Large studies link phototherapy to increase in childhood cancers

Last Updated: 2016-05-23

By Anne Harding

NEW YORK (Reuters Health) - Two large new studies confirm an association between phototherapy and an increased risk of some childhood cancers.

While it remains unclear whether the relationship is causal, and the absolute increase in risk is small, physicians should use phototherapy judiciously, especially in babies with Down syndrome, who have a higher baseline cancer risk, Dr. Thomas B. Newman of the University of California, San Francisco, who helped conduct both studies, told Reuters Health in a telephone interview.

Phototherapy has been shown to cause DNA damage in vivo, while epidemiological studies focusing on leukemia have had mixed results. The new studies by Dr. Newman and colleagues include a retrospective cohort study of nearly 500,000 children followed for at least 60 days after birth (LIGHT, the Late Impact of Getting Hyperbilirubinemia or Phototherapy study) and an analysis of a state-wide data set on 5.1 million babies (CLIPS, the California Late Impact of Phototherapy Study). Both studies were published online May 23 in Pediatrics.

In CLIPS, Dr. Newman, Dr. Andrea C. Wickremasinghe of Kaiser Permanente Northern California (KPNC) in Santa Clara, and their colleagues used ICD-9 codes to identify children who received phototherapy before 15 days of age, as well as discharge diagnoses of cancer at 61 to 365 days. The dataset included 5,144,849 infants born in 1998-2007, 1,100 of whom had a discharge diagnosis of cancer.

Among the 178,017 infants with phototherapy diagnosis codes, 58 were diagnosed with cancer, versus 1,042 of the 4,966,832 infants who did not receive phototherapy. Cancer rates were 32.6 per 100,000 person-years for the phototherapy patients, versus 21.0 per 100,000 for those who did not receive phototherapy (relative risk 1.6, p=0.002).

Propensity-adjusted analyses found an adjusted odds ratio of 1.4 for overall cancer in infants who received phototherapy, while the aOR was 2.6 for myeloid leukemia and 2.5 for kidney cancer. The absolute risk increase in the population was 9.4 per 100,000, for a number needed to harm of 10,638.

The cohort included 7,812 infants with Down syndrome, 19% of whom received phototherapy. Eighteen developed cancer, including lymphoid leukemia in four, myeloid leukemia in 10, other leukemia in two, brain/nervous system cancers in one, and other cancer in one. The marginal propensity-adjusted absolute risk associated with phototherapy for children with Down syndrome was 77.8 per 100,000, with a number needed to harm of 1,285.

"Phototherapy may slightly increase the risk of cancer in infancy, although the absolute risk increase is small," Dr. Wickremasinghe and her colleagues write. "This should be considered when making phototherapy treatment decisions, especially for infants with bilirubin levels below current treatment guidelines."

One limitation of the larger study was that it did not include information on bilirubin levels and other potential confounding variables.

In LIGHT, which included 499,621 children born at at least 35 weeks' gestation at KPNC hospitals in 1995-2011, the investigators had information on bilirubin levels and other key confounders, and up to 19 years of follow-up data.

In LIGHT, 60 of the 39,403 children who received phototherapy (25 per 100,000 person-years) developed cancer, versus 651 of the 460,218 unexposed children (18 per 100,000, incidence rate ratio 1.4). While phototherapy was linked to an increased risk of any leukemia (IRR 2.1), nonlymphocytic leukemia (IRR 4.0) and liver cancer (IRR 5.2), the associations were no longer statistically significant when the researchers adjusted for bilirubin levels, chromosomal disorders, congenital anomalies, and other factors.

Currently, Dr. Newman noted, clinicians often use phototherapy to treat children with bilirubin levels below the recommended guidelines. "We think without having people panic or stop using phototherapy, probably a bit more caution is warranted."

When possible, he added, doctors should wait for 12 to 24 hours before initiating light therapy for infants with jaundice, whose bilirubin levels may return to normal during that time period. It's also not always unreasonable to give phototherapy to infants with levels below the guideline recommendations, Dr. Newman pointed out, for example children who will be difficult to follow up after discharge.

In an editorial accompanying the studies, Dr. Lindsay Frazier of Dana-Farber Cancer Institute in Boston and colleagues note that the LIGHT study showed that the percentage of infants at KPNC hospitals receiving phototherapy increased from 2.7% in 1995 to 15.9% in 2011, in part because the health system introduced home phototherapy. However, they add, there is no evidence for a corresponding increase in the incidence of potentially phototherapy-related tumors.

"In the end, acknowledging that the information is imperfect, general pediatricians and neonatologists must make a choice," Dr. Frazier and her team conclude. "These data suggest that phototherapy may not be harmless, and that the risks as well as the benefits need to be weighed before flipping the switch."

The Agency for Healthcare Research and Quality supported one study. The Gerber Foundation and the American Academy of Pediatrics supported the other. No authors reported disclosures.

SOURCE: http://bit.ly/1TSzcud, http://bit.ly/1YSSKmw and http://bit.ly/1XOfCVK

Pediatrics 2016.