Among the most feared diagnoses in medicine, glioblastoma carries a median survival of roughly twelve to fifteen months even with the best available treatment, a combination of maximal safe surgical resection followed by radiotherapy with concomitant and adjuvant temozolomide known as the Stupp regimen. Clinicians have long relied on global measures such as the Karnofsky Performance Status to gauge how patients are likely to fare, alongside molecular markers like MGMT promoter methylation. Yet a deceptively simple bedside question may carry prognostic weight that these established tools do not fully capture: can this patient walk, and how much help do they need to do it? A new retrospective study published in the Journal of Neuro-Oncology suggests that walking ability measured at the moment of hospital discharge is one of the most informative functional markers in the entire peri-treatment course of newly diagnosed glioblastoma, adding predictive information that persists even after accounting for contemporaneous performance status.
The research, conducted at Kagoshima University in Japan, followed 185 adults with newly diagnosed glioblastoma treated between 2011 and 2020. Rather than assessing walking at a single moment, the investigators tracked it at six distinct points spanning the treatment trajectory: before surgery, on postoperative day five, before the start of chemoradiotherapy, at hospital discharge, and at one and four months after discharge. The instrument they chose was the Functional Ambulation Category, or FAC, a six-point ordinal scale that classifies walking according to the level of human assistance required, ranging from nonfunctional ambulation at level 0 to fully independent walking on all surfaces at level 5. Because the scale depends on observation of assistance rather than timed performance, it remains feasible even in patients whose motor impairment, reduced consciousness, aphasia, or cognitive dysfunction would make stopwatch-based gait tests impractical or impossible.
The study’s central finding is striking in its consistency. Higher FAC scores were associated with longer overall survival at all six assessment points, but the association was strongest at hospital discharge. When patients were grouped by discharge walking ability, median survival measured from surgery was 346 days for those with FAC 0 to 1, 590 days for those with FAC 2 to 4, and 879 days for those with FAC 5, a gradient that the log-rank test confirmed was highly unlikely to arise by chance. Using hospital discharge as a landmark origin for survival measurement, median post-discharge survival was 286, 531, and 823 days for the three groups respectively, again with a highly significant overall difference. Patients unable to walk or requiring constant human support separated clearly from those with partial independence, while the distinction between partial assistance and full independence narrowed once statistical corrections for multiple comparisons were applied.
The methodological care behind these numbers deserves attention. In the primary analysis, the researchers used a discharge-landmark design in 162 patients with 138 deaths, aligning the survival clock with the assessment itself and adjusting hazard estimates for age, sex, extent of resection, and the Karnofsky Performance Status measured at the same discharge moment. Against the reference group with FAC 0 to 1, the adjusted hazard ratio for death was 0.45 for patients with FAC 2 to 4 and 0.33 for those who walked fully independently, meaning that better walking categories were associated with mortality risks roughly half to a third of those seen in the most impaired group. Discharge Karnofsky Performance Status itself remained independently prognostic, with each ten-point increase associated with a fourteen percent lower hazard of death, which makes the persistence of the FAC signal all the more notable.
To determine whether the walking measure truly added information beyond global performance status, the team compared statistical models with and without discharge FAC using the Akaike information criterion, Harrell’s concordance index, and a likelihood-ratio test. Adding FAC improved model fit by 8.4 AIC units and improved discrimination by 0.022 on the C-index, a modest but statistically significant gain with a p value of 0.002. In a head-to-head comparison of the four in-hospital assessments on a common complete-case subset, discharge FAC produced the largest improvement over the clinical reference model, with an AIC gain of 20.1 and a C-index improvement of 0.043, while earlier assessments contributed less. Systematic landmark sensitivity analyses that shifted the survival origin to match each assessment reproduced the same temporal pattern, reinforcing that the discharge measurement sits at a uniquely informative moment in the clinical course.
The authors are careful to explain why discharge, rather than any earlier or later point, carries the strongest signal. Discharge FAC is measured only after surgery, early postoperative recovery, inpatient rehabilitation, and a substantial portion of the initial chemoradiotherapy course have already unfolded. It therefore functions as an integrated clinical marker, summarizing the net effect of tumor biology, treatment intensity, surgical sequelae, and functional recovery up to a meaningful transition point, rather than as a purely baseline prognostic factor. Walking ability in the cohort followed a characteristic trajectory: lowest immediately after surgery, then progressively recovering toward discharge and beyond. This pattern carries a practical clinical warning, namely that early postoperative deterioration should not automatically be read as a patient’s eventual functional ceiling, since meaningful recovery often continues throughout hospitalization.
The findings also illuminate the relationship between function and treatment delivery. Patients with higher discharge FAC were markedly more likely to begin post-discharge adjuvant temozolomide, with initiation rates of 60 percent in the FAC 0 to 1 group, 90 percent in the FAC 2 to 4 group, and 98 percent in the FAC 5 group. This suggests that walking ability may partly serve as a proxy for treatment readiness, the physiological and cognitive reserve required to tolerate continued intensive therapy. Importantly, statistical adjustment for adjuvant temozolomide did not materially change the FAC estimates, and the authors emphasize that because exact initiation dates were unavailable, the findings establish prognostic association rather than a causal effect of walking recovery or rehabilitation on survival. Subgroup analyses found the inverse association present in every stratum of age, sex, preoperative performance status, and extent of resection, with no significant interactions; the age interaction p value of 0.783 argues against the possibility that the effect simply reflects younger patients walking better.
The study Situates itself against a backdrop of mixed prior evidence. Earlier work had associated balance performance with subsequent loss of walking ability and mortality in glioblastoma, while timed tests such as the ten-meter walk test and the six-minute walk test failed to emerge as independent survival predictors in their respective cohorts. This new study extends the evidence to an assistance-based, non-timed measure, one that can be assigned at the bedside without equipment or patient cooperation with standardized testing. Reliability support came from a substudy of 71 patients in whom a second physical therapist independently assigned discharge FAC ratings, yielding a linear weighted Cohen’s kappa of 0.819, indicating strong agreement for chart-based assessment. Rehabilitation was delivered within an integrated institutional pathway that initiated inpatient rehabilitation typically within three days of surgery and continued it throughout chemoradiotherapy, a context that shaped both the long median hospital stay of 63 days and the timing of the discharge assessment itself.
The investigators are forthright about the limitations that temper interpretation. The retrospective, single-center design, the absence of systematically collected IDH and MGMT molecular data across the historical cohort, possible misclassification under contemporary WHO diagnostic criteria, heterogeneous treatment including upfront bevacizumab, and the inability to model corticosteroid exposure all leave room for residual confounding. The three-group FAC categorization requires external validation, and the incremental gain in discrimination, while real, was modest. The authors accordingly position discharge FAC not as a stand-alone prediction tool or a validated clinical model, but as a practical, complementary bedside marker available precisely at the transition from inpatient treatment to post-discharge care, before further survivor selection accumulates. Within that frame, the message is clear and clinically actionable: in newly diagnosed glioblastoma, how a patient walks on the day they leave the hospital encodes prognostic information that global performance scores alone do not fully convey, and it can be captured with nothing more than an attentive clinical eye. Future multicenter prospective studies, the authors conclude, should validate the finding and explore how walking independence integrates with neurological, oncological, and supportive-care factors in modern, molecularly stratified glioblastoma care.
Subject of Research: Prognostic value of walking ability at hospital discharge for overall survival in newly diagnosed glioblastoma
Article Title: Prognostic significance of walking ability at hospital discharge for overall survival in newly diagnosed glioblastoma: a comparison across six peri-treatment assessments
Article References: Prognostic significance of walking ability at hospital discharge for overall survival in newly diagnosed glioblastoma: a comparison across six peri-treatment assessments. (n.d.). https://doi.org/10.1007/s11060-026-05784-0
Image Credits: AI Generated
DOI: 10.1007/s11060-026-05784-0
Keywords: glioblastoma, walking ability, Functional Ambulation Category, hospital discharge, overall survival, prognosis, Karnofsky Performance Status, rehabilitation, chemoradiotherapy, brain tumor, functional outcomes, neuro-oncology
Cite Scienmag News
APA MLA Chicago
Nathaniel Bowman. (September 12, 2026). How Well You Walk at Hospital Discharge May Predict Glioblastoma Survival. Scienmag. https://scienmag.com/how-well-you-walk-at-hospital-discharge-may-predict-glioblastoma-survival/
Nathaniel Bowman. “How Well You Walk at Hospital Discharge May Predict Glioblastoma Survival.” Scienmag, 12 September 2026, https://scienmag.com/how-well-you-walk-at-hospital-discharge-may-predict-glioblastoma-survival/. Accessed 12 September 2026.
Nathaniel Bowman. “How Well You Walk at Hospital Discharge May Predict Glioblastoma Survival.” Scienmag. September 12, 2026. https://scienmag.com/how-well-you-walk-at-hospital-discharge-may-predict-glioblastoma-survival/
Copy citation Download RIS
Tags: assessment of patient mobility during brain tumor treatmentbrain tumorchemoradiotherapyFunctional Ambulation Categoryfunctional outcomesfunctional prognostic markers in glioblastomaGlioblastomaglioblastoma survival predictionhospital dischargeimpact of mobility on brain tumor outcomesinfluence of discharge walking ability on prognosisKarnofsky Performance Statuslong-term survival predictors in glioblastomaneuro-oncologyoverall survivalperi-treatment functional assessmentpostoperative mobility and survival in glioblastoma patientsprognosisrehabilitationretrospective study on glioblastoma biomarkersrole of performance status in glioblastoma prognosissignificance of walking ability in neuro-oncologywalking abilitywalking ability at hospital discharge
