Уникальный случай: бактерии вызвали скопление газа в черепе 18-летней пациентки, что привело к летальному исходу

Air Kept Filling an 18 Year Old Woman's Skull and Her Doctors Could Not Find a Hole On the day she was admitted, the 18-year-old's head CT showed a subarachnoid hemorrhage and nothing else remarkable. No skull fracture. No air inside the skull. She had arrived at the emergency department in cardiopulmonary arrest after a crushing injury to her chest and abdomen. By hospital day 19, the follow up scan showed gas distributed through the frontal, temporal and occipital lobes. By day 32, it had spread throughout the intracranial cavity. She had never had a neurosurgical procedure. Serial imaging found no cerebrospinal fluid leak, no pneumothorax, and no pneumomediastinum, which ruled out the usual explanations. The air was coming from the infection itself. The Bacteria Were Making Gas Inside Her Skull The case, published Aug. 25 in Frontiers in Medicine by clinicians at hospitals in Ningbo, Lanxi and Hangzhou, China, documents what the authors say is the first published report of pneumocephalus, meaning gas inside the cranial cavity, associated with an NDM-1-producing carbapenem-resistant Klebsiella pneumoniae infection. The mechanism they propose is metabolic. K. pneumoniae is a facultative anaerobe. In low oxygen conditions it ferments glucose through the mixed-acid pathway, producing carbon dioxide and hydrogen as byproducts. That process is already known to explain gas forming liver abscesses caused by the same organism, and pneumocephalus has been reported previously as a complication of multidrug-resistant Klebsiella meningitis. Several conditions converged to make her infected subarachnoid space an ideal fermentation chamber. Her blood glucose on arrival was markedly elevated at 17.98 mmol/L, roughly 324 mg/dL, despite no history of diabetes, supplying abundant substrate. Inflammation caused microvascular thrombosis and focal tissue death, creating poorly oxygenated pockets. The subarachnoid hemorrhage raised intracranial pressure, which reduced cerebral perfusion and deepened the local hypoxia. Gas accumulation then raised pressure further, feeding the loop. Genome Sequencing Traced the Bacteria from Her Blood to Her Brain Cultures from her blood on day 11 and from an internal jugular venous catheter on day 14 both grew multidrug-resistant K. pneumoniae. A lumbar puncture on day 23 yielded turbid, pale red cerebrospinal fluid with an opening pressure of 250 mmH2O, a nucleated cell count of 29,427 per microliter, glucose below 1.11 mmol/L and protein of 653 mg/dL. Culture confirmed the same organism. A second lumbar puncture on day 35 produced no fluid at all. What came back in the syringe was frank pus. The cell count had climbed to 44,800. Retrospective whole-genome sequencing established that this was one infection, not two. Blood and CSF isolates belonged to the same clonal lineage, sequence type ST34, and differed by only 14 single nucleotide polymorphisms. Both carried the blaNDM-1 carbapenemase gene on a roughly 54 kb IncX3 plasmid, along with an extended-spectrum beta-lactamase gene and resistance genes for aminoglycosides, tetracycline and fosfomycin. Notably, the strain was not hypervirulent. Screening for the hypervirulence markers rmpA, rmpA2, aerobactin and salmochelin came back negative. Previous pneumocephalus cases have mostly involved hypervirulent or hypermucoviscous strains. This was an ordinary, extensively drug-resistant one. Almost Nothing Could Reach the Infection The therapeutic problem the authors describe is a double bind. NDM-1 is a metallo-beta-lactamase that hydrolyzes essentially all beta-lactam antibiotics, including carbapenems. Ceftazidime-avibactam, which she received based on blood culture results, has no activity against NDM-1-producing isolates. That left polymyxin B and tigecycline as theoretical options. Both fail on the second constraint. Polymyxin B penetrates cerebrospinal fluid at only about 5% to 11% of serum concentrations, and tigecycline levels in the central nervous system are negligible. The authors describe the result as a pharmacological sanctuary: her bloodstream cultures turned negative by day 27 while bacteria kept replicating in her CSF, which still grew the organism on day 37. Aztreonam-avibactam and cefiderocol were not administered because of limited availability, delayed identification of the resistance mechanism, and rapid clinical deterioration. Intrathecal administration, which has cured NDM-1 Klebsiella meningitis in previously reported cases, was not feasible given her cerebral edema and critical condition. Her Glasgow Coma Scale score stayed at 3 throughout. She died on hospital day 47. Why One Death Matters Beyond One Patient Carbapenem-resistant K. pneumoniae is not a fringe organism. NDM-1 and NDM-5 are the most commonly identified NDM variants in China, and metallo-beta-lactamase producers are a recognized global problem in intensive care. The authors draw three clinical lessons. Negative blood cultures do not mean the central nervous system is clear, so CSF should be reassessed directly before de-escalating therapy. New intracranial air on imaging in a patient with a gas-forming organism should prompt microbiological workup even without a fracture or prior surgery. And when NDM production is suspected, early molecular testing and consideration of regimens that reach the central nervous system matter more than culture-guided adjustments made after the fact. There is movement on treatment. Ceftazidime-avibactam combined with aztreonam has achieved microbiological clearance in severe NDM-1 infections, a broad-spectrum inhibitor called xeruborbactam has entered clinical development, and monoclonal antibodies against pandrug-resistant Klebsiella showed protection in animal models in a 2025 paper. None of that had reached this patient's bedside. This is a single case report and does not establish how often pneumocephalus accompanies these infections. Key Questions Answered What is pneumocephalus? An abnormal accumulation of gas inside the cranial cavity. It usually follows head trauma or neurosurgery, which makes this case unusual because neither applied. Where did the air come from? The authors attribute it to the bacteria. Klebsiella pneumoniae ferments glucose under low oxygen conditions and releases carbon dioxide and hydrogen as byproducts. What is NDM-1? New Delhi metallo-beta-lactamase-1, an enzyme that breaks down nearly all beta-lactam antibiotics, including carbapenems, leaving very few treatment options. How did doctors prove the brain infection came from her bloodstream? Whole-genome sequencing showed the blood and cerebrospinal fluid isolates were the same clonal lineage, differing by only 14 single nucleotide polymorphisms. Why did antibiotics fail? Two reasons combined. The organism destroyed the beta-lactam drugs, and the remaining options penetrate the blood-brain barrier poorly, with polymyxin B reaching only about 5% to 11% of serum levels in CSF. Is this a common risk for patients? No. This is a single reported case in a critically injured patient with a highly resistant organism, and the authors describe it as exceptionally rare. Published by Medicaldaily.com