Standardisation of Centella asiatica (Pegaga) Extracts: Triterpene Marker Compounds, Extraction Effects and Human Study Evidence

Abstract

Background: Centella asiatica (L.) Urban, known as pegaga, is supplied as extracts standardised to pentacyclic triterpenes: the glycosides asiaticoside and madecassoside, and the aglycones asiatic acid and madecassic acid. Marker content varies with plant material and extraction method. The relationship between marker content and human exposure or outcome is not settled.

Objective: To review the published human evidence on Centella asiatica extracts, and the published evidence on how plant material and extraction affect the four triterpene markers.

Methods: PubMed was searched on 24 October 2026 using seven search phrasings. Human randomised controlled, open-label and pharmacokinetic studies were considered, together with laboratory studies of marker content and extraction. Ten records were included on the basis of their published abstracts.

Results: Six randomised controlled trials in humans were included, covering venous hypertension, diabetic microangiopathy, flight-related oedema, post-laser wound appearance, acute wound healing and cognition in healthy older adults. Two open-label pharmacokinetic studies in healthy Thai volunteers reported very low plasma exposure to madecassoside and asiaticoside, higher exposure to madecassic acid and asiatic acid, and accumulation of asiatic acid on repeated dosing. Two laboratory studies reported that marker content varied with plant part, harvest season and place of cultivation, and that 80% ethanol extraction by microwave or ultrasound gave high four-marker yields. Several of the human trials were small, and their outcomes differed.

Conclusion: The human evidence is at grade B: at least one randomised controlled trial exists, with material limits on size, duration and population. The record supports the four triterpenes as the characterised constituents and shows that their content varied with place of cultivation and harvest period, and that the percent of ethanol was the major factor for yield in microwave- and ultrasound-assisted extraction. It does not establish that marker content predicts human effect.

Review methodology

Database and date: PubMed, searched on 24 October 2026.

Search terms, as entered:

1. Centella asiatica randomised controlled trial (human clinical evidence filter on)
2. asiaticoside madecassoside (human clinical evidence filter on)
3. Centella asiatica extract standardisation triterpenes (human clinical evidence filter off)
4. Centella asiatica triterpene content plant material extraction method (human clinical evidence filter off)
5. Centella asiatica pharmacokinetics healthy volunteers asiatic acid madecassic acid (human clinical evidence filter on)
6. Centella asiatica cognitive function randomized placebo-controlled human (human clinical evidence filter on)
7. Centella asiatica hypertrophic scar topical clinical trial (human clinical evidence filter on)

Included: human randomised controlled trials of a Centella asiatica preparation tested on its own, not as one of several ingredients or treatments; human open-label pharmacokinetic and safety studies of a named standardised extract; and laboratory studies that report the content of the four triterpenes in relation to plant material or extraction method. Laboratory studies are labelled as such and are not presented as human findings.

Excluded, and why:

• 32279615: Centella was one of four herbal extracts added to a silicone gel, so any effect cannot be attributed to Centella alone.
• 32409760: a comparison of Centella with flavonoids in haemorrhoidal disease, so Centella was one of two phlebotonic treatments and was not tested on its own.
• 23377333: a multicomponent topical scar product compared with white petrolatum after shave biopsies. Excluded because its abstract does not name Centella asiatica.
• 35190633 and 36456663: beagle dog pharmacokinetic studies. These are animal studies and are not used as human findings.

How records were assessed: each included record was read as its PubMed abstract. Figures, doses, extract descriptions, durations and populations are restated as the abstract reports them.

Summary of study findings

Human randomised controlled trials

De Sanctis et al., 2001 (PMID 11666125): venous hypertension

Design: prospective, placebo-controlled, randomised, dose-ranging trial.
Participants: three groups of patients with venous hypertension (ambulatory venous pressure above 42 mm Hg), with 20, 20 and 12 patients, and 10 normal subjects. The abstract gives these group sizes.
Dose and form: total triterpenic fraction of Centella asiatica (TTFCA), 60 mg three times daily or 30 mg three times daily; placebo; and 60 mg three times daily in normal subjects.
Duration: 4 weeks.
Outcome, in the abstract’s wording: “After 4 weeks of TTFCA treatment, there was a significant decrease of CFR, AC, and AET time in patients. This was also greater in the higher dose group.” The abstract concludes that “180 mg/day is more effective in improving symptoms and CFR.”

Incandela et al., 2001 (PMID 11666119): diabetic microangiopathy

Design: prospective, placebo-controlled, randomised study.
Participants: patients with severe diabetic microangiopathy, neuropathy, and edema; patients with microangiopathy without neuropathy; and healthy subjects followed as a reference.
Dose and form: TTFCA tablets, 60 mg twice daily, or matching placebo tablets.
Duration: 12 months.
Outcome, in the abstract’s wording: “In conclusion, the decrease in capillary filtration and edema is associated with symptomatic improvement.” The abstract also states that “Treatment was well tolerated; no side effects were reported.”

Cesarone et al., 2001 (PMID 11666121): flight-related microcirculation and oedema

Design: prospective randomised study. The comparator group received no drug or other treatment; it was not a placebo group.
Participants: adults aged 30 to 50 travelling more than 3 hours in economy class, with mild to moderate superficial venous disease and varicose veins.
Dose and form: TTFCA 60 mg three times daily, for 2 days before the flight, on the day of the flight, and for one day after it.
Outcome, in the abstract’s wording: “RAS and edema were significantly lower in the TTFCA-treated group (p<0.025).” RAS is the rate of ankle swelling.

Damkerngsuntorn et al., 2020 (PMID 32310680): postlaser resurfacing wound appearance

Design: split-face, double-blind, randomised, placebo-controlled trial.
Participants: 30 individuals with facial acne scars treated with a 2940 nm Er:YAG laser.
Dose and form: 0.05% w/w ECa 233 gel, a standardised extract of Centella asiatica, or placebo gel on the other half of the face. Applied four times daily for 7 days, then twice daily for 3 months.
Duration: 3 months after the initial week.
Outcome, in the abstract’s wording: “The ECa 233 treated side exhibited significantly less erythema index over total follow-up by 0.03 U.” The abstract also reports that “Skin biophysics did not differ between the two groups.”

Wattanathorn et al., 2008 (PMID 18191355): cognition and mood in healthy older adults

Design: randomised, placebo-controlled, double-blind study.
Participants: 28 healthy elderly volunteers.
Dose and form: Centella asiatica plant extract at 250, 500 and 750 mg once daily.
Duration: 2 months.
Outcome, in the abstract’s wording: “The results showed that the high dose of the plant extract enhanced working memory and increased N100 component amplitude of event-related potential. Improvements of self-rated mood were also found following the Centella asiatica treatment.”

Changsan et al., 2023 (PMID 38063299): acute wound healing with a spray film

Design: multicentre, randomised, controlled trial, with the formulation’s stability and physicochemical properties also evaluated.
Participants: 60 volunteers with clean-contaminated (class 1) wounds, randomised to a control group (n=30) and a testing group (n=30).
Dose and form: a polymeric spray film solution containing Centella asiatica leaf extract. The abstract reports the product’s marker content: asiatic acid 0.20±0.02 mg/ml, madecassic acid 0.16±0.01 mg/ml, asiaticoside 0.32±0.03 mg/ml and madecassoside 0.10±0.00 mg/ml. The abstract reports that the product was stable for six months at 30±2°C and at 40±2°C.
Duration: PUSH Tool scores on days 0, 3, 5 and 7.
Outcome, in the abstract’s wording: “The tested product significantly reduced the total PUSH and exudate scores, indicating that the polymeric spray film solution containing Centella asiatica improved wound healing.” The mean healing recovery times were 4.6±1.1 days for the testing group and 4.87±1.0 days for the control group.

Human pharmacokinetic and safety studies (open-label)

Songvut et al., 2019 (PMID 30699457): phase 1 safety and pharmacokinetics

Design: open-label, two-sequence dosage, single- and repeated-dose phase 1 study under fasting conditions.
Participants: healthy Thai volunteers.
Dose and form: ECa 233 capsules, 250 mg and 500 mg.
Duration: single dose and repeated dose.
Outcome, in the abstract’s wording: “maximum plasma concentration and area under the curve of the parent compounds (madecassoside and asiaticoside) were very low, while their respective metabolites (madecassic acid and asiatic acid) demonstrated higher values.” The abstract also reports that “none of the volunteers discontinued their participation due to adverse effects during the study.”

Songvut et al., 2021 (PMID 33767223): pharmacokinetics and metabolomics of a modified formula

Design: two-sequence dosage open-label phase I clinical trial.
Participants: twelve healthy Thai volunteers.
Dose and form: a modified formula of ECa 233, at 250 mg and 500 mg, in single and multiple doses.
Duration: multiple doses for seven days.
Outcome, in the abstract’s wording: “The major parent compounds, madecassoside and asiaticoside, were rarely absorbed, instead undergoing biotransformation into active metabolites, madecassic acid and asiatic acid.” The abstract also reports “accumulation of asiatic acid after multiple oral administration for seven days.”

Laboratory studies of marker content and extraction (context only)

Puttarak and Panichayupakaranant, 2012 (PMID 22966846): plant part, cultivation and harvest

Design: laboratory comparative study of plant parts from several Thai growing sites and harvest periods. Dried powder was extracted with ethanol by microwave-assisted extraction, and the four triterpenes were determined by HPLC.
Outcome, in the abstract’s wording: “Among the various parts of C. asiatica, the leaves contained the highest amount of pentacyclic triterpenes with a total content of pentacyclic triterpenes of 19.5 mg/g dry powder.” The abstract reports that “the contents of the pentacyclic triterpenes in C. asiatica varied according to the place of cultivation and the harvesting period.” Total content reached 37.2 mg/g dry powder for material from Trang harvested in March, and 37.4 mg/g for material from Songkhla harvested in December.

Thong-On et al., 2021 (PMID 34764384): green extraction optimisation

Design: laboratory study using central composite design and response surface methodology. Four-month-old tetraploid plants were the raw material. Microwave-assisted extraction (MAE) and ultrasound-assisted extraction (UAE) were compared with maceration, and the four triterpenes were measured by a validated HPLC method.
Outcome, in the abstract’s wording: the abstract reports that “the percent of ethanol was the major factor that had a significant effect on C. asiatica yield and glycoside and aglycone content during MAE and UAE.” MAE with 80% ethanol at 100 W for 7.5 min gave, in the extract, 7.332% w/w madecassoside, 4.560% w/w asiaticoside, 0.357% w/w madecassic acid and 0.209% w/w asiatic acid. UAE with 80% ethanol at 48 °C for 50 min gave, from dry plant, 2.262% w/w madecassoside, 1.325% w/w asiaticoside, 0.082% w/w madecassic acid and 0.052% w/w asiatic acid. The abstract reports that MAE and UAE consumed energy 59% and 54% lower, respectively, than maceration, per kilowatt-hour per gram of total triterpenoids.

Evidence-quality assessment

Grade: B. At least one human randomised controlled trial exists, and the record has material limits on size, duration and population. Grade A was not met, for the reasons below.

• Consistency across studies. The six trials cover different conditions: venous hypertension, diabetic microangiopathy, flight-related oedema, post-laser erythema, acute wound healing and cognition in healthy older adults. Oedema is measured in three trials (PMIDs 11666119, 11666121 and 11666125), and capillary filtration in two (PMIDs 11666119 and 11666125). The three 2001 trials share an author group and all use the same preparation, TTFCA, so they point in one direction but cannot be treated as independent confirmation of one finding.
• Study design. Six randomised trials are included. One of them (PMID 11666121) used an untreated control rather than a placebo. Two open-label pharmacokinetic studies are included; they describe exposure and do not test outcomes.
• Size. Study sizes stated in the abstracts range from 12 participants (PMID 33767223) to 62 (PMID 11666125, the four groups of 20, 20, 12 and 10 combined).
• How directly the studies match the subject. Only PMID 38063299 reports the four-marker content of the product it tested. PMID 32310680 names ECa 233, a standardised extract. The 2001 trials used a total triterpenic fraction. The match to a marker-standardised extract is therefore partial for most of the human record.
• Transfer to the format or population in question. The populations are narrow: healthy Thai volunteers for pharmacokinetics, facial laser patients, acute wound patients, healthy older adults, and patients with venous or microvascular disease. The pharmacokinetic studies show that the glycosides named on a label are present at low plasma levels after oral dosing, and that the acid metabolites are the higher-exposure compounds. Label marker content therefore does not directly measure exposure. PMID 22966846 used plant material collected from sites in Thailand. The laboratory studies used specific plant material and extraction conditions, and their results may not transfer to other sources.

Limitations

Records were assessed from their published abstracts rather than from full texts.

• Several trials are small.
• The outcomes, conditions and preparations differ between trials. No pooled analysis was undertaken.
• The three 2001 trials used a total triterpenic fraction.
• The pharmacokinetic studies are open-label, with 12 participants in one study, in healthy Thai volunteers.
• No human study in this set compares extraction methods, plant parts or harvest conditions with a clinical outcome.
• The laboratory studies each use one plant population or a limited set of growing sites, and the extraction results are reported at laboratory scale.

Conclusion

The record supports the four pentacyclic triterpenes (asiaticoside, madecassoside, asiatic acid and madecassic acid) as the constituents that Centella asiatica extracts are characterised by. It supports that their content varies with plant part, harvest season and place of cultivation, and that 80% ethanol extraction by microwave or ultrasound can give high four-marker yields in laboratory conditions. It supports that small randomised trials of Centella preparations have been conducted in several settings, and that the oral pharmacokinetics of the glycosides differ from those of the acid metabolites.

The record does not support a conclusion that marker content predicts clinical effect, because no trial in this set relates marker content to outcome. It does not establish that the 2001 total triterpenic fraction is equivalent to a marker-standardised extract. It does not establish a dose–response relationship for the four markers across the trials. It does not establish any clinical benefit that would be generalisable beyond the populations studied, and it provides no basis for any statement that a product treats, prevents or cures a condition.

References

1. De Sanctis MT, Belcaro G, Incandela L, Cesarone MR, Griffin M, Ippolito E, et al. Treatment of edema and increased capillary filtration in venous hypertension with total triterpenic fraction of Centella asiatica: a clinical, prospective, placebo-controlled, randomized, dose-ranging trial. Angiology. 2001;52 Suppl 2:S55-9. PubMed: 11666125
2. Incandela L, Belcaro G, Cesarone MR, De Sanctis MT, Nargi E, Patricelli P, et al. Treatment of diabetic microangiopathy and edema with total triterpenic fraction of Centella asiatica: a prospective, placebo-controlled randomized study. Angiology. 2001;52 Suppl 2:S27-31. PubMed: 11666119
3. Cesarone MR, Incandela L, De Sanctis MT, Belcaro G, Geroulakos G, Griffin M, et al. Flight microangiopathy in medium- to long-distance flights: prevention of edema and microcirculation alterations with total triterpenic fraction of Centella asiatica. Angiology. 2001;52 Suppl 2:S33-7. PubMed: 11666121
4. Damkerngsuntorn W, Rerknimitr P, Panchaprateep R, Tangkijngamvong N, Kumtornrut C, Kerr SJ, et al. The Effects of a Standardized Extract of Centella asiatica on Postlaser Resurfacing Wound Healing on the Face: A Split-Face, Double-Blind, Randomized, Placebo-Controlled Trial. Journal of alternative and complementary medicine (New York, N.Y.). 2020;26(6):529-536. PMID: 32310680. doi:10.1089/acm.2019.0325
5. Songvut P, Chariyavilaskul P, Khemawoot P, Tansawat R. Pharmacokinetics and metabolomics investigation of an orally modified formula of standardized Centella asiatica extract in healthy volunteers. Scientific reports. 2021;11(1):6850. PMID: 33767223. doi:10.1038/s41598-021-86267-2
6. Songvut P, Chariyavilaskul P, Tantisira MH, Khemawoot P. Safety and Pharmacokinetics of Standardized Extract of Centella asiatica (ECa 233) Capsules in Healthy Thai Volunteers: A Phase 1 Clinical Study. Planta medica. 2019;85(6):483-490. PMID: 30699457. doi:10.1055/a-0835-6671
7. Wattanathorn J, Mator L, Muchimapura S, Tongun T, Pasuriwong O, Piyawatkul N, et al. Positive modulation of cognition and mood in the healthy elderly volunteer following the administration of Centella asiatica. Journal of ethnopharmacology. 2008;116(2):325-32. PMID: 18191355. doi:10.1016/j.jep.2007.11.038
8. Changsan N, Srichana T, Atipairin A, Sawatdee S. Wound healing efficacy of a polymeric spray film solution containing Centella asiatica leaf extract on acute wounds. Journal of wound care. 2023;32(Sup12):S22-S32. PMID: 38063299. doi:10.12968/jowc.2023.32.Sup12.S22
9. Puttarak P, Panichayupakaranant P. Factors affecting the content of pentacyclic triterpenes in Centella asiatica raw materials. Pharmaceutical biology. 2012;50(12):1508-12. PMID: 22966846. doi:10.3109/13880209.2012.685946
10. Thong-On W, Pathomwichaiwat T, Boonsith S, Koo-Amornpattana W, Prathanturarug S. Green extraction optimization of triterpenoid glycoside-enriched extract from Centella asiatica (L.) Urban using response surface methodology (RSM). Scientific reports. 2021;11(1):22026. PMID: 34764384. doi:10.1038/s41598-021-01602-x

Disclosure of commercial interest

Bionutricia Extract has a commercial interest in the subject of this review. Its product line lists Gotu Kola Pegaga (Centella asiatica) Extract Powder, which Bionutricia Extract supplies as an ingredient; the product page states its asiaticoside content as 0.3-0.8%. Gotu kola, pegaga and Centella asiatica are the same plant and the same material: the product line uses the names Gotu Kola Pegaga and Centella asiatica together, and this review uses Centella asiatica. The review was written from the published literature listed in the references, and its findings and evidence grade were not altered to favour any product. The review was not commissioned or paid for by a third party. Nothing in this review is medical advice, and nothing in it claims that a product treats, prevents or cures any condition.

Author biography

Vitthia Rama Murti, Registered Pharmacist (RPh), RPh 15632. Subject scope: formulation, excipients, stability, tolerability and safety; clinical evidence: human trials, indications, dosing and clinical outcomes.

Review history

Date written: 24 October 2026.
Reviewed and signed off by Dr Yong Yi Yi, 26 October 2026.
Date reviewed: 26 October 2026.

See our ingredient line