# Deep-dive SOP — Gel-clot LAL endotoxin testing of oil-based HRT injectables via aqueous liquid–liquid extraction

**Document type:** validated-style SOP for a non-compendial matrix workaround.
**Scope:** oil-vehicle injectable drug products (testosterone esters, estradiol valerate/cypionate/benzoate, progesterone, nandrolone decanoate, cyproterone acetate depot) tested by **gel-clot LAL** after **1:1 aqueous liquid–liquid extraction (LLE)**.
**Out of scope:** aqueous injectables (peptide reconstitutions, GnRH analogues in saline); these are tested neat-then-diluted per the round-1 workflow in `../03-microbial-pyrogen/home-lab-endotoxin-feasibility.md`.
**Last updated:** 2026-04-23.
**Compiled by an AI agent** using the structured claim format in `/workspace/overview/RESEARCH-BEST-PRACTICES.md`. Numeric claims carry C1–C5 confidence tiers; unflagged prose is inference and should be treated as C3 or worse.

---

## Key takeaways

- **LLE + gel-clot is a reasonable chemistry extrapolation — not a validated compendial method.** The cited source (Lester, Ramsey, Salsbury, *American Pharmaceutical Review* Nov/Dec 2019, PPD Labs) validated LLE on fractionated coconut oil using **kinetic turbidimetric** LAL, not gel-clot. You must revalidate your specific oil matrix with spiked CSE recovery before treating any pass result as trustworthy.
- **Every new oil matrix needs its own suitability study.** MCT, grapeseed, castor, cottonseed, sesame, and benzyl-benzoate/benzyl-alcohol co-solvent blends all partition differently. A one-time "I validated MCT" does not cover grapeseed estradiol valerate.
- **Vortex time is the extraction variable that actually matters.** The PPD paper (Lester et al. 2019, Table 1 feasibility study) shows **extraction efficiency** of 0%, 55%, 37%, and 69% at 5, 15, 30, and 50 min of high-speed vortexing; only 50 min is reliable. Their subsequent 1:100 method-verification (Table 2) showed **PPC recovery** of 100–138% (drug product) / 101–127% (placebo) — those are two different quantities (efficiency vs recovery), and the 69% / 138% figures should not be merged. A 2 min vortex — the round-1 shortcut — is insufficient for quantitative recovery. (Round-2 correction R2-M12.)
- **Dilution is your friend, up to MVD.** For HRT at typical concentrations, the MVD at λ=0.25 EU/mL is 400× to 5,600×. You need most of that headroom to dilute past both oil carryover and matrix interference.
- **Depyrogenation is 250 °C / 30 min (primary) or 200 °C / 60 min (Ph. Eur. alternative).** Not 180 °C. Not 4 h. Round-1 got this wrong in one place; the round-2 correction is canonical.

Cross-links: `../03-microbial-pyrogen/home-lab-endotoxin-feasibility.md`, `../03-microbial-pyrogen/endotoxin-lal.md`, `../03-microbial-pyrogen/endotoxin-limits-reference.md`, `../03-microbial-pyrogen/depyrogenation.md`, `../09-safety-waste/horseshoe-crab-ethics.md`.

---

## 1. Why LLE is required for oils

The LAL enzymatic cascade evolved in horseshoe crab hemolymph; it functions only in an aqueous environment. When the test article is an oil (MCT, grapeseed, castor, cottonseed, sesame, or a benzyl-benzoate/benzyl-alcohol blend), three distinct problems collide:

1. **Physical exclusion.** Lipopolysaccharide (LPS) forms micelles in both aqueous and organic solvents, but its polysaccharide chains make it more soluble in water. In a neat oil sample, endotoxin is sequestered at inverse micelle surfaces where the LAL enzymes — which are water-soluble — simply cannot reach it. USP <85> and Ph. Eur. 2.6.14 both treat this as "physical interference."
2. **Readout interference.** An oil droplet in the reaction tube scatters light (bad for turbidimetric and chromogenic kinetic readouts) and, worse for gel-clot, physically interrupts the gel network so that a properly-formed clot may slide on inversion even when endotoxin was present. You get false negatives.
3. **Sticky chemistry downstream.** The lipid-A anchor of LPS has high affinity for non-polar plastic pipette tips (low-retention polypropylene included). Once endotoxin is in an oil matrix, recovery through pipetting alone is poor.

The standard industry response is **liquid–liquid extraction**: partition endotoxin out of the oil and into an aqueous phase, discard the oil, and run LAL on the aqueous extract. The chemistry is driven by LPS's amphipathic nature — the polysaccharide head is hydrophilic and prefers water; the lipid-A tail anchors to whatever lipid is around but will migrate to the aqueous phase if given sufficient agitation, time, and a phase boundary to cross.

**Literature basis:**

**Claim:** Endotoxins can be recovered from fractionated coconut oil (MCT) by 1:10 LLE with LAL Reagent Water, 50 min vortex, 2,900 × g / 10 min centrifugation, with PPC recovery 100–138% at 1:100 dilution.
**Confidence:** C2.
**Source:** Lester CW, Ramsey JT, Salsbury JS. "Bacterial Endotoxins Testing in Lipid-Based Drug Formulations Using Liquid-Liquid Extraction." *American Pharmaceutical Review*, November/December 2019, pp. 52–55. https://www.ppd.com/wp-content/uploads/2020/02/11Nov19-APR-PPD-Labs-bacterial-endotoxins.pdf
**Date checked:** 2026-04-23.
**Notes:** Format used was **USP <85> kinetic turbidimetric**, not gel-clot. CSE was reconstituted in DMSO (not water) to enable spiking into oil. Vortex times of 5, 15, and 30 min gave 0%, 55%, and 37% extraction efficiency respectively — only 50 min was reliable. Drug product matrix was "a sterile solution of active drug substance in triglycerides" (Phase 1 clinical).

**Claim:** LPS spiked into ultrapure paraffin oil extracts into an equal volume of water with recoveries of 94.2–111%.
**Confidence:** C3.
**Source:** Chen D. "A New Method for the Analysis of Bacterial Endotoxins in Ultrapure Paraffin Oil." *Journal of Analytical Methods in Chemistry*, 2014, Article ID 575246. (Cited as reference 2 by Lester et al. 2019.)
**Date checked:** 2026-04-23 (via citation; primary not refetched in this round).
**Notes:** Paraffin oil is not a pharma vehicle. Cited for the general chemistry only, not as a matrix-specific result.

**Explicit limitation.** Both published studies used kinetic formats (turbidimetric or chromogenic) with automated absorbance readout. Gel-clot LAL reads a yes/no endpoint at a single sensitivity threshold (λ), so partial recovery that a kinetic format would flag as "57% efficiency" reads on gel-clot as an indistinguishable positive (if diluted extract still exceeds λ) or a false negative (if it does not). **A home lab running gel-clot after LLE is extrapolating kinetic-validated chemistry into a qualitative readout — the chemistry carries; the quantitative behaviour does not.** Every new matrix must be revalidated by a Positive Product Control (PPC) at 2λ performed on the specific oil after extraction. This is stated again throughout this document because it is the single most-likely error mode.

---

## 2. Per-matrix interference behaviour

The oil vehicle is the dominant variable in extraction quality. The table below summarizes expected behaviour for the oil matrices that show up in HRT supply chains. None of these are published gel-clot validations; they are chemistry-based expectations plus notes cross-referenced against the Lester et al. 2019 matrices and adjacent literature.

| Matrix | Typical HRT use | Viscosity @ 20 °C | LLE behaviour | Recommended centrifugation | Validation notes |
|---|---|---|---|---|---|
| **MCT oil (fractionated coconut / caprylic-capric triglyceride)** | Testosterone esters, some estradiol cypionate compounded products | Low (~25 cP) | Phase separation fast; clean boundary; minimal emulsion after centrifugation | 2,900 × g × 10 min (Lester 2019); 3,000 × g × 5 min adequate for home setups | **Closest match to published PPD Labs 2019 validation.** Still revalidate: Lester used fractionated coconut oil specifically; other MCT blends (Miglyol 812, neat C8/C10 triglyceride) may vary. |
| **Grapeseed oil** | Estradiol valerate (most UGL and some pharma), nandrolone decanoate | Medium (~50–60 cP) | Phase boundary less sharp than MCT; emulsion band often persists after short spins | 3,000–5,000 × g × 10 min; re-centrifuge if cloudy aqueous layer | No published BET validation. Expect longer centrifugation and possibly 2 × extraction steps. Spike recovery at 2λ is mandatory. |
| **Castor oil** | Estradiol cypionate (some depot products), historical testosterone undecanoate intramuscular formulations | High (~650 cP at 25 °C — ~25× MCT) | Very slow phase separation; stable emulsion is common; ricinoleic acid hydroxyl groups may bind LPS and reduce partition | 5,000 × g × 15 min; consider warming sample to 37 °C before extraction to reduce viscosity | **Castor is the worst-case matrix.** Expect low recovery on first pass. Validate with triplicate 2λ spike; if recovery <50%, consider 2-step extraction or switch to kinetic chromogenic with heptane addition as a modifier. |
| **Cottonseed oil** | Depo-Testosterone (Pfizer, vehicle), older depot testosterone products | Medium (~70 cP) | Phase separation moderate; historical gossypol residues in less-refined cottonseed may bind LPS or enhance LAL background | 3,000–5,000 × g × 10 min | Pharma-grade cottonseed is refined to remove gossypol; UGL cottonseed may not be. If the vial is from a pharma source (Pfizer Depo-T), expect behaviour close to MCT. UGL cottonseed — validate aggressively. |
| **Sesame oil** | Older progesterone and testosterone depots; Delatestryl (testosterone enanthate) | Medium (~55 cP) | Moderate phase separation; lignan content (sesamin, sesamolin) is pharmacologically active but not known to interfere with LAL | 3,000 × g × 10 min | Less common in contemporary HRT. Validate normally. |
| **Benzyl benzoate / benzyl alcohol co-solvents** | Ubiquitous in oil HRT — benzyl alcohol at 1–3% (preservative), benzyl benzoate at 5–20% (solubilizer) | — (co-solvents, not neat vehicles) | Partially water-miscible; partitions into aqueous phase during LLE; can carry into the tested extract | Not a centrifugation issue; it is an extract-composition issue | **Known LAL interferents at concentration.** Benzyl alcohol >0.5% in the aqueous extract can inhibit the LAL cascade; benzyl benzoate >2% similarly. Dilution past their interference threshold (typically 1:100 from the 1:1 extract → effectively 1:200 from the neat oil) is usually sufficient. PPC is mandatory to detect. |
| **Ethyl oleate** | Occasional UGL vehicle, some European testosterone products | Low (~5 cP) | Clean separation; behaves more like MCT than like castor | 3,000 × g × 5 min | Rarely encountered. Validate normally. |

**The Lester et al. 2019 matrix list as validated:** fractionated coconut oil (mock) and a "sterile solution of active drug substance in triglycerides" (Phase 1 clinical product, specific oil unnamed beyond "triglyceride vehicle"). **No other oil has a published BET LLE validation that this SOP author could verify.** Treat every non-MCT oil as a revalidation exercise.

**Per-matrix revalidation protocol — required before first real-sample run on any new matrix:**

1. Obtain a "blank" oil — ideally the same grade and supplier as the samples you will test (e.g., USP/NF grade MCT from Spectrum; USP grade grapeseed from Jedwards; pharma-grade cottonseed is hard to buy retail — accept UGL cottonseed as a proxy only with documentation).
2. Reconstitute CSE in DMSO (see §5) to ~100 ng/mL; spike 50 µL into 4.95 mL of the blank oil for ~1 ng/mL CSE in oil (Lester et al. method).
3. Run the LLE protocol of §3 on the spiked oil and on an unspiked blank oil.
4. Run the resulting aqueous extracts through gel-clot LAL at the intended sample dilution and at 2× that dilution.
5. **Acceptance:** the spiked oil extract must clot at the 2λ equivalent; the unspiked blank extract must not clot. If both clot, your oil is endotoxin-contaminated at the source (and every sample result on this matrix is uninterpretable until you source cleaner blank oil for the positive product control). If only the unspiked blank clots, your LRW or tubes are contaminated.
6. Document lot number, date, recovery outcome. Re-validate whenever you change oil supplier or LAL lot.

---

## 3. Step-by-step LLE protocol

### 3.1 Equipment required

| Item | Specification | Suggested source | Notes |
|---|---|---|---|
| Depyrogenated borosilicate reaction tubes, gel-clot | 10 × 75 mm, soda lime glass, certified endotoxin-free | ACC TS050-10 (52 tubes/pack × 10 packs) | **Not TB240** — TB240 is a 12 × 75 mm "dilutions only" tube per ACC. For the reaction, use **TS050** (soda lime, 10 × 75 mm) or **TK100-10** (borosilicate, 8 × 75 mm — note narrower geometry; verify fit for any tube-reader that requires 10–13 mm OD). Verify current catalog: https://www.acciusa.com/products-and-services/bet-products/accessory-products/pyrotubes |
| Depyrogenated dilution tubes | 12 × 75 mm borosilicate | ACC TB240-5 | For serial dilutions only. |
| Pyrogen-free centrifuge tubes, 15 mL | Polypropylene, certified non-cytotoxic and pyrogen-free (≤0.05 EU/mL or better) | Corning 430052 or equivalent, lot-certified | Lester et al. used 15-mL pyrogen-free centrifuge tubes. |
| Pyrogen-free 2 mL microcentrifuge tubes | Polypropylene, low-retention, pyrogen-free lot-certified | Axygen MCT-200-L-C or Eppendorf Biopur | For dilution series. |
| Pipettes + pyrogen-free tips | Adjustable 10–100 µL, 100–1000 µL; tips certified pyrogen-free and DNase/RNase-free | Rainin, Sartorius, or ACC pipette tips | Lipid A binds plastic — use the tips the vendor lot-certifies as pyrogen-free. Do NOT autoclave for pyrogen removal (121 °C does not destroy endotoxin). |
| Vortex mixer, variable speed | Must reach "high speed" and sustain it for 50+ min without auto-shutoff | VWR Analog Vortex Mixer 58816-121 or Fisher Digital Vortex | Many bench vortexers duty-cycle; confirm yours does not throttle under load. |
| Centrifuge | Bench, capable of ≥ 3,000 × g on a 15 mL tube rotor; ideally 5,000 × g for viscous oils | Eppendorf 5702, Thermo Sorvall Legend Micro 21 (2 mL), or used Beckman Allegra 6 | A $150–$400 used Eppendorf 5702 does the job for MCT and grapeseed. Castor may need a higher-g unit. |
| Dry heat block | 37 °C ± 1 °C, holds 10 × 75 mm tubes | Benchmark MyBlock (BSH1002) with 10×75 mm insert; or used Fisher Isotemp | **Use dry block, not water bath.** Water baths vibrate and cause clot false negatives (round-1 note M11 on failure modes). |
| Stopwatch / lab timer | Accurate to ±1 s over 60 min | — | Critical for 60 ± 2 min incubation. |
| Depyrogenation oven (optional) | 250 °C-capable, documented chamber uniformity | Used Despatch LCC, Memmert UFE, or Quincy Lab | Not required if you use vendor-certified depyrogenated disposables (ACC Pyrotubes + pyrogen-free disposable pipette tips). |
| PPE | Nitrile gloves (change frequently — they shed endotoxin once contaminated), lab coat, safety glasses | Standard lab supply | Nitrile is lower endotoxin than latex. Change gloves between samples. |

### 3.2 Reagents

Quantities shown for a single assay run (1 sample at 1 dilution, plus the mandatory NC, PC, and PPC controls — 4 reaction tubes + 2 extract handling tubes = ~$10–20 in consumables per run after fixed costs).

| Reagent | Use | Verified price (2026-04-22) | Catalog |
|---|---|---|---|
| Pyrotell lysate, 0.25 EU/mL, 5 × 5 mL pack (≈ 250 reactions) | LAL reagent | $595.35 | ACC G5250-5 / Fisher NC1241126. https://www.fishersci.com/shop/products/pyrotell-0-250-eu-ml-5ml-vial/NC1241126 |
| Control Standard Endotoxin (CSE), *E. coli* O113:H10, 0.5 µg/vial, 5 vials | Positive control & PPC spike | $251.13 for 6-pack (Fisher NC9263690) or single-vial pack via ACC | ACC E0005-5 (5-pack) or Fisher NC9263690 (Charles River-branded equivalent). https://www.fishersci.com/shop/products/control-std-endotox-500ng/NC9263690 |
| LAL Reagent Water (LRW), 30 × 50 mL case | Negative control, reconstitution, dilutions | $381.30 | ACC WP050C / Fisher. Verify current catalog: https://www.acciusa.com/products-and-services/bet-products/accessory-products/lal-reagent-water |
| — Alternative smaller size: 5.5 mL × 10-pack | LRW for low-volume programs | $183.27 (C3, Fisher-distributor) | **ACC W0051-10** (round-2 correction R2-M6; earlier draft cited "W055A" which is not a real ACC catalog code). |
| Depyrogenated reaction tubes, 10 × 75 mm, 520/pack (52 × 10) | Reaction vessels | Verify current catalog: https://www.acciusa.com/products-and-services/bet-products/accessory-products/pyrotubes (TS050-10). Round-1 cited ~$200–$400/1000 at ACC; confirm live. | ACC TS050-10 |
| DMSO, anhydrous, pharma grade or HPLC grade | CSE reconstitution for oil spiking (§5 validation only — not for routine samples) | $20–$40 per 100 mL | Sigma-Aldrich 276855 or equivalent; use only for CSE-in-oil spike experiments, not in assay reaction tubes. |

**Per-test reagent cost after fixed kit purchase:** ~$2.40 for Pyrotell lysate + ~$0.30 for LRW + ~$0.50 for tubes ≈ **$3.20/reaction**. A single complete assay run (NC + PC + PPC + sample, each in duplicate = 8 tubes) is ~$26 in reagents plus oil LLE consumables.

**Claim:** Pyrotell 0.25 EU/mL, 5 × 5 mL pack list price $595.35 via Fisher.
**Confidence:** C1.
**Source:** https://www.fishersci.com/shop/products/pyrotell-0-250-eu-ml-5ml-vial/NC1241126
**Date checked:** 2026-04-23.
**Notes:** ACC vendor part G52505; Fisher re-numbers to NC1241126. Direct ACC account pricing may differ.

**Claim:** ACC CSE catalog numbers include E0005-5 (0.5 µg × 5 vials, *E. coli* O113:H10); Fisher NC9263690 (500 ng × 6 vials) at $251.13.
**Confidence:** C2.
**Source:** https://www.acciusa.com/products-and-services/bet-products/lal-reagents/control-standard-endotoxin ; https://www.fishersci.com/shop/products/control-std-endotox-500ng/NC9263690
**Date checked:** 2026-04-23.
**Notes:** The Fisher part is Charles River-branded E. coli O113:H10 CSE at the same strain spec; either is compendially acceptable. Reconstituted CSE stability per ACC: **4 weeks at 2–8 °C for the 0.5 µg/vial size** (10 ng/vial is 7 days; 125 µg/vial is 3 months).

### 3.3 The numbered LLE protocol

> **Before you start:** confirm your reaction tubes, pipette tips, and centrifuge tubes are all vendor-certified pyrogen-free. Review §9 on depyrogenation. If you are reusing glass, it must have been baked at 250 °C / 30 min (or 200 °C / 60 min) within the last 6 months, foil-wrapped since.

1. **Label tubes.** For each sample, prepare:
   - One 15 mL pyrogen-free centrifuge tube marked "S-{SampleID}-LLE"
   - One pyrogen-free 2 mL microcentrifuge tube marked "S-{SampleID}-aqueous"
   - One dilution tube marked "S-{SampleID}-1:{MVD/4 dilution}" (see §4 for MVD math)
   - Four 10 × 75 mm reaction tubes: NC, PC, PPC, Sample (duplicate each in practice = 8 tubes for a duplicate run)

2. **Equilibrate the oil sample to room temperature** (20–25 °C). For castor-oil vehicles only, warm the sample to 37 °C for 10 min in the dry block to drop viscosity before extraction.

3. **Add oil to the extraction tube.** Pipette **0.5 mL of the oil sample** into the 15 mL centrifuge tube. Use a positive-displacement pipette or a glass volumetric pipette; standard air-displacement pipettes under-deliver viscous oils by 5–20%. If using an air-displacement pipette, pre-wet the tip and dispense slowly.

4. **Add LAL Reagent Water.** Pipette **0.5 mL of LRW** into the same tube. (This is the 1:1 LLE format — Lester et al. used 1:10, but for gel-clot qualitative readout, 1:1 conserves endotoxin in the aqueous phase at higher concentration, which aids detection at moderate sensitivities. If your intended test dilution is ≥1:100 anyway, use 1:10 LLE per Lester — 0.5 mL oil + 4.5 mL LRW — for the cleanest phase separation.) Match your LLE ratio to your validation; do not mix 1:1 with 1:10 across samples in the same run.

5. **Cap and vortex at high speed for 50 minutes continuously.** This is the single most important parameter.
   - Lester et al. Table 1 (feasibility CSE-in-oil extraction efficiency): 5 min → 0%; 15 min → 55%; 30 min → 37%; 50 min → 69%.
   - Table 2 validation (different metric) at 1:100 dilution: PPC recovery 100–138% drug product / 101–127% placebo. (Round-2 fix R2-M12: the earlier "69–138%" fused two different quantities.)
   - A 2 minute vortex is NOT sufficient. The round-1 shortcut ("vortex 60 s") is under-validated and will produce false negatives.
   - If your vortexer auto-stops before 50 min, tape the trigger or use a vortexer with a timed run mode. Inversion mixing is acceptable only as a supplement — not a substitute.
   - **Do not add glass beads** (Lester et al. found they produced a white precipitate from friction against the tube wall).

6. **Centrifuge to break the emulsion.**
   - **MCT, grapeseed, sesame, ethyl oleate:** 3,000 × g × 5 min (home setup) or 2,900 × g × 10 min (published PPD conditions).
   - **Cottonseed:** 3,000–5,000 × g × 10 min.
   - **Castor:** 5,000 × g × 15 min; inspect; re-centrifuge if aqueous layer is not clear.
   - **Do not use gravity settle overnight as a substitute unless you have validated it.** Overnight settle allows CSE to re-partition, and the published work shows "CSE settles out of solution over time, so the lipid and aqueous layers must be separated after extraction so that CSE can be resuspended" (Lester et al. 2019, p. 55).

7. **Visual inspection.** Aqueous phase should be clear, below the oil. An emulsion band at the interface indicates insufficient centrifugation — re-spin. A cloudy aqueous phase indicates carried-over oil droplets — re-spin.

8. **Aspirate the aqueous phase.**
   - Use a fresh pyrogen-free pipette tip.
   - Insert the tip past the oil layer to the bottom of the tube (aqueous is denser than oil; it sits below).
   - **Discard the first 50 µL** as an "oil-wetting" blank to clear any oil in the pipette lumen.
   - Transfer ~300–400 µL (from a 0.5 mL aqueous phase) into the 2 mL microcentrifuge tube labeled "-aqueous". Do NOT aspirate more than 80% of the aqueous phase — the last 20% almost always carries oil droplets.
   - If doing 1:10 LLE (4.5 mL LRW): transfer 200 µL of aqueous into 1.8 mL LRW in a fresh tube to achieve the 1:100 overall dilution Lester et al. validated.

9. **Re-vortex the aqueous extract for 30 seconds** to re-homogenize any CSE or native endotoxin that settled during aspiration. Lester et al. note this explicitly: "the aqueous layer is then vortexed to re-homogenize the CSE."

10. **Proceed to §4 for dilution to MVD.**

---

## 4. Maximum Valid Dilution — worked examples

The MVD defines the upper bound on dilution; beyond MVD, even a fully-contaminated product could falsely pass. Formula:

```
Endotoxin Limit (EU/mg) = K / M
Endotoxin Limit (EU/mL) = Endotoxin Limit (EU/mg) × product concentration (mg/mL)
MVD = Endotoxin Limit (EU/mL) / λ
```

Where **K = 5 EU/kg** for parenteral non-intrathecal routes (USP <85> / Ph. Eur. 2.6.14), **M = maximum recommended human dose per kg per hour**, and **λ = label-claim sensitivity of the LAL reagent** (0.25 EU/mL for Pyrotell G5250). For home-lab HRT work, use the 70 kg adult M-basis unless dosing for specific body weight: allowable load per dose = 5 × 70 = **350 EU**.

### 4.1 Worked example — Testosterone Cypionate 200 mg/mL

- Concentration: 200 mg/mL
- Typical max dose: 1 mL = 200 mg weekly (single hour bolus basis)
- Allowable load: 350 EU
- **Endotoxin Limit (EU/mg) = 350 / 200 = 1.75 EU/mg**
- **Endotoxin Limit (EU/mL) in the neat oil = 1.75 × 200 = 350 EU/mL**
- At λ = 0.25 EU/mL: **MVD = 350 / 0.25 = 1,400×**

**How the 1:1 LLE consumes part of the MVD:**
- 1:1 LLE effectively dilutes neat oil endotoxin 2-fold into the aqueous phase (**if** extraction is 100% efficient — which no one achieves; real recovery is 50–140% per Lester).
- Remaining dilution budget: 1400 / 2 = **700× further dilution allowed** from the aqueous phase.
- A practical dilution scheme: 50 µL aqueous extract into 450 µL LRW (1:10 from extract, 1:20 from neat oil). Still within MVD by 70×. This is the tube you test.
- Run PPC in parallel at the same dilution.

**How the 1:10 LLE (Lester et al. recipe) consumes MVD:**
- 1:10 LLE dilutes neat oil 10× if extraction efficiency = 100%.
- Remaining budget: 1400 / 10 = 140× further dilution allowed.
- Lester et al. took 200 µL into 1.8 mL → further 1:10 → cumulative 1:100 from neat oil. Still 14× under MVD. Tested dilution = 1:100.

### 4.2 Worked example — Estradiol Valerate 40 mg/mL

- Concentration: 40 mg/mL
- Typical max dose: 10 mg = 0.25 mL of 40 mg/mL, or 40 mg for pushing depot schedules; use the higher-exposure case — the monthly-dose argument is complex because USP uses single-hour max. For weekly dosing at 10 mg:
- Allowable load: 350 EU
- Endotoxin Limit = 350 / 10 = **35 EU/mg**
- Endotoxin Limit at 40 mg/mL = 35 × 40 = **1,400 EU/mL** in the neat vial
- MVD at λ = 0.25: **1400 / 0.25 = 5,600×**

For a monthly-dose basis (40 mg single shot) the calculation tightens:
- Endotoxin Limit = 350 / 40 = 8.75 EU/mg
- At 40 mg/mL = 350 EU/mL in the neat oil
- MVD at λ = 0.25 = **1,400×**

Use the more conservative (tighter) of the two — the monthly-dose basis with MVD = 1,400× — as your planning number. Your tested dilution should be well inside that.

### 4.3 Practical dilution schemes by matrix

| Matrix | Neat concentration | Conservative MVD (λ=0.25) | Recommended test dilution after 1:1 LLE | Headroom factor |
|---|---|---|---|---|
| Testosterone cypionate | 200 mg/mL | 1,400× | 1:200 (i.e., 1:1 LLE + 1:100 from extract) | 7× under MVD |
| Testosterone enanthate | 250 mg/mL | 1,400× | 1:200 | 7× under MVD |
| Estradiol valerate | 40 mg/mL | 1,400× (monthly) / 5,600× (weekly) | 1:200 | 7× / 28× under MVD |
| Estradiol cypionate | 5 mg/mL | 1,400× | 1:100 | 14× under MVD |
| Progesterone in oil | 50 mg/mL | 700× | 1:100 | 7× under MVD |
| Nandrolone decanoate | 200 mg/mL | 1,400× | 1:200 | 7× under MVD |
| Cyproterone acetate depot | 100 mg/mL | 467× | 1:100 | 4.7× under MVD |

Cross-reference: `../03-microbial-pyrogen/endotoxin-limits-reference.md` for the full cheat sheet including peptides.

Your tested dilution must be **≤ MVD**. Being close to MVD is legitimate if you have confirmed PPC passes at that dilution. Being 10× under MVD is more robust.

---

## 5. Positive Product Control (PPC)

**The PPC is mandatory every run.** No PPC = no valid result. This is where matrix interference is detected.

### 5.1 Preparation of CSE for PPC spike

1. Rehydrate a 0.5 µg vial of CSE with LRW per vendor insert (typically 5 mL LRW → 100 ng/mL; vendor-specific potency may shift this — read the Certificate of Analysis for the lysate–CSE lot pairing).
2. **Vortex ≥ 15 minutes** after rehydration to fully disaggregate LPS micelles. Lester et al. and ACC both emphasize this; under-vortexed CSE is the most common "my PC isn't working" root cause.
3. Aliquot into pyrogen-free microcentrifuge tubes (20 µL aliquots are common) and store at **2–8 °C for up to 4 weeks** (for the 0.5 µg/vial size — shorter for the 10 ng/vial size). Per ACC, do not freeze the 0.5 µg size. Track the date opened on a lab-notebook log.
4. Before each use, re-vortex the working aliquot for 1 min.

### 5.2 Running the PPC

Spike the PPC to a final concentration of **2λ = 0.5 EU/mL** (for λ = 0.25 lysate). Practical protocol:

- Working CSE stock at 50 EU/mL (confirmed against lysate lot CoA): add **10 µL stock** to **990 µL of diluted sample** (so that after mixing, spike contribution = 0.5 EU/mL).
- Take 100 µL of this spiked diluted sample + 100 µL lysate into the "PPC" reaction tube.
- Incubate alongside the sample.

**Pass criterion:** PPC must form a firm clot that survives 180° inversion. Same timing as the sample.

**Fail interpretation:**
- **PPC does not clot (and NC is clean, PC clots):** the matrix is inhibiting the LAL cascade at the tested dilution. Dilute the sample further (halving is typical) up to MVD and repeat. If still failing at MVD, switch methods (kinetic chromogenic after LLE is more tolerant) or use rFC (Factor C is less matrix-sensitive than the native cascade).
- **PPC clots but recovers weakly / unusual kinetics — not applicable to gel-clot readout.** (This is where kinetic formats add value.)

### 5.3 Extraction-verified PPC (for validation runs)

For the **per-matrix revalidation** described in §2, the PPC must be performed as a **pre-extraction spike**:

1. Reconstitute CSE in **DMSO** (not LRW) per Lester et al. / Schlösser — 5 mL anhydrous DMSO into 0.5 µg vial → 100 ng/mL in DMSO.
2. Spike 50 µL of DMSO-CSE into 4.95 mL of blank oil → ~1 ng/mL CSE in oil. Vortex 30 s to homogenize (oil and DMSO are miscible).
3. Run the full LLE workflow (§3).
4. Test the final diluted extract. This tube must clot at 2λ for the extraction to be validated for that matrix.
5. This is **not** a routine run PPC — it's a one-time (per matrix, per reagent lot) validation. Routine runs use the simpler LRW-reconstituted CSE spike of §5.2, added to the **aqueous extract post-extraction**.

---

## 6. Negative Product Control (NPC) and Negative Control (NC)

Two different controls, both required:

- **NC (Negative Control):** LRW only, 100 µL + 100 µL lysate. Must not clot. If it clots, your LRW, lysate, or tubes are endotoxin-contaminated — abort the run and investigate.
- **NPC (Negative Product Control / unspiked sample at test dilution):** your diluted sample extract, no CSE spike added, 100 µL + 100 µL lysate. Tests whether the sample itself exceeds λ.

Historical USP terminology sometimes collapses these into "Negative Control" broadly; modern practice (per USP <85> and ACC's protocols) separates LRW-only NC from the unspiked sample tube. Both must not clot for a "clean product" result.

**If NC clots:** water or lysate contaminated. Common causes:
- LRW left uncapped on bench (airborne contamination)
- LRW past its vendor-stated in-use window (typically 30–90 days once opened)
- Reaction tubes not depyrogenated (see §9)
- Lysate reconstituted with non-LRW water
- Cross-contamination from pipette tip reuse

**If NPC clots at MVD but PPC passes:** the sample is endotoxin-positive at ≥λ at that dilution. Back-calculate: endotoxin concentration in neat product ≥ (λ × cumulative dilution factor). For testosterone cypionate at 1:200 dilution with λ = 0.25, that's ≥ 50 EU/mL in the neat oil — which is 14% of the endotoxin limit (350 EU/mL) and still compliant. Run a higher dilution to find the threshold.

---

## 7. Incubation and readout

1. **Load tubes into the 37 °C dry heat block** in order: NC, PC, PPC, Sample (and duplicates). Stagger loading by 15–30 seconds if your pipetting is slow; record the start time for each tube or use a single timer from the last loaded.
2. **Incubate at 37 °C ± 1 °C for 60 minutes ± 2 minutes.**
   - Use a calibrated dry block. Cheap blocks drift; verify with a thermometer initially.
   - **Do not use a circulating water bath.** Vibration from the pump prematurely breaks the developing gel and causes false negatives. This is a known, documented failure mode.
   - Do not bump the block. Do not open it mid-incubation for long enough to drop the block temperature.
3. **Readout — the inversion test.**
   - At T = 60 min ± 2 min exactly, remove the tube one at a time.
   - Invert the tube through **180° in a single smooth motion**, completing in approximately 1 second.
   - Hold inverted for 2–3 seconds; observe whether the gel holds.
   - **Pass (negative for endotoxin at ≥λ):** solid gel that holds on inversion. No leakage, no slide.
   - **Fail (positive for endotoxin at ≥λ):** gel slides, breaks, drops, or forms a soft curd that does not hold.
   - Record result immediately. Photograph if possible (phone camera is fine; include sample ID in frame).
4. **Do not read before 60 min.** Gel formation in LAL typically develops in the last 10–20 min of the reaction; early reading is a classic false-negative source. The PPC is your confirmation that 60 min at 37 °C is sufficient for the reagent lot.
5. **Do not read after ~65 min** with the tube removed from the block. The gel can continue to firm up out of block, but prolonged ambient cooling changes the mechanical properties and can mask a marginal reaction.

---

## 8. Interpretation and reporting template

### 8.1 Result matrix

For each sample, four tubes minimum per dilution:

| Tube | Composition | Expected result | Interpretation if unexpected |
|---|---|---|---|
| NC | 100 µL LRW + 100 µL lysate | No clot | LRW / lysate / tube contamination. Void run. |
| PC | 100 µL 2λ CSE (in LRW) + 100 µL lysate | Clot | Lysate has degraded or was not reconstituted properly. Void run. |
| NPC | 100 µL diluted sample + 100 µL lysate | No clot (for a compliant product) | Clot = endotoxin present at ≥ λ at this dilution. Report as positive at that concentration back-calculated. |
| PPC | 100 µL diluted sample spiked to 2λ + 100 µL lysate | Clot | No clot = matrix inhibition. Dilute further, re-run. |

Run all four in duplicate. A run is valid only if NC and PC are unanimous (both NC tubes negative, both PC tubes positive) and both PPC tubes positive.

### 8.2 Lab notebook entry template

```
Sample ID:        __________________________
Date / time:      __________________________
Operator:         __________________________
Matrix (oil):     ☐ MCT  ☐ Grapeseed  ☐ Castor  ☐ Cottonseed  ☐ Sesame  ☐ Other: ____
Drug:             ____________ at ______ mg/mL (label claim / verify)
Endotoxin Limit:  ______ EU/mL (from §4, per K/M)
MVD:              ______× at λ = ______ EU/mL

Reagent lots:
  Lysate (Pyrotell G5250, cat #):    lot _______  exp _______
  CSE (ACC E0005 or CR equiv):       lot _______  exp _______  date reconstituted _______
  LRW (ACC WP050C):                  lot _______  in-use since _______
  Reaction tubes (ACC TS050-10):     lot _______

LLE parameters:
  Oil volume:     _____ mL
  LRW volume:     _____ mL    (ratio _____)
  Vortex time:    _____ min   (must be ≥ 50 min for validated recovery)
  Centrifuge:     _____ × g × _____ min
  Aqueous vol aspirated: _____ µL (discard first 50 µL)

Dilution scheme (math shown):
  Cumulative dilution from neat oil to reaction tube = ______× (must be ≤ MVD)

Incubation:
  Start: ______ °C, start time ______
  End:   ______ time
  Duration: ______ min (must be 60 ± 2)

Results:
  NC (LRW):        ☐ no clot  ☐ clot   (expect no clot)
  PC (2λ CSE):     ☐ clot     ☐ no clot (expect clot)
  NPC (sample):    ☐ no clot  ☐ clot
  PPC (sample+2λ): ☐ clot     ☐ no clot (must clot or run invalid)

  Replicate 2 of above: ______________________________

Interpretation:
  ☐ Valid run (NC neg, PC pos, PPC pos)
  ☐ Invalid run — void, cause: ______________________________
  ☐ Sample PASS at tested dilution = < _____ EU/mL in neat oil
  ☐ Sample FAIL at tested dilution = ≥ _____ EU/mL in neat oil

Retain sample: 2 weeks at 2–8 °C in original vial, for re-test if requested.

Operator signature: ______________________
```

### 8.3 Back-calculation worked example

Testosterone cypionate 200 mg/mL, tested after 1:1 LLE + 1:100 dilution from extract = 1:200 cumulative from neat oil, at λ = 0.25 EU/mL Pyrotell.

- NC: no clot. PC: clot. PPC: clot. NPC: **no clot.**
- Conclusion: endotoxin in the diluted tested sample is < 0.25 EU/mL.
- Back-calculated to neat oil: < 0.25 × 200 = **< 50 EU/mL in the neat vial.**
- Endotoxin Limit for this product = 350 EU/mL.
- **Result: PASS at 14% of limit.** (A kinetic chromogenic assay could refine "how far below 50 EU/mL"; gel-clot cannot.)

If NPC had clotted at 1:200, you would repeat at 1:400 and 1:800 (still within MVD = 1,400×) to find the lowest dilution at which the sample is negative. That bracketing gives you a range: "endotoxin in neat oil is between 100 and 200 EU/mL" (if 1:400 is negative but 1:200 is positive), which still evaluates against the 350 EU/mL limit. A product that fails at MVD — i.e., the neat-oil endotoxin ≥ 350 EU/mL — is non-compliant.

---

## 9. Depyrogenation of reusable glassware

**USP <1228.1> Dry Heat Depyrogenation** is the governing chapter (effective 2024-03-01 revision). Ph. Eur. 5.1.1 is harmonized on time/temperature combinations.

### 9.1 Validated cycles

- **Primary cycle: 250 °C for ≥ 30 minutes** (USP <1228.1>, Ph. Eur. 5.1.1).
- **Alternative cycle: 200 °C for ≥ 60 minutes** (Ph. Eur. 5.1.1 accepts; USP notes as a valid time/temperature trade-off).
- **220 °C / ~2 hours** (PDA TR3; valid but rarely used).

All cycles must demonstrate **≥ 3-log (1000-fold) reduction** of a challenge endotoxin inoculum. Validate with vendor endotoxin indicators (EI vials spiked with 10⁴–10⁶ EU) placed at worst-case chamber locations.

**Do NOT use 180 °C.** 180 °C is a dry-heat *sterilization* temperature. Endotoxin is roughly an order of magnitude more heat-stable than bacterial spores; at 180 °C, even multi-hour holds do not reliably deliver 3-log endotoxin reduction. The round-1 research once cited "180 °C / 4 h" as a Ph. Eur. alternative — **this was incorrect and is corrected in the round-2 pass.** Use 250 °C / 30 min or 200 °C / 60 min.

### 9.2 The "just buy depyrogenated disposables" alternative

Most home labs should skip the oven and buy vendor-certified disposable reaction tubes:

- **ACC TS050-10** (10 × 75 mm soda lime gel-clot tubes, pre-depyrogenated, 52/pack × 10) — canonical for this SOP.
- **ACC TB240-5** (12 × 75 mm borosilicate dilution tubes, 42/pack × 5).
- Pyrogen-free polypropylene microcentrifuge tubes (Axygen, Eppendorf Biopur).
- Pyrogen-free pipette tips (Rainin, Sartorius — lot-certified).

This removes the oven validation burden ($400–$2,500 used oven + endotoxin indicators + thermocouple data logger) for a program size below ~500 tests/year. Above that volume, an oven plus foil-wrapping reusable borosilicate tubes starts to pay back.

Cross-reference: `../03-microbial-pyrogen/depyrogenation.md` for oven selection, thermocouple mapping, and EI-based validation.

---

## 10. Common failure modes (specific)

1. **Using reconstituted lysate beyond its hold time.** Pyrotell reconstituted lysate is stable ~24 h at 2–8 °C, or up to 3 months single-freeze at ≤ −20 °C. Within a session, use within 1 hour of removing from refrigeration to maintain consistent activity. Beyond that → false negatives. Track reconstitution time on the vial label.
2. **Water bath condensation dripping into tubes.** If anyone on your team has substituted a water bath for a dry block, stop. Condensation on the tube wall can fall into the reaction volume and contaminate it; the water bath itself is almost never endotoxin-free.
3. **Vibrating bath / block.** Mechanical vibration at 37 °C breaks the developing gel. Dry blocks with loose-fitting tubes can vibrate enough to matter — use the correct 10 × 75 mm insert snugly. Place the block on a rubber mat; keep centrifuges and vortexers on a separate bench.
4. **Improperly depyrogenated tubes → NPC clots as false positive.** Symptom: NC is clean but NPC clots even on a sample you strongly suspect is low-endotoxin. Action: swap tube lot; run an all-LRW control series in the suspect tubes to confirm.
5. **Pipette-tip contamination.** Even pyrogen-free tips can contaminate if reused, laid on the bench, or handled with ungloved hands. Change tips between every reagent, every control, every sample. Gloves are not sterile — treat them as contaminating at the wrist.
6. **Cross-contamination during LLE.** Change tips between each extraction tube. Never touch the inside of a centrifuge tube with a pipette body. Do not reuse the pipette barrel across samples without barrier tips.
7. **Expired rehydrated CSE.** 0.5 µg/vial rehydrated CSE is stable **4 weeks at 2–8 °C per ACC**. 125 µg is 3 months. 10 ng is 7 days. Track on notebook log. Expired CSE → PC fails → run invalid.
8. **LRW left uncovered on bench.** LRW picks up airborne endotoxin and β-glucan from cellulose-lint environments. Re-cap immediately. Discard a bottle at 30 days opened unless you have validated longer.
9. **Insufficient centrifugation.** Aqueous phase that looks cloudy, has visible droplets, or sits over an emulsion band is contaminated with oil. Test results will be unreliable. Re-spin at higher g or longer time, or discard and re-extract.
10. **Reading gel before 60 min.** Gel forms in the last 10–20 min of the reaction for endotoxin concentrations near λ. Reading at 45 min on a marginal sample is a false negative. Set a hard timer. Do not eyeball "it looks done."
11. **Aspirating too much aqueous phase.** Taking > 80% of the aqueous layer almost guarantees oil carryover. Leave a ~100 µL reserve at the bottom. Practice this with dyed oil to train the visual.
12. **Undervortexed CSE (at reconstitution).** ACC and Lester et al. both emphasize: CSE must be vortexed ≥ 15 min after initial rehydration and 1 min before each reuse. Under-vortexed CSE is aggregated; the PC will fail or read inconsistently.
13. **Ignoring the 50-min vortex requirement in LLE.** This is the most load-bearing finding in the PPD paper and the most likely to be shortcut. 50 minutes is not negotiable unless you have performed your own validation showing shorter time gives ≥ 50% recovery in your matrix.
14. **Not running PPC every run.** "I validated this matrix last month" is not a substitute for a current PPC. Matrix composition varies lot-to-lot (co-solvent concentrations, API concentration, manufacturing impurities); PPC every run is the only way to detect drift.
15. **Benzyl alcohol / benzyl benzoate carryover.** If your aqueous extract tastes of the solvent (don't actually taste it — smell indirectly), benzyl alcohol has carried over. It inhibits LAL above ~0.5% in the reaction volume. Dilute past it or switch to a format with lower matrix sensitivity (rFC is more robust here).
16. **β-glucan cross-reactivity.** If your workflow uses cellulose membrane filters, glass-fiber filters, or paper-based wicks anywhere, (1→3)-β-D-glucan can cross-react with native LAL and cause false positives. Use ACC Glucashield reconstitution buffer or a glucan-blocked lysate. Or skip cellulose entirely.

---

## 11. When this SOP is the wrong tool

- **Quantitative result needed (e.g., "0.41 EU/mL"):** gel-clot can only give you a bracketed range by dilution series. Send out to a contract BET lab with kinetic chromogenic capability, or build out to kinetic on a heated microplate reader (used ~$500–$2,000; see `../03-microbial-pyrogen/endotoxin-lal.md`).
- **Intrathecal/epidural route products (K = 0.2 EU/kg):** endotoxin limits tighten 25×. Gel-clot at λ = 0.25 is marginal-to-inadequate. Use a more sensitive format (λ = 0.03 gel-clot pushes it, λ = 0.005 kinetic covers it).
- **Non-endotoxin pyrogens suspected:** LAL detects Gram-negative LPS only. Gram-positive lipoteichoic acid, mycobacterial components, fungal pyrogens are invisible to LAL. Use rabbit pyrogen or MAT (both infeasible at home; MAT discussion at `../03-microbial-pyrogen/endotoxin-mat.md`).
- **Horseshoe-crab-ethics concerns dominate procurement:** use rFC (recombinant Factor C). Lonza PyroGene or Hyglos EndoZyme. Cross-reference: `../09-safety-waste/horseshoe-crab-ethics.md` and `../03-microbial-pyrogen/endotoxin-rfc.md`. Note EP 2.6.32 effective 2021-01-01 (not 2016 — round-1 correction M1). USP <86> became official May 2025.
- **No mail-in commercial option for oils:** this is the operational gap. Janoshik explicitly excludes oils, powders, tablets, and capsules from their endotoxin service (per janoshik.com pricelist, verified 2026-04-22). There is no community-run mail-in endotoxin service for oil-based HRT as of this writing. This SOP exists because the mail-in fallback does not. Use the SOP carefully; the per-matrix revalidation is the whole point.

---

## Sources

- Lester CW, Ramsey JT, Salsbury JS. "Bacterial Endotoxins Testing in Lipid-Based Drug Formulations Using Liquid-Liquid Extraction." *American Pharmaceutical Review*, Nov/Dec 2019, 52–55. PDF: https://www.ppd.com/wp-content/uploads/2020/02/11Nov19-APR-PPD-Labs-bacterial-endotoxins.pdf (fetched 2026-04-23). Also hosted at americanpharmaceuticalreview.com (paywalled variant).
- Chen D. "A New Method for the Analysis of Bacterial Endotoxins in Ultrapure Paraffin Oil." *Journal of Analytical Methods in Chemistry*, 2014, Article ID 575246 (cited via Lester et al. 2019, reference 2).
- Schlösser A. "Endotoxin testing of oily substances." ECA Conference, 2009, Berlin (cited via Lester et al. 2019, reference 5; unpublished).
- USP <85> Bacterial Endotoxins Test — http://www.uspbpep.com/usp29/v29240/usp29nf24s0_c85.html
- USP <86> Bacterial Endotoxins Test Using Recombinant Reagents — USP publication, became official May 2025. https://www.usp.org/news/chapter-for-endotoxin-testing-using-non-animal-derived-reagents
- USP <1228.1> Dry Heat Depyrogenation — https://doi.usp.org/USPNF/USPNF_M8098_01_01.html (revision effective 2024-03-01, per https://www.uspnf.com/print/pdf/node/270811).
- Ph. Eur. 2.6.14 Bacterial Endotoxins; 2.6.32 Test for Bacterial Endotoxins Using Recombinant Factor C (effective 2021-01-01, per EP Supplement 10.3 published 2020-07-01); 5.1.1 Methods of Preparation of Sterile Products (depyrogenation cycles 250 °C / 30 min or 200 °C / 60 min).
- ACC Pyrotell product page — https://www.acciusa.com/products-and-services/bet-products/lal-reagents/pyrotell (fetched 2026-04-23).
- ACC CSE product page — https://www.acciusa.com/products-and-services/bet-products/lal-reagents/control-standard-endotoxin (fetched 2026-04-23).
- ACC Pyrotubes product page (TS050, TB240, TK100, TB013, TB16C catalog numbers) — https://www.acciusa.com/products-and-services/bet-products/accessory-products/pyrotubes (fetched 2026-04-23).
- ACC LRW product page — https://www.acciusa.com/products-and-services/bet-products/accessory-products/lal-reagent-water
- Fisher Scientific Pyrotell NC1241126 $595.35 — https://www.fishersci.com/shop/products/pyrotell-0-250-eu-ml-5ml-vial/NC1241126 (fetched 2026-04-23).
- Fisher Scientific CSE NC9263690 $251.13 — https://www.fishersci.com/shop/products/control-std-endotox-500ng/NC9263690 (fetched 2026-04-22).
- PDA Technical Report No. 3 (TR3): Validation of Dry Heat Processes Used for Depyrogenation and Sterilization (2013 revision; subscription).
- Round-1 cross-links: `../03-microbial-pyrogen/home-lab-endotoxin-feasibility.md`, `../03-microbial-pyrogen/endotoxin-lal.md`, `../03-microbial-pyrogen/endotoxin-limits-reference.md`, `../03-microbial-pyrogen/depyrogenation.md`, `../09-safety-waste/horseshoe-crab-ethics.md`.
- Round-2 critique-driven corrections: M1 (rFC date), M2 (depyrogenation cycle), M8 (Pyrotell vs PYROSTAR sensitivity confusion), M10 (LLE + gel-clot extrapolation caveat).