Advances in genome-wide CLL linkage

In this issue of Blood, Sellick and colleagues update a previous report¹  with a combined genome-wide linkage study in 206 families with chronic lymphocytic leukemia (CLL), making their study the largest effort so far to localize CLL predisposition genes. The findings of this study are important because they provide information to pursue familial predisposition to the most frequent type of adult leukemia, CLL.

Single nucleotide polymorphism (SNP)–based genotyping has matured quickly, and it is now possible to analyze hundreds of thousands of markers per genome in a single assay. “Chipping away” at the genome has become fashionable in the scientific community, and several studies are being published every month. However, without some clear concept of how to scrutinize the resulting candidate loci, many studies seem not to fulfill their promises completely. The major questions are not whether a locus is dominant or recessive, penetrant or merely modifying disease risk. It boils down to whether there is a gene or gene-controlling element (promoter, enhancer, or suppressor) that would make functional sense to be broken or dysfunctional in a given disease. Thus, when a genome-wide linkage analysis results in a map of candidate loci, the hard work follows in formulating testable hypotheses to guide the mappers to the right gene loci.

Sellick and colleagues have undertaken the first major step by analyzing which families contribute to which candidate loci. Unfortunately for all of us, that is where the story ends, without clear coordinates for which genes to implicate in the pathogenesis of familial CLL. The authors have speculated about a gene in 2q21, CXCR4, and SMAD7 in 18q21, but there are no significant results about these 2 genes or any other in these regions. This represents the challenge of studying such genes, in which correct identification of germ line mutations can take teams many years to fully validate. Responsible genetic factors were once more confirmed to be heterogeneous, which complicates hunting for familial CLL predisposition genes.^2,3  Publication of valuable, but not conclusive, work such as that of Sellick and colleagues is important because it provides all those working in the field of CLL familial predisposition with information to pursue further.

Despite the effort, why were no definitive loci identified for future study? The problem may, in part, be due to the study cohort; 122 families (approximately 60%) had only 2 affected individuals, and a further 54 (approximately 26%) contained only 3 affected individuals. These pedigrees provide only limited linkage power. On the other hand, there were 2 large families with 10 and 12 individuals with CLL, respectively. Such families should provide substantial linkage. The main drawback appears to be the lack of unaffected relatives in the study, as only 54 were included to reconstruct genotypes. For decades, geneticists have successfully studied single-gene disorders in relatively few large sibships and collected all obtainable samples. In cancer research, the notion of collecting samples from healthy relatives is still stigmatized because they may not (yet) present with a phenotype, and there is great potential to raise the anxiety of the uninvolved person. Nevertheless, such samples provide powerful means to exclude large chromosomal regions. People are understandably afraid to know what their genome holds in store for them in the future, and extensive counseling is required to provide patients with confidence in the value of contributing such samples. Additionally, collaborative efforts such as the one described by Sellick and colleagues are required so that a large number of extensively characterized families with CLL predisposition can be studied. The separate academic groups included in this work are to be commended for their selfless collaboration.