Peptide education
LL-37
Cathelicidin LL-37 · Cathelicidin · LL37
LL-37 is the only cathelicidin your body makes, cut from the hCAP18 precursor wherever innate defense is needed. It punches holes in microbial membranes, neutralizes endotoxin, recruits immune cells, and helps close wounds, which is a genuinely impressive resume for a molecule. It also binds your own DNA and RNA in ways that can drive psoriasis-like and lupus-like inflammation, and it shows up on both sides of cancer biology. The human trial record is real but deflating: small topical wound studies, with the largest controlled venous-ulcer trial missing its primary endpoint. The injectable Lyme-and-immunity market is built on none of that.
LL-37 is your own host-defense peptide with real topical wound trials, the biggest of which missed its endpoint. Everything the market claims beyond wounds has no human outcome data and runs into this molecule's double-edged immune biology.
Overview
Quick answer
LL-37 is not an FDA-approved therapeutic peptide. The name can refer to an endogenous human peptide, investigational topical wound products, an intratumoral oncology study drug, an oral recombinant Lactococcus product in a COVID-19 trial, compounded material, or research-use vendor vials. Each version changes the route, exposure, quality controls, and claim being tested.
What is LL-37?
The active peptide released when the human cathelicidin precursor hCAP18 is cleaved; the only cathelicidin humans have. It disrupts microbial membranes, binds endotoxin, and signals through receptors such as FPRL1, so it is host-defense machinery rather than a simple antibiotic.
Why is it discussed?
Because it sits at the intersection of several attractive ideas: antimicrobial and anti-biofilm activity, wound repair, and immune modulation. Topical wound programs took it into human trials. The market then stretched it into chronic infection, Lyme, mold, SIBO, and general immune support, which the trial record does not reach.
What route and dose patterns appear in published sources?
The wound trials used topical LL-37 at 0.5 or 1.6 mg/mL twice weekly: 4 weeks in the earlier trial, 13 weeks in the phase IIb. The melanoma registry used weekly intratumoral injections of 250 to 2000 micrograms per cubic centimeter of tumor for up to 8 weeks. A COVID study used an oral recombinant Lactococcus product, a different object entirely from a synthetic vial.
What do clinics and market pages add?
Injectable programs for chronic infections, Lyme, mold, SIBO, and immune resilience, plus nasal sprays and skin products, typically described at 50 to 100 mcg per day subcutaneously for 4 to 6 weeks. No human study supports those uses, and injecting an immune-active peptide raises an immunogenicity question the topical trials never had to answer.
How should the evidence be read?
Route by route, product by product. The topical ulcer trials apply to local wound care. The melanoma registry applies to direct tumor injection. The oral COVID study applies to a recombinant product. A research-use vial applies to none of the above; it mostly tells you about market access and product risk.
Reported practice
Commonly reported protocol
Community antimicrobial-peptide protocols. Community-reported patterns, not verified by controlled human trials and not a use recommendation. Full use-pattern detail
Evidence
Evidence snapshot
The cited wound literature includes a 34-participant randomized venous-leg-ulcer trial. Later wound examples add context, but the wound evidence still belongs in the investigational range.
FDA materials discuss LL-37 in compounding and safety-risk contexts, not in an approved therapeutic label. Market availability is not approval.
LL-37 has been studied by weekly intratumoral injection in cutaneous melanoma metastases, but the registry is too small and outcome reporting too limited to support a cancer-treatment claim.
FDA compounding materials and warning-letter context are important because public peptide listings, COAs, and research-use disclaimers do not show sterile-drug manufacturing, correct active content, endotoxin control, stability, or clinical equivalence.
Claims
Common claims vs evidence
| Claim | Human evidence | Mechanistic evidence | Anecdotal evidence | Verdict |
|---|---|---|---|---|
| LL-37 fights bacteria, fungi, viruses, and biofilms. | The strongest support for this claim is not human anti-infective outcome data. In vitro and preclinical papers describe antimicrobial and anti-biofilm activity. Infection treatment in people needs route- and product-specific human outcome data. | LL-37 is cationic and amphipathic, so it can interact with negatively charged microbial membranes and microbial products such as LPS. It also affects host receptors and immune signaling, which makes its effects broader than direct membrane disruption. | Clinics, forums, vendor pages, and online regimen posts discuss LL-37 for Lyme, chronic infections, mold, SIBO, immune defense, and biofilms. Those discussions show how it is marketed and discussed, not patient outcome data. | The preclinical host-defense rationale is real, but it has not shown human treatment benefit for infections, Lyme, mold illness, SIBO, or biofilm eradication. |
| LL-37 helps chronic wounds or skin repair. | This is the most credible human use case. The 34-participant venous-leg-ulcer trial reported healing results with topical LL-37. A later 148-participant phase IIb venous-leg-ulcer trial and a small diabetic-foot-ulcer trial add background wound context, but they do not set a dosing standard. | The wound rationale fits LL-37 biology: immune-cell recruitment, angiogenesis signaling, epithelial migration, and re-epithelialization are all part of the preclinical literature. | Clinic and vendor discussion often expands from chronic wounds to broader skin repair, recovery, and topical or injectable use. That expansion goes beyond the studied wound settings. | Discussable as topical investigational wound research. Not a general skin repair, infection, or injectable recovery claim. |
| LL-37 boosts immunity. | Human trials did not establish LL-37 as a general immune-support product. An oral recombinant Lactococcus product has been studied in COVID-19 research, which is not the same as synthetic LL-37 vials or broad wellness use. | LL-37 is immunomodulatory, not simply immune boosting. It can recruit immune cells and neutralize some microbial triggers, but it can also bind self-DNA or self-RNA and amplify inflammatory pathways tied to psoriasis and lupus-like biology. | Immune-support language is common in clinic, forum, and gray-market discussion because host-defense biology is easy to market. | A general immune-boosting claim is too broad. LL-37 immune biology is mixed, route-specific, and sometimes inflammatory. |
| LL-37 works as a general injectable wellness peptide. | The clearest injectable human source in the cited LL-37 material is a melanoma ClinicalTrials.gov registry, not a wellness or infection trial. The study used intratumoral injections, not routine subcutaneous personal-use injection. | Injecting a host-defense peptide changes the risk question. Local topical wound exposure, intratumoral oncology dosing, and systemic or subcutaneous gray-market use do not have the same exposure profile. | Online peptide discussion often treats injectable LL-37 as an immune or chronic-infection tool. Clinic and forum material does not provide a consistent dose, frequency, or cycle, much less a standard pattern. | The injectable human evidence belongs to the melanoma research context. Clinic and forum claims do not provide a general injection pattern. |
| LL-37 is safe because the body makes it. | Endogenous origin does not settle exogenous safety. The small venous-leg-ulcer trial was reassuring in its own topical setting, but it does not answer long-term systemic safety, injectable product risk, autoimmunity risk, reproductive concern, or cancer-direction questions. | LL-37 can be reparative or inflammatory depending on tissue and immune context. The same molecule has been discussed in antimicrobial defense, wound repair, autoimmunity, and cancer biology. | Vendor and clinic copy often uses natural-host-defense language to make LL-37 sound safer than a standard drug. That shortcut is not justified. | Natural origin does not remove formulation, route, sterility, endotoxin, immune activation, reproductive, or tumor-biology concerns. |
Bottom line
Main takeaway
LL-37 is a real human defense peptide with real wound trials behind it, but the trials are topical, the biggest one missed its endpoint, and the injection-and-immunity uses sold online are untested.
Never merge the products: topical wound gel, intratumoral oncology dosing, oral recombinant COVID product, and gray-market vials are four different situations with four different evidence values. A claim is only as good as the route it came from.
Start with Gronberg 2014 and Mahlapuu 2021 on venous ulcers, the NCT02225366 melanoma registry, and FDA's compounding-risk material. Then read the autoimmunity and cancer-duality papers, which are why "endogenous" does not mean benign.
Identity
What it is
LL-37 is the 37-amino-acid effector peptide of the human cathelicidin system, released from hCAP18 by proteinase 3 and named ropocamptide in regulatory naming materials. It is part of the innate immune toolkit, present in skin, neutrophils, and barrier tissues wherever first-line defense happens.
The direct action is membrane disruption against bacteria, fungi, and some viruses, plus endotoxin binding and anti-biofilm effects in the lab. The indirect action is immune signaling: recruiting immune cells, influencing angiogenesis and epithelial migration, which is the wound-repair rationale that took it into clinical development.
The same versatility is the caution. LL-37 binds self-DNA and self-RNA and can feed the interferon-heavy inflammatory loops implicated in psoriasis and lupus-like disease. The cancer literature runs in both directions: anti-tumor in some models, pro-growth and pro-metastatic in others. FDA's compounding-risk file flags male-reproduction concerns and pro-tumorigenic effects in some tissues. "The body makes it" is a description, not a safety argument.
Human testing happened anyway, mostly as a topical wound drug. The earlier venous-leg-ulcer trial reported healing signals at 0.5 and 1.6 mg/mL twice weekly; the larger phase IIb trial missed its full-cohort endpoint, with a post hoc signal only in ulcers of at least 10 cm2. An intratumoral melanoma dose-finding registry and an oral recombinant COVID trial round out a record that is real, small, and route-locked.
How people talk about it online
The online LL-37 conversation is dominated by chronic-infection territory: Lyme, mold, SIBO, recurring infections, biofilms, with immune-resilience framing throughout. Community protocols describe 50 to 100 mcg subcutaneously once daily for 4 to 6 weeks, conservative numbers relative to the preclinical work, but still unstudied.
The signature leap is from "antimicrobial in a dish" to "treats infections in people." The human trials that exist are topical wound studies, and injecting a peptide that can bind self-nucleic acids and amplify interferon pathways is a different risk proposition than putting it on an ulcer. Immunogenicity is a live concern for an immune-active peptide used this way.
Vendor listings add the second distortion: 5 mg research-use vials with HPLC purity and LC-MS documentation that read as reassurance. Those documents describe a sample's chemistry, not sterile manufacturing, endotoxin control, or any relationship to the material used in trials.
Use context
Routes, doses, and cycle patterns
It helps to separate topical wound research, intratumoral oncology research, oral recombinant infectious-disease research, and online peptide use. The topical wound studies give the most concrete chronic-wound numbers. The melanoma registry gives injectable numbers, but only for direct tumor injection. Clinic, forum, vendor, and online regimen discussion is broader and less consistent.
Human studies and product labels
Earlier venous-leg-ulcer trial
- Purpose
- Hard-to-heal venous leg ulcers
- Context
- Randomized placebo-controlled clinical trial
- Route
- Topical
- Amount
- 0.5 or 1.6 mg/mL
- Frequency
- Twice weekly
- Duration
- 4 weeks in the cited summary
The PubMed abstract reported healing results in the lower-exposure groups and did not report local or systemic safety concerns in that study. This is wound-specific topical research.
Later phase IIb venous-leg-ulcer trial
- Purpose
- Hard-to-heal venous leg ulcers with compression care
- Context
- Multicenter double-blind randomized phase IIb trial
- Route
- Topical
- Amount
- 0.5 or 1.6 mg/mL
- Frequency
- Twice-weekly local applications after placebo run-in
- Duration
- 13 weeks treatment with 16 weeks follow-up
The full cohort did not show a significant benefit. In a post hoc group with ulcers at least 10 cm2, 0.5 mg/mL showed stronger closure outcomes than placebo. That makes it useful wound-research context, but the full trial report would be needed before leaning on the subgroup result.
Diabetic-foot-ulcer trial
- Purpose
- Diabetic foot ulcer with mild infection
- Context
- Randomized double-blind controlled trial
- Route
- Topical cream plus standard wound care
- Amount
- The cited summary does not give a usable amount
- Frequency
- The cited summary does not give a usable frequency
- Duration
- The cited summary does not give a usable duration
Useful as wound-context background. The cited summary does not include enough detail for dose or outcome interpretation.
Melanoma phase I/II registry
- Purpose
- Cutaneous melanoma metastases
- Context
- ClinicalTrials.gov dose-finding registry
- Route
- Intratumoral injection
- Amount
- 250, 500, 1000, or 2000 micrograms per 1 cm3 tumor
- Frequency
- Weekly
- Duration
- Up to 8 weeks
This is the main injectable human study pattern in the cited human LL-37 material. It was designed around biologic dose, toxicity, and interferon response, not around personal injection use.
Oral recombinant COVID-19 trial
- Purpose
- Mild Omicron BA.5.1.3 COVID-19
- Context
- Single-center randomized trial
- Route
- Oral recombinant LL-37 Lactococcus lactis product
- Amount
- The cited summary does not give a usable amount
- Frequency
- The cited summary does not give a usable frequency
- Duration
- The cited summary does not give a usable duration
This only shows an oral recombinant product route. It cannot tell you how a synthetic peptide vial, topical product, nasal spray, or immune-support product would be absorbed, dosed, manufactured, or tolerated.
Real-world discussion
Community antimicrobial-peptide protocols
- Purpose
- Gut, skin, and infection-adjacent discussion
- Context
- Forums, clinics, and vendor listings
- Route
- Subcutaneous injection, with topical products also marketed
- Amount
- Community protocols commonly describe 50 to 100 mcg per day
- Frequency
- Once daily in most descriptions
- Duration
- Most often 4 to 6 week runs
LL-37 is a human antimicrobial peptide with real immunology literature but no established human self-use protocol; community ranges are conservative relative to what preclinical work used, and immunogenicity is a live concern for an immune-active peptide. Informational context, not advice.
What varies
- Route: topical chronic-wound use, intratumoral injection, oral recombinant delivery, nasal-spray discussion, and gray-market injection do not answer the same safety or efficacy question.
- Amount: 0.5 mg/mL topical wound exposure is not comparable to 2000 micrograms per 1 cm3 tumor or an unmapped vendor vial.
- Frequency: twice-weekly wound applications and weekly intratumoral injections came from different research settings; oral recombinant mentions only show that another product route has been studied.
- Product: regulated study material, compounded material, and research-use public-market material differ in identity testing, sterility assurance, endotoxin control, stability, and labeling.
Human data
Human evidence
The human record is small and mostly topical. An earlier randomized venous-leg-ulcer trial reported healing results at 0.5 and 1.6 mg/mL twice weekly over 4 weeks. The larger 148-participant phase IIb ulcer trial found no significant benefit in the full cohort, with a post hoc subgroup signal at 0.5 mg/mL in ulcers of at least 10 cm2. A small diabetic-foot-ulcer trial adds background without usable regimen detail. The intratumoral melanoma registry documents weekly dosing up to 8 weeks with tiny public denominators, and the oral recombinant COVID study concerns a different product. No controlled human outcome data exist for infection, Lyme, immune support, or injectable wellness use.
Evidence maturity
LL-37 has genuine host-defense biology and small human wound trials, but its largest controlled wound trial missed and every marketed claim beyond topical wound research lacks human outcome data.
LL-37 is the human cathelicidin effector peptide with antimicrobial, wound-repair, and immune-signaling roles.
Laboratory studies added anti-biofilm and immunomodulatory findings alongside autoimmunity and cancer-duality concerns.
Randomized topical venous-leg-ulcer studies, a diabetic-foot-ulcer trial, an intratumoral melanoma registry, and an oral recombinant COVID-19 study followed.
The larger phase IIb ulcer trial missed its full-cohort endpoint, and no controlled human outcome data exist for infection, Lyme, or immune claims.
No approved use exists; FDA flags compounded LL-37 safety risks while vendors sell research-use vials for unstudied uses.
| Study / evidence area | Population | Design | Product context | Main outcome | Limitations | Weight |
|---|---|---|---|---|---|---|
| Venous leg ulcer randomized trial | Adults with hard-to-heal chronic venous leg ulcers | Randomized placebo-controlled first-in-human trial | Investigational topical LL-37 | The PubMed abstract reported healing results in lower-exposure topical LL-37 groups after a placebo run-in and randomized treatment phase. | The trial was small and wound-specific. It does not address infection, immune support, cancer, nasal-spray use, or gray-market injection. | limited human |
| Phase IIb venous-leg-ulcer study | Adults with hard-to-heal venous leg ulcers | Multicenter double-blind randomized placebo-controlled phase IIb trial | Investigational topical LL-37 plus compression care | The full cohort did not show a significant benefit. A post hoc subgroup of ulcers at least 10 cm2 had stronger closure outcomes at 0.5 mg/mL. | The subgroup result was post hoc and the trial was not powered for subgroup efficacy testing. | limited human |
| Diabetic foot ulcer study | People with diabetic foot ulcers and mild infection | Randomized double-blind controlled trial | Investigational LL-37 cream plus standard wound care | Provides limited wound-context background, but it does not set dosing. | The exact regimen and outcomes need the full study before making specific claims. | limited human |
| NCT02225366 melanoma registry | People with cutaneous melanoma metastases | Phase I/II dose-finding registry | Investigational intratumoral LL-37 | The registry documents weekly intratumoral LL-37 dosing with small public outcome denominators. It establishes a studied injectable context, not cancer efficacy. | The registry is very small, and posted denominators are too limited for a broad oncology conclusion. | limited human |
| Oral recombinant LL-37 COVID-19 study | People with mild Omicron BA.5.1.3 COVID-19 | Single-center randomized placebo-controlled study | Oral recombinant LL-37 Lactococcus lactis product | Included to distinguish an oral recombinant product from synthetic LL-37 vials and topical wound products. | This was not a synthetic LL-37 vial, topical wound product, nasal spray, or injectable clinic product. | limited human |
Cautions
Safety and unknowns
- Local wound reactions matter. The later venous-leg-ulcer trial described local ulcer redness, edema, raised temperature, and adjacent-skin findings such as redness, scaling, papules, pustules, and vesicles. Most were mild or moderate in that study, but this does not settle other routes.
- Immune activation is a real concern. LL-37 can bind self-DNA or self-RNA and help drive interferon-rich inflammatory pathways linked to psoriasis and lupus-like biology.
- Cancer biology is not one-directional. Preclinical cancer literature describes anti-cancer activity in some models and growth, migration, angiogenesis, or metastasis-promoting findings in others.
- FDA compounding risk material flags LL-37 for possible significant safety concerns, including male-reproduction concern and pro-tumorigenic effects in some tissues.
- Long-term systemic safety, repeat injectable exposure, nasal use, reproductive safety, immunogenicity, and safety in inflammatory or malignant disease were not characterized in the reviewed material.
- For any non-approved peptide supply, contamination, counterfeit material, active-content variability, endotoxin, sterility failure, wrong ingredient, and storage degradation are practical risks, not abstract warnings.
Product quality
A vial label is only a starting point
Public COAs for LL-37 may show HPLC purity, NMR consistency, molecular weight, formula, or an endotoxin line, but those are partial release facts.
A research-use or clinic vial still has to establish formulation, strength, sterility assurance, stability, container-closure control, and chain of custody before it can be compared with investigational trial material.
Identifier inconsistency in public listings matters because salts, counterions, formula statements, and CAS references can be simplified or mismatched in sales copy.
Identity
The vial needs to contain the intended LL-37 material, not a related peptide, wrong salt form, degraded material, or mislabeled ingredient.
Sterility
Any injectable or wound-contact product raises sterile-manufacturing and contamination questions that a purity percentage does not answer.
Endotoxin
LL-37 interacts with microbial products, and endotoxin contamination can create inflammatory risk separate from the peptide itself.
Potency and concentration
Trial concentrations such as 0.5 or 1.6 mg/mL are not meaningful if a vial's actual active content, dilution, or post-mixing concentration is wrong.
Stability and storage
Peptides can degrade during storage, shipping, reconstitution, or repeat handling, which can change potency and impurity exposure.
Mechanism
How it is proposed to work
LL-37 is a positively charged host-defense peptide. It can interact with microbial membranes and microbial products, recruit immune cells, influence blood-vessel and epithelial-repair pathways, and change inflammatory signaling. That broad activity is why it is studied, and why simple "antibiotic," "anti-inflammatory," or "immune booster" claims are misleading.
Antimicrobial and anti-biofilm experiments are mostly preclinical. They help explain interest in infection and biofilm claims, but they do not show human anti-infective effects.
Wound-repair biology includes immune-cell chemotaxis, angiogenesis, and epithelial migration or closure pathways, which fits the topical chronic wound development history.
LL-37 can also bind host nucleic acids and feed endosomal immune sensing. That is one reason psoriasis, lupus-like inflammation, and other immune activation concerns matter.
The cancer data are mixed enough that both "anti-cancer peptide" and "safe because endogenous" are poor summaries.
LL-37 is the 37-amino-acid active form of the human cathelicidin antimicrobial peptide. It disrupts microbial membranes directly and also modulates immune signaling through receptors including FPRL1: a dual role that makes it both an antimicrobial candidate and an immunogenicity concern as an injected product.
FAQ
Common questions
Do broad antimicrobial or immune claims for LL-37 hold up?
No. The human evidence is a narrow chronic venous leg ulcer trial. Mechanistic antimicrobial or immune language does not substitute for direct human evidence in a specific disease setting.
Does LL-37 have human oncology evidence that changes the conclusion?
Only in a very small investigational melanoma registry. The official study record lists 4 actual enrollees, but the posted participant-flow, baseline, outcome, and adverse-event denominators report 3 analyzed participants. That registry is not enough for a cancer-treatment or broader LL-37 benefit claim.
What is the practical takeaway?
The evidence is a narrow wound category plus FDA risk limits. It does not make LL-37 a general recovery, antimicrobial, immune, or oncology peptide.
Details
Technical details
Sources
References
- 1.
FDA. Certain Bulk Drug Substances for Use in Compounding that May Present Significant Safety Risks 2026.
Accessed 2026-06-08.
FDA summarizes potential significant safety risks and missing safety information for several nominated peptide bulk substances.
- 2.
doi:10.1111/wrr.12211 PMID:25041740 Accessed 2026-06-09.
PubMed abstract describes a 34-participant venous leg ulcer trial with placebo run-in, randomized double-blind LL-37 exposure, healing signals in lower-exposure groups, and no local or systemic safety concerns reported in that study.
- 3.
ClinicalTrials.gov. ClinicalTrials.gov record NCT02225366 2026.
NCT02225366 Accessed 2026-06-09.
Completed phase 1/2 melanoma registry. The registry shows 4 actual participants, while posted participant-flow, baseline, outcome, and adverse-event denominators report 3 analyzed participants across two cohorts.
- 4.
FDA. Tailor Made Compounding LLC - 594743 - 04/01/2020 2020.
Accessed 2026-06-09.
FDA warning letter lists LL-37 among compounded bulk substances the agency said did not meet the cited section 503A conditions in the circumstances described.
- 5.
Overhage 2008 biofilm. Overhage et al. 2008 anti-biofilm study against Pseudomonas aeruginosa.
Preclinical anti-biofilm rationale.
- 6.
Scott 2011 and Hu 2014. Scott 2011 and Hu 2014 endotoxin and inflammasome modulation papers.
Endotoxin-neutralization and inflammasome mechanism context.
- 7.
Autoimmunity risk papers. Lande 2008, Ganguly 2009, Kahlenberg 2013, and Morizane 2012.
Autoimmunity and inflammatory-disease risk.
- 8.
Cancer duality papers. Weber 2009, Ji 2019, Zhang 2022 review, and Wang 2016.
Tissue-dependent cancer-biology context.
- 9.
Gronberg 2014 VLU trial. Gronberg et al. 2014 venous-leg-ulcer randomized trial.
Earlier human topical wound signal.
- 10.
Mahlapuu 2021 VLU trial. Mahlapuu et al. 2021 phase IIb venous-leg-ulcer trial.
Best direct human topical wound efficacy and safety data.
- 11.
Miranda 2023 DFU trial. Miranda 2023 diabetic-foot-ulcer trial and registry-linked sources.
Human diabetic-foot-ulcer evidence.
- 12.
Melanoma registry. ClinicalTrials.gov NCT02225366 melanoma registry.
Human oncology dosing and study-design context.
- 13.
Dolkar 2018 toxicity. Dolkar et al. 2018 dermatologic toxicity report.
Toxicity signal in the melanoma program.
- 14.
Zhao 2023 COVID trial. Zhao 2023 oral recombinant LL-37 COVID trial.
Non-wound human evidence that is highly formulation-specific.
- 15.
FDA LL-37 safety risk. FDA compounding safety-risk material referencing LL-37.
Male-reproduction and pro-tumor concern signal.
- 16.
Reddit SIBO and mold discussion. Reddit thread asking about LL-37 for SIBO and mold-associated symptoms.
Direct community evidence for SIBO, mold, stacking, and protocol-seeking claims
- 17.
Reddit Lyme experience discussion. Reddit thread containing firsthand and prospective LL-37 discussion in a Lyme community.
Direct community evidence for Lyme-oriented cycles, co-interventions, and uncertainty
- 18.
Bavaria Peptides LL-37 listing. Vendor listing for a 5 mg LL-37 research-use-only vial with HPLC and LC-MS documentation claims.
Direct vendor evidence for vial sales, RUO positioning, and analytical-document claims