Skip to content

Not yet independently audited

This page was written from verified sources but has not yet passed the independent citation audit. Treat its claims as provisional.

Curation log — degenerative disc disease

2026-08-28 — Seed

  • Created: four seed files, including an 18-page canonical plan, draft overview with 25 PMIDs, and 7 seed questions.
  • Searches run (live PubMed E-utilities esearch/esummary): asymptomatic and symptomatic imaging; low-back-pain GBD; twin/genetic and injury studies; disc anatomy, biology and oxidative stress; ACP guidance, exercise and NSAIDs; Modic changes and AIM antibiotics; biomarkers; basivertebral ablation; fusion versus disc replacement; rehabilitation; cell/regenerative therapy; spinal-stenosis boundary.
  • Verification: every PMID cited was returned live from PubMed in this session.
  • Next: full build queued per CONDITIONS-ROADMAP.md; preserve the sibling distinction from spinal-stenosis.

2026-08-29 — Full build

  • Canonical corpus completed: all 18 planned wiki pages were written and left at status: draft, spanning definition/diagnostic validity, burden, biology, risk, imaging, clinical differentiation, conservative care, procedures, surgery, regeneration, outcomes, guidelines, biomarkers, trials, patient experience and safety. Page lengths are 202–291 lines and each non-registry page contains 25–36 live-verified PubMed citations; the trials page additionally carries registry-verified NCT identifiers.
  • PubMed search and verification: 49 topic-focused PubMed E-utilities searches covered epidemiology, natural history, asymptomatic imaging, grading, Modic changes, pain validity, anatomy, biomechanics, nutrition, genetics, exposures, molecular pathways, biomarkers, conservative care, medicines, behavioral care, injections, ablation, surgery, outcomes, regenerative approaches, lived experience and red flags. Exact-record EFetch calls were used to complete and normalize metadata. The finished wiki corpus cites 243 unique PMIDs, each present in the live query corpus from this session.
  • ClinicalTrials.gov search and verification: two live API v2 searches—“Degenerative Disc Disease” and “intervertebral disc degeneration OR discogenic low back pain OR vertebrogenic low back pain”—were captured on 2026-08-29. Trial status tables distinguish recruiting, active-not-recruiting, completed and stopped studies; registration is not treated as evidence of efficacy.
  • Literature layer completed: generated the deduplicated, page-mapped bibliography; wrote six structured landmark evidence notes; built a 15-document international guideline registry; assembled the source-, year-, population- and method-qualified statistics bank; and created a four-file patient-voice layer.
  • Guideline and public-source checks: live official-site searches covered WHO, NICE, VA/DoD, NASS, ACP, ASIPP, ACOEM, KCE and national Australian, Danish and German sources. Patient-voice source checks covered public pain/back organizations and health-system resources in the UK, US, Australia, Canada and Europe; access dates and provenance are recorded in the source files.
  • Research frontier rewritten: OPEN-QUESTIONS.md now contains 20 Tier 1 questions designable now, 12 Tier 2 questions awaiting prerequisite methods, and 15 explicit “Dots not yet connected” hypotheses.
  • Boundary decisions: retained low-back-pain burden estimates as contextual rather than calling them DDD prevalence; separated radiographic degeneration, discogenic pain, vertebrogenic pain, radiculopathy and spinal stenosis; treated regenerative imaging signals as exploratory rather than validated surrogates; and did not infer guideline currency where official status was unclear.
  • Known evidence gaps: there is no independent gold standard for a painful disc or level; diagnostic and trial phenotypes remain inconsistent; treatment-effect evidence is difficult to compare across rehabilitation, procedures and surgery; long-term cell/device surveillance is incomplete; and patient-voice evidence remains geographically and linguistically concentrated.
  • Lifecycle: roadmap advanced from seeded to built. Independent citation audit and page promotion from draft to curated remain a separate future stage.

2026-08-30 — Page-by-page deepening

  • Canonical pages deepened: extended all 18 existing wiki pages in place; no verified content was removed and all pages remain draft. Added page-specific discriminating-evidence matrices covering population denominators, longitudinal thresholds, transport and diurnal mechanics, single-cell states, gene–environment effects, diagnostic alternatives, treatment effect sizes, surgical crossover/reoperation, guideline disagreement, biomarker validation, trial design, patient experience and red-flag likelihood ratios.
  • Citation density: the wiki now cites 348 unique live-resolved PMIDs, up from 243. Page-level unique-PMID counts are 31–42; the clinical-trials page contains 32 PMIDs plus 38 live-resolved NCT records. All pages remain within 150–400 lines.
  • PubMed searches: ran 144 live E-utilities searches—eight narrow searches per canonical page—then EFetch retrieval for 261 unique candidate records. Candidates were screened from titles and abstracts; unrelated disease papers, veterinary classifications, protocols used as if they were results and records without a resolvable PubMed article were excluded. The integrated additions emphasize systematic reviews, randomized trials, longitudinal cohorts, diagnostic-accuracy work and evidence that changes interpretation rather than citation count alone.
  • ClinicalTrials.gov: individually resolved all 35 NCT records already present through API v2, then ran a live query.cond=Degenerative Disc Disease sweep (100 returned records; 79 not already on the page). Added NCT06053242 as active-not-recruiting, NCT03461458 as terminated and NCT07137559 as expanded access; registration remains explicitly separated from efficacy evidence. The page now contains 38 live-resolved NCT IDs.
  • Literature layer: rebuilt the bibliography increment to 348 deduplicated PMIDs with page mapping; expanded the statistics bank to 162 method-qualified rows; expanded the guideline registry from 15 to 17 documents with ASPN 2022 and historical AANS/CNS fusion guidance; retained the disagreement between procedure-focused and general-care guidelines.
  • Research frontier: sharpened ten Tier 1 questions with explicit estimands, minimum discriminating design features and results that would redirect the question; added five cross-domain junctions to the “Dots not yet connected” table (20 total).
  • Quantitative additions: examples include occupational-loading ORs 1.6–3.3, health-personnel OR 2.45 (95% CI 1.41–4.26), diurnal height change 6.1–8.0%, HIZ–discography OR 8.65 (95% CI 6.01–15.23), WHO exercise SMD −0.33 (95% CI −0.58 to −0.08), adjacent clinical disease 1.8%/year (95% CI 1.3–2.4), fracture red-flag likelihood ratios and suspected-CES bladder-scan discrimination.
  • Unresolved/next pass: this was a build/deepening session, not the separate citation-by-citation audit required for promotion to curated. Current-status declarations for ASPN 2022 and the 2014 AANS/CNS fusion series were not established from an official supersession page and are labeled accordingly. No gold standard for a painful disc/level, product-level biologic class effect or validated quantitative-MRI surrogate was inferred.

2026-08-30 — Independent audit

  • Scope completed: audited all 18 canonical wiki pages and all 13 literature-layer artifacts (bibliography, six landmark notes, guideline registry, statistics bank and four patient-voice files). All pages remain within the 150–400-line target and contain the required TL;DR, open-questions, related-pages and full-reference sections.
  • PubMed verification: fetched all 348 unique cited PMIDs live through PubMed E-utilities in one EFetch corpus. Checked 924 claim-bearing PMID occurrences across wiki bodies and literature artifacts, plus 932 full-reference/bibliography metadata occurrences. Every identifier resolved; all final titles, first authors and publication years match the live records, tolerating publication-year offsets where applicable.
  • ClinicalTrials.gov verification: fetched all 38 pre-audit NCT records individually through API v2, re-ran broad live condition/term searches, then individually resolved the 48 NCT IDs in the final corpus. Added ten directly relevant records omitted from the earlier landscape: NCT03737461, NCT05066334, NCT05287867, NCT06144970, NCT06325566, NCT06615505, NCT07168603, NCT07254806, NCT07648927 and NCT07668466. Their recruiting, active-not-recruiting or not-yet-recruiting labels match the 2026-08-30 API records.
  • Errors found and fixed: corrected 100 inline author/year labels that named a non-first author (including one 2018/2017 year error); restored the missing GBD 2021 Low Back Pain Collaborators author in two page reference lists; corrected 11 bibliography tag/cited-by mappings; refreshed 26 guideline and patient-organization URLs/access dates; and updated the clinical-trials snapshot and tables from 38 to 48 live-resolved NCT records. No PMID pointed to a different paper, and no quoted quantitative result required deletion after abstract/record checking.
  • Stale-absence searches: reran live PubMed searches for externally validated DDD biomarkers, Pfirrmann-directed exercise selection, clinical gene/RNA therapy, DDD-specific patient-priority research and DDD-specific society guidelines; reran ClinicalTrials.gov searches for gene/RNA and disc-directed interventions; and reran web searches for DDD-specific patient organizations and official guideline status. Absolute claims were replaced with positively stated, dated evidence-gap sentences. The sweep identified one DDD-specific ODI content-validity study but no representative DDD patient-priority survey with a stated probability or validated sampling frame.
  • External-source verification: checked all 26 unique literature-layer URLs live. Twenty-one returned HTTP 200 directly; three rate-limited the command-line check and two Australian Commission pages produced an HTTP/2 transport error, but all five resolved through live search/open retrieval. No dead source was retained.
  • Final validation: zero unresolved-verification markers; zero unresolved PMID/NCT records; zero full-citation metadata mismatches; zero body/reference set mismatches; zero duplicate or unused bibliography entries; zero bibliography page-map mismatches; and zero broken local Markdown links. Remaining limitations are substantive evidence gaps, not audit flags: no independent painful-disc gold standard, no externally validated clinical biomarker or polygenic score, no human DDD gene/RNA-transfer therapy, and geographically narrow patient-voice evidence.
  • Lifecycle: promoted all 18 pages from draft to curated, updated the master index to audited, and advanced CONDITIONS-ROADMAP.md from built to audited.

2026-09-30 — Periodic literature sweep (Codex; independent audit pending)

  • Window and method: used the requested inclusive publication window 2026-08-28–2026-09-30, despite the later 2026-08-30 audit. Ran 16 live PubMed E-utilities ESearch queries (datetype=pdat, mindate=2026/08/28, maxdate=2026/09/30, retmax=1000) and EFetch retrieval. The base was ("intervertebral disc degeneration" OR "degenerative disc disease" OR discogenic OR vertebrogenic); broader LBP terms were added for care, safety, burden, guidance and experience. Exact queries, returned identifiers, selected citation metadata and all trial snapshots are in sweep evidence.
  • Search coverage (hits before deduplication): overview/nomenclature 132; epidemiology/natural history/burden 115; anatomy/biomechanics 34; cell biology 58; genetics/exposures 40; imaging/diagnosis 58; conservative care 95; injections/ablation 19; surgery 23; regeneration 77; outcomes 52; guidelines 41; biomarkers 49; trials 19; patient experience 13; safety/differential 50. Total 303 unique records (301 journal-article records and two book records). Screened titles and selected relevant abstracts; broad-query spillover included veterinary, cervical-only, stenosis, perioperative, protocol and case-report material. PubMed's publication-date filter also returned older online papers assigned to September issues; no claim is made that all 303 were newly published online during this interval.
  • Triage: retained 10 papers, bringing the bibliography to 358 unique records. Four wiki pages changed: conservative treatment (exercise-dose uncertainty and the negative pamidronate trial), injections/ablation (cohort safety versus MAUDE report denominators), guidelines (new acute guideline scope and ASIPP's companion summary), and trials (status/results refresh and newly tracked placebo-controlled fusion study). Pfirrmann reliability and H3K27me3 multi-omics were bibliography-only; the latter remains preclinical. Added the small qualitative care-pathway study to patient-voice sources and the smoking-attributable GBD estimate to statistics. No landmark warranted a deep note. The two July-online/September-issue papers retained were absent from the existing bibliography: Fischer's safety-report analysis and Soubrier's negative pamidronate RCT.
  • ClinicalTrials.gov: individually queried all 48 previously cited IDs through live API v2, with no failed records. DREAM (NCT05066334) changed from the local recruiting label to UNKNOWN; last posted update is still 2025-05-25, so this is not represented as a newly submitted sponsor change. HYDRAFIL (NCT06011551), allograft (NCT06778447) and phase-3 IDCT (NCT07254806) have September record updates and remain recruiting without results. Four tracked records contain results (NCT01299077, NCT01502644, NCT04544709, NCT05207813), all with record updates preceding the window; no tracked results-first-post date falls within it. No full historical JSON baseline was available for field-by-field amendment comparison.
  • Trial discovery: query.cond=Degenerative Disc Disease with AREA[LastUpdatePostDate]RANGE[2026-08-28,2026-09-30] returned 17 records, no further page. Added only “Fusion or Not” (NCT06709391), previously registered in 2024 and updated 2026-09-08, because its fusion-versus-surgical-exposure comparator directly addresses a central uncertainty. It is recruiting, 90 participants planned, six-month ODI primary endpoint, estimated primary completion December 2030, no posted results. The tracker now contains 49 records; other discoveries were cervical/perioperative/device cohorts or broader mixed musculoskeletal care and were not added merely for volume.
  • Guidelines: checked PubMed plus official/publisher WHO, NICE, VA/DoD, NASS, KCE, AAPM and ASIPP-related records. Catalogued Skelly's 2026-09-02 acute (<6-week) outpatient LBP guideline; verified scope/methods, not detailed recommendations. AAPM's hub still says “Coming Soon” despite the live publication. ASIPP's September analytical paper summarizes its existing 2025 guidance and was not counted as an independent new guideline or efficacy trial. Registry count is now 18 documents plus that companion summary. No replacement was found on the checked WHO/NICE/VA/DoD/NASS/KCE pages; this is a limited dated check, not a global absence claim.
  • Statistics and patient material: added four denominator-qualified statistics rows (166 total): smoking-attributable LBP DALYs, BVN cohort complications, MAUDE fracture report share and pamidronate outcome. Checked WHO burden material and the American Spine Registry 2026 report listing; did not substitute registry procedure volume or combined spinal/MSK burden for DDD prevalence. Added public U.S. Pain Foundation #OneVoiceForPain, BackCare's October booklet announcement and Painaustralia's September factsheet announcement. These are broader pain resources, with no private patient information recorded.
  • Research frontier and lifecycle: sharpened OQ-4 (prospective safety capture), OQ-6 (dose-by-phenotype evidence) and OQ-9 (specific fusion effect versus rehabilitation comparison), keeping stable identifiers and 32 questions. Updated the index and all edited Markdown frontmatter dates; four edited wiki pages are draft, the remaining 14 keep their prior audited status. All new PMID/NCT citations were returned by live E-utilities/API calls in this session. Local citation/link/status checks are recorded below; they are not an independent audit. Publishing remains outside this sweep as instructed.
  • Local validation: zero added PMID/NCT identifiers absent from this session’s live responses; zero duplicate bibliography entries (358 unique); all 49 tracked trial identifiers and table statuses match stored API fields; zero broken local links in edited files; exact body/reference PMID agreement on the four edited wiki pages; all four remain within the 150–400-line convention; edited frontmatter is dated 2026-09-30 and marked draft. The separate independent audit is still pending.
  • Next sweep: chase the full acute-guideline recommendation table and AAPM publication-status reconciliation; monitor DREAM's stale registry status and product-trial results; follow “Fusion or Not” without mistaking registration for outcomes; extract BackCare's booklet once released around 5 October; check VA/DoD/KCE replacements and NASS revision timing; assess the Chinese multidisciplinary consensus dated 2026-08-27 (discovered just outside the requested cutoff); retain the preclinical/clinical and DDD/LBP boundaries. Recheck independently ascertained fracture risk and whether exercise-dose estimates replicate within disc/endplate phenotypes.

2026-10-14 — Periodic literature sweep (Claude; independent audit pending)

  • Window and method: Window was 2026-08-28–2026-10-14, inclusive, as requested. Each of 19 PubMed E-utilities queries was run with both datetype=pdat and datetype=edat, and the ID sets were unioned. The 16 topic queries from 2026-09-30 were reused, with the base widened to add "disc degeneration" and "Modic changes". Three were new: a Cochrane query ("Cochrane Database Syst Rev"[jour] AND (back pain / intervertebral disc / lumbar / spinal / sciatica)), an LBP systematic-review/meta-analysis query and an LBP RCT query. The 303 IDs already screened on 2026-09-30 were excluded, leaving 239 unscreened unique records. These were screened by title/publication type, and 17 abstracts were read. Exact queries, translations, IDs and per-record checks are in sweep evidence.
  • Sources actually queried: PubMed (ESearch/ESummary/EFetch); Europe PMC REST (210 hits, 32 non-MEDLINE: 25 PMC-only, 7 preprints); Crossref /works/{doi} for retraction status; ClinicalTrials.gov API v2; the ISRCTN API; and official NICE, AAPM and BackCare pages. Europe PMC supplied the full text of the Chinese consensus (PMC13522480), and Nature served the Fillingim full text. Not searched: Cochrane Library/CENTRAL (HTTP 403); SciELO/LILACS (HTTP 403); WHO ICTRP (JavaScript portal, not searched programmatically); Embase/Scopus. Unpaywall was skipped because LIFEBANK_CONTACT_EMAIL is unset; Crossref was queried without mailto. OpenAlex and DOAJ were not used. Absences below are absences in the named sources only.
  • Cochrane: four CDSR records fell in the window: exercise in spinal cord injury, fracture prevention in men, cardiorespiratory training after stroke, and one previously screened. No new or updated Cochrane review bears on DDD or LBP in PubMed for this window, so nothing the condition asserts is superseded by Cochrane.
  • Most important finding: a correction. NICE NG59 reads "Last updated: 29 July 2026". NICE "withdrew the recommendations on psychological therapy … and combined physical and psychological programmes" (former 1.2.13–1.2.14) and removed psychological-therapy references from 1.1.3 and 1.2.7. The registry still said "last updated 2020", and the 2026-09-30 sweep repeated that status from a site check, so it was a stale claim that two passes did not catch. Corrected in REGISTRY.md (row, per-guideline entry, watch list) and in wiki/guidelines.md (major-families table, NICE section, exercise/psychological agreement table). It is framed as a withdrawal that leaves NG59 silent on psychological care, not as a recommendation against it. NG59's invasive recommendations (1.3.1, 1.3.9, 1.3.10) were re-read and are unchanged.
  • Guidelines: (1) The Chinese multidisciplinary consensus (Gao 2026, PMID 42660880), the follow-up queued on 2026-09-30, was added as registry document #19. Its recommendation wording was verified from the full text: fusion weak-for discogenic pain with concordant imaging (Level I, Grade B); TDR weak-for (Level I/II, Grade B); intradiscal steroid weak-for with Modic change/active discopathy (Level I, Grade B); MSC weak-for as investigational (Level I, Grade C); IDET and intradiscal radiofrequency thermocoagulation research/governance only. Verbatim checks found two things the abstract does not show. The PRP recommendation is internally inconsistent: the boxed statement says ankylosing spondylitis–related pain, while the body says chronic discogenic pain, and the evidence paragraph cites a sacroiliac PRP trial. The full text also contains no mention of basivertebral/intraosseous ablation or vertebrogenic pain (0 matches). (2) The AAPM hub reconciliation, a 2026-09-30 follow-up, is done: "Coming Soon" is gone, and the clinician key-recommendation summary is now recorded in guidelines.md and the registry. These are the society's summary bullets; graded recommendations from the full guideline are still not extracted. VA/DoD, NASS, KCE and WHO were not re-checked this sweep.
  • Triage — retained 6 of 239: Gao 2026 (consensus; guidelines, injections, registry, OQ-4/OQ-14). Madrazo-Ibarra 2026 (PMID 42852874): a meta-analysis of 11 PRP RCTs in 623 patients with no significant pooled pain/function difference at any time point, very low certainty, and replicable preparation in 1/11 trials. It is abstract-level only; the full text is Wiley HTTP 403 and absent from Europe PMC. It went into the injections TL;DR and PRP table, the regenerative PRP table, guidelines, STATISTICS and OQ-14. Fillingim 2026 (PMID 42834228; Nat Med pain reference curves: back pain ~40%, late-life low- vs high-HDI back-pain risk difference 20.2 pp, 95% CI 19.6–20.9, read from the full text) and Nkhata 2026 (PMID 42713377; pooled African LBP prevalence) went to STATISTICS only, labelled as LBP rather than DDD. Adegbenro 2026 (PMID 42863602; MR found no causal telomere-length effect on disc degeneration) is bibliography-only. Lindschouw Andersen 2026 (PMID 42821814; pre-fusion qualitative study, n=14) went to patient-voice sources. All six are clean on PubMed publication types/CommentsCorrections and on Crossref updated-by. No landmark deep note was warranted. The rest was mostly preclinical nanomaterial, ferroptosis and senescence work, cervical or herniation surgery, case reports, broad LBP rehabilitation and veterinary papers, and was not added for volume. Two of those were noted but left out: a small retrospective Modic-1 fusion cohort (PMID 42836528) and a 7-year M6-L follow-up (PMID 41729208).
  • ClinicalTrials.gov: all 49 tracked NCT records were re-queried individually. There were no status changes and no new results. ASCEND (NCT06615505) shows primary completion ACTUAL 2026-02-06 with no posted results. The enrollment field of Discogen (NCT06611397) now reads 5 (estimated) against "40 evaluable" in its summary. Discovery (query.cond=Degenerative Disc Disease, and a discogenic/vertebrogenic/IVDD/Modic term query, both with LastUpdatePostDate 2026-09-30→10-14) returned 7 and 10 records. Two were added: the VIA Disc NP Registry (NCT04652687), TERMINATED for "Sponsor decision.", 83 enrolled, no results; and the Intracept real-world registry (NCT06827262), recruiting, about 1,500 planned, 60-month ODI endpoint, the first prospective treated-denominator BVN harms source for OQ-4. The tracker now holds 51. ISRCTN (52 records) returned no lumbar-disc trial updated in the window.
  • Europe PMC preprints (not cited; watch list): one is a Research Square meta-regression (doi:10.21203/rs.3.rs-10748663/v1) reporting that antibiotic dose and modelled intradiscal exposure explain between-arm variation in Modic cLBP outcomes across four studies. It is relevant to OQ-3 and the AIM line, but it is a preprint and its authors' affiliations were not verified. The others are a culture-assay-sensitivity scoping review of disc microbiology and a Tanzanian Modic prevalence preprint.
  • Other edits: two mislabelled first authors in OQ-14 were corrected against live ESummary: PMID 32607308 is Hirase, not "Chang", and PMID 39238810 is Schepers, not "Zielinski". BackCare's Awareness-Week booklet was still not linked on 2026-10-14 and is recorded as pending.
  • Status changes: regenerative-and-biologic-therapy.md was demoted from curated to draft, because its PRP row rests on an abstract-level reading. Guidelines, injections and trials pages remain draft and are re-dated 2026-10-14. Counts are now 13 curated and 5 draft; 364 bibliography PMIDs; 51 trials; 19 guideline documents; 171 statistics rows.
  • Local validation: every PMID added in this sweep was returned by a live E-utilities call this session. Every added NCT ID was returned by the API v2 this session. Body and reference PMID sets agree on all four edited wiki pages. There are no broken relative links in edited files, and all pages are within 150–400 lines. This is not an independent audit; a different engine must audit both the 2026-09-30 and 2026-10-14 sweeps.
  • Next sweep: (1) Try to obtain the Madrazo-Ibarra full text, and check whether the pooled SMDs support the "trend favouring PRP" wording. (2) Watch for an erratum to the Chinese consensus PRP statement. (3) Extract the AAPM/Skelly graded recommendations from the full guideline. (4) Check whether NICE issues replacement psychological-care guidance, and re-check VA/DoD and KCE. (5) Follow ASCEND for results now that primary completion has passed, Discogen's enrollment discrepancy, and DREAM's UNKNOWN status. (6) Check whether the antibiotic exposure-response preprint is published. (7) Look for the BackCare booklet. (8) Run OpenAlex citation checks on the PRP meta-analysis.

2026-10-21 — Periodic literature sweep (Codex; independent audit pending)

  • Window and sources: used the requested inclusive publication window 2026-08-28–2026-10-21. Ran 18 topic-specific PubMed E-utilities ESearch queries with datetype=pdat, mindate=2026/08/28, maxdate=2026/10/21; paired each with a CDSR journal query and added a broad LBP/lumbar Cochrane safety net (37 main queries). Their union contained 549 records, including 72 absent from the two previous manifests. Fetched the new candidates plus the CDSR hit through EFetch, screened titles and selected abstracts, and retrieved additional Europe-PMC-discovered candidates through EFetch. Two final targeted PubMed follow-up queries checked the antibiotic preprint’s title/exposure-response terms and consensus errata; superseded broad queries are retained transparently but do not support absence claims. Exact strings, translations, identifiers and registry checks are in the search manifest.
  • Beyond PubMed: queried Europe PMC per topic (958 unique records; 230 non-MED records, including duplicate PMC representations and preprints, not 230 independent new studies). One PubMed-only search tag initially produced an unusable Europe PMC review query; reran it in native syntax and retained both requests. Queried Crossref for all three added records, ClinicalTrials.gov API v2, ISRCTN’s API, the CTIS ABBA portal and WHO ICTRP, and OpenAlex for the PRP review’s citation graph. Retrieved official WHO, NICE, VA/DoD, NASS, KCE, AAPM, American Spine Registry and BackCare material; AAPM’s page was accessible through browser retrieval despite a direct HTTP 403. No non-PMID record was retained, so no DOI-only citation was added.
  • Cochrane correction: Arienti’s September lumbar-support review (PMID 42680183) is relevant. Its ID was present in the 2026-10-14 manifest, but that log incorrectly said no relevant Cochrane review was found. Added a bibliography line and an OQ-6 full-text-appraisal follow-up. The abstract and Cochrane.org public summary were retrieved; PMC returned HTTP 403, Europe PMC’s XML endpoint returned 500, and Europe PMC reports no available full text. Its evidence synthesis has therefore not been promoted into the wiki. The manifest records per-topic CDSR queries plus a broad safety-net query and requires title/abstract checking before dismissing a returned review; this makes the previous screening omission inspectable.
  • Triage — three bibliography additions: (1) Arienti, above, pending full text; (2) Schol’s ISSLS VERTEBRA reporting standard (PMID 42900386), verified against the publisher’s full-text body and Table 2; (3) Dutra’s broad musculoskeletal sham-comparison review (PMID 42399214), discovered through Europe PMC and fetched independently through E-utilities, bibliography-only because Wiley full text remains HTTP 403 and procedure-specific extraction is pending. VERTEBRA was incorporated into the regenerative page, a separately labelled reporting-standard entry in REGISTRY, and OQ-14. It is not a treatment recommendation or a substitute for risk-of-bias assessment. Small MRS/discography and mixed cervical/lumbar microbiology cohorts and heterogeneous general-LBP exercise reviews did not change clinical claims. No landmark warranted a deep note.
  • Retraction checks: all three additions were checked against live PubMed publication types and correction links and Crossref updated-by; no retraction/expression-of-concern flag or Crossref update was found. Cochrane’s UpdateOf link is a protocol-to-review update, not a retraction. Root-record identifier checks exclude identifiers embedded in reference lists. No new PMID or NCT citation came from memory or an earlier manifest alone.
  • ClinicalTrials.gov: re-fetched all 51 previously cited records individually. The 49 with structured prior snapshots show no overall-status change and no newly posted results. The other two (added in the preceding sweep) match that log’s status/results descriptions, but have no prior structured snapshot for field-level comparison. IDCT phase 3 (NCT07254806) has a newer posted date, 2026-10-15, and remains recruiting without results; the limited prior snapshot cannot identify changes to protocol fields it did not preserve. ASCEND remains active, not recruiting without results; Discogen’s enrollment-field/summary discrepancy persists; DREAM remains unknown. Discovery queries for DDD and discogenic/vertebrogenic/IVDD/Modic terms, with last-update dates restricted to the requested window, returned 30 and 42 records. Added only ABBA (NCT06771492), active-not-recruiting placebo-controlled amoxicillin trial in Modic-1 chronic LBP, and CLARITY (NCT06661850), recruiting Nociscan disclosure study. CLARITY’s official title says randomized while its structured field says observational prospective cohort; the tracker and OQ-1 preserve this discrepancy. These were registered before this window and are newly tracked here, not newly initiated trials. Total tracker: 53.
  • International-registry limits: ISRCTN returned 52 records, seven updated in the window; their titles concern perioperative antibiotics, cervical devices, CES recognition, deformity or ophthalmology, not a newly reported intradiscal efficacy result. WHO ICTRP returned HTTP 403. ABBA’s named CTIS page was queried, but only a JS portal was retrievable; the CTIS number is supplied by its live ClinicalTrials.gov record and European status was not independently reconciled. No international absence claim is made. CENTRAL was not searched (Cochrane Library HTTP 403 limitation). SciELO and LILACS could not be searched (HTTP 403 limitation), so regional evidence completeness is not asserted. Unpaywall was skipped because LIFEBANK_CONTACT_EMAIL is unset; no personal address was sent. DOAJ and subscription databases were not searched.
  • Guidelines, statistics and patient material: NICE still lists its 29 July 2026 update; VA/DoD still lists the 2022 CPG, and NASS the 2021 LBP guideline. Corrected the registry’s unsupported present-tense “update in progress” for VA/DoD to what its current hub actually establishes. KCE lists study 2026-13 under planned projects, not as a published replacement. AAPM’s clinician summary remains available; full Skelly recommendation-grade extraction is still blocked by publisher HTTP 403. WHO’s fact sheet and ASR’s 2026 report listing were checked; no new numeric estimate survived triage and no report tables were extracted. STATISTICS records that bounded check without substituting general LBP or surgery counts for DDD prevalence. BackCare’s 2026 landing page now embeds its chronic-back-pain booklet, resolving the pending availability check; updated the patient-source row, explicitly without claiming full booklet appraisal.
  • Prior follow-ups: the PRP full text remains unavailable through the checked publisher/Europe PMC routes. OpenAlex’s live PRP record reports zero citing works; that is a database snapshot, not proof of no subsequent discussion. The corrected, date-limited PubMed follow-up queries returned no matching antibiotic exposure-response publication or consensus erratum; these are PubMed-only findings and do not resolve the preprint or the consensus’s wording conflict.
  • Changes and lifecycle: updated two existing wiki pages (regenerative therapy and trial landscape), BIBLIOGRAPHY, guideline registry, statistics check, patient-source entry, OQ-1/OQ-3/OQ-6/OQ-14 and INDEX; appended this single log entry and the manifest at the canonical path. Counts: 367 bibliography PMIDs; 53 tracked trials; 19 clinical guideline/policy documents plus a companion summary and a separate reporting standard; 171 unchanged statistics rows; six unchanged landmark notes. All edited Markdown frontmatter is dated 2026-10-21 and draft; the other 13 curated wiki pages retain their status. Local checks are not the independent audit. No publishing or global regeneration was run, and no file outside this condition was changed.
  • Local verification: all three added PMIDs and both added NCT IDs occur in this session’s live EFetch/API records; Crossref checks are recorded for all three papers. The bibliography contains 367 unique entries without duplicates, all 53 tracker IDs have current snapshots, the two edited wiki pages have matching body/reference PMID sets and remain within 150–400 lines, and edited-file relative links and frontmatter checks pass. Wiki statuses remain 13 curated and five draft. No global build was run because it would regenerate files beyond the authorized condition.
  • Next sweep: obtain full text for the Cochrane lumbar-support review and Dutra’s sham-comparison review before revising evidence synthesis; retrieve PRP pooled estimates and Skelly graded recommendations; clarify CLARITY’s design fields and independently reconcile ABBA’s CTIS record; monitor ABBA and ASCEND results, IDCT amendments, Discogen enrollment and DREAM status; watch for the Chinese consensus erratum and antibiotic-preprint journal publication; monitor KCE’s planned update and any NICE replacement; appraise BackCare’s actual booklet. Independently audit this sweep and the two earlier pending sweeps before page promotion.

2026-10-28 — Periodic literature sweep (Grok; independent audit pending)

  • Window and method: inclusive publication window 2026-08-28–2026-10-28, as requested. Ran 18 topic-specific PubMed E-utilities ESearch queries with both datetype=pdat and datetype=edat (mindate=2026/08/28, maxdate=2026/10/28), paired each topic with a CDSR journal query, and added a broad LBP/lumbar Cochrane safety net. Rate-limited (HTTP 429) queries were retried and succeeded. The union contained 673 unique records, of which 122 were absent from the three previous manifests. Those 122 were fetched by EFetch and screened by title, publication type and selected abstracts. Exact strings, translations, identifiers and registry checks are in the search manifest.
  • Sources actually queried: PubMed ESearch/EFetch; Europe PMC REST (19 topic queries; 544 unique IDs in the first page-set); Crossref /works/{doi} for retained records (no mailto; LIFEBANK_CONTACT_EMAIL unset); OpenAlex pmid:42852874 (PRP review cited_by_count=0); ClinicalTrials.gov API v2 (53 individual records plus three discovery searches); Cochrane.org public summary for Arienti; official WHO, NICE, VA/DoD, NASS, KCE, AAPM, ASR, BackCare, Painaustralia, U.S. Pain Foundation and ASIPP pages. Rejner full text from Europe PMC XML (PMC13706458); Mathieson full text from the PLoS publisher manuscript.
  • Not searched / limits: Cochrane Library/CENTRAL HTTP 403; SciELO HTTP 403; LILACS HTTP 200 of a ~6 KB JS shell, not a searchable index; WHO ICTRP HTTP 200 JS portal without a usable study-detail API; ISRCTN JSON/search endpoints returned JS shells this session, so no structured ISRCTN update list was parsed; Unpaywall skipped; DOAJ and Embase/Scopus/Web of Science not searched. Absences below are absences in the named sources only. Regional evidence completeness is not asserted.
  • Cochrane: the only CDSR record in PubMed for the per-topic and safety-net queries remains Arienti 2026 (PMID 42680183), already catalogued. Europe PMC’s looser CDSR-syntax query returned seven hits; six were congress/bibliometric records that are not Cochrane reviews. No new or updated Cochrane review of DDD or LBP appears in PubMed for this window, so nothing the condition asserts is superseded by Cochrane.
  • Triage — retained 2 of 122: (1) Rejner 2026 (PMID 42941526), a 210-person preoperative lumbar-DDD coping/affect cohort, verified from PMC full text including Table 7 prevalence and the adjusted catastrophizing βs; added to the patient-experience page, patient-voice sources/themes, STATISTICS and an OQ-20 note. (2) Mathieson 2026 (PMID 42945153), a PLoS Medicine systematic review of cost-effectiveness of guideline-recommended nonspecific-LBP care (20 papers, 24 studies, 55 comparisons, all high-income); verified from publisher full text; added to guidelines.md and the registry as an economic companion synthesis, not a CPG. No landmark deep note was warranted. Screened and not retained: nonspecific-LBP rehabilitation RCTs, physical-agent NMAs, single-center opioid-prescribing trends, China GBD 2021 reanalysis (already screened 2026-10-14), narrative intradiscal/orthobiologic reviews, small surgical series, case reports, and preclinical ferroptosis/senescence work.
  • Retraction checks: both additions have live PubMed publication types without retraction/erratum/expression-of-concern flags, empty CommentsCorrections, and empty Crossref updated-by. A Frontiers mechanotransduction correction (PMID 42719552) is not a paper this condition cites. The PRP meta-analysis DOI 10.1002/pmrj.70216 was re-checked via OpenAlex (zero citing works as of this snapshot); Crossref 404’d an outdated DOI guess and was not treated as a retraction signal.
  • ClinicalTrials.gov: re-fetched all 53 previously cited records individually. No overall-status change and no newly posted results versus the 2026-10-21 snapshots. ASCEND remains active, not recruiting without results after actual primary completion 2026-02-06; Discogen’s enrollment-field/summary discrepancy persists; DREAM remains unknown. Discovery queries returned 33 and 46 records in the requested last-update window. Added only NCT07895225 (not yet recruiting; single-blind PRP vs allogeneic NP graft in chronic discogenic LBP; 60 planned; estimated primary completion 2028-06). It is not sham-controlled and cannot answer OQ-14. Total tracker: 54. Other discoveries were cervical cages, herniation/radiculopathy devices, perioperative care or mixed MSK programs.
  • Guidelines, statistics and patient material: NICE still lists 29 July 2026; VA/DoD still lists the 2022 CPG; NASS still lists the 2021 LBP guideline (its “2026 Biologic Interventions” item is a November course, not a CPG); KCE 2026-13 remains planned; AAPM clinician summary remains live and the Skelly full text remains HTTP 403; WHO 2023 publication page HTTP 200. No published Chinese-consensus erratum in PubMed. ASR still lists a 2026 annual report; tables were not extracted. BackCare’s booklet remains embedded and unappraised. STATISTICS gained three Rejner rows (surgical-selected, instrument-specific) and did not treat LBP or surgery counts as DDD prevalence.
  • Prior follow-ups: PRP full text still publisher-blocked; OpenAlex citing-works count remains 0. Antibiotic exposure-response work remains the Research Square preprint (PPR1309929). Arienti and Dutra full texts still unavailable through PMC/Europe PMC. CLARITY’s observational-vs-randomized field conflict and ABBA’s CTIS status are unchanged.
  • Changes and lifecycle: updated three wiki pages (patient-experience, guidelines, trials), BIBLIOGRAPHY, guideline registry, STATISTICS, patient-voice sources and themes, OQ-14/OQ-20 and INDEX; appended this log entry and the manifest at the canonical path. Counts: 369 bibliography PMIDs; 54 tracked trials; 19 clinical guideline/policy documents plus a companion summary, a reporting standard and one economic-synthesis companion; 174 statistics rows; six unchanged landmark notes. Patient-experience was demoted from curated to draft because new claims were added; guidelines and trials remain draft. The other 12 curated wiki pages retain their status. No publishing or global regeneration was run, and no file outside this condition was changed.
  • Local verification: both added PMIDs and the added NCT ID occur in this session’s live EFetch/API records; Crossref checks are recorded for both papers. The bibliography contains 369 unique entries without duplicates; all 54 tracker IDs have current snapshots; the three edited wiki pages have matching body/reference PMID sets and remain within 150–400 lines (299 each). Wiki statuses are 12 curated and six draft. This is not an independent audit; a different engine must audit this sweep and the three earlier pending sweeps.
  • Next sweep: obtain full text for Arienti (lumbar supports) and Dutra (sham-comparison MSK review) before revising evidence synthesis; retrieve PRP pooled SMDs and Skelly graded recommendations; watch NCT07895225 enrollment and whether it reports a replicable PRP preparation; monitor ASCEND results, ABBA, IDCT amendments, Discogen enrollment and DREAM status; watch for a Chinese-consensus erratum and journal publication of the antibiotic-preprint; monitor KCE 2026-13 and any NICE psychological-care replacement; appraise BackCare’s booklet; independently audit the four pending sweeps before page promotion.