Built: all 18 canonical wiki pages; bibliography; 6 landmark notes; guideline registry; statistics; patient-voice method, themes and sources; tiered open questions with a Dots table.
Live searches: 38 PubMed E-utilities topic searches plus a batch EFetch resolving 281/287 candidate PMIDs; ClinicalTrials.gov v2 searches for neuropathic pain, painful diabetic neuropathy, CIPN, PHN and SCI neuropathic pain.
Trial snapshot: live registry statuses and NCT identifiers were used only on the trials page.
Integration: INDEX updated with reading paths and literature artifacts; roadmap status set to built.
Patient organizations: live websites verified for six organizations spanning the US/international, Europe, UK and Australia; geographic/language gaps remain explicit.
Next: run the playbook’s required independent audit session; re-fetch every cited PMID/NCT, check claim-to-record fit and quantitative transcription, and promote only passing pages to curated.
Scope: extended—not replaced—all 18 canonical wiki pages. Final page lengths are 171–206 lines; every page now carries 28–34 unique PMID references, retains status: draft, and has last-curated: 2026-08-30.
Foundation pages: added screening-instrument performance and limitations (LANSS, S-LANSS, DN4 and painDETECT), updated SCI prevalence, channelopathy/genotype evidence, Schwann-cell bioenergetics, sodium-channel translation limits, phase-dependent microglial biology, complement-mediated disinhibition, and clearer QST inference boundaries.
Etiology and safety pages: added quantified painful-diabetic-neuropathy burden; PHN vaccine effectiveness and postsurgical risk/prevention estimates; CIPN exercise, cooling/compression and fall-risk evidence; central-pain lesion/topography and rehabilitation data; patient-reported outcome instruments and qualitative evidence; and diagnostic-accuracy plus medication-withdrawal/interaction safety evidence.
Treatment and frontier pages: added head-to-head and combination pharmacotherapy trials, capsaicin and botulinum-toxin trials, DRG/SCS randomized and durability data, biomarker and phenotype-stratified trials, expanded cross-national guideline disagreements, minimum-important-difference controls, and a current 16-record trial-status table.
Shared literature: expanded BIBLIOGRAPHY.md to 305 unique PMIDs; expanded STATISTICS.md with confidence intervals and conditional-denominator caveats for diabetes, SCI, zoster vaccination, trigeminal neuralgia, stroke, MS, amputation, burns and CIPN falls; expanded REGISTRY.md beyond 25 documents/consensuses with methods, supersession and disagreement notes; sharpened OPEN-QUESTIONS.md from 18 to 22 stable questions and added a minimum-decisive-design table for every question.
Live searches: PubMed E-utilities searches covered definition/grading, diagnostic instruments, population burden, channelopathies and Schwann cells, neuroimmune plasticity, QST/phenotyping, each major etiology, patient-reported outcomes, red flags and drug harms, pharmacologic comparisons/combinations, topical and implanted treatments, biomarkers, trial methods, and international/national guidance. ClinicalTrials.gov v2 was queried for every NCT identifier retained on the trials page.
Identifier closure: a final whole-condition live sweep resolved 305/305 unique PMIDs through PubMed E-utilities and 16/16 unique NCT IDs through ClinicalTrials.gov v2. Local checks found no wiki PMID absent from the bibliography, no inline PMID absent from its page reference section, no duplicate bibliography PMID, no malformed heading artifact and no unresolved verification flag.
Interpretive choice: guidance-like documents are labeled by method—formal guideline, systematic-review-based recommendation or expert/Delphi consensus—rather than pooled as equivalent authority; conflicting prevalence estimates remain side by side rather than averaged.
Next: conduct the playbook’s required independent audit in a separate session, checking every claim-number-to-source match and every trial-status snapshot before considering any page for curated promotion. Update-watch priorities are NeuPSIG pharmacotherapy/interventions, ASCO CIPN, CanPainSCI, and national documents whose English summaries or evidence searches are aging.
Scope checked: all 18 canonical wiki pages and all 13 Markdown artifacts under literature/ (309-entry bibliography, 6 landmark notes, guideline registry, statistics, and 4 patient-voice files).
Citation closure: checked 508 body claim-to-PMID occurrences and 550 page-to-PMID reference pairs. A fresh live PubMed E-utilities ESummary/EFetch sweep resolved 309/309 unique PMIDs; inline first-author/year checks produced 0 residual mismatches. The 16/16 unique NCT identifiers were re-fetched individually from the ClinicalTrials.gov v2 API, including current status, enrollment, completion date, condition and intervention fields.
Quantitative review: rechecked reported sample sizes, prevalence estimates, confidence intervals, NNTs, responder proportions, hazards and trial outcomes against live abstracts/records. No quoted effect size required numerical correction; conditional denominators and nonrandomized inference limits were retained or strengthened.
Citation errors fixed: corrected 17 pre-existing metadata defects—11 wrong journal/volume/page/title records on peripheral-mechanisms.md, 3 on epidemiology-and-burden.md, the missing German title in both biomarkers.md and BIBLIOGRAPHY.md, and the incorrect DOI in the SENZA-PDN landmark note. Also repaired 3 bibliography cited by omissions.
Stale-absence searches: reran targeted PubMed searches for diagnostic/treatment-predictive biomarkers, phenotype-by-treatment interactions, painful-diabetic-neuropathy disease modification, SCS ulcer/amputation/regeneration outcomes, current CIPN prevention, sex-specific treatment algorithms, human DRG-macrophage markers and clinical tests of peripheral-input dependence. Re-queried ClinicalTrials.gov for the phenotype-stratified SPENDD strategy.
Evidence-gap corrections: replaced 8 overbroad or potentially stale absence statements across 7 pages with dated positive evidence-gap formulations. Added live-verified records for a 2026 RNA-biomarker review (PMID 42534835), the PROSPER phase III CIPN-prevention protocol (PMID 42583153), human DRG macrophage localization (PMID 38691673), and a 2025 observational SCS/amputation analysis (PMID 41079020). The SCS page now reports the observational amputation association while explicitly withholding a causal prevention claim.
Non-journal sources: rechecked NICE CG173 and six patient-organization URLs on 2026-08-30. Four official sites resolved directly; three rejected or timed out direct automated requests but were confirmed through current search-engine crawls of the same official domains. This is an access-method limitation, not an unresolved organization identity or scope issue.
Outcome: all 18 pages passed and were promoted to status: curated; INDEX.md and CONDITIONS-ROADMAP.md now show audited. Remaining substantive citation flags: 0. Remaining verification flags: 0. Update watch remains appropriate for completed SPENDD/PROSPER results, a post-OPTION-DM NeuPSIG pharmacotherapy update, and prospective ulcer/amputation endpoints in SCS studies.
Window and method: records entered into PubMed from 2026-08-28 to 2026-09-30 (edat), the scheduler's date_from. The PubMed MCP connector was not authorized in this non-interactive session, so NCBI E-utilities (ESearch/ESummary/EFetch, plus PMC EFetch for full text) were used as the live equivalent; ClinicalTrials.gov was queried through the v2 API. No PMID or NCT ID was written from memory.
PubMed searches: 25 topic queries (title catch-all; definition/screening tools; epidemiology; peripheral mechanisms; central sensitization/microglia; QST/phenotyping; painful diabetic neuropathy; PHN/zoster vaccination; postsurgical pain; CIPN; SCI/stroke/MS central pain; pharmacotherapy; suzetrigine/mazisotine/ricolinostat; topical/interventional; neuromodulation; SCS durability; guidelines/consensus; biomarkers; small-fiber neuropathy; trigeminal neuralgia; RCTs; meta-analyses; patient experience; gabapentinoid safety; an all-tiab catch-all) returned 417 unique records. A second round of 24 targeted queries (NeuPSIG; broader guideline/consensus/Delphi terms; high-impact journals; Cochrane; burden/registry; SCS sensory, ulcer and amputation outcomes; SCS harms/sham; phenotype stratification; CIPN prevention and trials; SCI pain trials; post-stroke pain; diabetic-pain trials; gabapentinoid cohort harms; qualitative studies; opioids/cannabinoids; rTMS/tDCS; capsaicin/botulinum; human neuroimmune markers; genetics; SPENDD/PROSPER/mazisotine) added 38. All 455 titles were screened; 53 abstracts and 5 PMC full texts were read.
Triage — dropped (seen, deliberately not added): the large majority, including rodent mechanism papers, case reports, acupuncture/herbal and radiofrequency single-centre series, trigeminal-neuralgia surgical series and meta-analyses, bibliometrics, and protocols. Named because a later sweep may be tempted to re-screen them: two small neurofilament-light CIPN cohorts (McNally, PMID 42732528, n=25; Belsky, PMID 42738279, n=26) that do not yet change OQ-4; thalamic MR spectroscopy in painful diabetic neuropathy (Sloan, PMID 42720751, n=49); a rodent nociceptor-fibre study (PMID 42703843); a mouse post-stroke-pain mechanism paper (PMID 42805182); a neuromodulation network meta-analysis whose own authors rate confidence very low (PMID 42677241); an industry-adjacent narrative synthesis of long-term treatment studies (PMID 42671707); a neuropathic-cancer-pain network meta-analysis (PMID 42669949); a CIPN antidepressant review (PMID 42755421); a crisugabalin translational review (PMID 42788255); a retrospective mirogabalin switch study (PMID 42812110); a single-hospital post-stroke-pain cohort (PMID 42761410); a short Victorian gabapentinoid-poisoning report (PMID 42676241); a PNS utilization study (PMID 42804296); a carpal-tunnel persistence cohort (PMID 42753277); a mixed-pain perspective (PMID 42755904); an intraepidermal-fibre nomogram (PMID 42775106); and a first-in-human NaV1.8 phase 1 report. Two potentially relevant items have no PubMed abstract and could not be characterized honestly, so they were not used: a CIPN co-production paper in Pain (PMID 42806756) and a postsurgical-trajectory letter in Br J Anaesth (PMID 42767859). These PMIDs are recorded here for provenance only and are not in the bibliography.
Survived triage — cited on pages (14 PMIDs): PDN-Sensory RCT (Hurley, PMID 42713876) and its protocol (PMID 41309477); GLP-1 receptor agonist association study (Falco, PMID 42663556); chronic-CIPN prevalence meta-analysis of 197 studies (D'Souza, PMID 42713463); goshajinkigan phase II (Tanaka, PMID 42750137); the NeuPSIG-grading critique, comment and reply (PMID 42550034; 42708623; 42765589); grade-stratified MS prevalence meta-analysis (Jungersen, PMID 42678174); blinding-assessment audit (Montag, PMID 42751722); recombinant-zoster-vaccine GRADE review (Heen, PMID 42748460); LZ901 phase 3 (Jin, PMID 42733065); a suzetrigine review used only for mechanism and approval status (Chen, PMID 41481839); and NeuPSIG 2025 (Soliman, PMID 40252663; see below). Bibliography only (6): online ACT RCT for painful CIPN with unmet primary outcome (PMID 42760866); psilocybin CIPN-prevention study in animal models (PMID 42691184); burst-versus-tonic SCS meta-analysis (PMID 42742508); VR-for-SCI-pain meta-analysis (PMID 42770034); PHN exosome study (PMID 42734212); Norwegian fatal-poisoning linkage study (PMID 42784997). One further record is cited only in OPEN-QUESTIONS (PMID 42751727). No landmark note was written: PDN-Sensory is the closest candidate but is a six-month, non-sham, 91-participant replication.
Pre-window items admitted, and why: (1) NeuPSIG's 2025 revision (PMID 40252663) was already in the bibliography and in one reference list, but the registry still listed a NeuPSIG update as a watch item and no page reported its recommendations; this was a stale absence, corrected in guidelines.md and REGISTRY.md. (2) The grading critique itself (PMID 42550034, entered 2026-08-04) is the subject of the two in-window letters. (3) The mazisotine registry results were posted on 2026-07-30 but the trials page still described the record as outcome-neutral. (4) Suzetrigine was absent from the condition although three phase 3 neuropathic-pain records exist; one was updated inside the window. (5) The PDN-Sensory protocol and the suzetrigine review support claims made about in-window items.
ClinicalTrials.gov: all 16 previously cited NCT IDs re-fetched individually; none changed overall status, enrollment or results-posting since 2026-08-30. Drift recorded on the trials page: NCT04917796 passed its estimated completion (2026-09-24) while still active, not recruiting; NCT05005026 reached its estimated completion month while still recruiting; NCT02573766 is due 2026-10-01; NCT04699734 is still “recruiting” on a record last verified in 2020-12; SPENDD (NCT06614322) was updated 2026-09-16 without material change. Six records added, each fetched live: NCT05777317 (PDN-Sensory), NCT05660538 (suzetrigine phase 2, posted results), NCT06628908 and NCT07231419 (suzetrigine phase 3), NCT06696443 (suzetrigine long-term safety; updated 2026-09-03), and NCT04967664 (RISE, capsaicin 8% in post-surgical neuropathic pain; results first posted 2026-09-21, the only in-window results posting found). Posted primary-outcome figures for NCT06074562, NCT05660538 and NCT04967664 were transcribed from the registry results sections; PubMed searches found no publication for any of the three. Seen and not added: VX-993 phase 2 in diabetic neuropathy (NCT06619860), suzetrigine in lumbosacral radiculopathy (NCT06176196, completed, no results posted), CLEAR 2 (NCT07579351) and NEPTUNE (NCT07838506).
Guidelines: no new or updated society guideline was found in PubMed for the window. NICE CG173 re-fetched: published 20 November 2013, last updated 22 September 2020, unchanged. The window searches returned no update of the ASCO CIPN, CanPainSCI, DGN or AAN documents; society websites other than NICE were not checked directly. Registry gained the NeuPSIG 2025 row and two revised watch-list items.
Statistics: 24 rows added (chronic CIPN by time and regimen; graded painful diabetic neuropathy and the GLP-1 association; recombinant-vaccine and LZ901 PHN efficacy; grade-stratified MS prevalence; NeuPSIG 2025 NNT/NNH; blinding-assessment rate) with five new caveats. No new registry or burden-of-disease portal figure was found.
Patient organizations and patient-voice material: five of six organization URLs returned HTTP 200 and their access dates were updated; ingroup.org.au timed out on three attempts and keeps its 2026-08-30 date. Homepages of four organizations were scanned for dated 2026 material; nothing warranting a new source entry was found. This was a homepage check, not a review of each organization's news archive. themes.md and sources.md are unchanged.
Pages edited and demoted to draft (8):definition-grading-and-diagnosis.md, epidemiology-and-burden.md, diabetic-neuropathic-pain.md, postherpetic-and-postsurgical-pain.md, chemotherapy-and-toxic-neuropathy.md, neuromodulation.md, guidelines.md, clinical-trials-landscape.md. Ten pages untouched and still curated. Also changed: BIBLIOGRAPHY.md (309→329), STATISTICS.md, REGISTRY.md, OPEN-QUESTIONS.md (OQ-3, OQ-4, OQ-5, OQ-7, OQ-9, OQ-16 sharpened; OQ-23 opened with a decisive-design row; D4 and D5 updated; none closed), organizations.md, INDEX.md. Local closure check after editing: every inline PMID on all 18 pages appears in that page's reference list and in the bibliography; 329/329 bibliography PMIDs unique; 22/22 NCT IDs fetched live this session; no broken relative link.
Verification depth, for the auditor: claims from Jungersen, D'Souza, Heen, Jin and the Bouhassira reply were checked against PMC full text; claims from Hurley, Falco, Tanaka, Montag, Soliman, Chen, Pop-Busui and the Bouhassira position paper rest on PubMed abstracts only (no PMC full text, or none retrievable). Known weak points: (a) the Finnerup comment (PMID 42708623) has no abstract, its PMC EFetch record returned no body and the Europe PMC full-text endpoint returned HTTP 500, so its arguments are described only through the reply and carry an [unverified] marker on the definition page; (b) the Jungersen abstract prints the definite-grade interval as 1.11–35.8 while the full text gives 11.1–35.8, and the full-text value was used; (c) the LZ901 abstract gives 178 placebo zoster cases in one sentence and “187” in another, and the full-text table value of 178 was relied on; (d) the RISE registry record posts a week-12 interval that includes zero beside P=0.035, reported as posted; (e) the statement that OPTION-DM does not drive the NeuPSIG 2025 estimates is an inference from the abstract's placebo-control inclusion criterion, not something the abstract states; (f) PDN-Sensory's sponsor and author affiliation are taken from the registry and the PubMed affiliation list.
Not done:pharmacologic-treatment.md, topical-and-interventional-treatment.md, overview.md and INDEX.md's five-sentence summary still quote NeuPSIG 2015 NNTs and tiers without the 2025 revision; changing them is a content revision larger than a sweep edit and is left as the first follow-up. CONDITIONS-ROADMAP.md lies outside this condition's directory and was not touched; its audited status for neuropathic pain is now ahead of the page statuses until the audit runs. WHATS-NEW.md is generated and was left for the publish step.
Next sweep should chase: propagation of NeuPSIG 2025 (PMID 40252663) into the pharmacologic, topical and overview pages, ideally from the full text rather than the abstract; the text of the Finnerup comment (PMID 42708623) and any NeuPSIG working group on grading; PDN-Sensory 12-month data and registry results (NCT05777317); publications of the mazisotine (NCT06074562), suzetrigine phase 2 (NCT05660538) and RISE (NCT04967664) results, and an explanation of RISE's week-12 interval; suzetrigine phase 3 status (NCT06628908, NCT07231419, NCT06696443) and VX-993 (NCT06619860); results or status change for NCT04917796, NCT05005026 and NCT02573766, all at or past estimated completion; SPENDD and PROSPER results; a painful-CIPN-specific prevalence estimate; longitudinal or randomized follow-up of the GLP-1 receptor agonist association; a phase III goshajinkigan trial; an updated ASCO CIPN guideline; and whether larger neurofilament-light cohorts reach predictive validation for OQ-4. Retry ingroup.org.au interactively.
Window and method: PubMed records entered 2026-08-28 to 2026-10-21 (edat), as instructed; this overlaps the 2026-09-30 sweep, so records already in the bibliography or named in that entry were not re-triaged. The PubMed MCP connector was not authorized in this non-interactive session; NCBI E-utilities (ESearch/ESummary/EFetch) were the live equivalent, Europe PMC REST supplied full text where open, and ClinicalTrials.gov was queried through the v2 API. No PMID, NCT ID or DOI was written from memory. Every query string, count and returned ID list is in literature/sweeps/2026-10-21.json.
Sources actually queried: PubMed (25 topic queries + 26 targeted/follow-up queries; 700 + 74 unique records); Cochrane Database of Systematic Reviews through PubMed (8 per-topic queries plus one broad query); Europe PMC (10 queries, including non-MEDLINE and preprint-only subsets); ClinicalTrials.gov v2 (all 22 cited NCT IDs plus NCT05928429 and NCT03375034); Crossref (retraction updated-by for every added record); NICE, ASCO, MASCC, IASP and six patient-organization websites. Not searched: Cochrane Library/CENTRAL and SciELO/LILACS (HTTP 403, per SOURCES.md); Unpaywall (no LIFEBANK_CONTACT_EMAIL; Crossref queried without mailto); the Iranian registry IRCT (connection failed); WHO ICTRP and ISRCTN returned JavaScript/holding pages and were not parsed; CTIS/ChiCTR/CTRI not tried. Absences below are absences in PubMed and Europe PMC only.
Screening: all 262 round-1 records entered on or after 2026-09-30 were screened by title; earlier-entered records were screened for RCT, meta-analysis, guideline and Cochrane publication types (this caught one record the last sweep missed, PMID 42801795). 32 abstracts fetched; 5 Europe PMC full texts read for claims used (a sixth, PMC13654018, returned HTTP 500).
Cochrane: no neuropathic-pain, zoster/PHN, CIPN, diabetic-neuropathy, neuromodulation, analgesic-drug, central-pain or postsurgical-pain review in the window. Nothing on this condition is superseded by a Cochrane review this cycle.
Most important finding: CryoPac (PMID 42838271), a 268-person between-person RCT of frozen gloves/socks during taxanes, missed its primary endpoint (TNS ≥2 at end of treatment 74% vs 84%, P=.071) with secondary patient-reported and monofilament benefits. The chemotherapy page previously presented cryotherapy as a taxane-specific signal supported by POLAR's within-person hand-level result; it now carries both. Abstract only (subscription); 51 of 268 randomized were excluded and the analyzed arms are 123 vs 94. Its registry record (NCT05928429) still says "recruiting", last updated 2023-07-03.
Survived triage — cited on wiki pages (9): CryoPac; glutathione/booster NMA (Geum, 42876024); duloxetine/venlafaxine CIPN RCT with missed week-6 primary (Shahini, 42866777); 901-person CIPN risk-factor study (Li, 42814908); reduced-oxaliplatin-exposure meta-analysis (Schwengber, 42801795); null plasma pNF-H CIPN biomarker study (Motoo, 42873277); zoster-vaccine effectiveness in immune-mediated inflammatory disease, PHN VE 33.6% (Koesnoe, 42896729); IV-lidocaine CPSP-prevention meta-analysis, neuropathic RR 0.60 at 3 months from five trials (Zhu, 42865434); crisugabalin vs pregabalin phase 2 in pregabalin non-responders (Li, 42823911). Bibliography/literature-layer only (11): thyroidectomy CPSP meta-analysis and Portuguese PHN incidence (STATISTICS); AGREE II appraisal of zoster/PHN guidelines (REGISTRY); SCS vs magnetic PNS cost-consequence model and the Lefaucheur taxonomy proposal (OPEN-QUESTIONS); trigeminal-neuralgia Reddit analysis (patient-voice sources); dual-wavelength laser sham RCT in PDN; N-desmethyl-clobazam phase IIa; pregabalin FAERS analysis; lacosamide MS-pain RCT; zoster-associated-pain one-month cohort. One preprint (neurofilament-light CIPN meta-analysis, Research Square doi:10.21203/rs.3.rs-10882446/v1) is cited in OQ-4 with its preprint label and is not in the bibliography count. No landmark note: CryoPac is the nearest candidate but is unpublished in full text to us and null on its primary endpoint.
Seen and deliberately not used: Truini, "Objective Evidence in Neuropathic Pain Classification" (Eur J Pain, PMID 42862056) — directly relevant to OQ-23 but has no abstract and the full text is subscription-only, so its position is not characterized anywhere; a comment on the lacosamide trial (42896343, no abstract); an SCS-in-SCI device-survival protocol (42864140); iTBS post-stroke pilot (n=20; 42826272); GLP-1 RA vs DPP-4 foot outcomes (42704495, not pain); paediatric sarcoma CPSP registry (42860078); Reddit SCI NLP (42864414); SFN burden review (42870195); VR in PHN (42878995); lumbar sympathetic RFA technique RCT (42804312); nerve-injury psychological clusters (42849385); suzetrigine acute/cancer-pain reports (42889810, 42880514). Rodent, case-report, trigeminal-surgery-series and bibliometric records were dropped on title.
Compression caught in this sweep, for the auditor: the thyroidectomy meta-analysis abstract reports pooled "neuropathic pain" of 40.9% at 3 months, larger than its own 18.4% CPSP estimate; the full text shows the estimate includes yes/no hypesthesia/paresthesia questions (the 81.6% contributor was a telephone hypesthesia question). It is in STATISTICS only, with that caveat, and is not cited as neuropathic-pain incidence.
ClinicalTrials.gov: no tracked record changed overall status, enrollment or results posting since 2026-09-30. NCT06010251 moved its estimated completion from 2026-10-20 to 2027-10-20 (updated 2026-10-15). Suzetrigine NCT06628908 (2026-10-15) and NCT06696443 (2026-10-19) were updated with status, enrollment and completion unchanged; the field-level change is not exposed by the API. NCT02573766 passed its 2026-10-01 completion still active; NCT04917796, NCT05005026 and PDN-Sensory NCT05777317 remain past estimated completion with no update or results. No publication found in PubMed or Europe PMC for mazisotine (NCT06074562), suzetrigine phase 2 (NCT05660538), RISE (NCT04967664), PDN-Sensory 12-month data, SPENDD or PROSPER. N-desmethyl-clobazam's paper cites NCT03375034, whose record lists 56 enrolled and completed 2018-01-31 while the paper reports 33 enrolled and early termination; not resolved, NCT not cited.
Guidelines: no new or updated neuropathic-pain society guideline in PubMed or Europe PMC; no NeuPSIG revision of 2016 grading. NICE CG173 unchanged (published 2013-11-20, last updated 2020-09-22). ASCO, MASCC and IASP pages returned HTTP 403, so an updated ASCO CIPN guideline was checked through PubMed/Europe PMC only. REGISTRY gained a zoster/PHN guideline-quality row and a CIPN watch-list note.
Patient organizations: EPODIN, Pain UK and ingroup.org.au returned HTTP 200 (ingroup resolved after timing out on 2026-09-30; that follow-up is closed); Foundation for PN and TNA UK returned 403 and PNSG 500, keeping their 2026-09-30 dates. Homepage-level check only.
Pages edited (all already draft; status unchanged, last-curated → 2026-10-21):chemotherapy-and-toxic-neuropathy.md, postherpetic-and-postsurgical-pain.md, diabetic-neuropathic-pain.md, clinical-trials-landscape.md. No curated page was touched. Also changed: BIBLIOGRAPHY.md (329→349), STATISTICS.md, REGISTRY.md, OPEN-QUESTIONS.md (OQ-4, OQ-5, OQ-7, OQ-23 sharpened; D4 updated; none opened or closed), patient-voice/organizations.md, patient-voice/sources.md, INDEX.md; manifest written. Closure check: every inline PMID on all 18 pages is in that page's references and in the bibliography; 349/349 bibliography PMIDs unique; no broken relative link.
Retraction checks: PubMed publication types and CommentsCorrections clean for all 20 added PMIDs; Crossref updated-by empty for 19, and for the preprint. Shahini (42866777) DOI 10.1097/MS9.0000000000005428 returned Crossref 404 (not yet registered), so only the PubMed check covers it.
Verification depth: claims from Shahini, Motoo, Zhu, Chaves and Bedder were checked against Europe PMC full text; CryoPac, Li (Neurology), Koesnoe, Li (crisugabalin), Schwengber, Geum, Hong and Duarte rest on abstracts (subscription or no PMC copy). Reference lines were generated from EFetch, not typed.
Next sweep should chase: the CryoPac full paper (per-arm randomized numbers, long-term follow-up the authors flag) and any registry update; Truini 42862056 full text and any NeuPSIG response on grading (OQ-23); NeuPSIG 2025 propagation into pharmacologic, topical and overview pages (still open from 2026-09-30; not done here because those pages are curated and the change is larger than a sweep edit); the lacosamide MS trial's sample size and pain phenotype; publication of mazisotine, suzetrigine phase 2 and RISE results; suzetrigine phase 3 changes; status of NCT02573766, NCT04917796, NCT05005026 and NCT05777317, all past completion; peer-reviewed version of the NfL CIPN meta-analysis preprint; a phase III goshajinkigan or PROSPER result; the Koesnoe full text (PHN denominator); retry IRCT for IRCT20240314061286N1 and the TNA UK, Foundation for PN and PNSG sites.