Familial Cancer Associated with a Polymorphism in ARLTS1
http://www.100md.com
《新英格兰医药杂志》
To the Editor: Calin et al. (April 21, 2005, issue)1 proposed that the truncating sequence change in ARLTS1, G446A (Trp149Stop), predisposes persons to a wide range of malignant conditions including chronic lymphocytic leukemia (CLL). The assertion that this variant predisposes persons to CLL was, however, based on observations from only 17 familial patients of mixed ethnic background.
We conducted an association study to determine if this variant or five other coding single-nucleotide polymorphisms (cSNPs) in ARLTS1 predispose to CLL. We analyzed germ-line DNA from 413 patients with CLL (259 men and 154 women; mean age, 58 years; range, 33 to 89) and 471 healthy population controls (281 men and 190 women; mean age, 58 years; range, 51 to 70) by direct sequencing. None of the controls had a history of cancer. To avoid population stratification, which may have affected the findings of Calin et al., all patients and controls were British whites, with no obvious differences in place of birth or residence.
We found all six cSNPs to be polymorphic in patients and controls, and there was no evidence of population stratification; the genotype distribution of all SNPs was in Hardy–Weinberg equilibrium. There was no significant association between any of the ARLTS1 cSNPs and the risk of CLL (odds ratio for Ser99Ser, 0.71; 95 percent confidence interval, 0.39 to 1.25; P=0.13; odds ratio for Pro131Leu, 1.47; 95 percent confidence interval, 0.84 to 2.57; P=0.09; odds ratio for Leu132Leu, 1.14; 95 percent confidence interval, 0.30 to 4.29; P=0.52; odds ratio for Cys148Arg, 0.99; 95 percent confidence interval, 0.82 to 1.20; P=0.47; odds ratio for Trp149Stop, 1.47; 95 percent confidence interval, 0.49 to 4.67; P=0.30; odds ratio for Glu164Lys, 1.14; 95 percent confidence interval, 0.15 to 89.6; P=0.72). In addition, there was no evidence of epistatic interactions between ARLTS1 haplotypes and the risk of CLL. Furthermore, stratification by family history of CLL (in 158 patients) did not significantly influence our findings. Our observations do not support the hypothesis that variants in ARLTS1 function as multiorgan susceptibility alleles with respect to the risk of CLL.
Gabrielle S. Sellick, Ph.D.
Daniel Catovsky, M.D., D.Sc.
Richard S. Houlston, M.D., Ph.D.
Institute of Cancer Research
Sutton, Surrey SM2 5NG, United Kingdom
References
Calin GA, Trapasso F, Shimizu M, et al. Familial cancer associated with a polymorphism in ARLTS1. N Engl J Med 2005;352:1667-1676.
The authors reply: In our paper we reported that G446A has a statistically significant association with the risk of familial cancer, as compared with sporadic cancer; we did not find a differential frequency when we compared sporadic cancers with healthy controls. Further independent studies reported statistically significant associations of Trp149Stop and Cys148Arg with the risk of familial breast cancer.1,2 For these reasons, it is quite surprising to us that Sellick et al. report their results in all patients with CLL without considering the family history. It would be interesting to see the odds ratio calculated from their data after stratification by family history. Also, it would be of interest to know if there were any cases among their patients who had CLL with ARLTS1 (Trp149Stop variant) of familial CLL with evidence of segregation of the malignant phenotype along with the genetic abnormality, as was the case in the CLL family we reported. In our article, we stated, "we propose that ARLTS1 is a low-penetrance tumor-suppressor gene that accounts for a small percentage of familial melanoma or familial CLL." In light of the multigenic predisposition to familial CLL, the letter from Sellick et al. does not negate this hypothesis.
George A. Calin, M.D., Ph.D.
Carlo M. Croce, M.D.
Ohio State University Comprehensive Cancer Center
Columbus, OH 43210
carlo.croce@osumc.edu
Donna Neuberg, Sc.D.
Dana–Farber Cancer Institute
Boston, MA 02115
References
Frank B, Klaes R, Burwinkel B. Familial cancer and ARLTS1. N Engl J Med 2005;353:313-314.
Frank B, Hemminki K, Meindl A, et al. Association of the ARLTS1 Cys148Arg variant with familial breast cancer risk. Int J Cancer (in press).
We conducted an association study to determine if this variant or five other coding single-nucleotide polymorphisms (cSNPs) in ARLTS1 predispose to CLL. We analyzed germ-line DNA from 413 patients with CLL (259 men and 154 women; mean age, 58 years; range, 33 to 89) and 471 healthy population controls (281 men and 190 women; mean age, 58 years; range, 51 to 70) by direct sequencing. None of the controls had a history of cancer. To avoid population stratification, which may have affected the findings of Calin et al., all patients and controls were British whites, with no obvious differences in place of birth or residence.
We found all six cSNPs to be polymorphic in patients and controls, and there was no evidence of population stratification; the genotype distribution of all SNPs was in Hardy–Weinberg equilibrium. There was no significant association between any of the ARLTS1 cSNPs and the risk of CLL (odds ratio for Ser99Ser, 0.71; 95 percent confidence interval, 0.39 to 1.25; P=0.13; odds ratio for Pro131Leu, 1.47; 95 percent confidence interval, 0.84 to 2.57; P=0.09; odds ratio for Leu132Leu, 1.14; 95 percent confidence interval, 0.30 to 4.29; P=0.52; odds ratio for Cys148Arg, 0.99; 95 percent confidence interval, 0.82 to 1.20; P=0.47; odds ratio for Trp149Stop, 1.47; 95 percent confidence interval, 0.49 to 4.67; P=0.30; odds ratio for Glu164Lys, 1.14; 95 percent confidence interval, 0.15 to 89.6; P=0.72). In addition, there was no evidence of epistatic interactions between ARLTS1 haplotypes and the risk of CLL. Furthermore, stratification by family history of CLL (in 158 patients) did not significantly influence our findings. Our observations do not support the hypothesis that variants in ARLTS1 function as multiorgan susceptibility alleles with respect to the risk of CLL.
Gabrielle S. Sellick, Ph.D.
Daniel Catovsky, M.D., D.Sc.
Richard S. Houlston, M.D., Ph.D.
Institute of Cancer Research
Sutton, Surrey SM2 5NG, United Kingdom
References
Calin GA, Trapasso F, Shimizu M, et al. Familial cancer associated with a polymorphism in ARLTS1. N Engl J Med 2005;352:1667-1676.
The authors reply: In our paper we reported that G446A has a statistically significant association with the risk of familial cancer, as compared with sporadic cancer; we did not find a differential frequency when we compared sporadic cancers with healthy controls. Further independent studies reported statistically significant associations of Trp149Stop and Cys148Arg with the risk of familial breast cancer.1,2 For these reasons, it is quite surprising to us that Sellick et al. report their results in all patients with CLL without considering the family history. It would be interesting to see the odds ratio calculated from their data after stratification by family history. Also, it would be of interest to know if there were any cases among their patients who had CLL with ARLTS1 (Trp149Stop variant) of familial CLL with evidence of segregation of the malignant phenotype along with the genetic abnormality, as was the case in the CLL family we reported. In our article, we stated, "we propose that ARLTS1 is a low-penetrance tumor-suppressor gene that accounts for a small percentage of familial melanoma or familial CLL." In light of the multigenic predisposition to familial CLL, the letter from Sellick et al. does not negate this hypothesis.
George A. Calin, M.D., Ph.D.
Carlo M. Croce, M.D.
Ohio State University Comprehensive Cancer Center
Columbus, OH 43210
carlo.croce@osumc.edu
Donna Neuberg, Sc.D.
Dana–Farber Cancer Institute
Boston, MA 02115
References
Frank B, Klaes R, Burwinkel B. Familial cancer and ARLTS1. N Engl J Med 2005;353:313-314.
Frank B, Hemminki K, Meindl A, et al. Association of the ARLTS1 Cys148Arg variant with familial breast cancer risk. Int J Cancer (in press).