Table How to Use the Claus Model With Various Combinations of

Document Sample
Table How to Use the Claus Model With Various Combinations of Powered By Docstoc
					Table: How to Use the Claus Model With Various Combinations of Relatives*

Affected Relatives (most      Selection for Claus Model      Instructions
severe)                       use
Mother                        1 first-degree relative        Enter output of model
Sister                        1 first-degree relative        “
Daughter                      1 first-degree relative        “
Father                        1 first-degree relative        “
Brother                       1 first-degree relative        “
Son                           1 first-degree relative        “
Only one of the following:    Model does not apply as        Enter “0”
aunt, grandmother,            configured for trial
granddaughter, uncle,
grandfather, grandson

Mother and Sister             Two first-degree relatives     Enter output of model
Mother and Daughter           Two first-degree relatives     “
Two sisters                   Two first-degree relatives     “
Mother and Brother            Two first-degree relatives     “
Sister and Daughter           Two first-degree relatives     “

Mother and maternal aunt      Mother and maternal aunt       Enter output of model
Mother and maternal           Mother and maternal aunt       “
Sister and granddaughter      Mother and maternal aunt       “
Daughter and paternal aunt    Mother and paternal aunt       “
Paternal aunt and paternal    Two second-degree              “
grandfather                   relatives
Sister and paternal uncle     Mother and paternal aunt       “
Maternal half-sister and      Maternal aunt and paternal     “
paternal aunt                 aunt
Sister and maternal cousin    One relative, first-degree     Use output considering only

Maternal aunt and paternal    None of the above              Enter “0”
Maternal cousin and           None of the above              Enter “0”
maternal aunt
Three cousins                 None of the above              Enter “0”
*Not all possible combinations are listed.

In our application of the model, when there are two or more relatives with breast cancer,
we ask you to choose the two most severe (closest relations and youngest ages at
diagnosis). The choices of specific relations are as follows:
                Two first degree relatives
               Mother and one maternal aunt (i.e. one first degree relative and one
               maternal second-degree relative)
               Mother and one paternal aunt (i.e. one first degree and one paternal
               second-degree relative)
               Two second-degree relatives from same side
               None of the above

First-degree relatives are the following, with each equivalent to “mother”: mother, sister,
daughter, father, brother, son. When there are two first-degree relatives affected, you will
enter the age at diagnosis of the relative affected at the younger age first.

Second-degree relatives are the following, with each equivalent to “aunt”: aunt, uncle,
grandmother, grandfather, granddaughter, grandson, half-sister, half-brother.

The following relations are not considered in the model: cousins, great-aunts or uncles,
or any other third or more remote degree relative.

The rationale for using this table is based upon a review of the Claus paper (Claus EB,
Risch N, Thompson WD, Cancer 1994;73:643-651) and model by Dr. Wendie Berg and
Dr. Mark Schleinitz. Claus et al explicitly state that the data are from the CASH study
and are derived from white women with family history in the mother or sister. Claus et al
find that risk estimates fit an autosomal dominant inheritance pattern, and generalize the
risk estimates to second degree relatives (e.g. aunt) as well. Specifically, on page 647 of
the paper, it is stated: “Once a likely model has been identified through the use of
segregation analysis, it may be extended to provide estimates of risk for classes of
relatives not originally available for inclusion in the reference population used in the
analysis, in this case, second-degree relatives.” Male relatives are not included in the
reference population, but male breast cancer is more likely genetic, and they should not
be excluded. Claus et al do not distinguish among different second-degree relatives (i.e.
aunt is same as grandmother or granddaughter). Claus et al further state that there is
likely underreporting of second degree family history so that it is particularly important
to use it when it is available. While the data derive from white women, the Claus model
has been successfully applied to minority women as well. Importantly, there is a higher
risk if two relatives are both on the same side of the family (e.g. maternal or paternal), so
we need to distinguish that aspect of family history. While the model includes tables
with only a single second-degree relative affected, or with two second-degree relatives on
different sides of the family, it is never possible to attain a lifetime risk of 25% in these
situations so that those data were not included in our application for the ACRIN 6666
protocol (