Tiger Milk Mushroom (Lignosus rhinocerus TM02) in Allergic Rhinitis: A Critical Reappraisal of the First Randomised Controlled Trial and the Human Evidence Base

1. Introduction

Allergic rhinitis affects an estimated 15% to 25% of people worldwide, and many patients remain symptomatic despite intranasal corticosteroids and antihistamines [1]. This unmet need has encouraged interest in botanical and fungal products as adjunctive treatments. The tiger milk mushroom, Lignosus rhinocerus, is a polypore whose sclerotium has long been used by indigenous communities in Southeast Asia for cough, asthma and other respiratory complaints. A cultivated strain, TM02, is registered as a traditional product in Malaysia and is sold widely as a supplement.

Preclinical evidence for an anti-allergic effect is consistent. In murine models of allergic asthma, extracts of the sclerotium reduced airway inflammation, mucus hypersecretion, airway hyperresponsiveness and remodelling [2,3,4], and its polysaccharides relaxed airway smooth muscle in vitro [5]. A 2026 systematic review concluded that these data are promising but that human evidence is very limited [6]. In 2026, the first randomised, double-blind, placebo-controlled trial of TM02 in allergic rhinitis was published, reporting that the product "significantly reduced the nasal symptoms" of patients and that the effect was associated with interleukin-5 downregulation [1].

Because this trial is likely to be cited as clinical support for TM02 in allergic rhinitis, its conclusions deserve close scrutiny. This review summarises the human studies of L. rhinocerus relevant to respiratory and immune outcomes, appraises the allergic rhinitis trial against established standards for trial conduct and reporting, recalculates its main comparisons from the published data, and sets out what a confirmatory trial would require.

2. Methods

Human studies of L. rhinocerus or L. rhinocerotis were identified by searching Europe PMC and Crossref to October 2026, and by screening the reference lists of the allergic rhinitis trial [1] and the 2026 systematic review [6]. The allergic rhinitis trial was appraised against the CONSORT 2010 statement [7], the ICH E9(R1) addendum on estimands [8], the requirement of the International Committee of Medical Journal Editors for prospective trial registration [9], and the 2018 United States Food and Drug Administration guidance on drug development for allergic rhinitis [10].

Between-arm differences in the primary outcomes were recalculated from the within-arm mean changes and 95% confidence intervals reported in the trial's Table 2. The standard error of each within-arm change was estimated as the width of its confidence interval divided by 3.92, and the standard error of the between-arm difference as the square root of the sum of the two squared standard errors, assuming independent arms and approximately normal sampling distributions. To interpret the reported standardised regression coefficients, each coefficient was multiplied by the pooled standard deviation of all visit scores for that outcome in the analysis set, approximating the standardisation described by the authors. Differences were compared with an anchor-based minimal clinically important difference of 0.55 units for total nasal symptom score [11] and with the threshold of 30% of the maximum score proposed by the United States Agency for Healthcare Research and Quality [12]. Sample sizes for a future trial were calculated for a two-sided α of 0.05 and 90% power.

3. The Human Evidence for Lignosus rhinocerus

Five human studies with respiratory or immune outcomes were identified (Table 1). Two had no placebo control [13,14]. The open-label study by Tan et al. reported improvements in respiratory symptoms, lung function and inflammatory markers over three months, but without a control group these changes cannot be separated from regression to the mean, seasonal variation or expectation [14]. A placebo-controlled trial in women after breast cancer treatment analysed quality of life on a per-protocol basis and reported within-patient changes and trends rather than significant differences from placebo [15]. A preliminary study in asthma has been published [16], but its design could not be assessed from accessible sources. The allergic rhinitis trial is therefore the only placebo-controlled trial of L. rhinocerus with a respiratory primary outcome [1].

Table 1. Human studies of Lignosus rhinocerus sclerotium relevant to respiratory or immune outcomes.

Study	Design and participants	Product and dose	Main reported findings
Chen et al. 2016 [13]	Randomised four-group trial of supplement and resistance training in young men; no placebo	L. rhinocerotis sclerotium, 500 mg daily for 8 weeks	Fitness and immune parameters; no respiratory outcomes
Tan et al. 2021 [14]	Open-label, single-arm, 50 volunteers	Tiger milk mushroom, 300 mg twice daily for 3 months	Reductions in interleukin-1β, interleukin-8, malondialdehyde and respiratory symptoms; increases in immunoglobulin A and lung function; no control group
Eng et al. 2025 [15]	Randomised, double-blind, placebo-controlled, 52 women after breast cancer treatment; per-protocol analysis	TM02, 1040 mg daily for 6 months	Within-patient change in global health status; trends in physical functioning and fatigue
Poh et al. 2025 [16]	Preliminary investigation in asthma	TM02	Cited by the trial authors as supporting modulation of allergic inflammation
Tan et al. 2026 [1]	Randomised, double-blind, placebo-controlled, 108 adults with persistent rhinitis on standard treatment; 56 analysed	TM02, 1200 mg daily for 30 days	No significant unadjusted difference from placebo; significant differences only after adjustment for baseline immunoglobulin E

Table 1. Human studies of Lignosus rhinocerus sclerotium relevant to respiratory or immune outcomes.

Study	Design and participants	Product and dose	Main reported findings
Chen et al. 2016 [13]	Randomised four-group trial of supplement and resistance training in young men; no placebo	L. rhinocerotis sclerotium, 500 mg daily for 8 weeks	Fitness and immune parameters; no respiratory outcomes
Tan et al. 2021 [14]	Open-label, single-arm, 50 volunteers	Tiger milk mushroom, 300 mg twice daily for 3 months	Reductions in interleukin-1β, interleukin-8, malondialdehyde and respiratory symptoms; increases in immunoglobulin A and lung function; no control group
Eng et al. 2025 [15]	Randomised, double-blind, placebo-controlled, 52 women after breast cancer treatment; per-protocol analysis	TM02, 1040 mg daily for 6 months	Within-patient change in global health status; trends in physical functioning and fatigue
Poh et al. 2025 [16]	Preliminary investigation in asthma	TM02	Cited by the trial authors as supporting modulation of allergic inflammation
Tan et al. 2026 [1]	Randomised, double-blind, placebo-controlled, 108 adults with persistent rhinitis on standard treatment; 56 analysed	TM02, 1200 mg daily for 30 days	No significant unadjusted difference from placebo; significant differences only after adjustment for baseline immunoglobulin E

4. Reappraisal of the Allergic Rhinitis Trial

4.1. Design

The trial randomised 108 adults aged 18 to 60 years with persistent rhinitis who were already using intranasal corticosteroids or antihistamines, to TM02 1200 mg daily or a brown-rice placebo for 30 days [1]. The primary outcomes were the total nasal symptom score over the past 12 hours and the past 2 weeks, each summing five items scored 0 to 3 (range 0 to 15), and the Allergic Rhinitis Control Test (range 5 to 25), each measured at baseline and day 30. Serum immunoglobulin E and plasma interferon-γ, interleukin-4, interleukin-5 and interleukin-13 were secondary outcomes.

4.2. Registration and Prespecification

The trial received ethics approval in June 2024, recruited from September 2024 to December 2025, and was registered in the International Traditional Medicine Clinical Trial Registry on 2 August 2026 [1]. The manuscript was received by the journal on 31 July 2026. Registration therefore followed the completion of the trial and the submission of its report. The International Committee of Medical Journal Editors requires registration before the enrolment of the first participant [9], precisely so that readers can verify that primary outcomes, analysis populations and covariate adjustments were fixed in advance. Without a prospective record, the choice of the modified per-protocol population as the main analysis, the definition of intercurrent events and the adjustment for baseline immunoglobulin E cannot be shown to have been prespecified.

4.3. Population

Half of the TM02 arm and 61% of the placebo arm were classified as having mixed rather than allergic rhinitis (Table 1 of the original report) [1]. Allergic sensitisation was not an inclusion criterion, and baseline immunoglobulin E varied widely. An anti-allergic mechanism would be expected to act mainly in allergic disease, so the inclusion of a majority with mixed rhinitis in one arm dilutes any specific effect and complicates interpretation of the immunological findings. Background treatment was used as needed by most patients, but the use of rescue medication during the trial was not reported, although changes in as-needed antihistamine or corticosteroid use could produce or mask symptom differences [10].

4.4. Analysis Population

The main analysis included 56 of 108 randomised patients (52%). Three TM02 patients were lost to follow-up, 27 patients had intercurrent events such as respiratory infections or compliance below 80%, 15 attended the final visit more than 7 days after finishing treatment, and 7 lacked baseline immunoglobulin E [1]. Exclusions were unequal: 29 of 54 TM02 patients (54%) against 23 of 54 placebo patients (43%). The analysis was therefore neither intention-to-treat nor a randomised comparison. Patients who complete treatment without infections and attend promptly may differ systematically from those who do not, and with unequal exclusions the comparability created by randomisation is lost [7]. The authors also describe a modified intention-to-treat analysis with multiple imputation, reported only in supplementary material as giving similar results. The ICH E9(R1) framework recommends that the handling of such intercurrent events be defined in the estimand before the trial, with sensitivity analyses reported alongside the main result [8].

The trial's sample size calculation required 45 patients per arm to achieve 90% power [1]. With 25 and 31 patients analysed, the trial had substantially less power than planned. The authors addressed this with post hoc power calculations from the observed effects. Post hoc power is a direct function of the observed p-value and adds no information about whether a result is true [17,18]; it cannot rescue an underpowered trial.

4.5. Unadjusted Results

Both arms improved. On the 12-hour score, the TM02 arm fell by 2.57 points and the placebo arm by 1.93 points; on the 2-week score, by 2.45 and 1.77 points; and the control test rose by 3.03 and 2.03 points [1]. The authors report that the unadjusted generalised estimating equation models, which tested the arm-by-time interaction, did not show TM02 to be superior to placebo for any primary outcome. Recalculation from the published confidence intervals agrees (Table 2): the between-arm differences were 0.64 and 0.68 points on the symptom scores and 1.00 point on the control test, with confidence intervals spanning zero and p-values of 0.30 to 0.48. For nasal itching over 2 weeks, the difference was 0.01 points.

Table 2. Between-arm differences in the primary outcomes of the allergic rhinitis trial, recalculated from the within-arm changes and 95% confidence intervals in Table 2 of the original report [1], compared with the reported adjusted coefficients. Positive differences favour TM02. The implied effect converts the standardised coefficient back to score units using the pooled standard deviation of all visit scores.

Outcome (score range)	Change, TM02 minus placebo	95% confidence interval (recalculated)	p-value (recalculated)	Reported adjusted standardised coefficient	Implied adjusted effect in score units
Total nasal symptom score, past 12 hours (0–15)	0.64	−1.15 to 2.43	0.48	−3.70	About 10.3 points
Total nasal symptom score, past 2 weeks (0–15)	0.68	−1.03 to 2.39	0.44	−4.44	About 11.6 points
Allergic Rhinitis Control Test (5–25)	1.00	−0.88 to 2.88	0.30	3.89	About 10.8 points
Nasal itching, past 2 weeks (0–3)	0.01	−0.48 to 0.50	0.97	−4.08	About 2.9 points
Rhinorrhoea, past 2 weeks (0–3)	0.32	−0.13 to 0.77	0.17	−3.88	About 2.7 points
Nasal congestion, past 12 hours (0–3)	0.23	−0.18 to 0.64	0.27	−3.75	About 2.6 points

Table 2. Between-arm differences in the primary outcomes of the allergic rhinitis trial, recalculated from the within-arm changes and 95% confidence intervals in Table 2 of the original report [1], compared with the reported adjusted coefficients. Positive differences favour TM02. The implied effect converts the standardised coefficient back to score units using the pooled standard deviation of all visit scores.

Outcome (score range)	Change, TM02 minus placebo	95% confidence interval (recalculated)	p-value (recalculated)	Reported adjusted standardised coefficient	Implied adjusted effect in score units
Total nasal symptom score, past 12 hours (0–15)	0.64	−1.15 to 2.43	0.48	−3.70	About 10.3 points
Total nasal symptom score, past 2 weeks (0–15)	0.68	−1.03 to 2.39	0.44	−4.44	About 11.6 points
Allergic Rhinitis Control Test (5–25)	1.00	−0.88 to 2.88	0.30	3.89	About 10.8 points
Nasal itching, past 2 weeks (0–3)	0.01	−0.48 to 0.50	0.97	−4.08	About 2.9 points
Rhinorrhoea, past 2 weeks (0–3)	0.32	−0.13 to 0.77	0.17	−3.88	About 2.7 points
Nasal congestion, past 12 hours (0–3)	0.23	−0.18 to 0.64	0.27	−3.75	About 2.6 points

These differences are similar in size to the anchor-based minimal clinically important difference of 0.55 units for total nasal symptom score [11], but the confidence intervals are compatible with no effect, and they are far below the threshold of 30% of the maximum score proposed by the Agency for Healthcare Research and Quality, which would be 4.5 points on a 15-point scale [12]. The large improvement in the placebo arm, which the authors attribute to expectation and regression to the mean [1], illustrates why only the between-arm comparison is informative.

4.6. The Adjusted Results

Statistical significance appeared only after adjustment for baseline immunoglobulin E. The authors fitted a full-interaction model containing time, arm and log-transformed immunoglobulin E, which was not centred [1]. They then reported the standardised arm-by-time coefficients from this model as treatment effects: −3.70 for the 12-hour score, −4.44 for the 2-week score and 3.89 for the control test, with similar values for individual symptoms.

These coefficients cannot be average treatment effects. Because the outcomes were standardised to the overall standard deviation, a coefficient of −3.70 corresponds to a reduction of about 10 points on the 12-hour score, more than twice the TM02 arm's baseline mean of 4.52. A coefficient of −4.08 for nasal itching corresponds to about 2.9 points on an item scored 0 to 3, where the observed between-arm difference was 0.01 points (Table 2). In a model that includes an interaction between treatment and a continuous covariate, the coefficient for treatment represents its effect when the covariate equals zero [19]. With log10 immunoglobulin E uncentred, that is the effect at an immunoglobulin E of 1 IU/mL, far below the interquartile range of the trial population, approximately 48 to 300 IU/mL. The reported coefficients are therefore most plausibly extrapolations to a patient who does not exist in the data. The average effect implied by the same model at typical immunoglobulin E levels would be close to the small unadjusted differences.

Covariate adjustment is a legitimate way to improve precision, but regulatory guidance requires that covariates be prespecified and that the treatment effect be estimated as an average over the population [10]. Centring the covariate, or reporting the treatment effect at the mean immunoglobulin E level with its confidence interval, would allow readers to judge whether any effect remains [19]. Until such estimates are published, the adjusted results should not be interpreted as evidence of efficacy.

4.7. Biomarkers

Neither immunoglobulin E nor any cytokine changed significantly in either arm [1]. Median interleukin-5 concentrations were 0.23 to 0.41 pg/mL, interleukin-4 medians were zero, and some values were below the lower limit of quantification or had coefficients of variation above 30%. The association reported between changes in interleukin-5 and in symptom scores came from a three-way interaction model with the same structure as the primary adjusted models, applied to multiple cytokines and symptom items without correction for multiple testing. It is an exploratory correlation within one arm, not evidence that TM02 lowers interleukin-5, and the title's reference to "IL-5 downregulation" is not supported by the cytokine results themselves.

4.8. Dose, Placebo and Safety

The dose of 1200 mg daily was chosen from a post-marketing survey of consumers rather than from dose-finding studies [1]. The placebo was matched for colour and size, but the masking of taste and smell, relevant for a mushroom powder, was not tested. Adverse events occurred in 29% of patients in both arms, mostly mild gastrointestinal symptoms, with no serious events. With 51 patients exposed for 30 days, serious adverse events with a true frequency above about 6% would probably have been detected, but rarer events or effects of longer use cannot be excluded.

4.9. Conflicts of Interest

The investigational product was supplied free of charge by its manufacturer, and the authors declared no conflicts of interest [1]. Both corresponding authors have previously published TM02 research with co-authors affiliated with the manufacturer, LiGNO Biotech [20]. Readers would be helped by a fuller statement of the relationship between the research group and the manufacturer.

5. What the Trial Supports

Table 3 compares the main claims of the trial report with the evidence presented. The trial provides useful information on short-term tolerability and on the size of the placebo response in this population, which is valuable for planning future studies. It does not demonstrate that TM02 reduces nasal symptoms more than placebo, that any benefit is clinically meaningful, or that it acts through interleukin-5.

Table 3. Claims made for TM02 in allergic rhinitis and the support provided by the trial.

Claim in the original report	What the data show	Assessment
TM02 significantly reduced nasal symptoms	Unadjusted comparisons with placebo were not significant; differences of 0.64 to 0.68 points on a 15-point scale	Not supported by the unadjusted comparison
Significant effects after adjustment for immunoglobulin E	Coefficients imply effects of 10 to 12 points, larger than baseline scores, consistent with an uncentred interaction term	Not interpretable as an average treatment effect
Clinically meaningful improvement	Point estimates near the anchor-based threshold of 0.55 units but confidence intervals include zero; far below the threshold of 30% of the maximum score	Not established
IL-5 downregulation	IL-5 did not change significantly in either arm; concentrations near the lower limit of quantification	Not supported; association is exploratory
Favourable safety	No difference in adverse events over 30 days in 105 patients	Supported for short-term use; rare events cannot be excluded
Adjunctive treatment option	Background therapy was not controlled and rescue use not reported	Hypothesis for a confirmatory trial

Table 3. Claims made for TM02 in allergic rhinitis and the support provided by the trial.

Claim in the original report	What the data show	Assessment
TM02 significantly reduced nasal symptoms	Unadjusted comparisons with placebo were not significant; differences of 0.64 to 0.68 points on a 15-point scale	Not supported by the unadjusted comparison
Significant effects after adjustment for immunoglobulin E	Coefficients imply effects of 10 to 12 points, larger than baseline scores, consistent with an uncentred interaction term	Not interpretable as an average treatment effect
Clinically meaningful improvement	Point estimates near the anchor-based threshold of 0.55 units but confidence intervals include zero; far below the threshold of 30% of the maximum score	Not established
IL-5 downregulation	IL-5 did not change significantly in either arm; concentrations near the lower limit of quantification	Not supported; association is exploratory
Favourable safety	No difference in adverse events over 30 days in 105 patients	Supported for short-term use; rare events cannot be excluded
Adjunctive treatment option	Background therapy was not controlled and rescue use not reported	Hypothesis for a confirmatory trial

6. Requirements for a Confirmatory Trial

A trial capable of answering the question would differ in several respects. It would be registered before enrolment, with the primary estimand, the handling of intercurrent events and any covariate adjustment fixed in advance [8,9]. It would enrol patients with allergic rhinitis confirmed by skin-prick testing or allergen-specific immunoglobulin E, use a run-in period with daily symptom diaries to establish a stable baseline, and record rescue medication use, as recommended for drug trials in allergic rhinitis [10]. Treatment would last at least 4 weeks and preferably cover a defined allergen season. The primary analysis would include all randomised patients, with multiple imputation and sensitivity analyses for missing data.

The sample size should reflect realistic effects. If the true between-arm difference were 0.6 points with a standard deviation of 2.8, as in this trial, about 460 patients per arm would be needed for 90% power. A standardised effect of 0.35, larger than any observed here, would still require about 170 per arm. Interleukin-5 and other biomarkers should be measured in a prespecified subgroup with assays validated for low concentrations, and a dose-ranging phase would establish whether 1200 mg is appropriate.

7. Limitations

This reappraisal relies on summary data from the published report and its main tables. The recalculated confidence intervals assume independent and approximately normal sampling distributions and may differ slightly from the authors' bootstrap and model-based intervals, although they agree with the authors' statement that unadjusted comparisons were not significant. The conversion of standardised coefficients to score units uses pooled standard deviations of the reported visit scores and is approximate. The interpretation of the adjusted coefficients as estimates at an immunoglobulin E of 1 IU/mL follows from the model described, but confirmation requires the authors' model output. Individual patient data would allow these questions to be resolved directly.

8. Conclusions

Lignosus rhinocerus has consistent anti-allergic activity in animal models, but its human evidence base is small. The first randomised trial in allergic rhinitis found no significant unadjusted difference from placebo, analysed only 52% of randomised patients with unequal exclusions, and was registered after submission. Its significant results depend on an uncentred interaction model whose coefficients imply effects larger than the symptom scores themselves, and its interleukin-5 findings are exploratory. The trial supports the short-term tolerability of TM02 but not its efficacy in allergic rhinitis. Clinicians, manufacturers and regulators should regard efficacy as unproven until a prospectively registered, adequately powered trial with an intention-to-treat analysis has been completed.