Available via license: CC BY 4.0
Content may be subject to copyright.
Citation: Schipper, J.A.M.; van
Laarhoven, C.J.H.C.M.; Schepers,
R.H.; Tuin, A.J.; Harmsen, M.C.;
Spijkervet, F.K.L.; Jansma, J.; van
Dongen, J.A. Mechanical
Fractionation of Adipose Tissue—A
Scoping Review of Procedures to
Obtain Stromal Vascular Fraction.
Bioengineering 2023,10, 1175.
https://doi.org/10.3390/
bioengineering10101175
Academic Editor: Cornelia Kasper
Received: 30 August 2023
Revised: 26 September 2023
Accepted: 28 September 2023
Published: 9 October 2023
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
bioengineering
Review
Mechanical Fractionation of Adipose Tissue—A Scoping
Review of Procedures to Obtain Stromal Vascular Fraction
Jan Aart M. Schipper 1, *,† , Constance J. H. C. M. van Laarhoven 2 ,† , Rutger H. Schepers 1, A. Jorien Tuin 1,
Marco C. Harmsen 3, Fred K. L. Spijkervet 1, Johan Jansma 1and Joris A. van Dongen 3,4
1
Department of Oral & Maxillofacial Surgery, University Medical Center Groningen, University of Groningen,
9713 Groningen, The Netherlands
2Department of Plastic, Reconstructive and Hand Surgery, Catharina Ziekenhuis Eindhoven,
5623 Eindhoven, The Netherlands
3
Department of Pathology & Medical Biology, University Medical Center Groningen, University of Groningen,
9712 Groningen, The Netherlands
4Department of Plastic, Reconstructive and Hand Surgery, University Medical Center Utrecht,
Utrecht University, 3584 Utrecht, The Netherlands
*Correspondence: j.a.m.schipper@umcg.nl; Tel.: +31-50-361-3840
†These authors contributed equally to this work.
Abstract:
Clinical indications for adipose tissue therapy are expanding towards a regenerative-
based approach. Adipose-derived stromal vascular fraction consists of extracellular matrix and all
nonadipocyte cells such as connective tissue cells including fibroblasts, adipose-derived stromal
cells (ASCs) and vascular cells. Tissue stromal vascular fraction (tSVF) is obtained by mechanical
fractionation, forcing adipose tissue through a device with one or more small hole(s) or cutting
blades between syringes. The aim of this scoping review was to assess the efficacy of mechanical
fractionation procedures to obtain tSVF. In addition, we provide an overview of the clinical, that is,
therapeutic, efficacy of tSVF isolated by mechanical fraction on skin rejuvenation, wound healing and
osteoarthritis. Procedures to obtain tissue stromal vascular fraction using mechanical fractionation
and their associated validation data were included for comparison. For clinical outcome comparison,
both animal and human studies that reported results after tSVF injection were included. We catego-
rized mechanical fractionation procedures into filtration (n= 4), centrifugation (n= 8), both filtration
and centrifugation (n= 3) and other methods (n= 3). In total, 1465 patients and 410 animals were
described in the included clinical studies. tSVF seems to have a more positive clinical outcome in
diseases with a high proinflammatory character such as osteoarthritis or (disturbed) wound healing,
in comparison with skin rejuvenation of aging skin. Isolation of tSVF is obtained by disruption
of adipocytes and therefore volume is reduced. Procedures consisting of centrifugation prior to
mechanical fractionation seem to be most effective in volume reduction and thus isolation of tSVF.
tSVF injection seems to be especially beneficial in clinical applications such as osteoarthritis or wound
healing. Clinical application of tSVF appeared to be independent of the preparation procedure, which
indicates that current methods are highly versatile.
Keywords:
tissue stromal vascular fraction; mechanical fractionation; adipose tissue; mechanical
dissociation; adipose-derived stromal cells
1. Introduction
Clinical indications for adipose tissue therapy are expanding [
1
]. Initially, fat grafting
has been widely used to restore volume loss due to trauma, ablative surgery, aging or
congenital defects [
2
]. Later, indications for grafting of fat or components of fat expanded
to more regenerative-based approaches such as improving wound healing or aged skin
as well as treating scars [
3
]. This shift was partly caused by the development of stromal
vascular fraction (SVF) of adipose tissue [
4
]. SVF is the fraction of adipose tissue without
Bioengineering 2023,10, 1175. https://doi.org/10.3390/bioengineering10101175 https://www.mdpi.com/journal/bioengineering
Bioengineering 2023,10, 1175 2 of 34
adipocytes. Adipose-derived SVF consists of an extracellular matrix and all nonadipocyte
cells such as connective tissue cells: fibroblasts, adipose-derived stromal cells (ASCs) and
vascular cells [
5
]. In the past few years, SVF has demonstrated its regenerative action by
stimulating angiogenesis and immunomodulation [6–8].
SVF can be obtained by enzymatic isolation or mechanical fractionation, which yield,
respectively, cellular SVF (cSVF) and tissue SVF (tSVF). cSVF is a single-cell suspension de-
void of both adipocytes and extracellular matrix [
9
]. Mechanical fractionation is performed
by forcing adipose tissue through a device with one or more small hole(s) or cutting blades
between syringes. The holes or blades disrupt adipocytes leaving only the SVF. Subsequent
fractionation, often by centrifugation, separates adipose tissue in an upper oily fraction
from disrupted adipocytes, a middle layer of stromal vascular fraction, a lower liquid
fraction with infiltration fluid, and a pellet of cellular debris. Mechanical fractionation
yields tSVF, in which the original tissue structure is maintained (including extracellular
matrix) but fragmented. The extracellular matrix functions
in vitro
as a scaffold for cells
and a release reservoir of already-bound paracrine factors to improve regenerative pro-
cesses [
10
,
11
]. It is thought mechanical fractionation holds higher therapeutic potential than
enzymatic isolation because of the additional advantages of the intact extracellular matrix
because it is less time consuming to make and less expensive [
5
]. Moreover, enzymes used
for isolation are legally forbidden for clinical application in many countries [12–14].
After fractionation of lipoaspirate, separation methods such as decantation, filtration
or centrifugation are used to separate the different fractions of lipoaspirate. The first devel-
oped procedure was the nanofat procedure [
15
]. The nanofat procedure uses decantated
lipoaspirate, forcing it through a device with one hole (a fractionator) multiple times
between syringes, and then filtrating it to obtain processed lipoaspirate. Later on, other
techniques were combined with centrifugation, for example, respectively, the fractionation
of adipose tissue (FAT) procedure or Lipocube [16,17].
All these fractionation procedures have been reported in both
in vitro
and clinical
studies. Meanwhile, the development of most mechanical fractionation procedures changes
continuously through small procedural optimizations. Due to this abundance in procedures
and associated publications, it is unclear what the differences between these mechanical frac-
tionation procedures are, for which clinical applications these procedures can be used and
whether certain procedures may be preferred over others for specific indications. Efficacy
of these procedures is determined by efficient removal of adipocytes, which can be assessed
by volume reduction and absence of (immuno)histochemical markers of adipocytes. Hence,
this review is conducted to assess the efficacy of mechanical fractionation procedures to
obtain tSVF. In addition, we provide an overview of the clinical, that is, therapeutic, efficacy
of tSVF on skin rejuvenation, wound healing and osteoarthritis.
2. Procedures
In general, three steps can be distinguished in the production of tSVF (Table 1): (1) ex-
cessive liquids such as oil, serum and infiltration liquid have to be removed (in jargon:
drying of fat); (2) adipocytes have to be adequately disrupted by mechanical fractionation,
that is, dissociation through a device with holes or sharp blades (a fractionator) by intersy-
ringe shuffling; (3) processed tissue has to be separated from the damaged adipocyte debris
and their contents, oil. These procedures can be categorized into three groups based on
their use of tissue separation methods: (1) filtration, (2) centrifugation or (3) a combination
of filtration and centrifugation (Table 2).
Bioengineering 2023,10, 1175 3 of 34
Table 1.
Overview of procedures. (Figure was created using BioRender.com, accessed at 24 July 2023).
Bioengineering 2023, 10, x FOR PEER REVIEW 3 of 36
Table 1. Overview of procedures. (Figure was created using BioRender.com, accessed at 24 July
2023)
Step 1
Tissue Separation
Step 2
Tissue Fractionation
Step 3
Tissue Separation
Filtration
Nanofat [15] F (cloth) One-hole fractionator F (cloth)
Nanofat 2.0 [18] F (cloth) One-hole fractionator -
Nanofat cell aggregates [19] F (mesh) One-hole fractionator, 2.4 mm then
1.4 mm F (mesh)
LipocubeNANO [20] D + F Undefined connector F (mesh)
Centrifugation
FAT procedure [16] C + d Three-hole fractionator (1.4 mm) C + d
FAT procedure 2.0 [21] C + d One-hole fractionator (1.4 mm) C + d
Mechanical micronization:
squeeze [22] C + d Slicer with spinning blade C + d
Step 1
Tissue Separation
Step 2
Tissue Fractionation
Step 3
Tissue Separation
Filtration
Nanofat [15] F (cloth) One-hole fractionator F (cloth)
Nanofat 2.0 [18] F (cloth) One-hole fractionator -
Nanofat cell aggregates [19] F (mesh) One-hole fractionator, 2.4 mm
then 1.4 mm F (mesh)
LipocubeNANO [20] D + F Undefined connector F (mesh)
Centrifugation
FAT procedure [16] C + d Three-hole fractionator (1.4 mm) C + d
FAT procedure 2.0 [21] C + d One-hole fractionator (1.4 mm) C + d
Mechanical micronization:
squeeze [22]C + d Slicer with spinning blade C + d
Lipocube [17,23] D Through blade grid
(1000–750–500 µm)
Add buffer + incubation
C
Centrifuge-modified
nanofat [24]C + d Fractionator
Evo-modified nanofat [24] C + d Fractionator
ARAT [25] C + d Mesh (4000–2400–1200–600–
400–200–100 µm)
MEST [25] C + d Mesh (4000–2400–1200–600–
400–200–100 µm) C
Bioengineering 2023,10, 1175 4 of 34
Table 1. Cont.
Step 1
Tissue Separation
Step 2
Tissue Fractionation
Step 3
Tissue Separation
Centrifugation and filtration
SVF gel [26] d + C + d + collect oil One-hole fractionator (1.4 mm)
F (mesh)
Add oil
C+d
Mechanical micronization:
Emulsification [22]C + d Three-hole fractionator (1.4 mm) C+d
F (mesh)
Supercharge-modified
nanofat [24](1) F (mesh) + C (2) fractionator Add (1) + (2)
Table 2. Detailed overview of processing steps of the procedures.
Step 1
Tissue Separation
Step 2
Tissue Fractionation
Step 3
Tissue Separation
Reduction in
Volume (%) *
Filtration
Nanofat [15]Filtering over nylon cloth
500 µm
30×through one-hole
(undefined size) connector
between 10 cc syringes
Filtering over nylon
cloth 500 µmNR
Nanofat 2.0 [18]Filtering over nylon cloth
500 µm
30×through one-hole
(undefined size) connector - NR
Nanofat cell
aggregates [19,20]
Fluids expelled by manual
pressure through filter
device (Fat press, Tulip
medical; undefined
mesh size)
1. 30×through one-hole
2.4 mm
2. 30×through one-hole
1.2 mm
between 20 cc syringes
Filtering through
600 µm to 400 µm
mesh screen
15%
LipocubeNANO: [20]
Decantation
1×through undefined
filter (port 1)
10×between undefined
holes (port 2 and 3)
1×through a
500 µm filter
(port 4)
NR
Centrifugation
FAT procedure [16]
Centrifugation (3000 rpm;
radius 9.5 cm; 2.5 min)
Oil + liquid discarded
30×through three-hole
1.4 mm
Centrifugation
(3000 rpm; radius
9.5 cm; 2.5 min)
Oil+liquid
discarded
90% [83–93%]
FAT procedure
2.0 [21]
Centrifugation (3000 rpm;
radius 9.5 cm; 2.5 min)
Oil + liquid discarded
30×through one-hole
1.4 mm
Centrifugation
(3000 rpm; radius
9.5 cm; 2.5 min)
Oil + liquid
discarded
89 ±4%
Mechanical
micronization:
squeeze [22]
Centrifugation (1200×g
for 3 min)
Oil + liquid discarded
Automated slicer with
a spinning sharp blade
Centrifugation
(1200×gfor 3 min)
Oil + liquid
discarded
52 ±6% **
Lipocube [17,23] Decantation
10×through blade grid
1000 µm holes
10×through blade grid
750 µm holes
10×through blade grid
500 µm holes
between 20 cc
metallic pistons
Add Ca-Mg
balanced buffer
solution (ratio 1:3)
Incubation 10 min
Centrifugation
(2000
×
gfor 10 min)
Collect the pellet
and resuspend
NR
Bioengineering 2023,10, 1175 5 of 34
Table 2. Cont.
Step 1
Tissue Separation
Step 2
Tissue Fractionation
Step 3
Tissue Separation
Reduction in
Volume (%) *
Centrifugation
Centrifuge-modified
nanofat [24]
Centrifugation (1300 rpm
for 10 min) and fat is
collected (pellet
is discarded)
30×through Luer lock
connector (undefined size) - NR
Evo-modified
nanofat [24]
Centrifugation (80 RPM ×
3 min) and fat is collected
(pellet was discarded)
30×through Luer lock
connector (undefined size) - NR
ARAT [25]
Centrifugation (500×g
for 2 min)
Lower liquid layer
was discarded
Shuffling 10 mL/s between
10 cc or 20 cc syringes
with a blade mesh
in between:
1. 4000 µm
2. 2400 µm
3. 1200 µm
4. 600 µm
5. 400 µm
6. 200 µm
7. 100 µm
- NR
MEST [25] Same as ARAT Same as ARAT
Centrifugation
(1200 g for 6 min)
and bottom two
layers are used
(Stromal cell
solution and
stromal cell
aggregate)
NR
Centrifugation and filtration
SVF gel [26]
Liquid discarded
Centrifugation (1200×g
for 3 min)
Liquid discarded
Collect oil layer
0.5–5 min through one-hole
Luer lock connector 1.4 mm
*** between syringes
(undefined size)
Filtering over mesh
filter 500-µm
Add 0.5 mL of
collected oil and
mix by shifting
3–5×
Centrifugation
(2000×gfor 3 min)
Discard oil layer
85–90%
Mechanical
micronization:
Emulsification [22]
Centrifugation (1200×g
for 3 min)
Oil+liquid discarded
30×through three-hole
1.4 mm Luer lock connector
between 2.5 cc syringes
Centrifugation
(1200×gfor 3 min)
Oil discarded
Filtering over mesh
filter 500 µm
(tissue in the filter)
61 ±5% **
(fluid after the filter)
90 ±3% **
Supercharge-
modified
nanofat [24]
Lipoaspirate divided in
two parts
Part 1
Automatic filtration
(120 µm filter)
Centrifugation (1300 rpm
for 10 min)
Pellet was collected
Part 2
30×through Luer lock
connector between
10 cc syringes
Part 1
is then added
to part 2 NR
Bioengineering 2023,10, 1175 6 of 34
Table 2. Cont.
Step 1
Tissue Separation
Step 2
Tissue Fractionation
Step 3
Tissue Separation
Reduction in
Volume (%) *
Other methods
Emulsified fat by
An et al. [27]
Washed with
phosphate-buffered
saline (PBS)
Minced with scissors
30×through three-hole
1.0mm connector between
2 cc syringes
NR
tSVF gel by
Wang et al. [28]
Washed with saline
Mincing with scissors
10 mL/s for 1 min between
syringes (undefined size)
Centrifugation
(2000×gfor 3 min)
Oil+liquid
discarded
NR
ECM/tSVF gel by
Li et al. [29]Mincing with scissors
90×through one-hole
2.0 mm connector between
10 cc syringes
Centrifugation
(2000×gfor 3 min)
Oil+liquid
discarded
NR
±
SD (range) NR = not reported. * End volume compared relatively to starting volume (start volume 1.0 and end
volume is a fraction of this). ** Only recorded relative to centrifuged fat. *** In their first article, Yao et al. mention
a 2.4 mm connector; however, in their follow up studies, a 1.4 mm connector is reported.
tSVF is characterized by (1) absence of adipocytes determined by (immuno)histochemical
staining of adipocyte markers and volume reduction, (2) presence of extracellular matrix by
(immune)histochemical staining and (3) tSVF composition based on expression of cell-specific
cluster of differentiation (CD) surface markers (Tables 3–5).
Table 3.
Overview of reported validation data. (Figure was created using BioRender.com, accessed at
24 July 2023).
Bioengineering 2023, 10, x FOR PEER REVIEW 7 of 36
tSVF is characterized by (1) absence of adipocytes determined by
(immuno)histochemical staining of adipocyte markers and volume reduction, (2) presence
of extracellular matrix by (immune)histochemical staining and (3) tSVF composition
based on expression of cell-specific cluster of differentiation (CD) surface markers. (Tables
3–5)
Table 3. Overview of reported validation data. (Figure was created using BioRender.com, accessed
at 24 July 2023)
Adipocytes Removal
ECM
Presence
SVF Composition, Cultured ASC Phenotype and
Characterization
Adipocytes Volume ECM
ASC
Phenotype
SVF
Composition CFU Assay Multilineage
Differentiation
Filtration
Nanofat - - -
✓ ✓ - Adipo+
Nanofat 2.0 - - - ✓ - - -
Nanofat cell
aggregates - ✓ - - - - Adipo+
LipocubeNAN
O - - - - ✓ - Adipo+
Centrifugation
FAT procedure Immunohistochemist
ry ✓ Histology ✓ ✓ * ✓ * Adipo+ Osteo+
SMC+
FAT procedure
2.0
Immunohistochemist
ry ✓ Histology - ✓ ✓ -
Mechanical
micronization:
s
queeze
Confocal microscopy ✓ SEM - ✓ - -
Lipocube - - -
✓ ✓ ✓ -
Centrifuge-
modified
nanofat
- - - - ✓ - -
Evo-modified
nanofat - - - -
✓ - -
ARAT Histology - - - ✓ ** ✓ ** -
MEST Histology - - - ✓ ** ✓ ** -
Adipocytes Removal ECM
Presence
SVF Composition, Cultured ASC Phenotype
and Characterization
Adipocytes Volume ECM ASC
Phenotype
SVF Com-
position
CFU
Assay
Multilineage
Differentiation
Filtration
Nanofat - - - X X - Adipo+
Nanofat 2.0 - - - X- - -
Nanofat cell
aggregates -X- - - - Adipo+
LipocubeNANO - - - - X- Adipo+
Centrifugation
FAT procedure Immunohistochemistry XHistology X X *X*Adipo+ Osteo+
SMC+
Bioengineering 2023,10, 1175 7 of 34
Table 3. Cont.
Adipocytes Removal ECM
Presence
SVF Composition, Cultured ASC Phenotype
and Characterization
Adipocytes Volume ECM ASC
Phenotype
SVF Com-
position
CFU
Assay
Multilineage
Differentiation
Centrifugation
FAT procedure 2.0 Immunohistochemistry XHistology - X X -
Mechanical
micronization:
squeeze
Confocal microscopy XSEM - X- -
Lipocube - - - XXX -
Centrifuge-modified
nanofat - - - - X- -
Evo-modified
nanofat - - - - X- -
ARAT Histology - - - X** X** -
MEST Histology - - - X** X** -
Centrifugation and filtration
SVF gel Confocal microscopy XSEM - X** - Adipo+ Osteo+
Chondro+
Mechanical
micronization:
emulsification
Confocal microscopy XSEM - X- -
Supercharge-
modified
nanofat
- - - - X- -
Legend: SEM = scanning electron microscopy. * data were reported in the FAT 2.0 paper. ** no quantification
was provided.
Table 4. Tissue validation data of procedures.
Histology/Immunohistochemistry Viability Other
Filtration
Nanofat [15] -
Fluorescence live/
dead staining:
no viable adipocytes
are visible
-
Nanofat 2.0 [18]
Oil-red O staining:
Showed marked damage at
cellular level.
- -
Nanofat cell
aggregates [19,20]---
LipocubeNANO: [20] - - -
Centrifugation
FAT procedure [16]
Quantified
(immuno)histochemistry:
Alpha-SMA: 6.2% ±5.5
vWF 7.0% ±4.2
Masson’s Trichrome SVF more
than control (no quantification)
Fluorescence live/
dead staining:
100%
-
FAT procedure one
hole [21]
Quantified
immunohistochemistry:
Alpha-SMA: 0.83% ±0.33
Perilipin A: devoid of adipocytes
More collagen than in control
- -
Bioengineering 2023,10, 1175 8 of 34
Table 4. Cont.
Histology/Immunohistochemistry Viability Other
Centrifugation
Mechanical
micronization:
squeeze [22]
BODIPY/Lectin/
Hoechst staining:
Showed structural damage
compared to normal fat and
exhibited occasional aggregations
of capillary fragments.
Perilipin/Lectin/
Hoechst staining:
Irregular-sized/shaped
adipocytes, and lipid droplets and
fragmented capillaries. The
irregular-sized/shaped
adipocytes seemed to be damaged
or dead, although they still
expressed perilipin.
-Scanning electron microscopy:
Damaged adipocytes.
Lipocube [17,23] - - -
Centrifuge-modified
nanofat [24]---
Evo-modified
nanofat [24]---
ARAT/MEST [25]
HE staining after 4000 micron,
2400 micron, 1200 micron,
600 micron, 400 micron and
200 micron adinizing:
Intact adipocytes could be seen
after all adinizing sessions.
-
Secretome after explant
culture: (pg/mL):
VEGF-A: 43.52 ±12.21
EGF-A: 16.44 ±2.67
FGF-2 8519.31 ±3122.42
PDGF: 64.60 ±21.43
NGF: 26.12 ±14.78
TGFB1: 840.94 ±115.77
Anti-inflammatory IL-10:
246.77 ±116.86
IL-1ra: 417.21 ±211.37
Inflammatory IFNg: 2.20 ±1.85
IL-1b: 1221.44 ±664.37
IL-6: 17,338.21 ±3224.60
TNFa: 68.12 ±21.44
Centrifugation and filtration
SVF gel [26]
Confocal Microscopy using
Hoechst(blue)/Lectin(red)/BODIPY(yellow):
After mechanical process multiple
lipid droplets. After
centrifugation most of the free
lipid drops were discarded
leaving very few flat,
fragmented adipocytes.
Also, density of vessel-associated
connective tissue increased.
-
Scanning electron microscopy:
Level of fragmentation of ECM
increased with the duration of
processing time (p< 0.05 for
5 min processing).
Bioengineering 2023,10, 1175 9 of 34
Table 4. Cont.
Histology/Immunohistochemistry Viability Other
Centrifugation and filtration
Mechanical
micronization:
Emulsification [22]
BODIPY/Lectin/
Hoechst staining:
Showed structural damage
compared to normal fat and
exhibited occasional aggregations
of capillary fragments.
Perilipin/Lectin/
Hoechst staining:
Irregular-sized/shaped
adipocytes, and lipid droplets and
fragmented capillaries. The
irregular-sized/shaped
adipocytes seemed to be damaged
or dead, although they still
expressed perilipin.
-
Scanning electron microscopy:
Filtrated fluid of emulsified fat
showed more damaged
adipocytes compared to
centrifuged fat. Filtrated fluid of
emulsified fat was filled with
extracellular matrix
fragments and
contained few intact adipocytes.
Supercharge-modified
nanofat [24]---
Other methods
Emulsified fat by
An et al. [27]---
tSVF gel by
Wang et al. [28]---
ECM/tSVF gel by
Li et al. [29]---
±SD (range).
Table 5. Validation data of procedures after digestion with collagenase.
Viability Cell Number Culture/
Differentiation
Colony
Formation
Assay
ASC
Phenotype
SVF
Composition Other
Filtration
Nanofat [15]
Fluorescence
microscopy:
Only very few
dead cells.
1.975 ×
106/100 mL
lipoaspirate
CD34+ 1 ×
105/100 mL
Adipogenic+ -
CD44
96.32% ±1.32
CD90
67.38% ±2.45
CD105
82.65% ±2.07
CD14
6.67% ±1.99
CD34
20.01% ±1.63
CD45
13.68% ±2.09
*
CD34+ 0.1 to
0.2 ×106
cells/100 mL
lipoaspirate
(4.5–6.5%)
-
Nanofat 2.0
[18]- - - -
CD44
98.88% ±0.71
CD90
65.58% ±2.95
CD105
75.83% ±2.88
CD14
5.45 ±1.87
CD34
14.86% ±2.09
CD45
3.45% ±2.72
-
MTT assay:
No difference
between micro-
fat/nanofat/
nanofat 2.0
Bioengineering 2023,10, 1175 10 of 34
Table 5. Cont.
Viability Cell Number Culture/
Differentiation
Colony
Formation
Assay
ASC
Phenotype
SVF
Composition Other
Filtration
Nanofat cell
aggregates
[19,20]
Image
cytometry:
76.80%
-
Muse flow
cytometer:
96.05%
DNA
quantification:
7.3 million
cells/g
Cell yield:
6.63 million
cells/g
lipoaspirate
Muse flow
cytometer:
1.44 ×106/cc
Adipogenic+ - - -
Gene
expression
profiles
(mRNA levels):
PPAR2 7.1-fold
higher than
NanoTransfer
Adiponectin
1.7-fold higher
than
NanoTransfer
LipocubeNANO:
[20]
Muse flow
cytometer:
96.75%
Muse flow
cytometer:
2.24 ×106/mL
Adipogenic+ - -
CD45−
CD90+ 7.92%
CD73+, CD90+
37.29%
CD45−CD31+
11.99%
CD45+ CD14+
2.43%
CD13+ 42.04%
CD73+ 53.5%
(extracted from
figure)
CD90+ 55.82%
CD146+ 53.2%
CD34+ 18.84%
-
Centrifugation
FAT procedure
[16]-
Bürker Turk
counting
chamber:
2.7 ×106±1.1
cells/mL
Adipogenic+
Osteogenic+
Smooth muscle
cell+
1.29% ±
0.045 *
CD90 CD31−
CD45−99.8%
±0.2; CD29
99.8% ±0.2
CD44 99.0% ±
0.7 CD105
95.9% ±4.5
CD45−CD90+
CD105+
41.4% ±16.5%
CD31+ CD34+
12.0% ±4.5%
CD45+ CD34−
5.3% ±
3.6% CD34+/
−
CD31
−
CD146+
0.3% ±0.3%
CD45+ CD34+
0.1 ±0.2%
CD34 bright
CD31−
CD146−
uncountably
low *
-
FAT procedure
one hole [21]-
Bürker Turk
counting
chamber: 2.35
×106±0.30
-1.29% ±
0.038 -
CD45−CD90+
CD105+
44.9% ±18.2%
CD31+ CD34+
19.1% ±2.3%
CD45+ CD34−
5.3% ±3.6%
CD34+/−
CD31−
CD146+ 0.5%
±0.5% CD45+
CD34+ 0.2% ±
0.3% CD34
bright CD31−
CD146−
uncountably
low
-
Bioengineering 2023,10, 1175 11 of 34
Table 5. Cont.
Viability Cell Number Culture/
Differentiation
Colony
Formation
Assay
ASC
Phenotype
SVF
Composition Other
Centrifugation
Mechanical
micronization:
squeeze [22]
Nucleocounter
NC-100:
89.9 ±4.6
percent
-
Culture:
After 1 week
normalized
number of
cultured
adipose-derived
stromal cells 1.1
±0.1 ×106
- -
Number of
cells/1mL of
fat
CD45−
CD31−
CD34+
1.5 ×105
CD45−
CD31+
CD34−
1.1 ×105
Mass fraction:
75% ECM
25%
adipocytes
Lipocube
[17,23]
85.82% ±
5.74% 97.55%
**
1.34 ×106/mL
±1.69
0.94 ×106**
-
Cell
proliferation
test:
On day 7,
more cluster
formation in
mechani-
cally
digested SVF.
A490 value
after 72 h
201 ±0.1
(p≤0.05)
CD90 11.39%
CD44 21.45%
CD105 9.0%
CD73 6.16%
CD45−
CD73+ 42.37%
CD90+ CD73+
52.08% CD45−
CD31+ 21.06%
CD45+
CD14+ 7.28%
Gene
expression
analysis:
PPAR2/
adiponectin
1.43- and
1.32-fold
higher in
mechanically
digested SVF;
Col1A PCR
test: 12.5 [1.2]
×
10
6
;
p≤0.05
,
(1.5-fold higher
compared to
enzymatically
digested SVF)
Centrifuge-
modified
nanofat [24]
-
Hemocytometer:
53,334 ±8000
nucleated
cells/mL
- - -
CD44
65.6% ±9.1%
CD90
59.7% ±6.9%
CD73
27.2% ±4.8%
CD34
55.3% ±7.5%
CD31
21.5% ±5.1%
CD146 19.7%
±3.9%
-
Evo-modified
nanofat [24]-
Hemocytometer:
125,000 ±
12,000
nucleated
cells/mL
- - -
CD44
68.3% ±7.8%
CD90
56.9% ±6.1%
CD73
32.3% ±5.5%
CD34
50.4% ±5.2%
CD31
17.8% ±4.8%
CD146
20.7% ±5.1%
-
ARAT/MEST
[25]
Flow
cytometry:
IP 1 94% ±2
IP 2 93% ±2
IP 3 93% ±2
IP 4 91% ±4
***
Flow
cytometry:
IP 1 0.91
106/mL
IP 2 0.88
106/mL
IP 3 1.61
106/mL
IP 4 1.47
106/mL
***
-
CFU-F assay:
Results not
reported
-Quantification
not reported -
Bioengineering 2023,10, 1175 12 of 34
Table 5. Cont.
Viability Cell Number Culture/
Differentiation
Colony
Formation
Assay
ASC
Phenotype
SVF
Composition Other
Centrifugation and filtration
SVF gel [26] -
0.5-min SVF
gel: 2.7
±
0.3
×
105cells/mL
1 min SVF gel:
4.1 ±0.3
×105
cells/ml
Adipogenic+
Osteogenic+
Chondrogenic+
- -
Quantification
not reported,
only figure.
-
Mechanical
micronization:
emulsification
[22]
Nucleocounter
NC-100:
residual tissue
90.6% ±2.8%,
filtrated fluid
39.3% ±9.1%
-
After 1 week of
culture, number
of cells in
residual tissue of
emulsified fat
(5.1 ±0.7 ×105)
- -
Number of
cells/1 mL of
fat
CD45−
CD31−
CD34+
1.4 ×105
CD45−
CD31+
CD34−
0.8 ×105
-
Supercharge-
modified
nanofat [24]
-
Hemocytometer:
200,000 ±
15,000
nucleated
cells/mL
- - -
CD44
71.2% ±8.0%
CD90
62.8% ±7.2%
CD73
30.1% ±5.4%
CD34
58.1% ±6.3%
CD31
19.9% ±4.4%
CD146
22.1% ±4.5%
-
Other methods
Emulsified fat
by An et al.
[27]
Trypan blue
staining: 58.2%
Nucleocounter
NC-100:
4.53 ×106- - -
CD45−
CD34+
12.40% ±0.86%
-
tSVF gel by
Wang et al. [
28
]
7-AAD
staining: 80% - - - -
CD34+
CD31−
CD45−64%
CD34+
CD31+
CD45−28%
-
ECM/tSVF gel
by Li et al. [29]- -
Chondrogenic+
Osteogenic+
Adipogenic+
-
CD29+ 97.3%
CD90+ 98.7%
CD105+ 99.7%
CD34+ 1.3%
CD45+ 1.2%
- -
* Quantification was derived from another manuscript describing these data. ** Two studies reported validation
data of the same procedure. *** Results were stratified based on reported indication protocols (IP).
(1) Mechanical fractionation of tSVF can only be obtained by disruption of adipocytes
through a fractionator. Their contents that is triglycerides (oil) are released after disrup-
tion of the cell membrane. Adipose tissue consists in volume of approximately 90%
adipocytes [
30
], although the volume of adipocytes also depends on donor adipocyte
cell hypertrophy which is influenced by factors such as body mass index [
31
]. Mechanical
fractionation should therefore result in a ratio of approximately 9:1 of oil:tSVF. Hence,
when most of the oil from the adipocytes is removed, the volume of the end product that
is tSVF should be 90% less than the starting volume of lipoaspirate. The end volume of
this fraction should therefore be reported to properly assess the efficacy of a mechanical
fractionation procedure.
Bioengineering 2023,10, 1175 13 of 34
(2) The presence of extracellular matrix can be confirmed using histochemical staining
such as Masson’s Trichrome staining.
(3) In 2013, a joint statement of the International Federation of Adipose Therapeutics
(IFATS) and International Society for Cellular Therapy (ISCT) was published regarding
the characterization of cultured ASCs by phenotype using flow cytometry and by function
using colony forming units as well as multilineage differentiation [9].
2.1. Filtration Procedures
In total, seven mechanical fractionation procedures are equivalent to the nanofat pro-
cedure [
15
,
18
–
20
,
24
] of which four are classified as filtration procedure using either a nylon
cloth or a filter device [
15
,
18
–
20
]. However, nanofat 2.0 and nanofat cell aggregates are
almost identical to the original nanofat procedure. Nanofat 2.0 lacks a last filtration step and
nanofat cell aggregates use a metal mesh screen instead of a nylon cloth. LipocubeNANO
is the other mechanical fractionation procedure that we categorized as a filtration-only
procedure (without the use of centrifugation).
Three out of the four nanofat procedures classified as filtration procedure do not
mention the end volume of the fraction of processed tSVF [
15
,
18
,
20
] or the number of
adipocytes that are still present in the processed tSVF. The authors of nanofat cell aggregate
reported a volume reduction of only 15%, while centrifugation methods such as the FAT and
SVF gel procedures acquire a volume reduction of 90%. Moreover, almost no histological
evaluation has been performed on tSVF isolated using nanofat procedures, that is, filtration
procedures. Authors of nanofat 2.0 only reported microscopic images using an oil-red O
staining showing a different tissue structure compared to lipoaspirate, although perilipin
is the required cytoplasmatic staining to evaluate whether the adipocyte cell membrane
is intact. Perilipin-positive cells would therefore suggest there are still intact adipocytes
present. The authors of the original nanofat procedure only performed viability staining
and showed that no viable adipocytes were present [
15
]. Due to the minimal amount of data
available, it remains unclear whether small modifications to the original nanofat procedure
as a filtration method resulted in improvements in the purity of tSVF.
In contrast to centrifugation, which separates the oily from the aqueous fraction,
filtration retains part of the adipocytes and therefore reduces the efficiency of the generation
of tSVF. A recent study found that filtration of lipoaspirate reduced the number of SVF
cells and extracellular matrix content compared non-filtrated lipoaspirate [
32
]. No studies
reported the measurement of extracellular matrix in the validation data of these filtration
procedures during the nanofat procedure.
None of the filtration studies reported CFU-F assay and only three studies reported
adipogenic differentiation capability [
15
,
19
,
20
], while at least three differentiation lineages
are warranted. Only the study describing the nanofat 2.0 procedure reported an ASC
phenotype after the third passage. SVF composition based on CD marker expression has
been evaluated in multiple studies; however, only the authors of LipocubeNANO used
combined CD marker expression to determine different cell populations [
20
]. Multiple CD
markers, both positive and negative, are needed to define different cell populations because
single CD markers are expressed by multiple cell types. It should be noted that in none
of the studies did the authors use adipocyte-specific markers, yet these cells are readily
scored by virtue of their size during FACS analyses. The use of too few distinct CD markers
results in an over- or underestimated number of a certain cell type. SVF cells can be di-
vided in two large cell populations: blood-derived (CD45+) and adipose-derived (CD45
−
).
Adipose-derived populations can be divided into endothelial-like cell types (CD31+) and
stromal cell types (CD31
−
). The authors of LipocubeNANO defined the CD45
−
/CD31+ as
endothelial cells (11.99%), the CD45
−
/CD90+ as well as CD73+/CD90+ populations as
ASCs (combined 45.21%) and the CD45+/CD14+ population as monocytes/macrophages
(2.43%) [
20
]. The actual number of ASCs is probably overestimated because both CD73+
and CD90+ (aka Thy-1) are expressed by endothelial cells (CD90 only on activated ECs)
as well, while CD90 is also abundantly expressed on other connective tissue cells such
Bioengineering 2023,10, 1175 14 of 34
as fibroblasts. The estimation of the number of ASCs would be more accurate if CD31+
was added to CD73+/CD90+ to differentiate between endothelial cells (CD31+) and ASCs
(CD31−). Of note, monocytes express CD31 too, albeit at lower levels than ECs.
2.2. Centrifugation Procedures
Mechanical fractionation was performed in four out of eight procedures using a Luer
lock fractionator/sizing transfer [
16
,
21
,
24
] and in four procedures using a blade-like frac-
tionator [
17
,
22
,
23
,
25
,
33
]. Centrifugation is a frequently used step in the fractionation of
adipose tissue to obtain SVF. Seven procedures use centrifugation as a first separation
step and six procedures use centrifugation as a final separation step [
16
,
17
,
21
–
23
,
25
,
33
].
Centrifugation as a final separation step results in several fractions: an upper oily layer,
a middle tissue layer rich in cells and with an extracellular matrix, a lower liquid layer
consisting of serum and infiltration fluid, and a pellet with debris and cells not bound to the
extracellular matrix (Table 1). Different authors use different layers for their end-product:
Lipocube and MEST use the pellet layer, but mechanical micronization—squeeze and the
FAT procedure use the middle tissue layer. Centrifuge-modified nanofat and evo-modified
nanofat use centrifugation prior to intersyringe processing and are therefore more compara-
ble to the centrifugation procedures than the original nanofat procedure, which is classified
as a filtration procedure [24].
Only four of eight studies reported the end volume (52% to 90% reduction) [
16
,
21
,
22
].
Volume reduction differs significantly between the types of fractionators. The authors of
the FAT procedure used a Luer lock one-hole 1.4 mm or three-hole 1.4 mm fractionator and
reduced the lipoaspirate by 90% in volume. Immunohistochemistry showed a reduced
number of adipocytes using a perilipin staining, which resulted in a significant enrichment
of small vessels and the extracellular matrix. Mechanical micronization—squeeze used an
automated device with sharp rotating blades in a syringe and reduced lipoaspirate by 50%
in volume. Sharp propellor-like blades in the piston spin electronically, moving up and
down twice [
22
]. Centrifugation force was 956
×
gfor 2.5 min for the FAT procedure and
1200
×
gfor 3 min for the mechanical micronization procedure. While the FAT procedure
resulted in tSVF that was virtually devoid of adipocytes [
16
], the authors confirmed the
presence of adipocytes in the mechanical micronization—squeeze procedure with perilipin
staining; however, no quantification was provided [
22
]. This large difference in tissue
volume reduction indicates that the one- or three-hole fractionator of the FAT procedure
more efficiently disrupts adipocytes than the blades of the mechanical micronization
procedure. However, no quantification of adipocytes or extracellular matrix was provided.
Authors describing the Lipocube and MEST procedures did not report the degree of
volume reduction and used only the cell pellet after centrifugation [
17
,
25
]. This cell pellet
contained 86–91% viable cells in Lipocube and 91–94% viable cells in the MEST procedure.
The adjustable regenerative adipose tissue transfer (ARAT) histologically showed intact
adipocytes; however, no quantification was provided [
25
]. Studies describing Lipocube,
centrifuge-modified nanofat and evo-modified nanofat did not report any histological or
immunohistochemical data. In the procedures without a final centrifugation step, that is,
the centrifuge-modified nanofat procedure or evo-modified nanofat procedure, no end
volume of the used fraction to assess the adequate removal of oil and debris was reported.
The authors of the FAT procedure showed immunohistochemical confirmation that
an extracellular matrix was present in tSVF and showed that there was enrichment of
the extracellular matrix compared to the control lipoaspirate [16,34]. No other procedures
provided validation data regarding extracellular matrix.
Centrifuge-modified nanofat and evo-modified nanofat reported only cell yield (num-
bers) and immunophenotypic characterization using flow cytometry. It is therefore not
clear if these modifications to the original nanofat procedure improved the original nanofat
procedure. Differentiation capacity to prove the capability of cells to differentiate into os-
teogenic, adipogenic or chondrogenic cell lineage and the colony formation assay to assess
clonogenic potential of these cells was only reported in the two FAT procedures [
16
,
21
]. Cul-
Bioengineering 2023,10, 1175 15 of 34
tured ASC phenotype was reported by the FAT procedure after 2–4 passages [
16
]. All other
procedures reported SVF composition [
16
,
17
,
21
–
25
]. SVF composition based on multiple
CD marker expression was reported for only two procedures: tSVF obtained using the FAT
procedure was composed of the following cell populations: CD45
−
/CD90+/CD105+ for
ASCs (41.4%), CD34+/-CD31–/CD146+ for pericytes (0.3%), CD31+/CD34+ for endothelial
cells (12.0%), CD45+/D34
−
for leukocytes (5.3%) and CD45+/CD34+ for hematopoietic
stem cell-like cells (0.1%) [
21
]. The amount of ASCs/pericytes in tSVF obtained using the
FAT procedure is comparable to LipocubeNANO. Although multiple cell populations are
determined, almost 40% of the cell types within isolated tSVF remains unknown. tSVF
derived from the Lipocube showed the following subpopulations: CD45
−
/CD90+ as
well as CD73+/CD90+ for ASCs (combined 94.45%), CD45
−
/CD31
−
for endothelial cells
(21.06%) and CD45+/CD14+ for macrophages/monocytes (7.28%). The total percentage of
distinct types of cells in tSVF obtained using Lipocube is 122.79%, which is impossible [
17
].
Hence, according to these data, 94.45% of the cells in tSVF would account for ASCs, which
is a gross overestimation.
2.3. Procedures Using a Combination of Centrifugation and Filtration
Three procedures used a combination of centrifugation and filtration: SVF gel, me-
chanical micronization emulsification and supercharge-modified nanofat [
22
,
24
,
26
]. SVF
gel and mechanical micronization emulsification both use a first centrifugation step and as
a final step both centrifugation and filtration [22,26].
SVF gel reduced 90% in volume, while mechanical micronization emulsification re-
duced 61% in volume when measuring the tissue in the filter and 90% when measuring the
fluid after filtration [
22
,
26
]. However, the fluid after filtration in the mechanical microniza-
tion procedure consisted of extracellular matrix fragments, very few intact adipocytes and
many dead cells (viability 39.3%
±
9.1). Supercharge-modified nanofat uses both a mesh
filter and centrifugation in which the pellet of the flowthrough is collected and then added
to mechanically processed tissue [
24
]. The diameter of the hole of the fractionator was
not described in supercharge-modified nanofat. The modified nanofat procedures did not
report the percentage of volume reduction. Confocal laser scanning microscopy showed
that after the SVF gel procedure, only a few adipocytes were present while capillaries
were fragmented but the density of vessel-associated connective tissue had increased. This
was a logical consequence of near-complete removal of adipocytes. The FAT procedure
is quite similar to the SVF gel procedure; the only difference is an extra filtration step in
the SVF gel procedure. The SVF obtained after either procedure showed an enrichment in
vasculature while presence of adipocytes was reduced to a bare minimum. The authors
showed that the mechanical micronization emulsification procedure resulted in irregular-
sized and irregular-shaped adipocytes and fragmented capillaries in the tissue before
filtration, but that the fluid resulting after filtration showed few live cells, cell remnants and
ECM fragments (viability 39%). Studies on supercharge-modified nanofat did not report
(immuno)histology or end volume of the product.
The authors of mechanical micronization emulsification and SVF gel confirmed the
presence of the extracellular matrix using scanning electron microscopy. Supercharge-
modified nanofat did not test for extracellular matrix content.
According to the minimal definitions of the IFATS statement, only SVF gel showed dif-
ferentiation capability [
26
]. SVF gel and mechanical micronization emulsification reported
SVF composition, but did not report quantitative data [
22
,
26
]. Authors of supercharge-
modified nanofat reported SVF composition [
24
]. There were no studies that reported CFU
assays or cultured ASC phenotype.
2.4. Studies Describing Direct Comparisons between Procedures
Osinga et al. studied the effects of intersyringe processing with a three-way stopcock
with 2 mm diameter for each hole [
35
]. This study shows that tissue viability was not
affected by intersyringe processing 1, 5 or 30 times. Although, other studies show that
Bioengineering 2023,10, 1175 16 of 34
a smaller diameter of the fractionator increases shear stress on tissue during processing,
which promotes adipocyte disruption resulting in oil formation [
36
,
37
]. To isolate and
condense SVF instead of emulsifying adipose tissue, an additional step of centrifugation
prior to fractionation is necessary [
38
]. A comparison study between nanofat and SVF gel
showed that volume is reduced 20% using the nanofat procedure and 80% in the SVF gel
procedure [
39
]. The number of viable SVF cells was also 10-fold higher in SVF gel compared
to nanofat. The increase in concentration of cells and volume reduction can be explained by
the fact that nanofat lacks two centrifugation steps. The first centrifugation step (instead of
decantation of the nanofat procedure) is necessary to remove excess liquid from the adipose
tissue to ensure adequate disruption of adipocytes upon usage of the disruptor mounted
on syringes [
38
]. The second centrifugation step is essential to separate oil from tSVF, as
described above earlier. Furthermore, the combination of centrifugation and fractionation
adds additional mechanical shear stress to SVF cells in comparison with fractionation alone.
Pallua et al. showed an increased concentration of SVF-associated cells: ASCs (CD31
−
,
CD34+, CD45
−
) and endothelial cells (CD31+, CD34+, CD45
−
) in the FAT procedure
(two times centrifugation) compared to the nanofat procedure (no centrifugation) [
40
].
bFGF, IGF-1, PDGFBB, VEGF-A and MMP-9 expression levels were similar between the
FAT procedure and the nanofat procedure. Banyard et al. confirmed that exposure to
mechanical shear stress by intersyringe processing of the nanofat procedure results in cells
with a higher expression of multiple non-SVF markers in processed lipoaspirate compared
to standard unprocessed lipoaspirate: CD34 (threefold), CD13 (threefold), CD73 (twofold),
CD146 (twofold), CD45 (twofold) and CD31 (twofold) [41].
3. Clinical Applications
As mechanical fractionation methods using intersyringe processing are convenient
and fast, these are broadly applied and also modified by various user groups. Therefore, the
variation in study and application areas is also continuously expanding [
1
]. Only research
with mechanical fractionation methods using intersyringe processing to provide tSVF as
stated above, and use of validated tools in analysis of clinical efficacy post-injection was
discussed for the current overview. Therefore, lipogems was not included, as this method
uses a device with metal balls inside that is manually shaken [
42
]. In addition, MyStem was
excluded, since it did not describe which type of connector is used for processing [43,44].
To improve readability, we categorized the literature in the following (pre)clinical
applications: skin and volume enhancement, wound healing, osteoarthritis and others.
In total, 45 articles (29 human, 16 animal) were eligible according to our search criteria
in which tSVF was mechanically isolated and clinical outcome was reported with validated
tools or histopathological analysis. Four RCTs (14%, 4/29) were conducted, and the majority
of articles was prospective (91%). In total, 1465 patients (median follow-up 12 months (range
0.5–12); Table 6and 410 animals were studied (median follow-up 3 months (range 0.5–6);
Table 7). As study design and outcome measurements varied widely, no direct comparisons
or meta-analysis of the included studies were performed.
Bioengineering 2023,10, 1175 17 of 34
Table 6. Overview of human clinical results in the literature.
Study Year Study
Design NMax. FU
(m)
Categorized
Method %
Additional
Product
Clinical
Endpoints
Clinical
Results *
(Index vs.
Control)
Overall
Result
Skin and volume enhancement
1
van
Dongen
et al.
(2021) [
45
]
2016–
2019 RCT 28 12 FAT
procedure
PRP + tSVF
vs.
PRP +
saline
VISIA,
FACE-Q,
complications
No
superior
result in
skin
quality or
satisfac-
tion.
No major
complica-
tions
+/−
2
Zhang
et al.
(2022) [
46
]
2018–
2020 RCT 63
(34 vs. 29) 12 FAT
procedure %
tSVF
vs.
Coleman’s
Volume
ratio on 3D
imaging,
retention
rate,
satisfaction
5p Likert
Increased
contour
ratio
0.87 ±
0.02 to
0.89 ±
0.03 *,
higher
retention *
41.2 ±
8.4% vs.
32.6 ±
8.8%,
higher sat-
isfaction
(4/5p)
79% vs.
62% *
+
3Cai et al.
(2019) [
47
]
? P 50
(28 vs. 22) 12
FAT
procedure %
(with 2.4 mm
fractionator)
tSVF
vs.
BTXa
Global
Aesthetic
Improve-
ment Scale
(GAIS),
patient
satisfaction,
histological
analysis
(n= 1)
Higher
GIAS and
satisfac-
tion in
high-
grade
wrinkles
group *,
increased
collagen
density
+
4
Wang
et al.
(2021) [
48
]
2017–
2019 P18
(6 vs. 12) 12 FAT
procedure %
tSVF
vs.
PBS
Ultrasonogram,
volume
measure-
ment
Similar
elasticity,
increased
volume of
2.3 ±0.3
mL
+/−
5Ding et al.
(2023) [
49
]
2020–
2021 P 31 15 FAT
procedure %tSVF
Survival,
volume,
GAIS
Reoperation
rate
65.3% ±
6.1
2.2 ±0.8
12.9%
+/−
Bioengineering 2023,10, 1175 18 of 34
Table 6. Cont.
Study Year Study
Design NMax. FU
(m)
Categorized
Method %
Additional
Product
Clinical
Endpoints
Clinical
Results *
(Index vs.
Control)
Overall
Result
Skin and volume enhancement
6Xia et al.
(2022) [
50
]
2017–
2021 P
33
(66
temples)
6FAT
procedure %tSVF
Hollowness
Severity
Rating
Scale,
satisfaction
3p Likert
91%
absent
hollow-
ness,
94%
satisfied
+
7Luo et al.
(2020) [
51
]
2017–
2018 P33 (66
eyes) 13 FAT
procedure %tSVF
GAIS,
depth mea-
surement,
retention
rate
2.5 [0.5]
Improvement
in all
depth
measure-
ments
73 ±10%
+
8Zhao et al.
(2021) [
52
]
2018 P 18 12
FAT
procedure, 1.4
mm %tSVF
Number of
inflamma-
tory lesions,
Investors
Global
Assessment
scale,
biopsies at
1-month FU
Decrease
in lesions,
7.3(2.7) vs.
0.7(0.7) *,
2.5(0.5) vs.
0.6(0.5) *,
decrease
in CD4+ T
cells after
4 weeks
+
9Cao et al.
(2022) [
53
]
2017–
2019 P 13 10 FAT
procedure %tSVF Satisfaction
Improvement
84 ±3 vs.
31 ±3 *
+
10
Liang
et al.
(2018) [
54
]
2014–
2016 P
231
(103 vs.
128)
24 Nanofat
tSVF + PRF
vs.
hyaluronic
acid
VISIA,
SOFT5.5,
satisfaction
Facial
skin
texture
improved
in both
groups *,
higher sat-
isfaction
rate
+
11 Wei et al.
(2017) [
55
]
2014–
2016 P139
(62 vs 77) 24 Nanofat
tSVF + PRF
vs.
Coleman’s
procedure
VISIA,
SOFT5.5,
satisfaction
Skin
quality
improve-
ment *,
higher sat-
isfaction
90% vs.
70% *
+
12
Menkes
et al.
(2020) [
56
]
2018 P 50 6 Nanofat Nanofat +
PRP
Satisfaction
(10p Likert),
biopsies
Improvement
in texture,
elasticity,
glow
Increase
in
collagen
and
elastic
fibers *,
higher
cellularity
and
vascular
density *
+
Bioengineering 2023,10, 1175 19 of 34
Table 6. Cont.
Study Year Study
Design NMax. FU
(m)
Categorized
Method %
Additional
Product
Clinical
Endpoints
Clinical
Results *
(Index vs.
Control)
Overall
Result
Skin and volume enhancement
13
Menkes
et al.
(2021) [
57
]
2017–
2018 P 50 18 Nanofat Nanofat +
PRP
Vaginal
health
index,
Female
Sexual
Distress
Scale
Revised
Improvement
in
VHI,
9.2 ±1.7
vs. 3.4 ±
1.5 *,
FSDS-R,
3.4±3.7
vs.
32.9±9.5*
+
14
Uyulmaz
et al.
(2018) [
58
]
2013–
2016 R 52 5 Nanofat -
Photographs,
satisfaction
(yes/no)
Improvement
in skin ap-
pearance
93%,
satisfactory
result 18%
rater
vs. 92%
patient
+
15 Zhu et al.
(2022) [
59
]
2016–
2020 R103
(58 vs. 48) 9FAT
procedure %
tSVF
(FAT %)
vs.
tSVF
(nanofat)
Clinical
data,
satisfaction
Comparable
improve-
ment
Less reop-
erations
in FAT-
treated
patients*,
higher sat-
isfaction
in FAT-
treated
patients *
+
16
Yao Yao
et al.
(2018) [
60
]
2015–
2017 R
204
(126 vs.
78)
11 tSVF gel
tSVF gel
vs.
Coleman’s
procedure
Photo
analysis,
satisfaction
(5p Likert
scale),
histological
analysis
(n= 1)
Higher
Likert
tSVF *,
lower rate
of 2nd
surgery
tSVF
10.9%
(11/101)
vs. 32.1%
(25/78) *
No cysts,
fibrosis or
calcifica-
tion
+
Bioengineering 2023,10, 1175 20 of 34
Table 6. Cont.
Study Year Study
Design NMax. FU
(m)
Categorized
Method %
Additional
Product
Clinical
Endpoints
Clinical
Results *
(Index vs.
Control)
Overall
Result
Wound healing
17
van
Dongen
et al.
(2022) [
61
]
2016–
2020 RCT 40
(20 vs. 20) 12 FAT
procedure
tSVF
vs.
saline
Histological
biopsies,
POSAS,
photographs,
blinded
analysis
with VAS
Equal
collagen
align-
ment,
depth,
width,
POSAS
patient
14.4 ±7.6
vs. 15.3 ±
9.0,
observer
14.5 ±6.4
vs. 14.6 ±
8.8,
no differ-
ences in
VAS
scores,
low agree-
ment
+/−
18
Abouzaid
et al.
(2022) [
62
]
2019–
2020 RCT 100
(50 vs. 50) 3
Centrifuged-
modified
nanofat %
Coleman’s
fat + tSVF
vs.
conventional
dressings
Photo +
clinical
analysis,
biopsies
Decrease
in
hospital
stay and
reopera-
tion *,
less con-
tractures
*,
rapid
collagen
deposi-
tion
*
+
19 Gu et al.
(2018) [
63
]
2014–
2016 P 20 6 FAT
procedure %-
POSAS and
pho-
tographs,
biopsies
Patient
total
28.8 (1.0)
vs. 12.2
(0.8) *,
observer
total
18.0 (0.7)
vs. 9.2
(0.4) *,
melanin
AOD
basal cell
layer 0.8
(0.1) vs.
0.7 (0.1),
no
difference
in elastic
fibers
+
Bioengineering 2023,10, 1175 21 of 34
Table 6. Cont.
Study Year Study
Design NMax. FU
(m)
Categorized
Method %
Additional
Product
Clinical
Endpoints
Clinical
Results *
(Index vs.
Control)
Overall
Result
Wound healing
20
Bhooshan
et al.
(2019) [
64
]
2015–
2016 P 34 3
Nanofat
(without first
filtration step)
-
POSAS and
pho-
tographs,
aesthetic
result (total
POSAS
score)
Patient
total
14 ±4.4
vs. 27.4 ±
7.5 *,
observer
total
18 ±6.8
vs. 31 ±
8.5 *,
77% good
outcome
+
21
Hung
et al.
(2022) [
65
]
2019 P 6 6 Nanofat Nanofat +
PRP
Pain-VAS,
PROM,
cystoscopy
Improvement
in PROMs
*,
100%
remission
of lesions
+
22
Rageh
et al.
(2021) [
66
]
? P 30 6 Nanofat
Vancouver
scar scale,
biopsies
Lower
VSS
scores in
height
and
pliability
*,
improved
epider-
mal
thickness
*,
increased
collagen
(52%) and
elastic
fibers
(50%) *
+
23
Huang
et al.
(2021) [
67
]
2017–
2020 R 44 12 Nanofat -
FACE-Q,
assessment
of pho-
tographs
Overall
satisfied,
30%
complete
healing,
41%
obvious
improve-
ment,
9% no
effect
+/−
24
Cantarella
et al.
(2019) [
68
]
? Pi 7 6
Centrifuged-
modified
nanofat %-
Videolaryngo-
stroboscopy,
max
phonation
time,
VHI,
EAT10
Improvement
in glottic
closure,
longer
phona-
tion *,
reduction
in VHI *,
improved
swallow-
ing
+
Bioengineering 2023,10, 1175 22 of 34
Table 6. Cont.
Study Year Study
Design NMax. FU
(m)
Categorized
Method %
Additional
Product
Clinical
Endpoints
Clinical
Results *
(Index vs.
Control)
Overall
Result
Wound healing
25
Tenna
et al.
(2017) [
69
]
2014–
2015 P30
(15 vs. 15) 12
Centrifuged-
modified
nanofat %
tSVF + PRP
vs.
tSVF + PRP
+ CO2laser
Ultrasound,
FACE-Q
Improvement
in subcu-
taneous
tissue of
0.67 vs.
0.63 cm,
comparable
FACE-Q
+
26
Deng et al.
(2018) [
70
]
2016–
2017 P20
(10 vs. 10) 0.5 SVF gel
tSVF
vs.
control
Wound
healing rate,
biopsies
35 ±11%
vs. 10 ±
3% *,
decreased
lympho-
cyte
infiltra-
tion,
more *
and
thicker
collagen
deposi-
tion,
more new
vessels *
+
Other indications
27
Stevens
et al.
(2018) [
71
]
2016–
2016 P 10 3 FAT
procedure tSVF + PRP Hair
density
30.7
hairs/cm
2
(range
5–59),
regrowth
observed
+
28
Gutierrez
et al.
(2022) [
72
]
2017–
2019 Pi 19
(9 vs. 10) 12 SVF gel
tSVF + PRP
vs steroid
injection
Skin
elasticity,
VAS,
QoL
(Skindex-
29),
biopsies
No
elasticity
improve-
ment.
Improvement
in symp-
toms, but
not in
pain *.
Improvement
in QoL *,
79.7 ±
33.2 to
59.7 ±
24.9.
Decrease
in all
inflamma-
tory cells
*.
+
29 Sun et al.
(2021) [
73
]
2017–
2018 P 22 18 SVF gel tSVF
Glottis
closure,
GRBAS
voice
quality
Improvement
in vocal
cord
shape and
closure,
improvement
in 19/22
patients *
+
Abbreviations: FU: follow-up (months);
%
: altered categorized mechanical method by the authors; *: statistically
significant, p< 0.05; RCT: randomized controlled trial; P: prospective; R: retrospective; Pi: pilot; FAT: fractionation
Bioengineering 2023,10, 1175 23 of 34
of adipose tissue; PRP: platelet-rich plasma; tSVF: tissue stromal vascular fraction PRF: platelet-rich fibrin; PBS:
phosphate-buffered saline; POSAS: patient and observer scar assessment scale; VAS: visual analogue scale; GRBAS:
grade of dysphonia, roughness, breathiness, authenticity, strain; VHI: Voice Handicap Index questionnaire; EAT10:
Eating Assessment Tool; AOFAS: American Orthopedic Foot and Ankle Society Ankle–Hindfoot Score; VISA-A:
Victorian Institute of Sport Assessment—Achilles; SF-36: Short-Form Health Survey; MIDAS: Migraine Disability
Assessment Score; PGIC: Patient Global Impression of Change.
Table 7. Overview of animal clinical results in the literature.
Study Year Study
Design NMax. FU
(m)
Categorized
Method %
Additional
Product
Clinical
Endpoints
Clinical
Results *
Overall
Result
Skin and volume enhancement
1
Akgul
et al.
(2018) [
74
]
2013–
2015 P14 rats/6
XG 1.5 FAT
procedure %
tSVF
enriched
with
adipocyte
fragments
vs. controls
Histological
biopsies
Viable
adipocyte
architec-
ture,
collagen
accumula-
tion,
CD68 +
CD44+
+
2Zhu et al.
(2021) [
75
]
NR P
60 mice
(15 vs.
45/17
XG)
3FAT
procedure %
tSVF
vs.
control
Histological
biopsies
Dermal
thickness
*
184.4 ±
2.8,
higher
collagen
deposi-
tion,
increased
TGF-b1
and Smad
2 expres-
sion *,
lower
MMP2/9
*,
more fi-
broblasts
+
3Yu et al.
(2018) [
76
]
2017 P
30
(20 vs. 10)
mice/5
XG
3 Nanofat
tSVF
vs.
control
Histological
biopsies,
integrity,
cysts/
vacuoles,
fibrosis,
inflammation,
capillary
density
(CD3+
vessels)
Better
survival
and
morphological
integrity,
3.6 ±0.5
vs. 2.7 ±
0.9 *,
2.6 ±0.7
vs. 3.2 ±
0.8 *,
2.1 ±0.6
vs. 2.9 ±
0.8 *,
2.1 ±0.6
vs. 2.6 ±
0.5 *,
24.6 ±4.7
vs. 10.4 ±
2.9 *
+
Bioengineering 2023,10, 1175 24 of 34
Table 7. Cont.
Study Year Study
Design NMax. FU
(m)
Categorized
Method %
Additional
Product
Clinical
Endpoints
Clinical
Results *
Overall
Result
Skin and volume enhancement
4An et al.
(2020) [
27
]
NR P 24 rats
(16 vs. 8) 1 Other %tSVF
vs.
control
Histological
biopsy,
collagen
AOD,
anti-PCNA
(cell prolif-
eration),
anti-CD31
(vascular-
ization
degree)
Higher
AOD in
1mL SVF
applica-
tion *,
102 ±12
vs. 55 ±8
*,
95 ±4.3
vs. 63 ±
2.7/mm2
*
+
5
VinayKumar
et al.
(2022) [
77
]
2018 P
9 guinea-
pigs
(9 vs. 9)
6 Nanofat tSVF vs.
control
Histological
biopsies,
polarized
light
microscopy
Similar
inflamma-
tory
infiltrate
and
collagen
fiber ori-
entation,
increase
in
collagen
distribu-
tion
*
+/−
6Xu et al.
(2018) [
78
]
NR P
18 mice (6
vs. 12,
10 XG)
2 Nanofat
tSVF vs
control and
enzymatic
tSVF
Histological
biopsies
Increased
dermal
thickness
*,
higher
capillary
density
and epi-
dermal
prolifera-
tion index
*;
high
VEGF,
EFG,
bFGF, IGF
and IL-6 *
+
7Liu et al.
(2021) [
79
]
NR P 12 rabbits 1.5 SVF gel
tSVF and
PRF
vs.
tSVF
Histological
biopsies
Slightly
more
volume
combined
with PRF,
larger
adipocytes
and more
ordered
fibroblast
distribu-
tion
+/−
Bioengineering 2023,10, 1175 25 of 34
Table 7. Cont.
Study Year Study
Design NMax. FU
(m)
Categorized
Method %
Additional
Product
Clinical
Endpoints
Clinical
Results *
Overall
Result
Wound healing
8
Zhang
et al.
(2017) [
80
]
NR P 10 mice
/NR XG 0.5 FAT
procedure %
tSVF
vs.
control
Photographs,
necrosis
rate,
histological
biopsies
Thicker
fatty layer,
22.1 ±0.1
vs. 53.8 ±
0.1% *,
7/10×
more
VEGF and
bFGF *,
more
human-
derived
vessels *,
43% more
CD31+
vascula-
ture
*
+
9
Chen et al.
(2019) [
81
]
NR P 10 rats 0.5 FAT
procedure %tSVF
vs. control
Wound
healing,
biopsies
Faster
and
complete
wound
healing *,
more
vesicular
structures
and
inflamma-
tory cells,
higher
capillary
density,
MCP-1
and
VEGF *
+
10 Sun et al.
(2017) [7]NR P 54
(18 vs. 36) 0.5 FAT
procedure %tSVF vs.
control
Wound
healing,
capillary
density,
inflammatory
reaction
Complete
healing at
14 days
FU,
more
vascular-
ization *,
sharp
increase
and later
decrease
in inflam-
matory
cells
+
11
Yao Yao
et al.
(2016) [
26
]
NR P
52
mice/17
XG
0.5 SVF gel
tSVF
vs.
control
Photographs
Wound
healing
and
closure
+
Bioengineering 2023,10, 1175 26 of 34
Table 7. Cont.
Study Year Study
Design NMax. FU
(m)
Categorized
Method %
Additional
Product
Clinical
Endpoints
Clinical
Results *
Overall
Result
Wound healing
12
Wang
et al.
(2019) [
28
]
NR P 15 rabbits 3 Other %tSVF gel
Size, color,
texture,
dermal
thickness,
histological
biopsies
Improvement,
0.5 ±0.3
vs. 1.4 ±
0.3 mm *,
CD206+
macrophages
dermal
layer,
lower
IL-6 and
MCP-1 *,
lower
collagen
density *,
less alpha-
SMA,
myofi-
broblasts
and
COL-1 *
+
Osteoarthritis
13 Li et al.
(2020) [
29
]
2019 P 30 rabbits 3 Other %tSVF
vs.
control
Radiology
(MRI),
histology,
immunohistochemistry,
total
histological
outcome
score,
ICRS
Cartilage
repair,
Filled
defect,
strong
gly-
cosamino-
glycan
staining,
COL-II
up, COL-I
down,
10.2 ±0.8
vs. 8.4 ±
1.1 *,
9.8 ±1.3
vs. 7.4 ±
1.1 *
+
Other indications
14 Ye et al.
(2021) [
82
]
NR P
50
(25 vs.
25)/7 XG
2 Nanofat
tSVF
vs.
Coleman’s
Histological
biopsies
Perilipin +
cell
density *
+
15
Weinzierl
et al.
(2022) [
83
]
NR P 16 mice 0.5 Nanofat Histological
biopsies
High
func-
tional
microves-
sel
density
+
16 Li et al.
(2020) [
84
]
NR P 6 mice
(6 XG) 0.5 FAT
procedure %
tSVF vs.
decellularized
tSVF
Hair growth
Biopsies
Increased
hair
growth,
increased
prolifera-
tion,
migra-
tion, cell
cycle pro-
gression
+
Abbreviations: FU: follow-up (months); NR: not reported;
%
: altered categorized mechanical method by the
authors; *: statistically significant, p< 0.05; XG: xenograft; RCT: randomized controlled trial; P: prospective; R:
retrospective; Pi: pilot. FAT: fractionation of adipose tissue; tSVF: tissue stromal vascular fraction; AOD: average
optimal density; PCNA: proliferating cell nuclear antigen; VEGF: vascular endothelial growth factor; bFGF: basic
fibroblast growth factor; ICRS: international cartilage repair society.
Bioengineering 2023,10, 1175 27 of 34
3.1. Skin and Volume Enhancement
Sixteen studies reported the effect of tSVF on skin and volume enhancement in human
study subjects, mainly in the facial area. Tissue SVF was obtained using the FAT procedure
(n= 10), nanofat (n= 5) and SVF gel (n= 1). Most commonly, higher satisfaction and
improvement in volume restoration were reported in study patients treated with tSVF
(Table 6).
Fifty-six percent (9/16) of the studies included a control study group, of which only
two (2/16, 13%) were conducted as RCT [
45
,
46
]. In three studies, tSVF was compared to
Coleman’s fat, and all three showed significantly higher patient satisfaction in the tSVF-
treated group [
46
,
55
,
60
]. One RCT (n= 63) also showed a higher retention rate (
41.2 ±8.4%
vs. 32.6
±
8.8%) and volume ratio (0.87
±
0.02 to 0.89
±
0.03) of the tSVF group compared
to Coleman’s fat [
46
]. tSVF was also compared to treatments with Botox in improving
horizontal neck wrinkles [
47
] or hyaluronic acid in improvement of wrinkles and skin
texture [
54
], in which patient satisfaction was higher after treatment with tSVF. In five
studies [
45
,
54
–
57
], tSVF was studied in combination with treatment with platelet-rich
plasma (PRP), of which the effect of adding tSVF to PRP was only studied in one RCT. This
study (n= 28) showed no difference of the combination of tSVF and PRP compared to PRP
alone in ageing skin quality or patient satisfaction (VISIA and FACE-Q) in terms of aging
skin rejuvenation [45].
In four studies [
46
,
47
,
52
,
56
], histopathological results from biopsies taken at the sur-
gical site pre- and post-injection were reported (baseline vs. 1/6/12 months FU). Besides
increased elastic and collagen density [
47
,
56
], a decrease in CD4+ T cell infiltration in the
basal layer was observed when comparing pre- and post-injection samples [
52
], and im-
provement in dermal cellularity (mainly fibroblasts) and vascular density of the superficial
layer [56].
A direct comparison of techniques was studied by Zhu et al. in patients treated with
tSVF for periorbital volume and skin rejuvenation, which was defined as correction of fine
wrinkles of the lower eyelid and infraorbital dark circles [
59
]. Significantly higher satis-
faction and fewer reoperations (secondary tSVF injection needed for volume restoration)
were observed for the group treated with tSVF using the FAT procedure instead of tSVF
harvested using a nanofat procedure. Nevertheless, in both tSVF-treated groups, clinical
improvement was reported.
Skin and volume enhancement was also studied in seven animal models (mice, rats,
rabbits and guinea pigs; Table 7). In four studies, xenografts from humans were used
in immunocompromised animals [
74
–
76
,
78
]. All seven studies compared tSVF treatment
with a control group. Histological biopsies showed that viable adipocyte architecture
and collagen accumulation were observed with use of tSVF. In addition, a mononuclear
infiltrate—predominantly macrophages—was seen on the periphery of the fat graft near
the blood vessels. Increased dermal thickness, higher collagen deposition and more fibrob-
lasts were also observed in mice, and increased mRNA-expression of TGF-
β
1 and Smad2
(nonphospho) expression was seen [
74
]. MMP-2 and MMP-9 were significantly lower in
tSVF-treated mice compared to controls [
75
]. Nanofat-treated mice also showed better cell
survival and integrity: fewer vacuoles, less fibrosis and inflammation, and the capillary
density measured with CD31+ vessels was significantly higher than controls (24.6
±
4.72
nanofat vs. 10.4
±
2.88 control) [
76
]. Volume retention was studied in twelve rabbit ears
by applying autologous tSVF (SVF gel) with plasma-rich fibrin (PRF) versus tSVF alone.
Histological biopsies showed no distinct effect, though tSVF and PRF combined showed a
more ordered fashion of adipocytes and fibroblast distribution, which was lacking in tSVF
monotherapy. This indicates PRF may induce formation of fat cells [
79
]. An et al. used
minced rat fat, which was fractionated by intersyringe shuffling through a fractionator
with three 1 mm diameter holes [
27
]. No post-fragmentation filtration or centrifugation
was applied. From histological biopsies, the average collagen density was higher in tSVF-
treated rats. Levels of PCNA and CD31, representing, respectively, proliferation and
vascularization, were also higher.
Bioengineering 2023,10, 1175 28 of 34
3.2. Wound Healing
Ten human studies reported improvements in wound healing after applying tSVF,
of which five included a control group, two were RCTs. Tissue SVF was obtained using
the FAT procedure (n= 2), nanofat (n= 4), centrifuge-modified nanofat (n= 3) and SVF
gel (n= 1). In four studies, tSVF was added to PRP (three studies) and Coleman’s fat (one
study). In general, the majority of studies showed positive results in usage of tSVF, such as
improvement in POSAS and wound healing (Table 6).
After mammoplasty, FAT procedure-derived tSVF improved wound healing compared
to in-patient saline-treated controls in an RCT with 40 female patients. A lower POSAS
score for tSVF was reported at six months, though after 12 months follow-up, differences
between tSVF and placebo treatment were less distinct (FAT vs. placebo; patient 14.4
±
7.6
vs.
15.3 ±9.0
; observer 14.5
±
6.4 vs. 14.6
±
8.8). In addition, similar collagen alignment,
depth and width were shown, though more organized post- (scar tissue) than preopera-
tively (normal skin). Another study reported significantly lower postoperative POSAS
and histological results in 20 facial scar patients treated with tSVF (FAT procedure) [
63
].
Histologically analyzed biopsies of the scar pre- and post-injection showed no difference
in elastic fibers, though higher melanin average optical density of the basal cell layer was
reported. In addition, development of sebaceous and sweat glands was seen, which were
absent in the baseline biopsy of the scar tissue. An RCT performed in 50 burn patients
treated with Coleman’s fat and topical tSVF (centrifuge-modified nanofat) showed faster
healing, less contractures and less reoperations when compared to 50 patients treated
with conventional silver sulphadiazine cream and split skin grafting therapy [
62
]. In addi-
tion, biopsies showed more rapid collagen deposition, which was equally leveled by the
conventional therapy group at 1–3 months after baseline.
In total, five studies [
61
–
63
,
66
,
70
] included histopathological biopsies in their results.
Taken together, collagen deposition seems to evolve more rapidly in tSVF treatment of
a wound compared to controls (saline or conventional dressings). No studies with direct
comparison between tSVF harvest techniques have been conducted.
Wound healing was additionally studied in five animal studies (mice, rats, rabbits), of
which two used xenografts to obtain tSVF [
60
,
80
] (Table 7). Tissue SVF was isolated using
the FAT procedure (n= 3), SVF gel (n= 1) and by shuffling and a single centrifugation step
(n= 1).
Tissue SVF accelerated dermal wound healing in all ten human and five animal studies.
Taken together, enzyme immunoassay results showed 7–10 times more gene expression
of vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF).
An increase in inflammatory cells and higher capillary density with higher expression
of MCP-1 and VEGF were shown
7
. An
in vivo
study with 15 rabbit models also showed
lower expression of IL-6 and MCP-1, and lower collagen density and less alpha-SMA,
myofibroblasts and COL-1, were observed post-injection with tSVF. These animal studies
show that subcutaneous administration of tSVF modulates inflammation.
3.3. Osteoarthritis
The application of tSVF in osteoarthritis is relatively new and yet unstudied in humans.
The available literature describes the usage of the Fastem and Fastkit systems [
85
,
86
], but
since these procedures do not use a form of intersyringe processing, these papers were not
included in the present review.
The study by Li et al. was conducted in a canine osteoarthritis model, in which
tSVF was obtained by mincing inguinal fat, followed by intersyringe processing 90 times
through a one-hole 2.0 mm fractionator, followed by a final centrifugation step [
29
]. MRI
and histological analysis at 3 months FU showed complete filling of the cartilage repair after
tSVF application versus poorly filled in the control group (purely minced fat graft). Gene
expression of collagen type II was upregulated and collagen type I downregulated, while
higher international cartilage repair society (ICRS) scoring was reported after SVF treatment
vs control. The restoration of cartilage injury after tSVF administration is promising and
Bioengineering 2023,10, 1175 29 of 34
should be studied further for future application in common arthritis manifestations, for
example, thumb carpometacarpal (CMC) joint osteoarthritis.
3.4. Other
Four clinical studies outside the categories mentioned in this scoping review (that
is skin rejuvenation, wound healing and osteoarthritis) are worth mentioning. A pilot
study (n= 10) was performed in alopecia patients who were treated with tSVF harvested
using the FAT procedure in combination with PRP. In all ten patients, hair follicle density
increased, and regrowth was observed. The restoration of hair follicles may implicate the
ability of tSVF and PRP to stimulate follicular regeneration through boosting resident
epidermal stem cells. Hair growth was also faster in tSVF (FAT procedure)-treated mice,
and
in vitro
analysis showed increased dermal papilla cell proliferation, migration and cell
cycle progression of the injected lipoaspirate contained adipose stem cells in combination
with the extracellular matrix [84].
Improvement in interstitial cystitis in six women was obtained with repetitive intrav-
esical injections with tSVF (nanofat). Both pain—VAS and related symptoms improved
significantly at 18-month follow-up after four intravesical injections. The intravesical mor-
phology was restored based on cystoscopic examination [
65
]. After four injections (1 year
FU) all inflammatory cells (T-lymphocytes, eosinophils and mast cells) were decreased in
vulvar lichen sclerosis. Associated symptoms and quality of life also improved significantly
in comparison with conventional topical clobetasol propionate cream. Glottis closure and
voice quality expressed in the GRBAS scale improved significantly in 22 tSVF-treated (SVF
gel) unilateral voice fold paralysis patients [
73
]. The effects were observed up to 18 months
of follow-up. Injection of tissue SVF (MyStem) was also proposed as treatment of venous
leg ulcers. In total, 58% had complete healing of the ulcer at 12 months FU, and ulcer size
was decreased in 96%
±
1.7. Between the 3 month and 6 month FU, the ulcer decreased
significantly (64.5%).
Volume retention of fat grafts (nanofat vs. Coleman’s fat) in mice were sampled after
being exposed to shear force [
82
]. The retention rate of the tSVF/ECM gel was higher with
a larger number of mesenchymal stem cells, supra-adventitial (SA) adipose stromal cells
(ASCs), and adipose-derived stem cells but a lower number of endothelial progenitor cells.
It was suggested that the pluripotency of adipose tissue-derived stem cells was increased by
intersyringe processing, which can improve graft retention in fat grafting. In sixteen mice,
nanofat injections showed high functional microvessel density in dermal sites after 14 days
post-injection; these results suggest tSVF administration to enhance tissue vascularization.
4. Discussion
The generation of tSVF through fractionation of lipoaspirate is achieved by mechanical
disruption of adipocytes. This results in a volume reduction after removing oil from
disrupted adipocytes and their cellular remnants. Since adipose tissue consists of 90%
(v/v) adipocytes on average, volume should be reduced by approximately 90% to obtain
the remaining 10% (v/v) tSVF fraction. SVF gel and the FAT procedure showed the most
volume reduction (approximately 90%) with an increase in SVF cells per volume unit
and high viability, while mechanical micronization squeeze filtration fluid obtained tSVF
with a 52% reduction in volume. Centrifugation as a first step removes fluid from the
lipoaspirate, which makes adipocytes more susceptible for disruption during intersyringe
shuffling and is therefore crucial for proper isolation tSVF [
38
]. In this review, many
procedures have not published the end volume of tSVF or the quantification of the number
of adipocytes to substantiate the loss of adipocytes. It is therefore impossible to conclude
that these procedures result in condensed tSVF. It is likely that these procedures generated
mere processed fat with an increased injectability, in other words emulsified fat, especially
when centrifugation was not applied before fractionation. Filtration as a final step seems
to improve injectability but does not reduce the voluminous oily layer. Moreover, part of
Bioengineering 2023,10, 1175 30 of 34
the extracellular matrix, with its bound cells and growth factors, retain in the filter, which
could lead to reduced clinical efficacy [32].
Technical improvements in these procedures are only relevant when they translate to
improved clinical outcomes. Various studies have shown beneficial effects of tSVF in the
field of skin rejuvenation, volume retention and wound healing. At this point, no superior
isolation method for tSVF and associated clinical outcomes can be identified, since the study
methods and endpoints vary widely and thus comparisons between studies are difficult to
make. Volume reduction and concentration of tSVF seem especially important in medical
conditions that necessitate injection in small volumes, such as small joints, scars or perianal
fistulas. Hypothetically, injection under these conditions could show improved clinical
outcomes when using concentrated tSVF instead of regular fat transfer. Furthermore, shear
stress of fractionation activates cells, which could stimulate their regenerative potential, as
was previously published [41].
In general, tSVF seems to have a more positive clinical outcome when diseases with
a high proinflammatory character are treated, such as osteoarthritis or (disturbed) wound
healing, in comparison with rejuvenation of the aging skin. Pathological processes result
in a disbalance in extracellular factors such as inflammation, excessive ECM deposition
and crosslinking, or a lack of angiogenesis. The difference in clinical outcome suggests that
tSVF needs a trigger such as inflammation to ‘re-educate’ damaged tissue.
There are a few limitations of this review. The scoping instead of systematic character
of this review might result in missing studies that would otherwise have been included de-
spite our considerable efforts in our searches. In this review, only fractionation procedures
based on the shuffling principle using a device with holes or blades were included. To
identify the appropriate isolation method of SVF, other mechanical isolation or enzymatic
isolation procedures should be included as well. On the other hand, the goal of this review
is to correlate differences in techniques between shuffling procedures to clinical efficacy.
Most of the validation studies of the reported procedures do not report the complete
set of validation data that is defined by the joint statement of the IFATS and ISCT. Multiple
studies reported differentiation capacity and phenotype using flow cytometry, but only
three procedures reported colony formation capacity. Conclusions cannot be drawn from
a comparison of cell number yield because cell numbers vary between donors [
20
]. Also,
differences in lipoaspirate handling protocols and quantification methods have a large
impact on cell yield, which makes absolute comparison almost impossible [
24
]. We recom-
mend that future validation studies of procedures report the complete set of validation data
as defined by the joint statement of the IFATS and ISCT [
9
]. They should focus on providing
evidence for the fractionation of stromal vascular fraction by reporting end volume, loss of
adipocytes and increase in number of SVF cells.
Future studies should be well designed and include validated endpoints that are mea-
sured and reported homogeneously. It would be of interest to compare different isolation
methods within a specific clinical field, preferably in a randomized fashion. As wound
and scar healing is known to expand over a minimum of twelve months, we suggest
monitoring follow-up results at least twelve months postoperatively. Adding therapies to
tSVF treatment, such as PRP, PRF or CO
2
laser, demand another control group in the study
design, as the exact effect of tSVF is still under debate.
Standardization of procedures is of paramount importance to analyzing and com-
paring the results of clinical studies. Many clinical studies were found that described
procedures that do not correspond with the original description and validation of the
methods of these procedures. It is recommended that authors strictly adhere to the meth-
ods of procedures because otherwise additional validation of their altered procedures are
necessary and clinical comparisons are ambiguous.
tSVF injection is a promising new therapy that is easy to use, can be injected during
the same surgical procedure and shows positive results. In our opinion, the future of tSVF
lies especially in clinical indications with a proinflammatory character such as osteoarthritis
Bioengineering 2023,10, 1175 31 of 34
and wound healing. Future well-designed clinical trials should focus on using validated
procedures with validated outcome measurement tools.
Author Contributions:
J.A.M.S.: Conceptualization, Writing—original draft, Project administration,
Visualization. C.J.H.C.M.v.L.: Conceptualization, Writing—original draft. R.H.S.: Writing—reviewing
and editing, Supervision; A.J.T.: Writing—reviewing and editing; M.C.H.: Writing—reviewing and
editing, Supervision; F.K.L.S.: Writing—reviewing and editing, Supervision; J.J.: Writing—reviewing
and editing, Supervision; J.A.v.D.: Methodology, Conceptualization, Writing—reviewing and editing,
Supervision. All authors have read and agreed to the published version of the manuscript.
Funding: Funding for this study was provided by the University Medical Center Groningen.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Ghiasloo, M.; Lobato, R.C.; Díaz, J.M.; Singh, K.; Verpaele, A.; Tonnard, P. Expanding Clinical Indications of Mechanically Isolated
Stromal Vascular Fraction: A Systematic Review. Aesthetic Surg. J. 2020,40, NP546–NP560. [CrossRef] [PubMed]
2.
Van Dongen, J.A.; Van Boxtel, J.; Harmsen, M.C.; Stevens, H.P. The Development of Facial Lipofilling from a Historical Point of
View. Facial Plast. Surg. 2019,35, 358–367. [CrossRef]
3.
Zuk, P.A.; Zhu, M.; Mizuno, H.; Huang, J.; Futrell, J.W.; Katz, A.J.; Benhaim, P.; Lorenz, H.P.; Hedrick, M.H. Multilineage Cells
from Human Adipose Tissue: Implications for Cell-Based Therapies. Tissue Eng. 2001,7, 211–228. [CrossRef] [PubMed]
4.
Eto, H.; Suga, H.; Matsumoto, D.; Inoue, K.; Aoi, N.; Kato, H.; Araki, J.; Yoshimura, K. Characterization of Structure and Cellular
Components of Aspirated and Excised Adipose Tissue. Plast. Reconstr. Surg. 2009,124, 1087–1097. [CrossRef] [PubMed]
5.
van Dongen, J.A.; Tuin, A.J.; Spiekman, M.; Jansma, J.; van der Lei, B.; Harmsen, M.C. Comparison of Intraoperative Procedures
for Isolation of Clinical Grade Stromal Vascular Fraction for Regenerative Purposes: A Systematic Review. J. Tissue Eng. Regen.
Med. 2018,12, e261–e274. [CrossRef]
6. Guo,