October 11, 2011 — Both BRCA1 and BRCA2 germ-line mutations are associated with early ovarian cancer, but new observational data suggest that each might require different treatment strategies, according to a report published in the October 12 issue of JAMA: The Journal of the American Medical Association.
The study found that women with high-grade ovarian cancer live longer and respond better to platinum-based chemotherapy when their tumors have BRCA2 genetic mutations than when they have wild-type BRCA. The same does not hold true for BRCA1 mutations.
The analysis looked at 316 cases of high-grade serous ovarian cancer; 29 tumors had BRCA2 mutations and 37 had BRCA1 mutations.
This suggests that stratification according to BRCA status will become more important in ovarian cancer clinical trials, senior author Wei Zhang, PhD, from the University of Texas M.D. Anderson Cancer Center in Houston, told Medscape Medical News. The findings also suggest that combining drugs that damage DNA (such as platinum) with newer drugs that block DNA repair might be particularly effective in these cancers, said Dr. Zhang.
However, an expert not involved with the study questioned the findings because of the small number of patients with BRCA mutations.
"It is fundamentally unsound to evaluate clinical epidemiology with sample sizes of 35 and 27 (BRCA1 and BRCA2) — or 37 and 29, depending on whether one looks at [the] abstract or table — and draw firm conclusions. The power is low," said Paul Pharoah, MD, senior clinical research fellow at Cambridge Cancer Center in the United Kingdom, in an email. He reviewed the study for Medscape Medical News.
The study is also partly in error, said Dr. Pharoah, who has conducted a similar study in these patients but with much a larger sample size.
"I know from our dataset (over 10 times the sample size) that both BRCA1 and BRCA2 are associated with improved survival [respectively, compared with wild-type BRCA]. This was statistically robust. So Yang's conclusion is partly wrong," he said, referring to study lead author Da Yang, PhD, who is also from M.D. Anderson Cancer Center.
Dr. Zhang responded to these criticisms by saying that his team used The Cancer Genome Atlas project (TCGA) as a data source, the power of which is not simply epidemiologic. "The comprehensive genomic and epigenomic data of TCGA allow us to seek mechanistic insight," he said. He also noted that 2 cases had both BRCA1 and BRCA2 mutations, which accounted for the varying sample sizes — a fact that "we explained in our paper." Finally, he said he looks forward to seeing Dr. Pharoah's paper when it is published. "There are many different types of ovarian cancer and they are quite different biologically and genomically. It will be interesting to see what Dr. Pharoah finds when he looks at the serous ovarian cancer subtype, like we did."
Lack of Statistical Significance With Overall Survival Differences
According to the authors, BRCA2 and BRCA1 are both tumor-suppressing genes that affect DNA repair, but in different ways. BRCA2 mutations change the RAD51 protein, which is required for the repair of double-strand DNA breaks by homologous recombination. Without RAD51, the tumor cell cannot repair DNA damaged by antitumor treatment.
In contrast, BRCA1 is involved in multiple functions (including DNA damage response and checkpoint control). A mutant BRCA1 can cause failure in a function and set the cell up for tumorigenesis without making it more vulnerable to drugs such as cisplatin.
"Uncovering the separate potential effects of BRCA1 and BRCA2 mutations takes us a step toward a more personalized approach to treating ovarian cancer, and perhaps other cancers," Dr. Zhang said in a press statement. "This paper suggests that those 2 genes, and the many others involved in DNA repair, are prime targets for further research."
This teasing apart of BRCA1 and BRCA2 roles in ovarian cancer is possible because of TCGA. TCGA reported more than 400 high-grade serous ovarian cancer cases in which an exhaustive analysis of each tumor's genome and comprehensive clinical data on each patient were combined.
"TCGA gave us enough analytical power to differentiate between BRCA1 and BRCA2 mutations and to conduct a survival analysis," said Dr. Yang.
The outcome measures in the study were overall survival, progression-free survival, and chemotherapy response. The analysis showed 5-year survival of 61% with BRCA2 mutations and 25% with wild-type BRCA. Three-year progression-free survival was 44% with BRCA2 mutations and 16% with wild-type BRCA2. Overall survival was not significantly different between BRCA2 and BRCA1 mutations (P = .17), but the progression-free survival difference was (P =.05).
BRCA1 mutations did not affect either overall survival or progression-free survival.
The response rate to platinum chemotherapy was 100% in patients with BRCA2 mutations, 82% in those with wild-type BRCA2, and 80% in those with BRCA1 mutations. Response duration was 18 months with BRCA2 mutations, 11.7 months with wild-type BRCA2, and 12.5 months for BRCA1 mutations.
In an accompanying editorial, Victor R. Grann, MD, MPH, and Ramon E. Parsons, MD, PhD, from the Columbia University Medical Center in New York City, write that the study "provides a major advance in the understanding of the use of new treatments for ovarian cancer among patients with BRCA mutations by demonstrating a difference in the response among patients with BRCA1 and BRCA2 mutations diagnosed with ovarian cancer." They and the study authors all say that the next step is to enroll these patients in randomized clinical trials to test whether BRCA1 or BRCA2 mutation carriers respond differently to treatment.
However, Dr. Pharoah's research represents a challenge to this opinion.
He and his group have investigated BRCA mutations and clinical outcomes using the same TCGA data plus additional cases collected worldwide — producing a much larger dataset of more than 1100 BRCA1 mutation carriers, 300 BRCA2 mutation carriers, and 2000 noncarriers. They reported their data at the 2011 meeting of the American Association for Cancer Research. That study has been submitted for publication.
Dr. Pharoah also emphasized the fact that the overall survival differences between the 2 mutation types in the study by Dr. Yang and colleagues was not statistically significant. "There is no significant difference between BRCA1 and BRCA2 (which is not highlighted in the abstract or press release). Thus, it is not a sound conclusion that BRCA2 but not BRCA1 is associated with improved survival. This apparent contradiction is all a problem with sample size and an overestimation of statistical significance. With just 35 BRCA1 cases, the power to detect a difference with noncarriers is limited, so the conclusion is not sound."
Dr. Pharoah added that "even where a result is statistically significant, it is more likely to be wrong when the sample size is small than when it is large — a fact that is very rarely appreciated."
The authors have disclosed no relevant financial relationships.
JAMA. 2011;306:1557-1565, 1597-1598. Abstract, Editorial
The study found that women with high-grade ovarian cancer live longer and respond better to platinum-based chemotherapy when their tumors have BRCA2 genetic mutations than when they have wild-type BRCA. The same does not hold true for BRCA1 mutations.
The analysis looked at 316 cases of high-grade serous ovarian cancer; 29 tumors had BRCA2 mutations and 37 had BRCA1 mutations.
This suggests that stratification according to BRCA status will become more important in ovarian cancer clinical trials, senior author Wei Zhang, PhD, from the University of Texas M.D. Anderson Cancer Center in Houston, told Medscape Medical News. The findings also suggest that combining drugs that damage DNA (such as platinum) with newer drugs that block DNA repair might be particularly effective in these cancers, said Dr. Zhang.
However, an expert not involved with the study questioned the findings because of the small number of patients with BRCA mutations.
"It is fundamentally unsound to evaluate clinical epidemiology with sample sizes of 35 and 27 (BRCA1 and BRCA2) — or 37 and 29, depending on whether one looks at [the] abstract or table — and draw firm conclusions. The power is low," said Paul Pharoah, MD, senior clinical research fellow at Cambridge Cancer Center in the United Kingdom, in an email. He reviewed the study for Medscape Medical News.
The study is also partly in error, said Dr. Pharoah, who has conducted a similar study in these patients but with much a larger sample size.
"I know from our dataset (over 10 times the sample size) that both BRCA1 and BRCA2 are associated with improved survival [respectively, compared with wild-type BRCA]. This was statistically robust. So Yang's conclusion is partly wrong," he said, referring to study lead author Da Yang, PhD, who is also from M.D. Anderson Cancer Center.
Dr. Zhang responded to these criticisms by saying that his team used The Cancer Genome Atlas project (TCGA) as a data source, the power of which is not simply epidemiologic. "The comprehensive genomic and epigenomic data of TCGA allow us to seek mechanistic insight," he said. He also noted that 2 cases had both BRCA1 and BRCA2 mutations, which accounted for the varying sample sizes — a fact that "we explained in our paper." Finally, he said he looks forward to seeing Dr. Pharoah's paper when it is published. "There are many different types of ovarian cancer and they are quite different biologically and genomically. It will be interesting to see what Dr. Pharoah finds when he looks at the serous ovarian cancer subtype, like we did."
Lack of Statistical Significance With Overall Survival Differences
According to the authors, BRCA2 and BRCA1 are both tumor-suppressing genes that affect DNA repair, but in different ways. BRCA2 mutations change the RAD51 protein, which is required for the repair of double-strand DNA breaks by homologous recombination. Without RAD51, the tumor cell cannot repair DNA damaged by antitumor treatment.
In contrast, BRCA1 is involved in multiple functions (including DNA damage response and checkpoint control). A mutant BRCA1 can cause failure in a function and set the cell up for tumorigenesis without making it more vulnerable to drugs such as cisplatin.
"Uncovering the separate potential effects of BRCA1 and BRCA2 mutations takes us a step toward a more personalized approach to treating ovarian cancer, and perhaps other cancers," Dr. Zhang said in a press statement. "This paper suggests that those 2 genes, and the many others involved in DNA repair, are prime targets for further research."
This teasing apart of BRCA1 and BRCA2 roles in ovarian cancer is possible because of TCGA. TCGA reported more than 400 high-grade serous ovarian cancer cases in which an exhaustive analysis of each tumor's genome and comprehensive clinical data on each patient were combined.
"TCGA gave us enough analytical power to differentiate between BRCA1 and BRCA2 mutations and to conduct a survival analysis," said Dr. Yang.
The outcome measures in the study were overall survival, progression-free survival, and chemotherapy response. The analysis showed 5-year survival of 61% with BRCA2 mutations and 25% with wild-type BRCA. Three-year progression-free survival was 44% with BRCA2 mutations and 16% with wild-type BRCA2. Overall survival was not significantly different between BRCA2 and BRCA1 mutations (P = .17), but the progression-free survival difference was (P =.05).
BRCA1 mutations did not affect either overall survival or progression-free survival.
The response rate to platinum chemotherapy was 100% in patients with BRCA2 mutations, 82% in those with wild-type BRCA2, and 80% in those with BRCA1 mutations. Response duration was 18 months with BRCA2 mutations, 11.7 months with wild-type BRCA2, and 12.5 months for BRCA1 mutations.
In an accompanying editorial, Victor R. Grann, MD, MPH, and Ramon E. Parsons, MD, PhD, from the Columbia University Medical Center in New York City, write that the study "provides a major advance in the understanding of the use of new treatments for ovarian cancer among patients with BRCA mutations by demonstrating a difference in the response among patients with BRCA1 and BRCA2 mutations diagnosed with ovarian cancer." They and the study authors all say that the next step is to enroll these patients in randomized clinical trials to test whether BRCA1 or BRCA2 mutation carriers respond differently to treatment.
However, Dr. Pharoah's research represents a challenge to this opinion.
He and his group have investigated BRCA mutations and clinical outcomes using the same TCGA data plus additional cases collected worldwide — producing a much larger dataset of more than 1100 BRCA1 mutation carriers, 300 BRCA2 mutation carriers, and 2000 noncarriers. They reported their data at the 2011 meeting of the American Association for Cancer Research. That study has been submitted for publication.
Dr. Pharoah also emphasized the fact that the overall survival differences between the 2 mutation types in the study by Dr. Yang and colleagues was not statistically significant. "There is no significant difference between BRCA1 and BRCA2 (which is not highlighted in the abstract or press release). Thus, it is not a sound conclusion that BRCA2 but not BRCA1 is associated with improved survival. This apparent contradiction is all a problem with sample size and an overestimation of statistical significance. With just 35 BRCA1 cases, the power to detect a difference with noncarriers is limited, so the conclusion is not sound."
Dr. Pharoah added that "even where a result is statistically significant, it is more likely to be wrong when the sample size is small than when it is large — a fact that is very rarely appreciated."
The authors have disclosed no relevant financial relationships.
JAMA. 2011;306:1557-1565, 1597-1598. Abstract, Editorial
Δεν υπάρχουν σχόλια:
Δημοσίευση σχολίου