Predictors of Cognitive Performance Among Infants Treated for Brain Tumors: Findings From a Multisite, Prospective, Longitudinal Trial

Abstract

Purpose: Infants treated for CNS malignancies experience a significantly poorer response to treatment and are particularly at risk for neuropsychological deficits. The literature is limited and inconsistent regarding cognitive outcomes among this group. We investigated predictors of cognitive outcomes in children treated for brain tumors during infancy as part of a large, prospective, multisite, longitudinal trial.

Patients and methods: One hundred thirty-nine infants with a newly diagnosed CNS tumor were treated with chemotherapy, with or without focal proton or photon radiation therapy (RT). Cognitive assessments were conducted at baseline, 6 months, 1 year, and then annually for 5 years. The median length of follow-up was 816 days (26.8 months). Neurocognitive testing included assessment of intellectual functioning (intellectual quotient [IQ]), parent ratings of executive functioning and emotional and behavioral functioning, and socioeconomic status.

Results: At baseline, IQ, parent-reported working memory, and parent-reported adaptive functioning were worse than normative expectations. Baseline cognitive difficulties were associated with younger age at diagnosis and lower socioeconomic status. Linear mixed models did not demonstrate a decline in IQ over time. There were increased parent-reported attention and executive problems over time. Increased concerns were related to supratentorial tumor location and CSF diversion. There were no differences in cognitive outcomes based on treatment exposure (chemotherapy-only v chemotherapy with RT and proton v photon focal RT).

Conclusion: Even before adjuvant therapy, young children with brain tumors experience cognitive difficulties that can affect quality of life. Changes in cognitive functioning over time were dependent on tumor location and surgical factors rather than adjuvant therapy. These findings may serve to guide treatment planning and indicate targets for cognitive monitoring and intervention.

Trial registration: ClinicalTrials.gov NCT00602667.

FIG 1.

Determination of low-, intermediate-, and high-risk groups. ATRT, atypical teratoid rhabdoid tumor; GTR, gross total resection; M, metastatic disease; M+, any metastatic disease; M0, no metastatic disease; PNET, primitive neuroectodermal tumor.

FIG 2.

Planned treatment on the basis of risk group. ^aThe guidelines for RT specified a 0.5-cm clinical target volume and 0.3-cm planning target volume margins surrounding the postoperative tumor bed and a prescribed dose of 54 Gy using photons or 54 Gy (RBE) using protons. Carbo, carboplatin; CDDP, cisplatin; CPM, cyclophosphamide; CSI, craniospinal irradiation; MTX, methotrexate; PO, by mouth or gastrostomy tube; RBE, relative biological effectiveness; RT, radiation therapy; Topo, topotecan; VBL, vinblastine; VCR, vincristine; VP16, etoposide.

FIG 3.

Neurocognitive performance at baseline and 2 years postdiagnosis. ^aSignificance calculated at P < .01. BASC Attn Prob, BASC-II Attention Problems scale; BRIEF GEC, BRIEF-P Global Executive Composite scale; BRIEF WM, BRIEF-P Working Memory scale; IQ, intellectual quotient; SD, standard deviation.

FIG A1.

Flow diagram.

FIG A2.

Profile plot for BRIEF Global Executive Functioning Scores over time. Blue lines show individual profiles. The red line shows the fitted line estimated by a random coefficients model.

FIG A3.

Profile plot for BASC-II Attention Composite Scores over time. Blue lines show individual profiles. The red line shows the fitted line estimated by a random coefficients model.