First-degree relatives of pediatric cancer patients are two to four times as likely to develop a malignancy when compared with first-degree relatives of children without cancer, according to a study scheduled for publication in the Nov. 15 issue of the International Journal of Cancer ( 2013;10:2444-2453 ).
“One of the first questions that families ask at diagnosis is, ‘will my other children get cancer?,’” said the study's lead investigator, Joshua Schiffman, MD, Medical Director of the Huntsman Cancer Institute High-Risk Pediatric Cancer Clinic and the Edward B. Clark Endowed Chair in Pediatric Research at the University of Utah.
“Previously, physicians didn't have any tools to provide a reliable answer, but now after assessing the Utah Population Database, we can say it depends on the patient's family history of cancer.” However, even with an increased risk, a sibling is not necessarily going to develop a malignancy, he added.
Based on the findings, the authors recommend collection of three generations of family medical history for all newly diagnosed pediatric cancer patients and referral of families with a history of early-onset cancers in children or adults for genetic counseling. Additionally, parents of children diagnosed with cancer before age five with a family history of cancer should be advised of the potential for increased risk to other children in the family.
While the oncologists interviewed for this story noted that the study underscores the importance of obtaining a detailed and, perhaps, more expanded family history of children with cancer, not all agreed with the research authors regarding what to tell families about cancer risk or when to refer them for genetic counseling.
Using the database, the researchers—first author was Karen Curtin, PhD—compared the family medical history of 4,482 children diagnosed with cancer from 1966 to 2009 with matched population controls in a 5:1 ratio to determine the cancer risk in their relatives.
The results showed that when children were diagnosed with any kind of cancer at age 18 or younger, the risk to first-degree relatives was double that of families with no pediatric cancer patients. Second-degree relatives also had nearly double the risk.
First-degree relatives of children diagnosed before age five had a 3.6-fold increased risk of developing pediatric cancer, while second-degree relatives had a 2.5-fold risk and third-degree relatives were at a two-fold risk. A higher risk associated with younger age may be an indication that genetics play a large role in the etiology of pediatric cancer, Schiffman said.
“What's driving these numbers are the families who have an increased risk of cancer—those presumably with hereditary cancer syndromes,” he said, referring to Li-Fraumeni syndrome (LFS). “When we looked closely at cancer in pediatric patients, nearly 20 percent had a family history of cancer, which may indicate an underlying genetic syndrome,” he said. In comparison, the literature indicates that five to 10 percent of adult cancers are due to a hereditary disorder.
Unique Database, But Questions
This study is unique because the Utah Population Database “provides more detailed and more accurate information than was previously available on familial cancer history of patients with pediatric cancer,” said Denise Rokitka, MD, MPH, Director of the Pediatric Long-Term Follow-Up Clinic at Roswell Park Cancer Institute.
Added Laura L. Worth, MD, PhD, Professor and Center Medical Director in Pediatrics Patient Care at the University of Texas MD Anderson Children's Cancer Hospital: “Most of the time, when we obtain history and physicals, families rely on memory to provide information.”
The Utah Population Database “is one of the most unique resources in the world for conducting genetic, epidemiology, and public health studies,” Schiffman said. “It contains more than seven million individuals who are connected through marriage, birth, and death certificates, as well as driver licenses, and is record-linked to the Utah Cancer Registry.”
While the National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) Program database and hospital cancer registries are extremely useful, they capture only a diagnosed population without linking relatives, said Kate Gowans, MD, a pediatric hematologist/oncologist at Beaumont Children's Hospital in Michigan.
Still, Worth said, while the database is a valuable resource, “you can't necessarily apply it to the rest of the U.S. population.” Oncologists need to know whether the results illustrate a true trend.
One reason the study may not be generally applicable overall is that the average child in Utah has more siblings and relatives than do children living in other states, Gowans said. The population may also be somewhat more homogenous than the rest of North America, noted Alan Gamis, MD, Chief of the Section of Oncology at Children's Mercy Hospitals and Clinics in Kansas City.
Schiffman said that questions about the homogeneity of the database often arise, especially given Utah's large Mormon population—“It is actually a common misconception that the Utah population is genetically homogenous,” he said.
Indeed, Schiffman said, data from several papers have shown that the inbreeding rate in the Utah Mormon population is about the same as in the rest of the United States. “Mormons are also genetically much like other northern European populations, he said. “Moreover, while it's true that there may be large families in Utah, this should not affect our findings, especially given the 5:1 controls,” he said.
“Our study was done mostly in Caucasians, the majority of the state, but also the majority of the United States,” Schiffman continued. “Therefore, we believe it will be widely generalizable.” However, the study will still need to be replicated in other ethnicities, he said.
Expanding Family History
The study indicates that physicians need obtain a more detailed, three-generation family history for patients and update that history annually, Rokitka agreed.
Surya Rednam, MD, Director of the Childhood Cancer Prevention and Screening Clinic at Texas Children's Cancer Center, cautioned, though, that in Houston simply limiting the family history interview questions to cancers in parents or siblings is not enough. Establishing a comprehensive pedigree on all patients with childhood cancer and maintaining an up-to-date family history clearly has value, he said
“While oncologists are aware of the presence of familial cancer syndromes, most of our questions in the course of daily visits tend to focus on immediate issues, with questions about family history often neglected, leading to conditions like Li-Fraumeni Syndrome and other hereditary cancer predisposition syndromes being underdiagnosed.”
Obtaining a comprehensive family history can help oncologists raise the likelihood of patients being referred for genetic testing, identifying any underlying cancer susceptibility, and evaluating family members who are also may be at risk, he continued.
“It's imperative to start asking about a family history—not necessarily right when a child is diagnosed, but at some point during or shortly after treatment is complete,” Schiffman said.
While the study emphasizes that a comprehensive family history of cancer should be taken and regularly updated, “what we don't want to do is cause parents to panic about the risk in their other children,” Gowans said. When a child is diagnosed, parents want to know whether they should be worried about their other children, and the answer is generally no.”
Also, Gowans continued, what are families to do with the information? “If you had a cancer screening process that could be put in place, that would be a different situation. But that's not the case, and it's good to have this information only if you can do something about it.”
Oncologists also need to make sure that families are informed about risk in a very accurate way, Gamis said. For example, 1 in 300 males and 1 in 333 female children will develop cancer before age 20, according to SEER. If relatives, specifically siblings in this case, have a two-fold risk for cancer development, this means that 2 in 300 males and 2 in 333 females will have to contend with a malignancy.
The Schiffman paper also recommends referral of families with a history of early-onset cancers in children or adults for genetic counseling.
Each family should be assessed carefully, Gowans said. For example, a toddler with leukemia or a brain tumor who has a mother with breast cancer before age 50, sounds like a family who may benefit from a genetics referral. “Still, there is always the chance it may uncover nothing, or something of unclear significance.”
If a family has a young child with cancer and a positive family history or adults who have developed malignancy at younger age than normal and there are other young children to consider, Gamis suggests having a discussion about genetic testing. Oncologists should also assess whether families want to undergo screening for a rare genetic syndrome.
The current criteria for when to test for LFS p53 mutations, for example, focuses on individuals and first-degree relatives with cancer but does not include a more extensive family history, Rokitka said.
Patients with genetic family syndromes such as LFS can benefit from surveillance and early treatment, making genetic evaluation especially necessary, Schiffman said. Also a recent study found that up to 29 percent of pediatric cancer survivors are eligible for genetic evaluation for hereditary cancer syndromes, although she said that given the new results, she suspects this number may be even higher.