrence within 5 years, compared with 20% of their counterparts with a high-risk result. The data were presented during a teleconference held Jan. 18 in advance of a meeting on gastrointestinal cancers sponsored by the American Society of Clinical Oncology (ASCO).

"Our results confirm previous validation studies of ColoPrint," lead investigator Dr. Robert Rosenberg said. "ColoPrint is able to predict the development of distant metastases of stage II patients and facilitates the identification of patients who might not need additional chemotherapy."

"This study constitutes the second independent validation of the profile ColoPrint, and that makes it unique in itself," he added. "No other prognostic genomic profile in colon cancer has been submitted yet to a second independent validation."

The investigators used the microarray-based test to analyze fresh frozen tumor tissue from 135 patients who underwent curative resection of stage II colorectal cancer at the University Hospital of the Technical University in Munich between 1987 and 2003.

The 18-gene profile includes some genes already known to be associated with carcinogenesis and metastasis, explained Dr. Rosenberg, an assistant professor and surgeon at the hospital, "but at least five genes have not been described yet."

Agendia, the test manufacturer, ran ColoPrint on samples while blinded to patient data. The investigators in Munich assessed the association of test results with clinical outcome.

Study results showed that the test classified the majority of patients (74%) as having a low risk of recurrence, and indeed, just 5% of these patients developed distant metastases within 5 years.

The remaining patients (26%) were classified as having a high recurrence risk, and 20% of this group developed distant metastases during this time period.

In univariate analyses, the ColoPrint result was the only factor significantly predicting the risk of distant metastases, with a 4.1-fold greater likelihood for patients with a high-risk, compared with a low-risk, result (P = .009).

In contrast, none of a variety of clinical and pathologic factors – age, sex, tumor location, grade, number of lymph nodes assessed, pathologic T stage, or the combination of factors in the ASCO risk score (T4 stage, perforation, fewer than 12 lymph nodes assessed, and high grade) – were significant predictors.

In multivariate analyses, a high-risk ColoPrint test result independently predicted a more than quadrupling of the likelihood of recurrence (hazard ratio, 4.3; P = .01). Again, the ASCO risk score was not significantly associated with this outcome.

When patients were stratified by both ColoPrint result and ASCO risk score, agreement was limited. "ASCO risk factors and ColoPrint identify different patients. There is overlap but also a clear discordance," Dr. Rosenberg said. "It will therefore be interesting to find out if and how clinical factors can add to the value of ColoPrint."

Those questions are being addressed in an ongoing international study, called PARSC (Prospective Analysis of Risk Stratification by ColoPrint), which is prospectively comparing the genomic profile with clinical risk factors in 600 patients with stage II colon cancer. "We expect first results from PARSC this year," he commented.

Other tests and markers are also being used to assess recurrence risk in patients with stage II colorectal cancer, acknowledged Dr. Rosenberg. One such marker is microsatellite instability (MSI), with previous research showing that an MSI-high status is associated with a good prognosis.

"We were able to see that the good prognosis of MSI-high patients was recognized by ColoPrint in most of the patients," he said: fully 86% of those having a low-risk test result had a high MSI status. "Only one of the MSI-high patients experienced a relapse, and this patient was correctly identified as ColoPrint high-risk, so the results are independent of the MSI status."

The Oncotype DX colon cancer assay (manufactured by Genomic Health) is also being used to estimate prognosis in these patients and, like ColoPrint, taps a tumor’s genetic information, "but the tests are completely different," Dr. Rosenberg asserted.

Key differences include the ways in which the genomic signatures were developed, the genes they assess, the technology used for the assay, and the nature of the test result, with ColoPrint giving a binary result and Oncotype Dx having an intermediate category, he said.

Offering a clinical perspective, Dr. Jennifer C. Obel, moderator of the teleconference and a medical oncologist with the NorthShore University HealthSystem in Evanston, Ill., noted that the two tests "are very similar in some regards in terms of the questions they answer to identify stage II patients at higher risk of recurrence." Hence, clinicians’ choice between them might be driven by logistic factors, such as the steps required for tissue preparation. At the same time, she cautioned that, thus far, both tests have been shown to provide only prognostic information.

"Those tests help to further identify patients at higher risk for recurrence, though neither test helps to ... predict benefit from chemotherapy," said Dr. Obel, who did not report any relevant conflicts of interest.

Dr. Rosenberg reported that he had no relevant conflicts of interest. Some of the other coinvestigators have employment or leadership positions with or own stock in Agendia.

An investigational test based on an 18-gene profile predicts recurrence risk and may help inform decisions about the need for adjuvant chemotherapy in patients with localized colorectal cancer.

The test, called ColoPrint, was validated in a study using tumor samples from 135 patients who underwent curative surgery for stage II colorectal cancer.

Just 5% of patients with a low-risk test result had a distant metastatic recurrence within 5 years, compared with 20% of their counterparts with a high-risk result. The data were presented during a teleconference held Jan. 18 in advance of a meeting on gastrointestinal cancers sponsored by the American Society of Clinical Oncology (ASCO).

"Our results confirm previous validation studies of ColoPrint," lead investigator Dr. Robert Rosenberg said. "ColoPrint is able to predict the development of distant metastases of stage II patients and facilitates the identification of patients who might not need additional chemotherapy."

"This study constitutes the second independent validation of the profile ColoPrint, and that makes it unique in itself," he added. "No other prognostic genomic profile in colon cancer has been submitted yet to a second independent validation."

The investigators used the microarray-based test to analyze fresh frozen tumor tissue from 135 patients who underwent curative resection of stage II colorectal cancer at the University Hospital of the Technical University in Munich between 1987 and 2003.

The 18-gene profile includes some genes already known to be associated with carcinogenesis and metastasis, explained Dr. Rosenberg, an assistant professor and surgeon at the hospital, "but at least five genes have not been described yet."

Agendia, the test manufacturer, ran ColoPrint on samples while blinded to patient data. The investigators in Munich assessed the association of test results with clinical outcome.

Study results showed that the test classified the majority of patients (74%) as having a low risk of recurrence, and indeed, just 5% of these patients developed distant metastases within 5 years.

The remaining patients (26%) were classified as having a high recurrence risk, and 20% of this group developed distant metastases during this time period.

In univariate analyses, the ColoPrint result was the only factor significantly predicting the risk of distant metastases, with a 4.1-fold greater likelihood for patients with a high-risk, compared with a low-risk, result (P = .009).

In contrast, none of a variety of clinical and pathologic factors – age, sex, tumor location, grade, number of lymph nodes assessed, pathologic T stage, or the combination of factors in the ASCO risk score (T4 stage, perforation, fewer than 12 lymph nodes assessed, and high grade) – were significant predictors.

In multivariate analyses, a high-risk ColoPrint test result independently predicted a more than quadrupling of the likelihood of recurrence (hazard ratio, 4.3; P = .01). Again, the ASCO risk score was not significantly associated with this outcome.

When patients were stratified by both ColoPrint result and ASCO risk score, agreement was limited. "ASCO risk factors and ColoPrint identify different patients. There is overlap but also a clear discordance," Dr. Rosenberg said. "It will therefore be interesting to find out if and how clinical factors can add to the value of ColoPrint."

Those questions are being addressed in an ongoing international study, called PARSC (Prospective Analysis of Risk Stratification by ColoPrint), which is prospectively comparing the genomic profile with clinical risk factors in 600 patients with stage II colon cancer. "We expect first results from PARSC this year," he commented.

Other tests and markers are also being used to assess recurrence risk in patients with stage II colorectal cancer, acknowledged Dr. Rosenberg. One such marker is microsatellite instability (MSI), with previous research showing that an MSI-high status is associated with a good prognosis.

"We were able to see that the good prognosis of MSI-high patients was recognized by ColoPrint in most of the patients," he said: fully 86% of those having a low-risk test result had a high MSI status. "Only one of the MSI-high patients experienced a relapse, and this patient was correctly identified as ColoPrint high-risk, so the results are independent of the MSI status."

The Oncotype DX colon cancer assay (manufactured by Genomic Health) is also being used to estimate prognosis in these patients and, like ColoPrint, taps a tumor’s genetic information, "but the tests are completely different," Dr. Rosenberg asserted.

Key differences include the ways in which the genomic signatures were developed, the genes they assess, the technology used for the assay, and the nature of the test result, with ColoPrint giving a binary result and Oncotype Dx having an intermediate category, he said.

Offering a clinical perspective, Dr. Jennifer C. Obel, moderator of the teleconference and a medical oncologist with the NorthShore University HealthSystem in Evanston, Ill., noted that the two tests "are very similar in some regards in terms of the questions they answer to identify stage II patients at higher risk of recurrence." Hence, clinicians’ choice between them might be driven by logistic factors, such as the steps required for tissue preparation. At the same time, she cautioned that, thus far, both tests have been shown to provide only prognostic information.

"Those tests help to further identify patients at higher risk for recurrence, though neither test helps to ... predict benefit from chemotherapy," said Dr. Obel, who did not report any relevant conflicts of interest.

Dr. Rosenberg reported that he had no relevant conflicts of interest. Some of the other coinvestigators have employment or leadership positions with or own stock in Agendia.