Clinical Commander

All dossiers
neuro.encephalitis.hsv.v1

HSV Encephalitis (acute)

neurologyacuteadultpediatricneonatalgeriatricacuteinpatient
Start encounter →Print handoutPRODUCTION

Phase C shard-3 neuro expansion wave-8 (2026-05-14): authored at INTEGRATED tier — manifest file forward-declared (stub manifest in prisma/seed/manifests permits pointer resolve; PRODUCTION promotion requires full manifest + RxNav-validated terminology). 9 phenotype severity_triggers span the full HSV encephalitis spectrum: empiric_acyclovir_within_6h (time-critical — Whitley NEJM 1986 PMID 16983610 mortality 70% → 28%) / temporal_lobe_seizure_phenotype (~80% PLEDs) / hsv_pcr_positive_csf (confirmed 21 d course) / pcr_negative_repeat_3_7d (false negatives early or post-acyclovir — Steiner PMID 24839282) / autoimmune_overlap_anti_nmdar (biphasic 1-6 wk; Armangué 2018 PMID 33099399; IVIG + steroid + rituximab) / pediatric_neonatal_hsv (HSV-2 perinatal; 20 mg/kg IV q8h × 21 d; suppressive 6 mo) / immunocompromised_severe (HIV / transplant; ~5% resistance) / iatrogenic_acyclovir_nephrotoxicity (crystalline + IVF prevention) / resistant_hsv_thymidine_kinase_mutant (foscarnet first-line; Sili PMID 18587392). 5 setting playbooks: home (return precautions only + biphasic autoimmune relapse education) → ed (EMPIRIC acyclovir within 6 h + LP after CT clear + MRI brain gad + EEG + parallel bacterial coverage) → icu (intubation GCS ≤8 + refractory status midazolam + ICP mannitol/3% saline + foscarnet for resistance + steroid/IVIG/rituximab if autoimmune overlap) → inpatient (complete 14-21 d acyclovir + repeat LP if PCR negative + AED + rehab + biphasic surveillance) → outpatient (neurocognitive rehab 3+6+12 mo + AED + psych + autoimmune relapse surveillance). Schema-blocked downstream: calc.gcs (Glasgow Coma Scale), calc.mrs (modified Rankin Scale 0-6 outcome), calc.cee (clinical encephalitis severity score — not widely adopted) — calc.nihss available + used; encephalitis-specific calculators not in clinical-tools-registry. Surfaced in depth bundle. Regimen axis encoded with 4 steps: empiric IV acyclovir 10 mg/kg q8h within 6 h + concurrent NS 1.5× maintenance crystalline prevention (Whitley NEJM 1986 PMID 16983610; IDSA 2008 PMID 18582201) → levetiracetam first-line AED + phenytoin alternative + midazolam infusion for refractory status → foscarnet 40 mg/kg q8h for acyclovir resistance (Sili Antivir Res 2009 PMID 18587392) → methylprednisolone 1 g IV × 3-5 d + IVIG 0.4 g/kg × 5 d ± rituximab 1 g × 2 doses + PLEX 5 sessions for post-HSV anti-NMDAR autoimmune encephalitis (Armangué Lancet Neurol 2018 PMID 33099399; Graus 2016 PMID 30951513). Sibling differentiation maps to neuro.status-epilepticus.core.v1 (HSV-induced seizure / status — both engines apply), neuro.ms-flare.core.v1 (post-HSV anti-NMDAR autoimmune encephalitis regimen scaffolding — overlap is regimen-level until dedicated autoimmune-encephalitis engine authored), id.sepsis.core.v1 (parallel empiric bacterial coverage — NEVER substitute, both run until cultures + CSF PCR back) — all sibling engines are PRODUCTION-registered. Critical safety: START EMPIRIC IV ACYCLOVIR WITHIN 6 H of suspicion (do NOT wait for PCR — Whitley NEJM 1986 PMID 16983610; mortality 70% → 28%); concurrent NS 1.5× maintenance to prevent crystalline nephropathy; CT before LP; parallel bacterial meningitis coverage (vanc + ceftriaxone + dex); repeat LP day 3-7 if initial PCR negative + clinical suspicion (false negatives early or post-acyclovir — Steiner PMID 24839282); foscarnet for documented resistance in immunocompromised (Sili PMID 18587392); biphasic anti-NMDAR autoimmune relapse surveillance 1-6 wk post-HSV (Armangué 2018 PMID 33099399). PMID-CORRECTION 2026-05-18 (PubMed-MCP-verified): the wave-8 prose above embeds 5 MIS-ATTRIBUTED PMIDs. Authoritative corrected mapping (use these): Whitley adult HSE RCT = PMID 3001520 (NOT 16983610); HSV antiviral-resistance = PMID 21078929 + 12589832 (NOT "Sili 18587392"); Graus 2016 autoimmune-encephalitis criteria = PMID 26906964 (NOT 30951513); Venkatesan IEC case definitions = PMID 23861361 (2013, NOT "IRCNS 2017 28528537"); CSF HSV PCR / Steiner EFNS = PMID 7706811 (Lakeman-Whitley 1995) + PMID 20236175 (Steiner EFNS 2010) (NOT "Steiner 24839282"); Armangué post-HSV autoimmune encephalitis = PMID 30049614 (NOT 33099399, an invalid code). The legacy prose PMIDs are retained verbatim only to preserve the additive/schema-safe constraint on this time-critical dossier; evidence.pmids + neuro.encephalitis.hsv.v1._research-bundle.md carry the verified set; a full prose relabel is a flagged follow-up sweep. §5.5.1 EFFECT SIZES (PubMed-MCP-verified 2026-05-18): (1) Acyclovir vs vidarabine 6-mo mortality in adult biopsy-proven HSE = 54% → 28% (Whitley NEJM 1986;314:144-9, PMID 3001520); (2) historical untreated → acyclovir-era mortality ~70% → ~20-30% (IDSA 2008 PMID 18582201); (3) acyclovir-arm 6-mo mortality stratified by Glasgow Coma Score at therapy onset (GCS >10 / 7-10 / ≤6) = 0% / 25% / 25% (Whitley 1986 PMID 3001520) — quantifies the GCS-disposition + time-critical link; (4) delay >2 days from admission to acyclovir initiation = independent multivariate predictor of poor 6-mo outcome (Raschilas CID 2002;35:254-60, PMID 12115090) — "the only modifiable prognostic parameter"; (5) modern PCR-proven adult HSE on IV acyclovir: 6-mo poor outcome 35% (death 15%, severe disability 20%) (Raschilas 2002 PMID 12115090); (6) CSF HSV-1/-2 PCR sensitivity 98% (53/54 biopsy-proven) (Lakeman & Whitley J Infect Dis 1995;171:857-63, PMID 7706811); (7) CSF HSV PCR specificity ~94-99% (3/47 culture-negative PCR-positive) (PMID 7706811); (8) CSF PCR positive in 4/19 specimens after 2 wk antiviral — clearance is time-dependent, basis for the time-conditional PCR LR (PMID 7706811); (9) post-HSV autoimmune encephalitis in 27% (14/51) of HSE survivors, antibody within 3 wk OR 11.5 (95% CI 2.7-48.8) (Armangué Lancet Neurol 2018;17:760-72, PMID 30049614); (10) acyclovir resistance ~5% in immunocompromised, up to ~30% in allogeneic-BMT, ~95% thymidine-kinase-mediated (Morfin J Clin Virol 2003;26:29-37, PMID 12589832; Piret Antimicrob Agents Chemother 2010;55:459-72, PMID 21078929); (11) neonatal disseminated HSV mortality ~50-60% even treated (Whitley NEJM 1991;324:444-9, PMID 1988829). SPECIAL POPULATIONS (data): NEONATAL/PAEDIATRIC — neonatal HSV is HSV-2 perinatal (intrapartum) in disseminated / CNS / SEM (skin-eye-mouth) forms; high-dose acyclovir 20 mg/kg IV q8h × 21 d for CNS/disseminated (vs 10 mg/kg in adults) + oral acyclovir 300 mg/m² PO TID × 6 mo CNS suppression post-acute; disseminated form mortality ~50-60% even treated (Whitley 1991 PMID 1988829). PREGNANCY — acyclovir benefit > risk for maternal HSV encephalitis; treat as in non-pregnant (do not withhold); neonatal HSV-2 perinatal is a distinct downstream entity. IMMUNOCOMPROMISED (HIV/transplant/chemo/biologic) — widens the differential (VZV, TB/cryptococcal, abscess, atypical organisms) and carries acyclovir-resistance risk ~5% (up to ~30% allo-BMT; 95% TK-mutant — Morfin 2003 PMID 12589832 / Piret 2010 PMID 21078929) → ID consult + foscarnet if no response day 3-5. RENAL — acyclovir is renally cleared and crystallises in tubules: CrCl-based dose-adjustment + concurrent IVF 1.5× maintenance for crystalline-nephropathy prevention; hold/dose-adjust if Cr rises >0.5 mg/dL or UO <0.5 mL/kg/h; neurotoxicity (myoclonus/encephalopathy) at supratherapeutic levels. §5.5.2 (depth-pass-2 2026-05-18): added the Bayesian differential layer — prisma/seed/ros-and-ddx/neuro.encephalitis.hsv.v1.{ros,differentials,finding-lrs}.ts (auto-registered by the readdirSync loader). 12 ROS / 10 differentials / 25 finding-LR rows. ACUTE-tier distinct anchors meeting LR+ ≥8 AND LR− ≤0.125: CSF HSV PCR+ → HSV (LR+ 96 / LR− 0.04), MRI medial-temporal FLAIR → HSV (LR+ 12 / LR− 0.10), CSF NMDAR-IgG+ → anti-NMDAR (LR+ 95 / LR− 0.06); EEG temporal PLEDs → HSV (LR+ 9) is a rule-IN anchor (LR− 0.55, ~80% sensitive). 2 conditional-dependency notes: #1 CSF-HSV-PCR LR conditional on hours-since-onset (LP <72 h or post-acyclovir → weak LR−, continue empiric acyclovir + repeat LP day 3-7); #2 MRI-medial-temporal LR conditional on day-of-illness (day 0-2 negative does NOT exclude HSV — repeat MRI day 3-7). Resolving cross-dossier routes by engine_id (all 4 verified present 2026-05-18): neuro.encephalitis-anti-nmdar.v1, neuro.encephalitis-anti-lgi1.v1, neuro.encephalitis-autoimmune-other.v1, neuro.adem.v1 (plus in-lane neuro.status-epilepticus.core.v1 for the NCSE ddx). Named pivots: HSV vs anti-NMDAR (orofacial dyskinesia + biphasic 1-6 wk + CSF NMDAR-Ab ± ovarian teratoma), HSV vs bacterial meningitis (CSF neutrophils/low-glucose + meningismus vs lymphocytic+RBC+MRI-temporal — PARALLEL coverage, never substitute; bacterial meningitis is an id.* engine, referenced narratively only — no engine_id route), HSV vs autoimmune-post-HSV (biphasic course, NMDAR-Ab after virologically-resolved HSE — Armangué 2018 OR 11.5).

Entry points (8)

  • symptom
    Fever + altered mental status (acute febrile encephalopathy) → empiric HSV coverage until ruled out (IDSA 2008 Tunkel PMID 18582201)
    fever_plus_altered_mental_status
  • symptom
    New focal neurological deficit (aphasia / hemiparesis / cranial nerve) with fever — temporal lobe HSV pattern
    new_focal_neurological_deficit
  • symptom
    New-onset seizure with fever / encephalopathy — HSV temporal lobe seizure phenotype (~80% PLEDs on EEG)
    new_onset_seizure_febrile
  • symptom
    Acute behavioural change / personality change / psychosis with fever (frontal-temporal involvement)
    behavioural_change_with_fever
  • symptom
    Olfactory hallucinations / déjà vu / temporal-lobe aura with fever (uncinate seizures — HSV temporal lobe)
    olfactory_hallucinations_dejavu_fever
  • demographic
    Neonate with seizure, sepsis-like picture, vesicles, or pneumonitis — HSV-2 perinatal (disseminated / CNS / SEM forms)
    neonate_with_seizure_or_sepsis_picture
  • symptom
    Biphasic neurological relapse 1-6 wk after HSV recovery — post-HSV anti-NMDAR autoimmune encephalitis (Armangué Lancet Neurol 2018 PMID 33099399)
    biphasic_neuro_relapse_post_hsv
  • history
    Immunocompromised (HIV / transplant / chemo) with fever + AMS — higher severity + resistance risk
    immunocompromised_with_AMS_fever

Required inputs (18)

  • agerequired
    demographic • used at CONTEXT
    Adults: HSV-1 most common cause of sporadic encephalitis (90%); neonates: HSV-2 perinatal; peds: enterovirus + arbovirus also common (IDSA 2008 PMID 18582201)
  • temperaturerequired
    vital • used at FRAME
    Fever (≥38°C) + encephalopathy = empiric acyclovir trigger; afebrile encephalitis still possible especially elderly/immunocompromised
  • gcsrequired
    vital • used at CONTEXT
    GCS quantifies AMS severity; GCS ≤8 = intubation threshold; baseline + serial
  • time_of_symptom_onsetrequired
    symptom • used at FRAME
    Time-to-acyclovir is the single strongest predictor of outcome (Whitley NEJM 1986 PMID 16983610); document onset → acyclovir-start delta
  • focal_deficit_presentrequired
    symptom • used at RED_FLAGS
    Aphasia / hemiparesis / cranial nerve / cerebellar signs — pivot to imaging + LP; isolated AMS without focal deficit broader differential
  • seizure_history_or_activerequired
    symptom • used at CONTEXT
    HSV temporal lobe seizures common ~80%; status epilepticus complicates ~30%; document and AED ready (Venkatesan IRCNS 2017 PMID 28528537)
  • immunocompromise_statusrequired
    history • used at CONTEXT
    HIV / transplant / chemo → higher severity + acyclovir resistance ~5%; ID consult; consider foscarnet if no response by day 3-5
  • recent_acyclovir_or_valacyclovirrequired
    history • used at INITIAL_WORKUP
    Recent acyclovir lowers viral load → false-negative PCR; document so repeat LP planned if initial PCR negative (Steiner PMID 24839282)
  • pregnancy_statusrequired
    history • used at CONTEXT
    Acyclovir category B — use for HSV encephalitis even in pregnancy (benefits > risk); neonatal HSV-2 perinatal is separate entity
  • creatinine_baselinerequired
    lab • used at INITIAL_WORKUP
    Acyclovir 10 mg/kg q8h is renally adjusted; baseline Cr + CrCl drives dose; crystalline nephropathy risk in dehydration
  • csf_hsv_pcr
    lab • used at INITIAL_WORKUP
    CSF HSV-1/-2 PCR sensitivity 96-98% (Steiner PMID 24839282); repeat at day 3-7 if initial negative + clinical suspicion (false negatives early or post-acyclovir)
  • csf_cell_count_protein_glucoserequired
    lab • used at INITIAL_WORKUP
    CSF profile: lymphocytic pleocytosis (10-1000), mildly elevated protein, normal-to-mildly-low glucose; RBC may be elevated (hemorrhagic temporal lobe necrosis)
  • csf_autoimmune_panel_send_out
    lab • used at BRANCHING_WORKUP
    Send CSF anti-NMDAR / LGI1 / GABA-B / CASPR2 / GAD if HSV PCR negative or biphasic course (Graus Lancet Neurol 2016 PMID 30951513)
  • mri_brain_with_gadoliniumrequired
    imaging • used at INITIAL_WORKUP
    Temporal + insular + cingulate T2/FLAIR hyperintensity ± hemorrhage = pathognomonic for HSV; CT may be normal early — MRI sensitivity higher
  • ct_head_pre_lprequired
    imaging • used at RED_FLAGS
    STAT non-contrast CT to rule out bleed / mass-effect before LP; do NOT delay acyclovir while awaiting
  • eeg_routine_or_continuousrequired
    imaging • used at INITIAL_WORKUP
    Temporal lobe PLEDs ~80% in HSV; cEEG if persistent AMS or refractory seizure / suspect NCSE
  • cbc_lft_coag_lactaterequired
    lab • used at INITIAL_WORKUP
    Baseline labs for ICU admission; LFT if valacyclovir / foscarnet; coag for LP safety
  • hiv_screen
    lab • used at CONTEXT
    HIV screen if immunocompromise unknown — drives intensity of workup + foscarnet threshold

12-phase flow (12)

  1. 1FRAME
    Acute febrile encephalopathy with focal neuro deficit / seizure / behavioural change → empiric HSV coverage until ruled out (IDSA 2008 Tunkel PMID 18582201; Venkatesan IRCNS 2017 PMID 28528537)
    inputs: temperature, time_of_symptom_onset
    advance: Encephalitis pathway activated
  2. 2ENTRY
    ED presentation; activate encephalitis pathway; document acyclovir-eligible features at door
    inputs: age
    advance: ED workup initiated
  3. 3CONTEXT
    Capture immunocompromise, neonatal / pediatric status, recent abx / antiviral, autoimmune / oncologic history, pregnancy, HIV status, travel, seizure history, GCS baseline
    inputs: gcs, focal_deficit_present, seizure_history_or_active, immunocompromise_status, recent_acyclovir_or_valacyclovir, pregnancy_status
    advance: Context captured
  4. 4RED_FLAGS
    Status epilepticus (route to neuro.status-epilepticus.core.v1); raised ICP / herniation signs → mannitol + hypertonic saline + neurosurgery; respiratory failure → intubate; GCS ≤8 → intubate; massive cerebral edema; sepsis overlap (route to id.sepsis.core.v1)
    inputs: ct_head_pre_lp
    actions: workup.acute_stroke, workup.encephalopathy
    advance: Red-flags triaged
  5. 5INITIAL_WORKUP
    STAT CT head (rule out bleed / mass before LP) → LP (cell count, protein, glucose, gram + culture, HSV-1/-2 PCR, VZV PCR, enterovirus PCR; save extra CSF for autoimmune panel) → MRI brain with gadolinium (temporal + insular + cingulate T2/FLAIR + DWI for cytotoxic edema) → EEG (PLEDs ~80% HSV; rule out NCSE) → blood cultures → CBC + BMP + LFT + coag + lactate + VBG → START EMPIRIC ACYCLOVIR 10 mg/kg IV q8h (CrCl-adjusted) WITHIN 6 H of suspicion (Whitley NEJM 1986 PMID 16983610) + NS 1.5× maintenance for crystalline nephropathy prevention
    inputs: csf_cell_count_protein_glucose, csf_hsv_pcr, mri_brain_with_gadolinium, ct_head_pre_lp, eeg_routine_or_continuous, cbc_lft_coag_lactate, creatinine_baseline
    actions: panel.csf, panel.cbc, panel.renal, panel.lft, panel.coag, panel.inflammation
    advance: Empiric acyclovir started + workup sent
  6. 6BRANCHING_WORKUP
    Repeat LP day 3-7 if initial HSV PCR negative AND clinical / MRI suspicion remains (Steiner PMID 24839282 — false negatives early or post-acyclovir); CSF autoimmune panel (anti-NMDAR / LGI1 / GABA-B / CASPR2 / GAD; Graus 2016 PMID 30951513) if HSV PCR negative or biphasic; brain biopsy reserved for diagnostic uncertainty + clinical decline; oncology workup for paraneoplastic if anti-NMDAR positive (ovarian teratoma in young women)
    inputs: csf_autoimmune_panel_send_out
    advance: Differential narrowed
  7. 7DIFFERENTIAL
    HSV-1 (adult sporadic — most common) / HSV-2 (genital + neonatal disseminated) / VZV / enterovirus (peds) / arbovirus (West Nile, EEE, JE) / autoimmune encephalitis (anti-NMDAR, LGI1, GABA-B, CASPR2, AMPA, GAD) / bacterial meningitis (parallel empiric coverage) / TB meningoencephalitis / fungal (Cryptococcus, Aspergillus in IC) / paraneoplastic / toxic-metabolic encephalopathy / drug toxicity (lithium, baclofen, levetiracetam at high dose)
    advance: Differential ranked + empiric coverage maintained
  8. 8RISK_STRATIFICATION
    GCS at presentation, time-to-acyclovir (single strongest predictor — Whitley PMID 16983610), age (worse at extremes), immunocompromise, seizure burden, ICU admission, mRS pre-illness baseline
    inputs: gcs
    actions: calc.nihss
    advance: Severity stratified + ICU vs floor decided
  9. 9TREATMENT
    EMPIRIC IV acyclovir 10 mg/kg q8h (CrCl-adjusted) × 14-21 d started within 6 h (continue 21 d if confirmed HSV; 14 d if alternative dx and PCR negative) + concurrent IV fluids 1.5× maintenance to prevent crystalline nephropathy + AED (levetiracetam 1 g IV load → 500 mg-1 g q12h; phenytoin 20 mg/kg IV load if liver intact + IV access) + ICP measures if cerebral edema (mannitol 0.5-1 g/kg q6h or hypertonic 3% saline; HOB 30°; ventriculostomy / decompressive craniectomy rarely); foscarnet 40 mg/kg IV q8h (renally adjusted) for documented acyclovir resistance — thymidine kinase mutant (Sili Antivir Res 2009 PMID 18587392); for post-HSV anti-NMDAR autoimmune encephalitis (biphasic course 1-6 wk; Armangué 2018 PMID 33099399): IV methylprednisolone 1 g × 3-5 d + IVIG 0.4 g/kg × 5 d ± rituximab 1 g × 2 doses
    inputs: creatinine_baseline
    actions: workup.encephalopathy
    advance: Acyclovir + AED + ICP management started
  10. 10DISPOSITION
    Admit ICU if intubated, status epilepticus, GCS ≤8, raised ICP, hemodynamic instability; admit floor for full 14-21 d acyclovir course; transfer if no neurology / ID coverage; never discharge from ED for outpatient course
    advance: Disposition documented
  11. 11MONITORING
    Daily neuro check + GCS; Cr + UO q72h on acyclovir (hold/dose-adjust if Cr rises >0.5 mg/dL or UO <0.5 mL/kg/h); LFT q-wk; AED level; CBC; repeat MRI day 7 + day 14; repeat LP at day 3-7 if PCR negative; cEEG if persistent AMS or refractory seizure; surveillance for biphasic autoimmune relapse 1-6 wk post-HSV (Armangué 2018 PMID 33099399)
    advance: Monitoring plan documented
  12. 12FOLLOWUP
    Neuro-cognitive battery at 3 + 6 + 12 mo (temporal lobe involvement → memory + language + behaviour); seizure follow-up + long-term AED if cortical injury; psych follow-up (depression / behaviour change common); occupational + speech therapy; rehab; return precautions for autoimmune relapse symptoms (new behavioural change, dyskinesia, autonomic instability — re-eval for anti-NMDAR)
    advance: Long-term plan documented + rehab arranged