Inflammatory brain fluid may drive structural growth in preterm hydrocephalus

One of the most feared complications of extreme prematurity is acquired hydrocephalus, the abnormal accumulation of cerebrospinal fluid within the brain’s ventricles. In infants born very early, the condition most often follows intraventricular hemorrhage, bleeding into the fluid-filled spaces of the brain, or a central nervous system infection during the newborn period. Both insults disproportionately strike the smallest and most immature infants, and both can leave survivors with lifelong neurological disability. Yet only a subset of babies who suffer a severe hemorrhage actually progress to persistent hydrocephalus requiring permanent surgical diversion of fluid, and clinicians currently lack reliable tools to predict which infants will deteriorate. A new study published in iScience now offers a mechanistic explanation for that progression, pointing to an unexpected player: the choroid plexus, the delicate tissue lining the ventricles that manufactures cerebrospinal fluid, and its apparent tendency to physically remodel when exposed to inflammatory fluid during a narrow developmental window.

The research team, led by Leandro Castañeyra-Ruiz and colleagues at CHOC Children’s Research Institute, integrated three lines of evidence: quantitative magnetic resonance imaging in a cohort of forty-two preterm and term infants, multiplex profiling of inflammatory proteins in cerebrospinal fluid, and laboratory experiments using primary human choroid plexus cells. Their central hypothesis was that inflammatory cerebrospinal fluid, rich in growth factors such as vascular endothelial growth factor and epidermal growth factor, could reactivate proliferative programs that are normally active only during fetal development, causing the choroid plexus to enlarge in infants who go on to develop hydrocephalus.

The imaging findings were striking in their consistency. On T2-weighted MRI scans, choroid plexus thickness increased in a stepwise fashion across clinical categories. In the axial plane, mean thickness measured 3.44 millimeters in control infants, 4.53 millimeters in preterm infants with intraventricular hemorrhage managed without surgery, and 5.39 millimeters in infants who ultimately required cerebrospinal fluid diversion. Sagittal measurements showed an even more pronounced gradient, rising from 3.51 millimeters in controls to 6.19 millimeters in the diversion group. Statistical analysis revealed that clinical category alone explained roughly 38 to 41 percent of the variance in thickness, and the differences between the most severely affected infants and controls were highly significant in both imaging planes.

Critically, the researchers went beyond simple group comparisons. Because infants requiring diversion were also the most premature at birth, with a mean gestational age of just over 27 weeks compared with nearly 38 weeks in controls, the team used multivariable regression to adjust for gestational age at birth and postmenstrual age at imaging. The association between clinical progression and choroid plexus thickness survived these adjustments, indicating that the enlargement was not simply a byproduct of earlier birth or a different maturational stage at the time of scanning. The team also controlled for ventricular size using the frontal and occipital horn ratio, a standard measure of ventricular dilation, and for hemorrhage severity grade. Even after full adjustment, clinical category remained independently associated with thickness, while ventricular size itself showed no independent relationship with axial thickness. In other words, the choroid plexus appeared to be enlarging as part of the disease process itself, not merely being stretched by expanding ventricles.

To probe the cellular basis of this remodeling, the researchers turned to laboratory experiments with primary human choroid plexus epithelial cells and endothelial cells, the two principal cell types of the tissue. They exposed the cells for seventy-two hours to different cerebrospinal fluid conditions: an artificial fluid control, a low-inflammatory fluid obtained from patients with spinal muscular atrophy, and a growth factor-enriched inflammatory fluid derived from patients with severe traumatic brain injury. Multiplex cytokine analysis confirmed that the inflammatory fluid carried significantly elevated concentrations of vascular endothelial growth factor, and that hemorrhage-derived and injury-derived inflammatory fluids shared this growth factor signature.

The proliferative response was substantial. In epithelial cells, the proportion of nuclei staining positive for proliferating cell nuclear antigen, a marker of active DNA synthesis, rose from 23.7 percent under artificial fluid conditions to 42.1 percent under inflammatory exposure. In endothelial cells the shift was even larger, climbing from 44.26 percent to 77.22 percent. When the researchers supplemented the culture media with recombinant epidermal growth factor, epithelial proliferation surged to 60 percent in artificial fluid and 71.7 percent in low-inflammatory fluid, while recombinant vascular endothelial growth factor likewise boosted endothelial proliferation. These results demonstrated that growth factors alone are sufficient to drive robust proliferation in these immature tissue cells.

The team then tested whether the response depended on specific growth factor receptors. Blocking the epidermal growth factor receptor with erlotinib numerically reduced epithelial proliferation back toward baseline, though the reduction did not reach statistical significance, so the authors were careful to note that epidermal growth factor receptor dependence was not formally established. In contrast, pharmacological inhibition of the vascular endothelial growth factor receptor significantly reduced endothelial proliferation, cutting it by nearly 40 percentage points relative to inflammatory fluid alone and restoring it toward control levels. Taken together, the pharmacological evidence was stronger for vascular endothelial growth factor pathway involvement than for the epidermal growth factor arm, but both experiments supported the general principle that inflammatory fluid acts through growth factor signaling.

Independent published datasets lent further biological plausibility to the model. Reanalysis of a publicly available RNA sequencing dataset from a neonatal rat hemorrhage model revealed a 12.7-fold elevation of Mki67, a marker of actively cycling cells, in choroid plexus tissue from hemorrhaged animals compared with controls. Separate work has shown that hemoglobin, the primary injury mediator in intraventricular hemorrhage, directly induces choroid plexus epithelial proliferation at concentrations 37 to 65 percent above control levels. Clinical proteomic studies have also documented a temporal arc: inflammatory cytokines such as interleukin-6 and tumor necrosis factor alpha are elevated during the acute phase of post-hemorrhagic hydrocephalus and decline with treatment, while transthyretin, a constitutive secretory protein of the choroid plexus, is sharply suppressed during the acute phase and recovers as treatment progresses. This pattern is consistent with a two-stage process in which acute inflammatory injury triggers proliferative activation and altered epithelial identity, followed by persistence of a structurally remodeled tissue state.

The developmental timing of these events may be the key to why prematurity matters so much. During late gestation, roughly 24 to 28 weeks of human development, the epithelial and endothelial compartments of the choroid plexus remain structurally immature and retain proliferative competence. The tissue expresses receptors for insulin-like growth factor 1, epidermal growth factor, and vascular endothelial growth factor, and it responds to these signals. In a full-term infant, these growth programs have largely quieted. But in a baby born at the edge of viability, an inflammatory surge from hemorrhage or infection may find a tissue still poised to grow, effectively prolonging developmental programs that should have been time-limited and producing structural expansion that persists beyond the acute inflammatory phase.

The authors are careful about what their findings do and do not show. Cerebrospinal fluid secretion rates and transporter activity were not directly measured, so choroid plexus enlargement on MRI should not be read as evidence of fluid overproduction, and hydrocephalus remains a multifactorial disease involving impaired fluid resorption, scarring, and pathway obstruction. The in vitro experiments used injury-derived inflammatory fluid as a model of growth factor-enriched exposure rather than hemorrhage-specific neonatal fluid, and the cross-sectional imaging design cannot establish whether enlargement persists over time. Still, if the findings are validated prospectively, choroid plexus thickness could become an adjunct imaging biomarker for identifying which infants with hemorrhage are most likely to progress to shunt dependency, and modulation of growth factor pathways within defined developmental windows could emerge as a testable therapeutic strategy for a condition that currently offers clinicians few options beyond surgery.

Subject of Research: Inflammatory cerebrospinal fluid exposure and choroid plexus remodeling in preterm acquired hydrocephalus

Article Title: Choroid plexus remodeling associated with inflammatory CSF exposure in preterm acquired hydrocephalus

Article References: Jison, G., Ramos, N., Cotrim-Gomes, F., Lee, S., Hanak, B. W., Muhonen, M., & Castañeyra-Ruiz, L. (2026). Choroid plexus remodeling associated with inflammatory CSF exposure in preterm acquired hydrocephalus. iScience, 29(10), Article 117759. https://doi.org/10.1016/j.isci.2026.117759

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.117759

Keywords: hydrocephalus, choroid plexus, preterm infants, intraventricular hemorrhage, cerebrospinal fluid, inflammation, VEGF, EGF, MRI, proliferation, neonatology, iScience