The measurement in question is the Evans index, a ratio that radiologists have used for more than eight decades. First described by William Evans in 1942 as a way to estimate ventricular enlargement on pneumoencephalograms, the index is calculated by dividing the width of the frontal horns of the brain’s lateral ventricles by the maximum internal width of the skull at the same level. A value above 0.3 has traditionally signaled enlarged ventricles, whether from hydrocephalus or cerebral atrophy. What is new, and what the latest research confirms, is that this same ratio carries powerful prognostic information for patients whose cancer has spread to the leptomeninges.

In a study published in the Journal of Neuro-Oncology, researchers at the Instituto Nacional de Cancerología in Mexico City analyzed records from 345 patients with neoplastic meningitis from solid tumors who were seen between January 2010 and December 2024 and had imaging studies available for review. The cohort, one of the largest of its kind reported to date, was dominated by breast cancer, which accounted for 66 percent of cases, followed by lung cancer at 12 percent, genitourinary cancers at 6 percent, and head and neck cancers at 5 percent. This distribution reflects the epidemiology of leptomeningeal disease in the modern era, where breast cancer patients, living longer with better systemic therapies, increasingly face this devastating complication.

The headline finding is stark. Across the entire cohort, median overall survival was 6.9 months, with a 95 percent confidence interval of 5.5 to 8.3 months, underscoring how lethal the condition remains even at a major cancer center. But when the researchers stratified patients by their Evans index, the survival curves separated dramatically. Patients with an index of 0.3 or below had a median survival of 7.36 months, while those whose index exceeded 0.3 survived a median of only 2.99 months, a difference that reached statistical significance with a P value of 0.020. In practical terms, a single ratio measured on a scan that every patient already undergoes more than doubled the expected survival gap.

What makes this finding scientifically credible rather than merely suggestive is the rigor of the statistical adjustment. Enlarged ventricles could simply be a marker of older age, worse neurological status, or more advanced disease, all of which independently shorten survival. To rule out such confounding, the investigators built a multivariable model that accounted for age, sex, primary tumor type, Karnofsky Performance Score, neurological symptoms, activity of extracranial disease, and cerebrospinal fluid levels of both protein and lactic dehydrogenase, a marker of tissue breakdown. Even after all of these adjustments, an Evans index above 0.3 remained independently associated with mortality, carrying a hazard ratio of 1.95 with a 95 percent confidence interval of 1.09 to 3.50 and a P value of 0.024. In other words, patients with enlarged ventricles faced nearly twice the risk of death at any given time compared with those whose ventricles were normal in size.

The biological logic behind this association is plausible. Neoplastic meningitis can impair the resorption of cerebrospinal fluid, leading to a form of communicating hydrocephalus in which the ventricles gradually swell. Ventricular enlargement on imaging may therefore be a visible signature of disrupted fluid dynamics, rising intracranial pressure, and accumulating tumor burden within the nervous system. It may also capture the cumulative damage that cancer inflicts on the brain’s ability to manage its own fluid environment. If so, the Evans index functions less as a generic measure of brain shrinkage and more as a window into the pathophysiology of the disease itself, explaining why it retains predictive power even after accounting for clinical variables.

The study is particularly valuable because it provides what scientists call external validation. The prognostic role of ventricular size in leptomeningeal metastasis had been proposed before, most notably in a large European study published in Neurology in 2024 that examined ventricular size and its dynamics in patients with solid tumors. But prognostic biomarkers in oncology have a notorious history of failing when tested outside the populations in which they were discovered. By independently confirming the association in a distinct cohort, drawn from a different continent and health system, and by demonstrating robustness in a fully adjusted model, the Mexican team has strengthened the case that the Evans index is a genuine and transferable prognostic tool rather than a statistical fluke.

The clinical implications could be substantial. Decisions in neoplastic meningitis are among the most difficult in neuro-oncology. Options range from intrathecal chemotherapy and focal radiation to novel systemic agents that penetrate the blood-brain barrier, yet all carry toxicity, and evidence for survival benefit is limited to selected patient groups. Guidelines from the European Association of Neuro-Oncology and the European Society for Medical Oncology emphasize that prognosis should guide treatment intensity, but the prognostic tools available to date, such as performance status and cerebrospinal fluid markers, are imperfect. A ratio that requires nothing more than calipers on an existing magnetic resonance or computed tomography scan could be integrated into clinical trials as a stratification variable, into prognostic scores alongside existing clinical factors, and into everyday conversations between oncologists and patients about what lies ahead.

There are, of course, caveats. The study was retrospective and conducted at a single institution, which raises the possibility of selection bias, and the data could not be made publicly available because they contain potentially identifiable patient information. The Evans index is also a static measurement, capturing ventricular size at a single moment, whereas some researchers argue that serial changes in ventricular caliber over time may be even more informative. Prospective studies that track the index dynamically, and that test whether it improves decision-making in real time, would be the natural next step. Still, the effect size observed here, nearly a doubling of mortality risk after adjustment, is large enough that few would dismiss it.

For a disease that has resisted decades of therapeutic progress, the emergence of a cheap, universally available prognostic marker is a meaningful advance. The Evans index was never designed for this purpose; it was conceived in an era of air encephalography as a crude gauge of ventricular dilation. That it now appears to forecast survival in patients with cancer invading their central nervous system is a reminder that sometimes the most powerful tools in modern medicine are the simplest ones, hiding in plain sight on scans that clinicians already order every day. For the thousands of patients diagnosed with neoplastic meningitis each year, and for the physicians wrestling with how best to treat them, this humble ratio may soon become part of the standard vocabulary of care.

Subject of Research: Prognostic value of the Evans index, a ventricular size ratio on brain imaging, in patients with neoplastic meningitis from solid tumors

Article Title: Prognostic value of the Evans index in neoplastic meningitis: a real-world cohort study

Article References: González-Vázquez, A., Lorenzana-Mendoza, N. A., Reyes Pérez, J. A., & Cacho-Díaz, B. (2026). Prognostic value of the Evans index in neoplastic meningitis: a real-world cohort study. Journal of Neuro-Oncology, 179(2), Article 71. https://doi.org/10.1007/s11060-026-05780-4

Image Credits: AI Generated

DOI: 10.1007/s11060-026-05780-4

Keywords: neoplastic meningitis, leptomeningeal metastasis, Evans index, prognostic biomarker, brain imaging, hydrocephalus, ventricular enlargement, overall survival, breast cancer, neuro-oncology, cerebrospinal fluid, validation study

Post-hemorrhagic ventricular dilatation, commonly abbreviated as PHVD, is one of the most consequential neurological complications affecting premature infants. It arises when bleeding within the germinal matrix — a highly vascularized region near the ventricles that is particularly vulnerable in preterm babies — extends into the ventricular system itself. Blood products accumulating in the cerebrospinal fluid can impair the normal circulation and reabsorption of this fluid, causing the ventricles to expand progressively. As the ventricles enlarge, they compress the surrounding developing brain tissue, and without effective intervention, this pressure and distortion can contribute to significant long-term neurodevelopmental impairment, including cerebral palsy, cognitive deficits, and sensory problems.

The clinical management of PHVD follows a carefully staged logic. Because many infants will not ultimately require permanent cerebrospinal fluid diversion, neurosurgeons typically begin with temporizing measures designed to relieve pressure and drain excess fluid while the infant’s brain and body mature. Among these temporizing strategies is the ventriculosubgaleal shunt, or VSGS. In this procedure, a small catheter is inserted through a tiny opening in the skull into one of the lateral ventricles, and the other end is tunneled under the scalp into the subgaleal space — the potential space between the scalp’s galea aponeurotica and the periosteum covering the skull. Cerebrospinal fluid then drains from the ventricles into this pocket under the scalp, where it accumulates and is gradually reabsorbed by the body. The technique offers an external-free, closed-system approach that reduces infection risk compared with external ventricular drainage, while buying the infant’s brain valuable time.

The alternative temporizing approaches — including serial lumbar punctures, tap of the ventricular reservoir, or external ventricular drainage — each carry their own limitations, and the VSGS has gained favor in many centers precisely because it provides continuous drainage without repeated needle procedures or external tubing. However, the fundamental question confronting every clinical team remains the same: which infants will recover adequate independent cerebrospinal fluid dynamics, and which will demonstrate that their condition is permanent and requires definitive surgical treatment? When an infant’s ventricles continue to enlarge or fail to decrease in size despite temporizing drainage, and when cerebrospinal fluid protein levels and clinical findings support it, the child undergoes conversion to a ventriculoperitoneal shunt. The VPS is the gold standard for permanent cerebrospinal fluid diversion: a valve-regulated catheter runs from the ventrles, beneath the skin, all the way into the peritoneal cavity of the abdomen, where the fluid is permanently reabsorbed into the circulation.

Conversion from VSGS to VPS is extremely common — in most published series, a majority of infants treated with ventriculosubgaleal shunts ultimately receive permanent shunts — but the proportion varies considerably between studies and institutions. Understanding what distinguishes infants who can be spared permanent hardware from those who cannot has been an enduring clinical puzzle. Identifying reliable predictors early in the course would allow clinicians to counsel families more accurately, plan surveillance more intensively, and potentially refine treatment protocols to minimize unnecessary procedures. This is precisely the first question the new study set out to answer: the research team systematically analyzed factors associated with ventriculoperitoneal shunt conversion following VSGS placement in infants with post-hemorrhagic ventricular dilatation.

The second question the researchers addressed is arguably even more pressing from a safety standpoint. Intraparenchymal hemorrhage — bleeding directly into the substance of the brain, rather than into the ventricles or the subdural space — is a recognized but poorly quantified complication that can occur in the context of shunt-related interventions. Rapid or excessive drainage of cerebrospinal fluid can, in theory, alter pressure gradients across brain tissue and the bridging veins that traverse it, potentially precipitating tearing of these vessels or collapse-related bleeding. In the fragile, incompletely myelinated brain of a premature infant, whose tissue is more deformable and whose vasculature is more delicate than that of older children, these mechanical shifts are of particular concern. An intraparenchymal hemorrhage can be devastating, adding direct tissue injury on top of the ventricular dilatation the treatment was intended to relieve.

By identifying factors associated with intraparenchymal hemorrhage following VSGS, the study aims to give clinicians an evidence-based framework for anticipating and potentially mitigating this risk. The investigative approach implicit in the study design is a comparative analysis of infants who underwent VSGS placement, examining patient characteristics, disease severity markers, imaging findings, and treatment parameters against the two outcomes of interest. Neuroimaging plays a central role in this kind of analysis: cranial ultrasound is the workhorse for monitoring ventricular size in premature infants, while magnetic resonance imaging, interpreted by neuroradiologists such as the study’s senior imaging author, provides detailed characterization of hemorrhage extent, parenchymal injury, and post-treatment complications. Distinguishing pre-existing parenchymal injury from new post-procedural bleeding is a critical methodological consideration, and the involvement of experienced neuroradiology expertise in the study reflects the diagnostic precision this requires.

The clinical significance of this line of investigation extends well beyond the walls of the neurosurgical operating room. Roughly speaking, among very low birth weight infants, intraventricular hemorrhage occurs in a substantial fraction of those born extremely preterm, and of those who experience hemorrhage, a smaller but meaningful proportion progress to progressive ventricular dilatation requiring intervention. With survival rates for extremely premature infants continuing to improve worldwide, the number of children entering this treatment pathway is substantial, and the long-term neurological outcomes of this population carry enormous personal, familial, and societal weight. Every refinement in the ability to predict which infants need permanent shunts — and every insight into avoiding procedural complications — translates into concrete differences in developmental trajectories for these children.

The study also speaks to a broader and ongoing debate in pediatric neurosurgery regarding the optimal temporizing strategy for PHVD. Different centers favor different interventions, and head-to-head comparisons across modalities remain limited. Temporizing approaches have each been associated with varying rates of complications, including infection, CSF leak, over-drainage, and the need for repeat procedures. The VSGS occupies a particular niche in this landscape: it is simpler to place than some alternatives, avoids the infection risk of an externalized system, and can sometimes be revised or exchanged at the bedside or with minimal anesthesia. Yet questions persist about its failure modes — including obstruction, inadequate drainage, and the very complications the current study investigates. By systematically characterizing the predictors of both VPS conversion and IPH, the research contributes a piece of evidence that can inform these protocol-level decisions.

Methodologically, studies of this type typically rely on retrospective cohort designs drawing on institutional databases spanning years of clinical care, with statistical modeling used to isolate variables independently associated with each outcome. Candidate predictors in the PHVD literature have included the severity of the initial hemorrhage as graded on imaging, ventricular dimensions at the time of intervention, cerebrospinal fluid protein and red blood cell content, the rate of ventricular expansion, gestational age and birth weight at delivery, the presence of parenchymal involvement of the original bleed, and details of the shunt procedure itself such as the duration of drainage and the volume of fluid evacuated. When such variables are linked to the two outcomes examined in this study, they can be incorporated into clinical decision-making — for example, flagging infants with particular risk profiles for closer imaging surveillance or more cautious drainage protocols.

The timing of publication in Pediatric Research, a journal dedicated to advancing the understanding of childhood disease and development, situates the work within a clinical community actively seeking to harmonize practice. Variation in shunt management protocols between neonatal intensive care units is well documented, and multi-factorial outcome analyses such as this one provide the building blocks from which consensus guidelines can eventually be assembled. For parents of infants facing VSGS placement, the findings offer a clearer vocabulary for understanding the questions their clinical teams are weighing: how severe was the initial hemorrhage, what does the imaging show about the brain’s response to drainage, and what is the realistic likelihood that a permanent shunt will ultimately be needed.

What emerges from this research is a portrait of a treatment pathway that is highly effective at its central task — relieving dangerous ventricular dilatation in a vulnerable population — but one whose outcomes are shaped by identifiable clinical variables. The identification of factors associated with ventriculoperitoneal shunt conversion helps clinicians and families anticipate the most likely long-term course, while the characterization of intraparenchymal hemorrhage risk addresses one of the most feared procedural complications. As survival of extremely preterm infants continues to rise globally, studies of this kind — meticulously linking imaging findings, clinical parameters, and surgical outcomes — represent an essential investment in the neurological futures of thousands of children each year. The full analysis, published open access with detailed methodological documentation, is available to clinicians and researchers through the journal, and it is expected to inform ongoing discussions about the standardization of PHVD management across neonatal and pediatric neurosurgical centers.