Anesthesia’s Next Frontier: Personalized Dosing, Brain Monitoring and Greener Operating Rooms

General anesthesia is one of medicine’s quiet triumphs. Every year, tens of millions of patients drift off under a carefully balanced cocktail of drugs and wake up safely hours later, often giving little thought to the pharmacological tightrope walk that kept them unconscious, pain-free, still, and physiologically stable. Yet according to a new specialty grand challenge article published in Frontiers in Anesthesiology by Dr. Tomohiko Aoe of Teikyo University in Japan, the field is now confronting a set of deeper questions that go far beyond simply getting patients through surgery alive. The central challenge, he argues, is to transform anesthetic practice from a standardized, protocol-driven discipline into a personalized, outcome-focused science.

The modern practice of general anesthesia is built on the concept of balanced anesthesia, in which separate pharmacological components deliver hypnosis, analgesia, amnesia, immobility, and autonomic stability while preserving a smooth recovery. In a typical adult case, induction begins with an intravenous hypnotic such as propofol, analgesia is provided by opioids like fentanyl or remifentanil, rocuronium relaxes the muscles to facilitate airway management, and maintenance follows with a volatile agent such as sevoflurane, desflurane, or isoflurane. Each element has been refined over decades, and short-term safety is remarkably high. But Aoe contends that the discipline’s success inside the operating room has exposed a new frontier of problems outside it: how to predict drug effects in individual patients, how to anticipate interactions in increasingly complex multimodal regimens, and how the molecular footprints of anesthetics might shape outcomes days, months, or even years after the drugs wear off.

One of the most persistent debates in the field pits volatile anesthetic-based anesthesia against total intravenous anesthesia, or TIVA, in which propofol or the newer benzodiazepine remimazolam, usually paired with a remifentanil infusion, maintains the anesthetic state without any inhaled agent. TIVA holds clear advantages for selected patients: those at high risk of postoperative nausea and vomiting, those susceptible to malignant hyperthermia, and situations where rapid, clear emergence is desired. Recent systematic reviews suggest the two approaches are broadly comparable in overall safety, with propofol-based TIVA tending to reduce nausea and emergence agitation in some populations. But the deeper question, whether TIVA improves major long-term outcomes such as cognition or cancer recurrence, remains unresolved. Observational studies have sometimes favored propofol for neurological and oncological endpoints, yet confounding by patient selection, surgical factors, and institutional practice is difficult to exclude. Aoe’s verdict is pragmatic: technique should be chosen according to patient risk, surgical context, and recovery goals, not an assumed universal superiority of either method.

A key technical limitation of TIVA illustrates why personalization is so hard. Unlike volatile agents, whose end-tidal concentrations can be measured directly in exhaled breath, intravenous hypnotic levels are only estimated. Target-controlled infusion systems use microprocessors and population pharmacokinetic-pharmacodynamic models to adjust infusion rates toward a predicted plasma or effect-site concentration, making titration more systematic than manual dosing. But these are predictions, not measurements. Model misspecification, obesity, frailty, low cardiac output, organ failure, and propofol-opioid interactions can all produce substantial prediction error, meaning the same target concentration can deliver very different depths of anesthesia in different patients. Refining these models for elderly, frail, obese, critically ill, pediatric, and organ-failure populations is one of the review’s most urgent priorities.

The drug arsenal itself is evolving. Propofol remains the dominant induction agent thanks to its rapid onset, smooth recovery, and antiemetic properties, but it can cause hypotension and, rarely, propofol infusion syndrome during prolonged high-dose use. Remimazolam, an ultrashort-acting benzodiazepine broken down by tissue esterases and reversible with flumazenil, has emerged as a compelling alternative: meta-analyses suggest it causes less hypotension than propofol, though delirium outcomes in older patients appear broadly similar. Ketamine, acting through NMDA receptor antagonism, remains invaluable in trauma, shock, and opioid-tolerant patients, while etomidate preserves hemodynamic stability in fragile patients at the cost of adrenal suppression and myoclonus. Notably, a recent multicenter randomized trial of 2,365 critically ill adults undergoing emergency intubation found that ketamine did not reduce 28-day mortality compared with etomidate, and cardiovascular collapse during intubation was actually more frequent with ketamine, at 22.1 percent versus 17.0 percent, a finding that challenges some long-standing assumptions about the two drugs.

At the same time, the field is deliberately retreating from opioid-heavy practice. Regional techniques such as epidural analgesia, peripheral nerve blocks, and fascial plane blocks, combined with non-opioid adjuncts including acetaminophen, nonsteroidal anti-inflammatory drugs, ketamine, dexmedetomidine, and intravenous lidocaine, can substantially reduce opioid exposure. The goal is not strictly opioid-free anesthesia, Aoe emphasizes, but opioid-sparing, rational multimodal analgesia that limits respiratory depression, nausea, ileus, urinary retention, opioid-induced hyperalgesia, and persistent postoperative opioid use. The transition from intraoperative to postoperative analgesia is a particularly vulnerable moment: remifentanil’s rapid esterase metabolism allows exquisite titration during surgery, but inadequate transitional analgesia after it stops can contribute to severe early pain and acute tolerance.

Some of the field’s most consequential controversies concern effects that unfold long after emergence. Postoperative delirium and neurocognitive impairment are multifactorial syndromes in which age, frailty, pre-existing cognitive impairment, hypotension, sleep disruption, pain, and systemic inflammation may matter more than the anesthetic drug itself. Regulatory warnings have also raised concern about repeated or prolonged anesthetic exposure in very young children, although current evidence suggests a single short exposure is unlikely to cause major neurodevelopmental harm, and necessary surgery should not be withheld. On cancer, mechanistic studies suggest that volatile anesthetics, opioids, propofol, and local anesthetics may influence immunosurveillance and tumor biology, but a recent randomized multicenter trial found no clear oncological benefit of propofol-based maintenance over volatile anesthesia, and the evidence remains insufficient to select a technique on oncological grounds alone. Similarly, dexamethasone, routinely given to prevent nausea, has produced conflicting signals on chronic postsurgical pain: a prespecified analysis of the PADDI trial found slightly more wound pain at six months with an 8-milligram dose, while a propensity-matched analysis of ENIGMA-II data found no increased risk at twelve months.

Beneath these clinical questions lies a molecular frontier. Anesthetics interact with calcium homeostasis, mitochondrial function, oxidative stress, the unfolded protein response, neuroinflammation, apoptosis, and autophagy. Volatile agents may disturb endoplasmic reticulum calcium handling and protein folding, activating a coordinated stress network involving PERK-eIF2alpha-ATF4-CHOP, IRE1-XBP1, and ATF6 pathways; transient activation may be adaptive, but prolonged stress can drive mitochondrial dysfunction, inflammatory signaling, synaptic injury, and cell death. Different drugs act differently on this network: propofol has shown antioxidant and anti-inflammatory effects experimentally, dexmedetomidine may attenuate perioperative stress while preserving immune function, and ketamine and local anesthetics modulate neuroimmune signaling through NMDA receptor and sodium channel mechanisms. Crucially, traditional endpoints such as loss of consciousness and hemodynamic stability capture none of this. Aoe calls for future studies to incorporate biomarkers including inflammatory cytokines, neuronal injury markers, oxidative stress measures, metabolomics, and cerebrospinal fluid or neuroimaging endpoints in high-risk populations.

Pharmacogenetics is beginning to convert rare, unpredictable anesthetic responses into actionable preoperative decisions. Pathogenic variants in RYR1 or CACNA1S identify malignant hyperthermia susceptibility and justify trigger-free TIVA, while BCHE variants explain prolonged paralysis after succinylcholine. A mitochondrial DNA variant, m.11232T>C, has been linked to severe neurological deterioration after sevoflurane exposure, and associations involving CYP2B6 with propofol disposition, CYP2D6 with codeine and tramadol response, and OPRM1 or COMT with opioid sensitivity may eventually refine drug selection. Still, Aoe cautions that broad panel testing for all surgical patients is not yet justified given variable effect sizes, cost, and turnaround times; targeted testing for patients with relevant family history or prior unexplained reactions should come first, integrated with pharmacokinetic modeling and real-time monitoring.

The final grand challenge is planetary. Desflurane’s high global warming potential has led many institutions to phase it out in favor of low-flow anesthesia and reduced nitrous oxide use, and anesthetic gas capture technologies are likely to expand. But TIVA is not environmentally neutral either, given its plastics, syringes, infusion pumps, and propofol waste. The reasonable path, Aoe concludes, is to prioritize patient safety while adopting the lowest-impact clinically appropriate methods. Taken together, his roadmap sketches an anesthesia of the near future that is continuously monitored with quantitative neuromuscular and electroencephalographic feedback, dosed by refined models rather than fixed protocols, built on interaction-aware multimodal analgesia, judged by delirium, quality of recovery, persistent pain, and cancer outcomes rather than intraoperative stability alone, and accountable to the environment as well as the patient. The drugs that put patients to sleep, in other words, are being asked to answer for everything that happens after they wake up.

Subject of Research: Grand challenges in anesthetic pharmacology and personalized, outcome-focused anesthesia practice

Article Title: Grand challenges in anesthetic pharmacology: toward personalized, outcome-focused practice

Article References: Aoe, T. (2026). Grand challenges in anesthetic pharmacology: toward personalized, outcome-focused practice. Frontiers in Anesthesiology, 5, Article 1934044. https://doi.org/10.3389/fanes.2026.1934044

Image Credits: AI Generated

DOI: 10.3389/fanes.2026.1934044

Keywords: anesthetic pharmacology, general anesthesia, total intravenous anesthesia, volatile anesthetics, pharmacogenetics, postoperative delirium, neuroinflammation, multimodal analgesia, malignant hyperthermia, environmental sustainability, target-controlled infusion, pediatric neurotoxicity