Efficacy and safety of combined microneedling therapy for androgenic alopecia: A systematic review and meta‐analysis of randomized clinical trials

To provide dermatologists with more clinical experience in treating androgenetic alopecia, we evaluated the effect and safety of combined microneedling therapy for androgenetic alopecia.


| INTRODUC TI ON
AGA is a hereditary alopecia disease with progressive reduction of hair quantity. 1The main pathogenesis of AGA is due to elevated expression of androgen receptor and/or type II-5α reductase genes in scalp hair follicles, which increases the sensitivity of hair follicles to dihydrotestosterone formed after androgen conversion, leading to progressive miniaturization of hair follicles and hair loss. 2 Clinical manifestations include a progressive decrease in the number of hairs, progressive thinning of hair diameter, and often accompanied by increased scalp oil spillage.Gender differences contribute to differences in hair loss site and prevalence.Male patients are more likely to be seen in the frontal hairline and temporal angle, while females are mostly seen at the top of the head. 3Epidemiological data abroad show prevalence rates of approximately 80% in males and 50% in females. 4AGA can seriously affect the appearance of patients and cause psychological effects such as low self-esteem, anxiety, and depression. 5,6Therefore, effective treatment is extremely important for patients with AGA.
MN is a novel physically facilitated penetration technique that uses microneedle-like instruments to puncture the skin to enhance the concentration of topical medications.or achieve cosmetic benefits.
Initially, Orentreich 7 used the damage repair mechanism of MN to treat scars and named it "Subcision."Fernandes 8 then applied MN to treat perioral wrinkles and named the technique "Percutaneous Collagen Induction (PCI)."Since Henry 9 used MN to study transdermal drug delivery techniques, MN is now widely used for various skin diseases such as acne scar, pigmentary disorders, skin aging, and hair loss. 10,11wever, there is still no systematic evaluation of MN for AGA.We evaluate the studies on microneedling for hair loss to provide dermatologists with a novel therapeutic technique for treating hair loss.

| Search strategy
The databases searched included PubMed, Excerpta Medica Database, and the Cochrane Library Database.The literature search spanned the period from 2012 to 2022.The search terms are androgenic alopecia and microneedling and their respective free words.The literature searched needed to meet the following criteria to be included.

| Data extraction
The extracted data include (1) basic features of the included studies; (2) MN device and treatment protocols: MN device (type, depth, diameter, number of needles), treatment interval, treatment time, adverse event, and follow-up.

| Study risk of bias assessment
The Cochrane Bias Risk Tool was used to evaluate the quality of studies, which was negotiated and summarized upon completion.If there is any dispute, please discuss or decide with a third researcher.
We used Revman 5.3 software to plot the risk and bias summaries of the literature.

| Statistical analysis
We used Revman 5.3 in conjunction with Stata 15.1 for statistical analyses.The heterogeneity of each RCT was judged by the Q test and the I 2 test.If there was no significant heterogeneity among the studied results (I 2 ≤ 50% and Q test p > 0.1), the fixed-effect model was used.Otherwise, the random-effect model was used.Fixed-or random-effect model was selected based on the results of heterogeneity.Finally, the stability of the results was evaluated by funnel plot and Egger's test.

| Study selection
We retrieved a total of 280 studies from databases.One hundred and six duplicate studies were excluded.After reading the titles and abstracts, 74 researches irrelevant to the research purpose and 42 reviews, 11 case series/reports, 7 animal experiments, 2 clinical conferences, and 1 letter were excluded.Thirty-seven studies were initially screened.After reading the full-text, 15 studies for inconsistent group design, 1 retrospective study, and 8 studies for which full-text and available data were not available were excluded.

K E Y W O R D S
androgenic alopecia, combined microneedle therapy, system evaluation

| Changes in hair density
Four studies [12][13][14]24 reported specific changes in hair density in combined MN group and the MN alone group. Sine the heterogeneity test revealed a large heterogeneity (I 2 = 82%, p = 0.0007), we performed a sensitivity analysis results showed that Bao2020 14 had a | 1565 large effect on the heterogeneity of this meta-analysis (Figure 3A).
Heterogeneity was reduced by excluding Bao2020, 14 so a fixed- The changes in hair density were derived from the data from nine included studies [13][14][15][16][17][18][19][20]23 which compared combined MN group with single medication group. It hould be noted that the combined MN group of Faghihi2021 16 was designed with two different depths, so the results of Faghihi2021 16 were considered as two studies to be meta-analyzed.For the chi-square test: I 2 = 98%, p < 0.01, indicating heterogeneity among studies.We continued with sensitivity analysis and found that Faghihi2021 16 may have interfered with the meta-analysis (Figure 3B).There was no statistically significant change in the pooled MD after the exclusion of Faghihi2021. 16Therefore, a random-effect model was ad-

| Changes in hair diameter
14]24 The chi-square test showed In six included studies, 13,14,16,18,20,23 changes in hair diameter between combined MN group and single medication group were reported.Because the Faghihi2021 16 had two MN groups with different depths, the results of the Faghihi2021 16 were meta-analyzed as two studies.Due to high heterogeneity (I 2 = 78%, p < 0.01), we excluded the literature one by one to find sources of heterogeneity and found no overall heterogeneity (I 2 = 0%, p = 0.81) after removing Ramadan2020. 19Therefore, the fixed-effect model was adopted, and results showed that relative to single medication group, combined MN group significantly increased the diameter of hair (MD = 2.50, 95% CI = [0.99,4.02], Z = 3.23, p = 0.001) (Figure 5B).

| Other outcomes
Eight studies 13,14,[18][19][20][21]23,24 have also reported other outcomes regarding hair growth including VAS pain assessment, pull test, telogen hair count, ratio of vellus hair/terminal hair (RVHTH), follicular unit density, change in scalp tissue structure, and expression of molecules in the Wnt/β-catenin pathway. Howevr, because the number of studies reporting these outcomes indicators is too small to perform a meta-analysis, we have created a table to describe the results (Table 3).

| The adverse events
A fixed-effect model was adopted as the heterogeneity was less than 50% (I 2 = 36%, p = 0.19).The results revealed no statistical difference in the rate of adverse reactions between the two groups (RR = 0.83, 95% CI = [0.62,1.12], Z = 1.22,p = 0.22) (Figure 7).But, the asymmetric funnel plot and the p = 0.002 in the Egger's test suggest a possible publication bias in hair density in the combined MN group versus the single medication group(Figures 8B and   9).The results of the trim and fill method suggested that one additional studies comparing combined with MN and medication alone on hair density need to be included in the future to reduce publication bias (Figure 10).

Outcome indicator Result
Bao 2017 13 VAS pain score The mean VAS score for all subjects treated by MN therapy was 4.52 ± 3.7 Bao 2020 14 Expression of molecules in the Wnt/β-catenin pathway The MN + M group showed further upregulation of FZD3, β-catenin, and LEF-1 expression levels at both mRNA and protein levels in the treated areas, and the differences were statistically significant (p < 0.05) needles for subcutaneous injections.Such subcutaneous injections not only increase patient pain but also reduce the efficiency of clinicians. 31MN only penetrates the stratum corneum and does not reach the nerve endings or blood vessels.Therefore, patients feel significantly less pain compared to traditional subcutaneous drug delivery methods. 32It can be seen that MN not only significantly improve the transdermal rate and absorption of drugs, but also enable drugs to directly act on the treatment area to avoid systemic side effects.In addition, MN can also reduce the pain caused by traditional injection methods to a certain extent.
A total of 13 RCTs on the treatment of AGA with MN combined with other promoting hair growth therapies were included in this meta-analysis.Meta-analysis results showed (1) in terms of objective efficacy evaluation, compared with the monotherapy group, the combined MN group showed significant improvement in hair density and hair diameter after treatment, suggesting that combined MN can significantly increase the efficacy of single treatment.The combined MN group also showed some advantages in improving the telogen hair count, follicle number, RVHTH, and hair pull test; (2) in terms of subjective efficacy evaluation, physicians in the monotherapy group on the assessment of hair growth significantly lower than physicians in the combined MN group.
However, there was no statistically significant difference in patients' assessment between the two groups.This may be because each individual thinks that the standard of how much hair growth can produce satisfaction is different; (3) For safety, there was no statistical difference in adverse events between the combined MN group and the monotherapy group, and no serious adverse reactions were reported.Scalp tingling was the most common adverse event in all groups involving microneedling.The pain level was tolerated by all patients, and the pain could disappear automatically after treatment in the majority of subjects.A small number of patients present with localized lymph node enlargement and scalp erythema with pruritus, but the magnitude of these adverse events is mild and their duration is short.All of the patients resolved on their own, showing that microneedling has a good safety profile.| Because AGA is a genetically related polygenic recessive disorder, it has the potential for recurrence.Therefore, follow-up is particularly important in the treatment of AGA.Some of the studies included in this meta-analysis reported on long-term efficacy and recurrence.Dhurat 15 performed a follow-up 8 months after the end of treatment, and all patients in the combined MN group reported a sustainable response.Bao 14 returned to all subjects treated with microneedling 6 months after completion of the study: 20% had a further increase in hair density, 70% had a small amount of hair loss, and 10% had a loss of all new hairs.Aggarwal 12 visited the long-term effect of treatment by telephone.Most patients did not report relapse, but 12 patients reported that they started hair loss again.Ramadan 20 followed up patients 3 months after the last treatment, and all patients remained in an improved state of clinical symptoms and dermoscopy.Although the effect of combined MN is encouraging, there are still a few still patients who relapse after stopping MN therapy.Therefore, we believe that the efficacy can be maintained with a long-term treatment modality such as extended intervals of MN after the condition is significantly improved.
This systematic review and meta-analysis also had several limitations.First of all, since MN is an invasive operation, it is impossible to perform blinding and allocation concealment when comparing combined MN therapy with single minoxidil.The sample size included in the study was relatively small, and there were differences in the country, race, gender, and severity of disease among the subjects.So the results of this meta-analysis may not represent the global population perfectly.Secondly, different MN device types and different treatment protocols may be the reasons for the heterogeneity of individual outcome indicators.We used sensitivity analysis in an attempt to find out the heterogeneity of combined MN and single medication in hair density, but did not find a significant change in the pooled effect size of the meta-analysis after excluding a particular study.The number of MN parameters (number of needles, size, depth) and treatment interval and duration of treatment varied greatly from study to study, resulting in a small number of withingroup studies that could not be subgroup analyzed to further explore the reasons for heterogeneity.Therefore, more large RCTs are needed to formulate the best standardized MN regimens, which will benefit most AGA patients.

| CON CLUS ION
In conclusion, the results of this meta-analysis showed that the combination of MN therapy significantly increased hair density and hair diameter in patients with androgenetic alopecia compared to monotherapy without significant adverse effects.Patients and doctors are also more satisfied with the clinical efficacy of MN therapy.
Therefore, the use of MN for the treatment of androgenetic alopecia is worth promoting in clinical practice.
Inclusion criteria included (a) AGA patients; (b) experimental group: combined MN therapies; (c) control group: monotherapy for hair growth except MN; (d) outcomes: hair density, hair diameter, doctors, and/or patient satisfaction for effect; and (e) type of study: randomized controlled trials.Exclusion criteria included (a) non-AGA type of hair loss; (b) repeated published studies; and (c) studies with unavailable data.
effect model was selected (I 2 = 50%, p = 0.13).Results indicated that the combined MN group significantly increased hair density relative to the single MN group (MD = 13.36,95% CI = [8.55,18.16], Z = 5.45, p < 0.00001) (Figure 4A).F I G U R E 2 (A) Risk of bias graph and (B) risk of bias summary.F I G U R E 3 Sensitivity analysis (A) combined MN versus single MN and (B) combined MN versus single medication.F I G U R E 4 Forest plot of hair density (A) combined MN versus single MN and (B) combined MN versus single medication.
indicating the heterogeneity is high.After excluding Aggarwal2020 12 heterogeneity is low (I 2 = 48%, p = 0.14), so the fixed-effect model was chosen for meta-analysis.The pooled result indicated that combined MN group had a significantly higher compared with single MN group in the changes in hair diameter (MD = 13.36,95% CI = [8.55,18.16], Z = 5.45, p < 0.00001) (Figure5A).
cacy.Monotherapy group refers to the use of other hair growth promoting medications or therapies other than MN.Because there are different criteria involved in doctor's efficacy assessment including a 7-point scale, a 10-point scale, and so on.So the doctors' assessment of 50% improvement in hair growth was defined as a good response for analysis.The chi-square test indicated the heterogeneity was high (I 2 = 77%, p = 0.0002).There was low heterogeneity after excluding the Dhurat2013 15 and Ozcan2021 19 (I 2 = 38%, p = 0.17), and a fixed-effect model was chosen.The pooled result showed that the doctors considered the good response of the combined MN group significantly better than the monotherapy group (RR = 2.03, 95% CI = [1.62,2.53], Z = 6.24, p < 0.00001) (Figure6A).A total of six studies13,15,19,[21][22][23] evaluated patients selfassessment of efficacy.Due to the inconsistent subjects selfassessment criteria, 50% improvement in patients self-assessment hair growth was also considered a good satisfaction.The chisquare test revealed high heterogeneity (I 2 = 97%, p < 0.00001).Therefore, a random-effect model was selected.The results showed that there was no statistical difference in the good F I G U R E 5 Forest plot of hair diameter (A) combined MN versus single MN and (B) combined MN versus single medication.satisfaction of the patients between the monotherapy group and the combined MN group (RR = 3.44, 95% CI = [0.67,17.59], Z = 1.49, p = 0.14) (Figure 6B).

Funnel
plot that combined MN VS. single MN in hair density and combined MN VS.Monotherapy in hair diameter displayed a generally symmetrical funnel shape (Figure 8A, C, D), and the Egger test also showed low risks of publication bias, respectively (p = 0.986, 0.467, 0.961).
Traditional treatments for AGA include topical medications, systemic medications, and surgery.Currently, FDA-approved treatments include minoxidil, finasteride, low-level laser therapy (LLLT), and hair transplantation, but these single treatments have limited efficacy and some adverse effects.As a result, AGA treatment often requires more combination therapy to mitigate the occurrence F I G U R E 6 Forest plot of good response/good satisfaction (A) doctors' efficacy assessment: combined MN versus Monotherapy and (B) patients' efficacy assessment: combined MN versus monotherapy. of adverse effects while achieving and maintaining optimal efficacy.In recent years, the emerging microneedling technology has shown promising results in improving hair growth and has become a safer and more effective new adjunctive therapy for treating AGA.It consists of several micron-level needle tips arranged on a base.Most microneedle instrument body is 150-1500 μm in length and 50-250 μm in width.25 Microneedle mainly plays the role of hair growth through the following three principles: first, microneedle puncture in the epidermal stratum corneum can form a large number of micro-channels to increase the penetration and absorption of topical drugs, thus improving the bioavailability of topical drugs; second, the micro-injury formed by microneedle can stimulate the growth and regeneration of stem cells of hair follicle in the area of hair loss, and at the same time, the micro-injury can improves the microcirculation in the follicles, providing more nutrients for the hair follicles to grow; third, the micro-injury can be up-regulated with the growth of hair-related Wnt/β-catenin pathway, promoting increased expression of growth factors associated with hair growth, such as vascular endothelial growth factor, while micro-injury itself in the process of repair, the body can also release a variety of growth factors, together to promote hair growth.26Why does MN combined with such as minoxidil, type II-5α reductase inhibitors, PRP and other hair generating drugs or treatment methods achieve better clinical treatment effect?The specific advantages of the combined MN are shown in the following aspects.The barrier function of the stratum corneum is a major obstacle to the transdermal absorption of topical drugs.27Studies have found that the topical absorption of minoxidil is only 0.3%-4.5% of the dose used.28MN, by forming a large number of micro-pores in the skin's stratum corneum, can significantly increase the absorption rate and speed up the drug's onset of action.Dhurat 15 randomly assigned 100 AGA patients to the single minoxidil group and the MN combined with minoxidil group and found that the hair count was significantly higher in the combined group (p = 0.039).Interestingly, hair regrowth occurred in the combination group at only Week 6 of treatment, while it occurred in the single minoxidil group at Week 10.Finasteride is the only oral drug approved by the FDA for the treatment of AGA.It reduces the concentration of dihydrotestosterone in serum and scalp tissues and mitigates the damage to hair follicles by blocking the conversion of dihydrotestosterone.However, long-term application of finasteride may lead to sexual dysfunction.29Compared to oral finasteride, MN allows the drug to work directly on the scalp for optimal efficacy, which largely TA B L E 3 Summary of results for other outcome indicators.

F I G U R E 7
2022 18 Change in scalp tissue structure The epidermal thickness was significantly increased only in MN + M group (p < 0.001) The dermis thickness was significantly increased in two groups (p < 0.001) The average follicle diameter was significantly increased only in MN + M group (p < 0.001) Ozcan 2021 19 Pull test, telogen hair count The change of pull test was statistically significant in two groups (p < 0.05) The mean telogen hair count decreased in MN + PRP group, increased in PRP group (p < 0.05) Ramadan 2020 20 Pull test After treatment, the pull test became negative in more than 95% of subjects Sánchez-Meza 2022 21 Ratio vellous hair/terminal hair(RVHTH) The difference in RVHTH in the MN + 0.01% dutasteride group was statistically significant compared to the MN + NS group (p = 0.005) Yu 2018 23 VAS pain score The average VAS score for all subjects treated by FRM therapy was 3.63 ± 1.38 Yu 2021 24 Follicular unit density (/cm 2 ) The follicular unit density was statistically significant only in MN + FGF (p = 0.031) Forest plot of adverse events.| 1569 PEI et al. avoids the systemic side effects of finasteride and improves patient compliance.PRP is rich in various growth factors such as vascular endothelial growth factor (VEGF), epidermal growth factor (EGF), transforming growth factor (TGF), and platelet-derived growth factor (PDGF), which can improve hair follicle microcirculation and promote hair growth. 30However, dermatologists mostly use insulin F I G U R E 8 Funnel plot of the primary outcomes: (A) combined MN versus single MN in hair density; (B) combined MN versus single medication in hair density; (C) combined MN versus single MN in hair diameter; and (D) combined MN versus single medication in hair diameter.F I G U R E 9 Egger's test for publication bias: combined MN versus Single medication in hair density (p = 0.002).

Zhou 33
systematically evaluated the efficacy and safety of minoxidil combined with finasteride or LLLT or MN therapy versus single minoxidil for AGA.Global photographic assessments and changes in hair density were used as outcome indicators.Results showed that overall photographic scores in all three combination therapy groups were better than in single minoxidil group (p < 0.05).Hair density was significantly increased with minoxidil combined with LLLT or MN compared with single minoxidil (p < 0.05).More interestingly, the efficacy of minoxidil combined with MN is the most significant, and the difference is greater than all other combined group.With regard to safety, the three combination treatment groups did not exacerbate any adverse events compared to single minoxidil.The same high level of heterogeneity was present in Zhou's meta-analysis due to differences in microneedling devices and treatment protocols.In addition, Zhou's meta-analysis lacks a comprehensive evaluation of hair growth indicators such as hair diameter, follicle number and RVHTH, as well as follow-up reports.The strength of our meta-analysis was a comprehensive search of all clinical trials on the use of combination MN for AGA, including MN in combination with minoxidil, PRP, growth factor solution, spironolactone, finasteride, or dutasteride.To our knowledge, this meta-analysis is the first systematic evaluation of the efficacy and safety of combined MN versus monotherapy for AGA.We evaluated not only hair thickness and hair density, which represent objective outcome indicators of hair growth, but we also evaluated the satisfaction surveys of the researchers and subjects about hair growth.Because hair loss directly affects personal image and has a great impact on the mental health and quality of life of patients, the assessment of physicians and patients is a more realistic reflection of clinical efficacy.F I G U R E 1 0 Filled funnel plot: Nine included articles for combined MN versus single medication in hair density.