Highlights
– Personalized, remotely coached prehabilitation produced measurable improvements in physical performance and cognitive function before major elective surgery and was associated with fewer moderate-to-severe postoperative complications compared with a paper-based standard program.
– High-dimensional mass cytometry revealed cell type–specific dampening of inflammatory intracellular signaling after personalized prehabilitation (AUROC 0.88), changes not observed with standard prehabilitation (AUROC 0.63).
– Immune changes included reduced phosphorylated protein kinase R-like endoplasmic reticulum kinase 1/2 signaling in classical monocytes and myeloid-derived suppressor cells after IL-2/4/6 stimulation and lower phosphorylated CREB signaling in Th1 cells—pathways plausibly linked to postoperative infection risk and neurocognitive decline.
Background
Prehabilitation—structured interventions delivered in the interval before major surgery to improve physical, nutritional, and psychological readiness—is increasingly integrated into perioperative care pathways. Proponents propose that enhancing physiologic reserve reduces postoperative complications, shortens recovery, and improves functional outcomes. However, evidence remains heterogeneous and implementation variable, with many trials differing in intervention content, intensity, and timing. A critical gap has been linking clinical improvements to underlying biologic mechanisms: does prehabilitation simply improve functional capacity, or does it meaningfully modulate perioperative biology, including immune responsiveness, that contributes to surgical risk?
The randomized clinical trial by Cambriel and colleagues (JAMA Surg. 2025) addresses this gap by comparing a personalized, remotely coached multimodal prehabilitation program to a standard paper-based program and by applying a 47-plex mass cytometry immunoassay to characterize preoperative immune signaling changes.
Study design
This single-center, single-blinded randomized interventional trial enrolled 58 adult patients undergoing major elective surgery between June 2020 and September 2022. Block randomization assigned participants to either personalized prehabilitation (n = 28) or a standard paper-based program (n = 30). Key exclusions were contraindication to exercise, ASA score ≥4, palliative care, or having fewer than 14 days between screening and surgery. Fifty-four patients completed the study (27 per group).
Interventions
– Personalized prehabilitation: two weekly one-on-one remote coaching sessions tailored to progress across four domains—physical activity, nutrition, cognitive training, and mindfulness.
– Standard prehabilitation: paper-based materials covering the same domains without individualized coaching.
Primary clinical outcomes included cognitive testing and physical performance (wall squat, timed-up-and-go, and the 6-minute walk test [6MWT]). The primary immunologic outcomes were frequencies of major innate and adaptive immune cell populations and intracellular signaling responses assessed using a 47-plex mass cytometry panel, including stimulated signaling readouts.
Endpoints were analyzed with the goal of identifying whether personalized prehabilitation produced distinct immune modulation and improved perioperative outcomes compared with the standard approach. Data analysis occurred April 2023–May 2025.
Key findings
Clinical outcomes
– Completion: 54/58 patients completed the protocol (27 per arm).
– Physical performance: The personalized group showed statistically significant improvement in the 6MWT: median (IQR) increased from 496 (340–619) to 546 (350–728) meters after prehabilitation (P = .03). Other physical measures and cognitive assessments improved in the personalized arm, whereas the standard arm had only moderate clinical improvements.
– Postoperative complications: The personalized prehabilitation group experienced fewer moderate-to-severe postoperative complications (Clavien-Dindo grade >1): 4 events versus 11 in the standard group (P = .04). The report does not show differences in mortality or length-of-stay data in the provided summary; the primary signal was lower complication burden in the personalized arm.
Immunologic outcomes
– High-dimensional modeling: Multivariable models discriminated immune states before versus after personalized prehabilitation with an AUROC of 0.88 (95% CI, 0.79–0.97; P < .001) using leave-one-out cross-validation, indicating a robust and reproducible shift in the circulating immunome. In contrast, samples from the standard prehabilitation arm did not show a significant pre-to-post immunologic change (AUROC 0.63; 95% CI, 0.48–0.78; P = .12).
– Cell type–specific signaling alterations: Personalized prehabilitation was associated with dampened intracellular signaling responses in specific cell populations after ex vivo cytokine stimulation. Notable findings included reduced phosphorylated protein kinase R-like endoplasmic reticulum kinase 1/2 signaling in classical monocytes and myeloid-derived suppressor cells (MDSCs) after IL-2, IL-4, and IL-6 stimulation, and decreased phosphorylated cyclic adenosine monophosphate response-element binding protein (pCREB) signaling in Th1 cells. These pathways are implicated in inflammatory activation, monocyte/macrophage responses, and T helper cell transcriptional programs.
Interpretation of immunologic signals
– The observed ‘‘dampening’’ denotes reduced magnitude of stimulus-induced phosphorylation within selected signaling axes, suggestive of a lower pro-inflammatory reactivity setpoint. In the perioperative context, exaggerated innate immune activation has been linked to complications such as surgical site infection and postoperative delirium; conversely, excessive immunosuppression also increases infection risk. The authors interpret the observed modulation as movement toward a more regulated immune responsiveness that may be protective.
– The cell-specific nature of changes—classical monocytes and MDSCs vs Th1 cells—strengthens biologic plausibility. Monocytes and MDSCs are central to early inflammatory responses and immunoregulation, while Th1 signaling affects adaptive cytotoxic responses and neuroinflammatory pathways.
Strength and statistical rigor
– The study used high-dimensional single-cell proteomic readouts with multivariable modeling and cross-validation to reduce overfitting risk. The strong AUROC for the personalized arm suggests a coherent immune signal accompanying the clinical improvements.
– The randomized design strengthens causal inference that individualized coaching, rather than mere exposure to educational materials, drove the observed differences.
Expert commentary and mechanisms
Mechanistic plausibility
– Exercise, nutrition, sleep/cognitive training, and stress reduction are known to influence immune function. Exercise can reduce systemic inflammation, alter monocyte phenotypes, and improve T cell function. Nutritional optimization can restore micronutrient-dependent immune competency. Psychological interventions influence neuroimmune signaling and hypothalamic–pituitary–adrenal axis activity, which modulates cytokine responses. The trial’s multimodal personalized approach likely combined these effects to shift immune signaling toward a less pro-inflammatory profile.
– The specific reduction in PERK/ERK-type signaling (as reported) and pCREB in Th1 cells maps onto pathways involved in cellular stress responses, cytokine signaling, and transcriptional regulation relevant to infection susceptibility and neuroinflammation. While the study reports phosphorylation-level changes, downstream functional consequences (for example, altered cytokine secretion, phagocytic function, or antigen presentation) require further validation.
Limitations and generalizability
– Sample size and single-center design: With 58 randomized and 54 completers, the study is relatively small and from a single academic center. Findings are hypothesis-generating and require replication in larger, multicenter cohorts.
– Participant selection and intervention intensity: Patients with ASA ≥4 and those with very short waits to surgery were excluded; the benefit in more frail or time-constrained patients remains uncertain. The personalized program used twice-weekly remote coaching, a resource-intensive approach that may be difficult to scale without infrastructure and reimbursement models.
– Interpretation of immune changes: Mass cytometry provides rich, correlative information on intracellular signaling but does not by itself prove causal links between specific pathway modulation and clinical outcomes. The reported dampening could reflect beneficial immune recalibration or, conversely, a transient hyporesponsive state; clinical context and additional functional assays are needed.
– Outcomes measured: While complication rates favored personalized prehabilitation, the trial was not powered for hard endpoints such as mortality or infection rates, and the absolute number of events is small.
Clinical implications and next steps
– Translational potential: This trial provides proof-of-concept that individualized prehabilitation can deliver measurable biologic changes in immune responsiveness before surgery alongside functional gains. For clinicians, the study supports investment in tailored preoperative programs—particularly where remote coaching is feasible—to potentially reduce complication burden.
– Biomarker-driven prehabilitation: The demonstration that the circulating immunome can be monitored and discriminates responders suggests a future in which biologic monitoring could guide personalization (e.g., intensify interventions in patients without an immunologic response) or stratify perioperative risk.
– Future research priorities: Larger, multicenter randomized trials powered for clinical endpoints (infections, delirium, readmission, length of stay, cost-effectiveness) are needed. Mechanistic studies should pair signaling readouts with functional immune assays and tissue-level outcomes. Implementation science work will be required to determine scalable coaching models and reimbursement pathways.
Conclusion
Cambriel et al. provide compelling randomized-trial data that personalized, remotely coached multimodal prehabilitation can improve preoperative physiologic and cognitive metrics, produce reproducible cell type–specific dampening of pro-inflammatory signaling in the circulating immunome, and reduce moderate-to-severe postoperative complications compared with a paper-based standard program. The findings bridge the gap between functional preoperative optimization and measurable biologic change, supporting a path toward biologically driven personalization of perioperative care. Larger confirmatory trials and mechanistic follow-up will be essential before widespread adoption as standard practice.
Funding and ClinicalTrials.gov
Trial registration: ClinicalTrials.gov Identifier: NCT04498208.
Funding details are reported in the original publication (Cambriel A et al., JAMA Surg. 2025). Please consult the paper for full funding and conflict-of-interest statements.
References
1. Cambriel A, Tsai A, Choisy B, Sabayev M, Hedou J, Shelton E, Singh K, Amar J, Badea V, Bruckman S, Ganio E, Einhaus J, Feyaerts D, Stelzer I, Sato M, Langeron O, Bonham TA, Gaudillière D, Shelton A, Kin C, Gaudillière B, Verdonk F. Immune Modulation by Personalized vs Standard Prehabilitation Before Major Surgery: A Randomized Clinical Trial. JAMA Surg. 2025 Nov 12:e254917. doi: 10.1001/jamasurg.2025.4917. Epub ahead of print. PMID: 41222945; PMCID: PMC12613092
(Readers seeking broader context on prehabilitation and perioperative immune modulation should consult contemporary systematic reviews and enhanced recovery guidelines; the present summary focuses on the randomized trial above.)