Therapeutic Potential of Adipose-Derived Stem Cells and Their Secretome in Reversible Alopecias: A Systematic Review

ABSTRACT

Androgenic alopecia (AGA) and alopecia areata (AA) are two highly prevalent conditions, affecting both men and women of a wide range of ages, which strongly impact their quality of life and self-esteem. Both pathologies are deemed to be reversible, although conventional therapies have shown limited scope and efficacy. New therapeutic approaches, focusing on the degenerative changes that take place in the hair follicle, are needed to achieve better outcomes. For instance, adipose-derived stem cells (ADSC), abundant and easy to obtain, hold great potential in follicular regeneration. ADSCs can be isolated as stromal vascular fraction (SVF) by the enzymatic digestion of the lipoaspirate or as nanofat by the mechanical breakdown of adipocytes. In addition, commercial preparations of the conditioned medium of the ADSCs secretome (ADSC-conditionate medium [CM]) have entered the market as an appealing alternative because of their comparatively lower cost and accessibility. A search was conducted, crossing relevant terms, on PubMed Central and Google Scholar. Criteria for inclusion were studies in the past 10 years on humans with AGA or AA, where either SVF, nanofat, or ADSC-CM was tested as the main treatment. Eleven publications qualified: two studied nanofat, three, ADSC-CM, and six, SVF, either individually or in combination with other therapies. Only one randomized controlled trial (RCT) was found and classified as evidence 2b according to the Sackett scale. The rest were case–control studies or case series with small samples and no control, graded as evidence 3b and 4. A meta-analysis could not be conducted due to the heterogenicity of the study designs. Given the evidence obtained, Level D NICE recommendation was established. However, we consider that the positive findings are sufficiently consistent to support the elaboration of further RCTs that share criteria and methods.

INTRODUCTION

Alopecia, regardless of its etiology, is a condition that arouses public interest, mainly due to its high prevalence in both sexes and its strong psychosocial impact. Affected individuals often refer to a decline in their quality of life and self-esteem and some degree of alteration in their social relationships.[1,2,3]

It has been classified into two large categories: scarring and nonscarring. Subtypes of the first are defined by irreversible atrophy of hair follicles which leads to permanent hair loss. As for nonscarry causes, androgenic alopecia (AGA) accounts for 95% of all types, and alopecia areata (AA) is second in prevalence.[4]

AGA is genetically determined and causes progressive hair loss in two different patterns, male AGA (MAGA) or female AGA (FAGA). In males, the loss is conditioned by the miniaturization of follicles mediated by dihydrotestosterone (DHT). FAGA does not always coexist with abnormal levels of DHT, leading many authors to differentiate between androgen-dependent and androgen-independent FAGA.[5] About 50% of Caucasian males are affected by the age of 50 and 38% of females, by 70.[6,7] Validated scales to assess severity are Norwood-Hamilton, in men, and Ludwig, in women.[8]

The development of androgen-dependent AGA has been related with an inhibitory effect of DHT over adipose-derived stem cells (ADSCs) of the scalp,[9] as well as the induction of premature senescence on dermal papilla cells (DPCs).[10] This inhibition of ADSCs, together with the increase in the number of senescent DPCs, may condition a significant loss of activity over the human follicle stem cells (HFSCs). Consequently, follicles may lose their regenerative capacity, which leads to shorter anagen and telogen phases, resulting in a progressive reduction in diameter, density, and total hair count.[11,12,13,14,15,16,17] It has been demonstrated that an androgen-mediated inhibition of the Wnt/β-catenin pathway in DPCs leads to a loss in the differentiation capacity of HFSCs.[18] This knowledge has driven numerous researchers to look for therapies focusing primarily on follicular regeneration.[19]

In autoimmune scarring alopecia, there is an inflammatory infiltrate around the bulge area which destroys HFSCs, resulting in irreversible hair loss.[20] Conversely, AA and AGA are regarded as reversible since, in these pathologies, HFSCs are only partially damaged.[16]

Subjects with MAGA have a depletion in HFSC activity.[14,21] Similarly, in AA, the expression of CD200, an HFSC marker, is diminished. This may indicate a loss of the immune privilege these cells possess in physiologic conditions, contributing to the self-reactive lymphocyte reaction typical of AA.[22,23]

TREATMENT ASSESSMENT METHODS

Of all the proposed methods to assess hair variables, phototrichograms, and Trichoscan^® (Tricholog and Datinf GmbH, Freiburg, DE) are the most highly valued. Both allow the noninvasive in vivo study of density (N. º hairs/cm²), diameter (μm), length (mm), anagen/telogen ratio, and growth rate (mm/day). Variables are usually calculated on a previously delimited area of the scalp of 1 cm² for 2 days. By combining epiluminiscence microscopy and automated image analysis software, Trichoscan^® is capable of analyzing all biological parameters of hair growth on a hair-to-hair basis.[24,25] According to some authors, results of Trichoscan^® and phototrichograms, carried out by the same examiner, display 91% correlation.[26] To facilitate follow-up over long periods of time, a dot is tattooed on the experimental area. Both techniques are simple, sensitive, reproducible, and generally well tolerated by patients, so much so they have replaced invasive methods such as unit area trichogram and scalp biopsy to monitor the success of hair loss interventions.[24]

Another common assessment tool is global photography, consisting of the Canfield validated technique. Despite not providing information on hair variables, it allows observers to visualize changes in AGA and AA severity scales.[24]

Pull test is another widely used technique that consists of the count of the number of hairs after mild traction carried out by an examiner after a washout and brushing of the hair. It lacks validation, and there is no standardized protocol.[24]

Subjective self-evaluation consists of validated questionnaires for MAGA and FAGA. Their ease of comprehension and performance makes them practical and useful to correlate outcomes with objective techniques.[27,28]

Existing androgenic alopecia treatments

Current treatment strategies are centered in stimulating remaining follicles to produce new hair, in blocking the androgenic stimulus, or in surgical transplant.[29]

Topical minoxidil, a peripheral vasodilator, increases anagen/telogen ratio, hair size, and hair count.[30] Small oral doses are also proving effective[31] although its mechanism of action is unclear. Adverse effects include hypertrichosis and rash, and it cannot be used orally in patients with cardiovascular risk. Recurrences are common after treatment interruption, and lack of response is not rare.[32,33,34]

Finasteride and dutasteride are I and II 5α-reductase inhibitors, blocking the conversion of testosterone into DHT. Applied topically, intradermically, and orally, they have proven to increase hair count and diameter.[35] Risks include birth defects in male fetuses if used in pregnancy, decreased libido, sexual dysfunction, gynecomastia, menstruation disorders, gastrointestinal discomfort, acne, and dizziness.[36,37]

Autologous transplant of follicular units has become the gold standard for the treatment of AGA. However, its use is limited to the reduced number of follicles in the donor area and the lack of viability of transplanted follicles.[38,39]

Platelet-rich plasma (PRP) has been used for years in regenerative medicine,[40] including the treatment of alopecia or as a coadjuvant to hair transplant surgery. Its effect is due to the release several growth factors which stimulate HFSCs ad angioneogenesis in the perifollicular space and inhibit apoptosis in DPCs.[41,42] Recent reviews have shown it is an effective treatment for patients with AGA, who increased diameter and density and reduced pull test numbers.[43,44]

Existing alopecia areata treatments

Most common treatments include topical, intradermal, and oral corticosteroids, contact immunotherapy, and systemic immunosuppressors, depending on disease extension. In addition, laser and PUVA have proved helpful in some cases,[45] and PRP has been tested, achieving several degrees of response.[46,47,48]

The promising role of adipose-derived stem cells

Mesenchymal stem cells are a type of multipotent percursor first identified in bone marrow, and later found in connective and adipose tissue, where populations reach 40-fold those in bone marrow.[49] It was recently demonstrated that bone marrow mesenchymal cells were an effective, safe, and well-tolerated alternative for the workup of both resistant AA and AGA.[22]

There is evidence that ADSCs and their conditionate medium (ADSC-CM) or secretome promote follicular growth in vitro, in vivo, and ex vivo.[50,51] ADSC’s secretome is composed by several growth factors with paracrine effects.[52]

It is known that ADSCs can be used to regenerate hair by three mechanisms: (1) increase of proliferation rate of follicular cell; (2) protective action over DPCs against cytotoxic damage mediated by androgens and reactive oxygen species; (3) induction of anagen and elongation of the hair shaft in murine hair and ex vivo.[53]

ADSCs lack immunogenic properties, hold a high capacity for differentiation and immunomodulation, and have a pro-angiogenic effect. Its elevated number and ease of obtention make them the preferred candidate for clinical research in individuals with AGA and AA.[54]

Obtention of adipose-derived stem cells and adipose-derived stem cells conditionate medium

There are currently two ways of ADSC isolation: enzymatic digestion and mechanical disgregation of adipose tissue. Enzymatic digestion consists of the application of collagenase into the lipoaspirate, which separates its contents in two phases: adipocyte fraction and aqueous fraction. The aqueous fraction is centrifuged, giving as a result a dark red pellet, known as stromal vascular fraction (SVF). SVF is estimated to contain 2%–10% of ADSCs, whereas the rest are fibroblasts, adult endothelial cells and their precursors, macrophages, lymphocytes, pericytes, pericytes, erythrocytes, and smooth muscle cells.[53]

Mechanical disaggregation is achieved through emulsification of the lipoaspirate by its transference between two syringes, which break the adipocytes.[55] After filtering the emulsion, viable adipocytes are eliminated and a suspension of triglycerides and cells known as nanofat is obtained. Nanofat is rich in CD34+ populations, which correspond to ADSCs.[56,57]

Although the number of viable ADSCs recovered from mechanical disaggregation is inferior to that obtained by enzymatic lysis,[58] several case series have shown positive results using nanofat for the treatment of skin aging, scars, and chronic wounds.[55,59]

In the past years, commercial preparations of ADSC-CM have made their appearance in the market. Some examples are ASC-CCM^® (Lonza, Walkersville, MD, USA) and AAPE^® (Prostemics Research Institute, Sungnam, South Korea).

AAPE^® is obtained from enzymatic digestion and cell expansion. When a determined cell density is achieved, ADSCs are harvested in hypoxic conditions to induce the release of growth factors. Ultimately, ADSC-CM is collected, filtered, and lyophilized.[60] Active principles of AAPE® are detailed in Table 1.[61]

Table 1.

Main components in AAPE^® and their concentration, analysed by flow cytometry.[61]

Active principle	Concentration (per 5 ml saline solution)
PDGF	44.41±2.56 pg/mL
bFGF	131.35±30.31 pg/mL
KGF	86.28±20.33 pg/mL
TGF-b1	103.33±1.70 pg/mL
HGF	670.94±86.92 pg/mL
VEGF	809.53±95.98 pg/mL
Collagen	921.47±49.65 pg/mL
Fibronectin	1466.48±460.21 pg/mL

Some studies have shown efficacy of these preparations for retinal disease,[62] induction of angiogenesis,[63] wound healing,[64] and skin aging.[65] Their availability and low cost in comparison to cell therapy might make ADSC-CM a suitable alternative for regenerative medicine practitioners. At present, AAPE^® lacks FDA approval for the treatment of AGA.

The aim of this review is to know what current body of evidence supports the use of ADSCs and ADSC-CM for the treatment of adult subjects with AGA or AA.

MATERIALS AND METHODS

A cross-search of keywords was carried out in PubMed Central. Inclusion criteria were all articles published in the past 10 years, which contained the keywords shown in Table 2.

Table 2.

search parameters for PMC and preliminary results

Keywords		No of results
Hair regeneration + Adipose-derived Stem Cells.		909
Alopecia +	Adipose-derived Stem Cells	424
Baldness +		437
Hair loss +		781
Alopecia +	Nanofat	8
Baldness +		8
Hair loss +		11
Alopecia +	Stromal Vascular Fraction	202
Baldness +		207
Hair loss +		962
ADSC +	Hair growth	113
Nanofat +		6
SVF +		104
Total		4.172

Only experimental or quasi-experimental studies in humans with AGA or AA were included. A second search with the same parameters was carried out on Google Scholar to complete the results. Assessment of clinical evidence was performed using the Sackett scale.[66] Recommendation level was assessed according to NICE guidelines for therapeutically interventions.[67]

RESULTS

Results are summarized in Tables 3 and 4. A total of 11 publications met the criteria, the characteristics of which are detailed in Table 5. All studies used intradermotherapy except for Study 4, which used transdermal microneedling, and Study 7, which infiltrated using a cannula. Studies 6 and 9 mention a control group that was not described on the methods and were, therefore, considered to have none. Based on the evidence levels obtained, NICE recommendation Level D was established for therapies using ADSCs and ADSC-CM in patients with AGA.

Table 3.

Publications included in this review.

Title	Year	First author	Type of study	Indication	Object of study	Obtention method	Level of evidence[66]
The Use of Nanofat in Androgenic Alopecia. A Prospective Blinded Study.	2017	Vestita M.	Case-control study. SB.	AGA	Nanofat	Mechanical emulsification and filtration of fat graft♥.	3b
Nanofat+PRP Vs PRP in Androgenic Alopecia. An Intra-Patient Case-Control Study.	2018	Vestita M.	Case-control study. SB.	AGA	Nanofat + PRP Vs. PRP	Nanofat: ♥ PRP: Centrifugation of blood sample and extraction of PRP fraction ■.	3b
The Latest Advance in Hair Regeneration Therapy Using Proteins Secreted by Adipose-Derived Stem Cells.	2012	Fukuoka H.	Case series. UM.	AGA & AA	ADSC-CM + multivitamins & other substances.	AAPE®	4
Clinical use of conditioned media of adipose tissue-derived stem cells in female pattern hair loss: a retrospective case series study.	2015	Shin H.	Retrospective case series. UM.	FAGA (Ludwig I)	ADSC-CM (transdermally with microneedling).		4
Hair Regeneration Treatment Using Adipose-Derived Stem Cell Conditioned Medium: Follow-up with Trichograms.	2015	Fukuoka H.	Case-control study. SB.	AGA	ADSC-CM		4
Hair quality improvement in alopecia patients following adipose-derived stem cell treatment.	2016	Anderi R.	Case series. UM.	AGA	SVF	Enzymatic digestion of fat graft, filtration, centrifugation and cellular resuspensionΩ.	4
Hair follicle growth by stromal vascular fraction enhanced adipose transplantation in baldness.	2017	Pérez-Meza D.	Case series. UM	AGA	SVF + nanofat	SVF: Enzymatic digestion. Nanofat: ♥	4
Autologous Adipose Derived Stem Cell versus Platelet Rich Plasma Injection in the Treatment of androgenetic Alopecia: Efficacy, Side Effects and Safety.	2018	Hamed Kadry M.	RCT. SB.	AGA	SVF & PRP	Ω + ■	2b
Cellular therapy with human autologous adipose-derived adult cells of stromal vascular fraction for alopecia areata.	2018	Anderi R.	Retrospective case series. UM.	AGA & AA	SVF	Ω	4
Introducing Platelet-Rich Stroma: Platelet-Rich Plasma (PRP) and Stromal Vascular Fraction (SVF) Combined for the Treatment of Androgenetic Alopecia.	2018	P. Stevens H.	Case series. UM.	AGA	SVF + PRP	SVF: Mechanical disgregation off at graft, centrifugation and decantation of SVF PRP: ■	4
Stromal Vascular Fraction Enhanced Adipose Transplantation in Hair Loss: Early Experience & Active Phase II FDA Investigation.	2016	A. Aronowitz J.	Case series. SB.	AGA	Adipose graft enriched with SVF	Ω + Enrichment off at graft with SVF.	4

SB – Single-blinded; UM – Unmasked; RCT – Randomized clinical trial

Table 4.

List of exclusion criteria reported on each of the 11 publications

Study No	Exclusion criteria
1	Individuals under treatment with oral finasteride, laser, and infiltration therapies during 6 months prior to the study.
	Patients under topical Minoxidil were given a 60-day washout period.
2	Same as in study 1.
3	None.
4	Subjects who had used any products that could interfere with hair growth during the 6 months prior to the treatment.
5	None.
6	History of alcohol, smoking, loss of weight during the last 2 years, chemotherapy, immunosuppression, and abdominal surgeries.
7	Same as in study 4, plus:
	Treatment with an investigational product or procedure within 30 days or plans to participate in another clinical study.
	Subject deemed nonresponsive to a previous experimental hair loss treatment
	History of autoimmune disease, organ transplantation, cancer, active infection, chronic antibiotic/steroidal/anticoagulant treatment, coagulopathies, phychiatric disorders, surgeries in the treatment areas, pregnancy, lactation, diabetes, irritated or abraded scalp, allergy to treatment components.
	Family/close friends of the study’s staff.
8	Same as in study 4, plus:
	Pregnancy, bleeding disorders, history of keloids and malignancy.
9	Same as in study 4, plus:
	History of disease other than alopecia, alcohol, smoking, recent weight loss, chemotherapy, immunesupressants, hormonal imbalances, medication potentially causing hair loss, altered metabolic panel, allergies.
10	Subjects treated for male pattern hair loss in the previous 12 months.
	Previous hormone replacement therapy.
	Platelet disorders, cancer, sepsis, antiaggregating therapy, and smokers.
11	None.

Table 5.

Characteristics of the 11 studies

Sample size (n)	Mean age	Groups	Control	No of Treatments	Assessment method	Followup time (months)	Results	Observations
12 (♂)	Ø	1	Yes (intra-patient)	1	Phototrichogram on marked area. VAS (patients).	12	Increase in density and diameter superior to control at 3 and 6 months. Recurrence (n=6) at 12 months. Global satisfaction increased from 6 months onward.	Blind examiner. Nanofat samples were not characterised. The obtention process, injection procedure and volume injected were not described.
24 (♂)	Ø	1	Yes (intra-patient)	1		12	Increase of density and diameter superior to control. Partial recurrence (n=24) at 12 months. Patient satisfaction increased similarly, with peak at 6 months.
25 (♂=13; ♀=12)	Ø ♂ (20-49) ♀ (30-69)	1	No	4 (one every 3-4 months)	Phototrichogram & global photography (some patients) VAS (patient & examiner).	12	Statistically significant improvement in VAS scores compared to first treatment. Patients’ perception of improvement was higher than that of examiners.	One case had concomitant AA. Five patients under oral finasteride were not excluded. Results were maintained at least 1.5 years from first treatment.
27 (♀)	41,9±13,4	1	No	12	Phototrichogram on marked area.	3 (n=26); 6 (n=9); 12 (n=1)	Increased density and diameter. No recurrences.	No correlation between age, evolution and variations in the measure parameters was found.
22 (♂=11; ♀=11) Control: 10 (♂=8; ♀=2)	Ø 20-70	2	Yes	6 (one every 3-5 weeks)	Phototrichogram on marked area.	3	Increased hair count in experimental group compared to before treatment, without differences between sexes. Increased hair count in control group superior to placebo. Placebo area also showed increased hair count.	No significant difference was observed in the experimental group between subjects who recieved finasteride and those who did not.
20 (♂=11; ♀=9)	Ø 23-63	1	No	1	Pull test. Phototrichogram on marked area. Global photography.	6	Increase in density and diameter. Decrease in pull test numbers. Without differences between sexes.	Two patients did not experiment significant changes.
9 (♂=8; ♀=1)	29 (19-54)	1	One patient with control area.	1	TrichoScan® on marked area. Global photography.	3 (n=9); 6 (n=6) 8 (n=1)	20.5% increase in mean hair count at 6 months. 18.3% increase in width at 6 months.	Control patient: the experimental area showed a greater increase in density, cumulative thickness and anagen/telogen ratio than the control area (descriptive data).
60 (PRP: ♂=15; ♀=15) (SVF: (♂=13; ♀=17)	PRP: 27,67±4,62 SVF: 26,8±3,65	2 PRP (n=30) SVF (n=30)	No	3 (one every 3 months)	TrichoScan®. Global photography.	3	SVF group: increase in density and diameter of intermediate and terminal hair. Four patients improved their AGA stage. PRP group: increase in density of terminal hair. No changes in diameter and stage.	Study groups were statistically comparable. Degree of improvement was determined by changes in Norwood-Hamilton/Ludwig stages.
20 (♂=11; ♀=9)	38,3±2,3 (23-63)	1	No	1	Pull test. Phototrichogram (without area-marking). Global photography.	6	32% mean increase in diameter at 6 months (superior in men) in 19 patients. 36% mean increase in density at 6 months in 18 patients. Significant improvement in pull test at 3 and 6 months without differences by sex.	Cell viability and number were studied.
10 (♂=10)	45.2±14.5 (25-72)	1	Yes (intra-patient)	1	Follicle-by-follicle TrichoScan® analysis on marked area. Global photography.	3	Increase in hair density and demonstration of hair growth in previously inactive follicles in frontal, midscalp and vertex areas superior to control.	The occipital area acted as control. Authors suggest that an optimal treatment regime should include several sessions.
9 (♂=8; ♀=1)	Ø	1	No	1	TrichoScan®. Global photography.	6 (n=6)	14% mean increase in hair count at 6 months. 34% increase in anagen ratio.	Cell no and graft volume were not reported.

VAS: Visual Analogic Scale questionnaire. Ø: unreported data.

DISCUSSION

All studies, except one, had quasi-experimental designs. Only Study 8 assigned subjects randomly to the two groups. Expectancy and measurement bias were present in those studies that did not assess changes by objective means, such as phototrichograms or TrichoScan^® on shaved, tattoo-marked areas, and participant–expectation bias could arise in studies that compared before and after satisfaction, due to the lack of randomization, double-blinding, and placebo-control group. Only Study 5 was properly controlled by placebo with a separate control group, whereas Studies 3, 4, 6, 8, 9, and 11 had no control group at all, and Studies 7 and 10 had control groups, yet poorly designed. Study 7 included only one patient with a control area and, therefore, only descriptive data were provided, and Study 10 took the occipital region as control. Given its androgen-resistant nature, this region remains relatively unaffected by AGA and cannot be a reliable control.[68,69]

Selection bias occurred in Studies 3, 5, 6 and 11, due to the failure to exclude subjects who previously received (or were under) hair loss treatment. Attrition bias caused by dropouts in the long-term follow-up occurred in Studies 4, 7, and 11. Finally, the small samples compromised both the internal and external validity of the studies and exposed them to random errors.

All studies, however, obtained positive results, either in increased hair growth or increased patient satisfaction with the use of nanofat, SVF, and AAPE^®.

Nanofat

Studies 1 and 2 had strong exclusion criteria and included a washout period to rule out minoxidil as a confounding variable. The PRP variable in Study 2 was isolated by treating control areas with PRP and experimental areas with PRP plus nanofat.

The concordance between the increases in hair variables and patient satisfaction adds to the validity of results. However, the 12-month follow-up detected recurrences in both studies. This fact signals the need for study of long-term outcomes in ADSC-based therapies.

As for negative aspects, authors did not stage AGA severity and, in Study 1, they did not inform whether experimental and control areas had the same dimensions, which potentially made them incur measurement bias. Another oversight of both studies was the absence of characterization of the nanofat samples and the number of viable cells transplanted to each patient. The way nanofat was obtained was not described, nor was the injection procedure. Another omission was the volume injected.

In contrast, authors of Study 3 described the injection procedure and made sure all subjects received the same volume of AAPE^®. Disease severity was not staged, and the changes were only assessed by subjective means. Regrettably, subjects under finasteride and with concomitant AA were not excluded, and AAPE^® was used in combination with other active principles that might affect hair growth. Both are confounding variables that could have easily been avoided.

Adipose-derived stem cells-conditionate medium

Study 4 was the only one that used microneedling mesotherapy. There is current evidence that this technique on its own induces hair growth via activation of Wnt/β-catenin and VEGF.[70,71] Therefore, in the absence of a control group to exclude this confounder, the association between AAPE^® and the changes observed was unreliable.

Study 5 included the shaving and tattooing of treatment areas, which increased the reliability of results. However, due to the lack of randomization, the six patients under oral finasteride were assigned to the experimental group. Improvements in these patients compared to those receiving only AAPE^® showed no statistical difference, but this fact could be explained by random errors caused by the small sample size. Authors did not address this confounder, either by randomizing the sample or by giving subjects under finasteride a washout period. This was probably the greatest flaw of this study.

Stromal vascular fraction

Study 6 described the obtention process of SVF, characterized by the SVF samples, and analyzed ADSC viability. However, the volume of SVF and number of ADSCs that each subject received were not reported.

The correlation between the improvement in density and diameter, the reduction in pull test numbers, and the general increase of density in global photographs clearly strengthens the study’s internal validity. Unfortunately, long-term changes were not studied.

Study 7 had strict exclusion criteria and included basal AGA stages. TrichoScan^® with shaving and area tattooing provided objective data and the standardization of injection depth and volume guaranteed a homogenous intervention. The small sample was a limitation considering the variability in AGA stages, as was the fact that the single control subject did not allow for a statistical comparison between SVF and placebo.

Study 8 had a representative sample of 60 subjects, distributed in two homogeneous and statistically comparable groups. The increases in density and thickness on the TrichoScan^® were supported by the improvements seen on the Norwood/Ludwig scales. Major limitations were the short follow-up and the lack of a placebo-controlled group.

Study 9 had strict exclusion criteria. Unfortunately, the sample was small, all patients had concomitant AA, in undetermined stages, and phototrichograms were not performed on tattooed areas. These factors reduced the reliability of the changes observed.

Interestingly, cell viability and SVF apoptotic index were studied, and total volume was standardized. Likewise, the number of cells injected per patient was estimated, and doctors ensured that the time from lipoaspiration to SVF injection did not exceed 3 h so as to maximize cell viability. It was the only study using SVF to take this into consideration.

Study 10’s sample was very small and had great variability in AGA severity, which may have led to random errors. Due to design limitations, it was impossible to establish a causal relationship between the injection of SVF and the changes observed, since SVF and PRP were assessed in combination, and there is evidence that PRP alone improves hair growth in AGA.[41,42,43,44]

Study 11 was a very small case series with only one female patient and no baseline AGA stages reported.

Results of Studies 2 and 8 suggest the superiority of ADSCs over PRP on AGA. Despite PRP and ADSCs having a similar growth factor profile,[62] ADSCs have additional regenerative and protective effects on the follicle,[50,51,52,53] which might explain these results. As for differences between ADSC-CM and ADSC injection, both seemed to produce similar changes. ADSC-CM, however, can be used in multiple sessions with no need for repeated liposuctions or cryopreservation of the fat graft. This may be a suitable maintenance therapy, given the chronic course of AGA.

Regarding nanofat and SVF, no studies have as yet compared their efficacy at inducing hair growth, although nanofat should not be as effective, provided that it contains 8–20-fold fewer ADSCs than SVF.[72,73]

Respect to the efficacy of ADSCs/ADSC-CM on AA, the present review did not find enough body of evidence to extract conclusions. There were no studies assessing exclusively AA, rather AA was a concomitant finding in patients with AGA. The inclusion of cases affected by both diseases in the same group of study was misguided, since the natural history of AA, as opposed to that of AGA, can lead to spontaneous remissions, and therefore, create confounding.

Clinical trials assessing the efficacy of ADSC-based therapies on reversible alopecias should exclude patients with concomitant hair disease, as well as those who recently received any treatment for hair loss, to limit confounding. Monitoring changes in disease stages, as done in Study 8, is fundamental so that an accurate indication can be established. Patient satisfaction and phototrichograms/TrichoScan^® on a shaved, marked area should be considered as complementary assessment tools to study correlations between subjective and objective changes that increase the reliability of results. In trials studying SVF and nanofat, description of the obtention process, together with cell counting and characterization become indispensable to ensure their reproducibility. Besides, although there is currently no consensus on optimal doses and depth of injection in infiltrative therapies for hair stimulation,[74,75] the description of the injection technique and volume might prove useful.

CONCLUSION

The quasi-experimental designs had poor internal validity, and therefore, there are no strong grounds for a relationship of causality between the interventions and the changes seen. Studies with no exclusion criteria were vulnerable to selection bias, and confounders were present due to the lack of randomization, except in Study 8. Studies with a small sample size were open to random errors. Despite the low level of evidence gathered, the abundance of positive results supports the hypothesis that ADSC-based therapies may have therapeutic value in patients with AGA. This should encourage researchers to further develop larger randomized controlled trials that foresee and overcome the methodological pitfalls outlined in the present literature, with well-defined and easy-to-replicate designs.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.