Available online at www.sciencedirect.

com

Microvascular Research 75 (2008) 119 – 129

www.elsevier.com/locate/ymvre

Regular Article
Quantitative analysis of the microvasculature growing in the fibrin interface
between a skin graft and the recipient site
Xuemei Wu a , Neil Kathuria a , Charles W. Patrick a , Gregory P. Reece b,⁎
a
Reparative Biology and Bioengineering, University of Texas M.D. Anderson Cancer Center, Houston, TX, USA
b
The University of Texas, M. D. Anderson Cancer Center, USA
Received 6 March 2007; revised 26 April 2007; accepted 27 April 2007
Available online 8 June 2007

Abstract

Current tissue engineering techniques have failed to provide an established microvasculature in skin substitutes, a requisite for the maintenance
of graft viability and rapid revascularization subsequent to graft transplantation in vivo. To improve outcomes for both conventional skin grafts
and skin substitutes, the existing knowledge gap concerning the spatio-temporal mechanisms of skin graft revascularization must be abrogated.
The current knowledge gap is due, at least in part, to a lack of appropriate diagnostic methods to quantify skin graft revascularization. To enhance
the understanding of skin graft revascularization, we quantitatively evaluated revascularization of autologous skin grafts in a rat model by
quantifying 2- and 3-dimensional vascular metrics in the fibrin interface 3, 7, and 10 days after transplantation. In this study, the fibrin interface
appeared to be completely replaced with fibrovascular tissue by postoperative day 10. Although the mean vessel diameter was about 10 μm for the
time points sampled, the mean vessel number, area, and volume fraction increased about 2.5-fold from postoperative day 3 to 7 and then decreased
about 1.27-fold at postoperative day 10. There was no significant difference between 2- and 3-dimensional vascular metrics based on Bland–
Altman analysis. In conclusion, these data establish a standard for metrics of vessels growing in the fibrin interface of a rat autologous skin graft
and its donor site and suggests that once the blood supply has been restored to a viable transplant, the number, area, and volume fractions of
vessels decrease to levels found at postoperative day 3.
© 2007 Elsevier Inc. All rights reserved.

Keywords: Microvasculature; Quantitative; Revascularization; Skin graft; Three-dimensional; Transplantation

Introduction including a limited supply of donor skin, donor site morbidity,

Skin grafting has been a part of the physician's armamentar-
ium for over 2 millennia (Converse et al., 1975) and is the
second most common method of wound closure, after primary
closure. Indeed, the prevalence of skin grafting is growing: in
1996, the approximate number of skin grafts performed for
inpatients (excluding ambulatory procedures) in the United
States was 98,000 (U.S. Department of Health and Human
Services, Center for Disease Control and Prevention/National
Center for Health Statistics, 1998), compared with 240,000 in
2002 (U.S. Department of Health and Human Services, Center
for Disease Control and Prevention/National Center for Health
Statistics, 2005). However, skin grafting has drawbacks,
⁎ Corresponding author.
E-mail address: greece@mdanderson.org (G.P. Reece).
0026-2862/$ - see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.mvr.2007.04.012
and expense.
To minimize donor site morbidity and to increase the
availability of skin grafts, some investigators have turned to
tissue engineering to develop skin substitutes (Jones et al., 2002;
Mansbridge, 2002). These substitutes have been somewhat
successful in small animal models but have not been widely
accepted for patients. One fundamental constraint of partial- or
full-thickness skin substitutes is that they lack the three-
dimensional vascular spatial structures needed to support tissue
viability (Boyce et al., 1995; Reece and Patrick, 1998; Neumann
et al., 2003). Although attempts to use endothelial cells to
establish a microvasculature in skin substitutes have shown
some promise (Black et al., 1998; Soejima et al., 1998), the
development of a clinically translatable skin substitute remains
elusive. We believe that a primary obstacle to the development of
clinically useful suitable skin substitutes is a lack of information
about the biological mechanisms of skin graft revascularization,
120 X. Wu et al. / Microvascular Research 75 (2008) 119–129

vessel counts (Soejima et al., 1998). However, these techniques

are largely descriptive, and most do not provide the quantitative
information needed to fully understand the revascularization of
conventional autologous skin grafts or to design a clinically
useful skin substitute.
To adequately quantify microvascular growth in the fibrin
interface between a skin graft and its recipient site, the
microvasculature must be visualized at the capillary level and
ideally quantified in a three-dimensional format. Toward that
goal, our group previously described a method of quantitative,
three-dimensional imaging of the microvasculature that inte-
grates digital microscopy, immunohistochemistry, digital pro-
cessing, computational alignment and registration, and volume
rendering to produce high-resolution, three-dimensional, quan-
tifiable images of the microvascular structures growing in a
fibrin gel (Brey et al., 2002, 2004). In the present study, we
sought to determine the ability of this method to quantify split-
thickness skin graft (STSG) revascularization in a rat model.

Methods

Animals

All animal procedures were performed in facilities accredited by the

Association for Assessment and Accreditation of Laboratory Animal Care and
the National Institutes of Health. Fifteen male Sprague–Dawley rats, each
weighing 200–250 g and aged 63–70 days, were randomly assigned to five
groups (three animals/group). All rats underwent transplantation of autologous
STSGs (described below). No more than one rat per group underwent
autologous STSG application during an operative session, and a maximum of
two animals underwent the procedure during an operative session. Animals had
free access to food and water and were housed in individual cages in rooms that
were maintained at 21 °C, with 12-h light and dark periods. Depending on other
group assignments, the rats were killed on postoperative day 3, 7, 10, 14, or 21.

Skin graft procedure

Fig. 1. (A) Split-thickness skin grafts sutured to the back musculature using a Before any grafting procedures, all animals received an anesthetic cocktail
microsurgical technique. (B) Boltster dressing and protective cover placed over (64 mg/mL ketamine HCl, 3.6 mg/mL xylazine, and 0.07 mg/mL atropine
the grafts. (C) Healed skin grafts before postmortem harvesting. sulfate), which was injected intramuscularly. Isoflurane 0.5–2% was used as a

owing in part to the lack of appropriate diagnostic tools to

quantify this process. Since the first published description of
skin graft revascularization in 1863 (Bert, 1865), the process has
been studied using various examination techniques, including
vital microscopy (Taylor and Lehrfeld, 1953; Edgerton et al.,
1957; Markmann, 1966; Zarem and Zweifach, 1967; Birch and
Brånemark, 1969a,b; O'Donoghue and Zarem, 1971); injection
of various dyes followed by stereomicroscopy, histology, or both
(Davis and Traut, 1925; Scothorne and McGregor, 1953; Rolle
et al., 1959; Clemmensen, 1964; Smahel, 1968; Smahel and
Ganzoni, 1970; Converse et al., 1975); conventional histology
(Medawar, 1944; Hynes, 1954; Kamrin, 1960, 1961; Henry et
al., 1961, 1962; Haller and Billingham, 1967); scanning electron
microscopy of microvascular casts (Okada, 1986); microangio-
graphy (Bellman et al., 1964; Ljungqvist and Almgård, 1966;
Birch and Brånemark, 1969a,b; Ueda et al., 1986); in situ Fig. 2. Representative tissue section stained with mouse anti-rat CD31
hybridization and immunohistochemistry (Young et al., 1996; monoclonal antibodies and hematoxylin counterstain. Magnification is
Kunzi-Rapp et al., 1999); autoradiography (Lambert, 1971); and 200×.
 X. Wu et al. / Microvascular Research 75 (2008) 119–129 121

supplementary perioperative anesthetic if required. Each animal's back was Histology and immunohistochemistry
shaved from the shoulder to the hip area and from flank to flank. Cellophane
tape was used to strip away the loose epidermal layer. A depilatory agent (Nair, Six specimens were obtained from each animal at each time point. Using a
Carter-Wallace, Inc., New York, NY) was applied to the shaved skin; after 7 min, cryostat (Leica CM3050 S, Leica Microsystems, Nussloch, Germany) at
the product was wiped clean to remove all hair from the skin. The animal was
then placed on an operating table maintained at 37 °C, and its back was then
prepared with warm povidone–iodine solution and sterilely draped.
The same location was used for the STSG donor site and the graft recipient
site of each rat. The skin of the left or right flank between the costal margin of the
ribs and the hip was marked with a circle measuring 1.5 cm in diameter. An
electric dermatome (Padget, Kansas City, MO), set to cut at a thickness of
0.02 in. (∼ 0.3 mm), was used to remove a STSG from the marked area of the
flank. The skin graft donor site was then excised down to the back musculature.
The areolar tissue overlying these muscles was removed, and the skin around the
edge of the circular defect was sutured to the back muscles circumferentially
with 5-0 polypropylene (Prolene, Ethicon, Somerville, NJ) sutures. Desiccation
of the recipient site was prevented by intermittent irrigation of the muscle with
saline solution (37 °C) during the procedure.
After removal, the STSG was pinned out to its original dimensions on a
silicone block and cut into six grafts, each measuring 4 × 4 mm, with the aid of
a plastic template. The skin grafts were wrapped in a sterile, saline-moistened
gauze pad to prevent desiccation and to maintain their viability. All grafts
were applied to the exposed muscle within 30 min after removal by using a
microsurgical technique (Fig. 1A). The STSGs, spaced 3 to 4 mm from the
surrounding skin and from each other, were sutured to the back muscles with
8-0 nylon or polyglactin sutures (Ethicon, Somerville, NJ), placed at each
corner of each graft with the aid of an operating microscope (Leica Wild
M690, Heerbrugg, Switzerland). A bolster dressing was then used to apply
even pressure to the grafts and to prevent the animal from damaging the grafts
(Fig. 1B). The bolster was made from a round (2-cm diameter) piece of sterile
polyurethane foam that was covered with a water vapor-permeable
polyurethane membrane barrier on the graft side (Kinetic Concepts Inc.,
San Antonio, TX) such that the adhesive side faced the foam. The foam
bolster was held in place with a whiffle ball (Sell, Inc., China) that had been
cut in half and was sutured to the skin around the graft site with 3-0 stainless
steel wire sutures. The mean operating time was about 1 h per animal. All
animals received an antibiotic (gentamicin 5 mg/kg; or enrofloxacin, 5–
10 mg/kg) intramuscularly upon induction of anesthesia and then daily for the
first three postoperative days.

Removal, preparation, and snap freezing of specimens

All animals were killed by CO2 asphyxiation on the assigned postoperative

day. The bolster dressing was carefully removed to avoid stripping either the
epidermal layer of the graft or the layer of fibrin entrapped between the graft and
recipient site (Fig. 1C). Because skin debris and follicles containing hair shafts
can cause cutting artifacts on histologic examination, especially with serial
sectioning of the specimens, the skin grafts were first inspected under an
operating microscope (Carl Zeiss, Inc., Obekochen, Germany), and any debris
was carefully removed with a jeweler's forceps. If hair was present, a depilatory
agent (Nair) was applied to the outer surface of the STSG for 7 min, and the hair
was gently removed with cotton-tipped swabs.
The full thickness of the abdominal wall and overlying undisturbed STSGs
in the graft site area were dissected en bloc from the rat's back and flank. Each
of the six STSGs with its underlying muscle was cut into individual specimens.
With the aid of the operating microscope and microsurgical instruments, muscle
and other tissues around each specimen were trimmed to within 1 mm of the
border of the STSG. A 6-0 Prolene suture was then placed through the muscle
just outside the graft area so that the specimen could be oriented in a cryomold
(Tissue-Tek, Sakura Fineteck, USA, Inc., Torrance, CA) perpendicular to the
sectioning plane. To avoid bubbles, OCT (Optimal Cutting Temperature)
embedding medium (Tissue-Tek, Torrance, CA) was placed around the Fig. 3. Digital tiling and segmentation progress flow for a single section. Eight
specimen by using a 10-mL syringe with an 18-gauge needle. Each cryomold 200× magnification images were automatically digitally tiled into a 4 × 2 matrix
was inspected under the operating microscope and all bubbles manually measuring 0.888 mm × 3.552 mm. The upper and lower boundaries of the fibrin
removed from the OCT medium. Each specimen was then snap frozen in liquid interface were selected as the boundaries for the region of interest. Finally,
nitrogen vapor in a Dewar flask. The specimens were stored in a − 80 °C freezer CD31-positive cells in the region of interest were automatically segmented
until ready for cryosectioning and staining. Specimens were sectioned in a (Brey et al., 2002). This was repeated for all 25 serial sections of each skin graft
direction perpendicular to the interface between the graft and recipient site. sample.
122 X. Wu et al. / Microvascular Research 75 (2008) 119–129

Table 1
Equations used to determine vascular metrics using measurements from segmented digital images
Metric Equation Units
No. vessels No. vessels (image 1) + No. vessels (image 2) Vessels
Area CD31 Total area stained (image 1) + Total area stained (image 2) Pixels2
Vessel diameter sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  Micrometers
ð4Þðarea CD31Þ 0:3703 Am
ðstained No: vesselsÞðpÞ pixel
Area interface Total area stained ðimage 1Þ Total area stained ðimage 2Þ Pixels2
h i þ h i
% of ROI ðimage 1Þ % of ROI ðimage 2Þ
100 100
Area fraction area CD31 No units
area interface
No. fraction  2   Vessels/mm2
No: vessels pixel 1000 Am 2
area interface 0:3703 Am 1 mm
Volume fraction X
25 No units
area CD31
1
X
25
area interface
1

ROI = fibrin interface.

− 20 °C, a 1- to 1.5-mm-wide portion of each frozen specimen was removed and
discarded before two or three sections, each 8 μm thick, were cut and stained
with hematoxylin and eosin to determine which specimens were acceptable for
serial sectioning. Specimens were considered acceptable if the entire graft was
held to the underlying muscle by an intact layer of fibrin without any defects or
curling of the specimen. At least three specimens were identified as acceptable
from each animal from each time point. Then 25 serial sections, each measuring
8-μm thick, were cut from each specimen and stained with hematoxylin and
mouse anti-rat CD31 monoclonal antibody (Serotec, Raleigh, NC) using a Dako
(Carpinteria, CA) autoimmunostainer. A section thickness of 8 μm was chosen
because it is consistent with the capillary diameter in a rat and because many
changes in capillary geometry (e.g., changes in intercapillary distances and
capillary loops) are found within 20 μm (Patan et al., 2001a,b; Norrby, 1998;
Less et al., 1991). The minimum number of sections required to accurately
represent the volume percentage of blood vessels in the interface between the
STSG and muscle was found to be 25 (data not shown). The remaining
specimens were stored in a −80 °C freezer to be used as replacements, if needed.
Fig. 2 illustrates a representative tissue section stained with mouse anti-rat CD31
monoclonal antibodies and a hematoxylin counterstain.
dermis of the graft and the surface of the muscle. The automated measurements
of values from the segmented images have been previously validated (King et
al., 2002; Brey et al., 2002; Brey et al., 2004). Two-dimensional “vascular
density” was measured according to the two most commonly reported
descriptors, area fraction and number fraction. Three-dimensional vascular
density was measured according to the descriptor volume fraction. Vessel
metrics are presented as the mean ± standard error of the mean (S.E.M.).
Statistical analysis
Differences in vascular metrics were evaluated using analysis of variance
(ANOVA) or two-tailed Student's t test, as appropriate. Statistical significance
was defined as P ≤ 0.05. Bland–Altman analysis was used to compare the two-
dimensional metric of area fraction with the three-dimensional metric of volume
fraction. Distributions of metrics were described using the statistical descriptors

Light microscopy and imaging

The mounted histologic sections were imaged along the fibrin interface with a
20× objective using an Olympus IX-70 inverted bright field microscope
(Olympus, Mellville, NY) coupled to a color CCD camera (RT-Spot, Diagnostic
Instrument, Sterling Heights, MI), an XYZ stage (Ludl Electronic Products,
Hawthorne, NY), and a G4 Macintosh computer. IPLab image analysis software
(Scanalytics, Inc., Fairfax, VA) was used to control the microscope stage during
digital tiling (Fig. 3). A 4 × 2 image matrix was programmed to automatically
capture the interface of each section at high magnification. This imaging system
has a pixel resolution of 0.6 μm; hence, the resulting digitally tiled two-
dimensional images were 2.368 mm × 0.888 mm. A 3.3 Gb capacity was required
to store each set of 25 serial sections for a sample. Blood vessels in the interface
were segmented from the background by color thresholding of the histochemical
and immunohistochemical stains (vessels were brown, and surrounding tissue
was blue-green). A 1.65 Gb capacity was required to store each segmented
sample. Alignment, registration, and visualization of the serial sections were
performed as previously described (Brey et al., 2002). A total of 99 Gb was
required to store the study's entire set of digital histology and segmented images.

Metrics
Fig. 4. Diameters of vessels in the fibrin interface at postoperative days 3, 7, and
Table 1 lists the equations used to evaluate the fibrin interface in the 10. Data are the mean ± S.E.M. There was no significant difference in any
segmented images. The fibrin interface was defined as the area between the comparison.
 X. Wu et al. / Microvascular Research 75 (2008) 119–129 123

of skewness and kurtosis. All statistical analyses were performed using Table 2
Microsoft Office Excel 2003 (Microsoft Corporation, Redmond, WA). Skewness and kurtosis values for distributions presented
Postoperative Postoperative Postoperative
Results day 3 day 7 day 10
Vessel diameter
The rats tolerated the surgery well, with only one death due Skewness 0.199 0.145 1.460
Kurtosis −0.454 − 0.728 2.157
to an anesthetic overdose in the early postoperative period. All
skin grafts survived transplantation, and no graft “sloughing” Number fraction
was observed. Histological processing was conducted with a Skewness 1.187 3.705 0.933
low rate of missed sections (2.2%). Only data for postoperative Kurtosis 0.754 14.275 0.327
days 3, 7, and 10 are presented because no fibrin interface was
Area fraction
discernable in any sample from day 14 or 21; that is, by day 14,
Skewness 0.901 0.918 0.415
the fibrin had been replaced by collagen, and tissue remodeling Kurtosis −0.067 0.475 − 0.462
had occurred.
Kurtosis is the degree of peakedness (N0, leptokurtic; =0, mesokurtic; b0,
The mean vessel diameter for postoperative days 3, 7, and 10 platykurtic).
was consistently 10–11 μm, the diameter of a capillary (Fig. 4). Skewness is the degree of asymmetry.
However, the distribution of vessel diameters differed for the
three days (Fig. 5 and Table 2). From day 3 to day 7, the number
of vessels b8 μm and N 16 μm decreased, resulting in a more The second two-dimensional vascular density metric
platykurtic distribution at day 7. From day 7 to day 10, the determined was area fraction. The trend was similar to
number of vessels N 16 μm increased, resulting in a leptokurtic that observed with number fraction, namely a maximum
and asymmetric distribution. value at day 7 (Fig. 8). The area fraction was 0.009 at day
The first two-dimensional vascular density metric deter- 3, increased 2.3-fold at day 7 to 0.021, and decreased to
mined was number fraction, or the number of vessels/mm2 0.012 at day 10. Fig. 9 and Table 2 present the distribution
(Fig. 6). At day 3, the mean number fraction was 116 vessels/ of area fractions. The distributions at day 3 and 10 were
mm2. A 2.5-fold increase was observed at day 7, resulting in a platykurtic and bimodal, whereas day 7 had a leptokurtic
significantly higher (P = 0.027) mean of 291 vessels/mm2. The and unimodal distribution. It is interesting that the low area
number fraction decreased at day 10 to a mean of 148 vessels/ fractions (≤0.008) at day 3 disappeared at day 7 and
mm2. There was no significant difference between number reappeared at day 10. This finding mirrors the trend
fractions at day 3 and day 10 (P = 0.54). Fig. 7 and Table 2 observed with densities ≤ 100 vessels/mm2 in the number
present the distribution of number fractions. Approximately fraction distributions.
60% of the number fraction distribution at day 3 possessed a Although the majority of studies have used one of the
density ≤ 100 vessels/mm2. This decreased dramatically to 2% two-dimensional metrics described above to define vascular
at day 7, when the number fraction distribution became highly density, the term “vascular density” is actually a three-
leptokurtic and asymmetric with densities in excess of 500 dimensional measurement. Therefore, we also determined
vessels/mm2. At day 10, the distribution shifted back to volume fraction (volume vessels/volume of tissue sample).
approximate the day 3 distribution. As with number fraction and vessel number fraction, volume

Fig. 5. Distributions of vessel diameters at postoperative days 3 (A), 7 (B), and 10 (C). The distributions possess identical ordinate and abscissa ranges and binning.
The black vertical lines denote the mean values.
124 X. Wu et al. / Microvascular Research 75 (2008) 119–129
Fig. 6. Number fraction in the fibrin interface at postoperative days 3, 7, and 10.
Data are mean ± S.E.M. ∗Denotes a significant difference between day 7 and any
other postoperative day.
fraction also had a maximum value at day 7 (Fig. 10). The
volume fraction increased 2.3-fold from day 3 to day 7,
followed by a decrease at day 10 to levels approximating
day 3. In determining whether there was any significant
difference between the two- and three-dimensional vascular
density measurements (Fig. 11), we found that the majority
of data fell within the Bland–Altman limits, with only 4.7%,
4.8%, and 2.3% of data falling outside the limits for
postoperative days 3, 7, and 10, respectively. Hence, a two-
dimensional vascular metric can adequately represent vas-
cular density during healing between a recipient site and a
skin graft.
Fig. 7. Distributions of vessel number fraction at postoperative days 3 (A), 7 (B), and 10 (C). The distributions possess identical ordinate and abscissa ranges and
binning. The black vertical lines denote the mean values.
Fig. 8. Area fraction in the fibrin interface at postoperative days 3 (A), 7 (B), and
10 (C). Data are mean ± S.E.M. ∗Denotes a significant difference between day 7
and any other postoperative day.
Discussion
In this study, we quantified microvascular growth in the
fibrin interface between autologous STSGs and the recipient
site (muscle) of Sprague–Dawley rats over the first 10
postoperative days. We noted that the fibrin interface appeared
to be completely replaced with fibrovascular tissue by post-
operative day 10, that there was no statistically significant
change in mean vascular diameter of the microvasculature of the
interface from postoperative days 3 through 10, and that the
vascular density in the interface increased 2.5-fold between
postoperative days 3 and 7 before returning to levels near those
of postoperative day 3 by the 10th post-grafting day.
 X. Wu et al. / Microvascular Research 75 (2008) 119–129
Fig. 9. Distributions of area fraction at postoperative days 3 (A), 7 (B), and 10 (C). The distributions possess identical ordinate and abscissa ranges and binning. Area
fractions on the abscissa are multiplied by 100 for ease of viewing. The black vertical lines denote the mean values.
To our knowledge this is the first study that has specifically
quantified the microvascular growth in the interface between a
STSG and its underlying recipient site. Other investigators
recently performed a quantitative study of the microvascular
growth in skin grafts (Capla et al., 2006; O'Ceallaigh et al.,
2006), but looked at the vascular density in the skin and tissue
around the graft (O'Ceallaigh et al., 2006) and in the dermis of
the graft itself (Capla et al., 2006; O'Ceallaigh et al., 2006).
In a study of full-thickness skin grafts cross-transplanted
between wild-type FVB/N and transgenic tie2/lacZ mice, Capla
et al. (2006) harvested grafts at postoperative days 3, 5, 7, 14,
and 21, which are similar to the time points in our study. The
number of lacZ-stained vessels in three random fields from four
nonconsecutive paraffin-embedded tissue sections was counted
Fig. 10. Volume fraction in the fibrin interface at postoperative days 3 (A), 7 (B),
and 10 (C). Data are mean ± S.E.M. ∗Denotes a significant difference between
day 7 and any other postoperative day.
125
at 200× magnification. Unlike the vascular densities noted in our
study, Capla et al. noted a progressive increase in vascular
density from postoperative days 3 through 21 (Table 3).
Although their units of measure differed from ours, the
proportional increase in number fraction was identical to ours
from postoperative days 3 to 7 but differed slightly from
postoperative days 7 to 14. The differences in vascular density
between their study and ours may be explained by different rates
of revascularization for split- versus full-thickness grafts, differ-
ences in species, vessel density in the interface versus the dermis
of the graft, and the use of different histological techniques.
In a study of autologous full-thickness skin grafts in male
C57/B16 mice, O'Ceallaigh et al. (2006) excised the grafts and
underlying muscle en bloc at 48 h, 60 h, 3 days, 5 days, or
14 days (three animals per time point) after grafting. To
distinguish unperfused and perfused blood vessels, the
vasculature of each animal was injected with fluorescein and
2% porcine gelatin before harvest of the specimens. Using
frozen sections, the investigators quantified the number of
vessels stained with anti-CD31 antibodies and the number of
fluorescein-stained vessels by their location in the graft relative
to its center and relative to their depth within the graft, that is,
the papillary dermis, reticular dermis, and underlying muscle;
the investigators did not specifically state whether their analysis
included the fibrin interface. The authors noted only a 1.4-fold
increase between postoperative days 3 and 5 and a 1.2-fold
increase between postoperative days 5 and 14 in the number
fraction compared with a 2.5-fold and 1.3-fold increase in the
number fraction in our study from postoperative days 3 to 7 and
7 to 10, respectively (Table 3). Because O'Ceallaigh et al. used
different methods and time intervals and looked at different
parts of the graft than we did, we cannot directly compare their
results with ours.
The temporal pattern of vascular growth into the skin graft in
our and Capla et al.'s study was also noted in a previous wound
healing study that we performed using an in vivo fibrin gel
model of angiogenesis (Brey et al., 2004). In that study, we
126 X. Wu et al. / Microvascular Research 75 (2008) 119–129

Fig. 11. Bland–Altman analyses comparing the vascular metrics of area fraction and volume fraction for postoperative days 3 (A), 7 (B), and 10 (C). The horizontal
lines denote the mean of the differences, and the dashed horizontal lines denote the Bland–Altman limits of agreement.
 X. Wu et al. / Microvascular Research 75 (2008) 119–129 127

Table 3 Table 5
Number fraction or equivalent in three studies Comparison of vessel diameter or equivalent in two studies
Author Postoperative Mean value Proportional (times) Author Postoperative Mean value Proportional (times)
of study day increase from of study day increase from
postoperative postoperative
day 3 value day 3 value
Wu et al. (current study) Wu et al. (current study)
3 116 vessels/mm2 3 10.25 μM
7 291 vessels/mm2 +2.5 7 10.89 μM +1.1
10 148 vessels/mm2 +1.3 10 10.56 μM +1
Capla et al. (2006) Brey et al. (2004)
3 2 ± 0.6 vessels/hpf 3 8.70 ± 1.94 μM
7 5 ± 2.2 vessels/hpf +2.5 7 10.70 ± 0.40 μM +1.2
14 9 ± 1.4 vessels/hpf +4.5 14 10.70 ± 1.40 μM +1.2
21 10 ± 1.8 vessels/hpf +5 21 0
O'Ceallaigh et al. (2006)
⁎3.2 vessels/70,000 μm2 Vessel diameter was not measured in the studies of Capla et al. (2006) and
3
O'Ceallaigh et al. (2006).
5 *4.4 vessels/70,000 μm2 +1.4
14 *3.7 vessels/70,000 μm2 +1.2
hpf = high power field. architecture and metrics and require inferring three-dimensional
⁎ Estimated from central graph 5 of the study.
information from two-dimensional results, creating obvious
errors in quantitation (Artacho-Perula et al., 1999; Bullitt et al.,
determined the vascular metrics for the microvasculature 1997). In fact, the absence of a standard for the three-
growing through a fibrin gel construct (Tisseel Fibrin Sealant, dimensional study of microvascular geometry has resulted in
Baxter Bio Science, Deerfield, IL) at postoperative days 3, 7, controversial conclusions about tumor angiogenesis (Maniotis
14, and 21 (Table 4). The proportional increase in area fraction et al., 1999; McDonald et al., 2000). Although we found no
between postoperative days 3 and 7 in that study was almost statistically significant difference between volume fraction and
identical to the area fraction increase in our current study; number and area fractions, we do believe that three-dimensional
however, the earlier study noted an increase in area fraction imaging is necessary to accurately delineate microvascular
between postoperative days 7 and 14, whereas the current study geometry, such as three-dimensional topology, tortuosity, and
noted a drop from postoperative day 7 to 10. The earlier study vessel length (Brey et al., 2004).
did not reveal a significant change in vessel diameter between Although the vast majority of microvascular growth
postoperative days 3 and 14, which was similar to our current appears to occur from the underlying recipient site, other
findings (Table 5). These findings suggest that once the blood studies have noted that the native skin around or bordering the
supply has been restored to a viable transplant, such as a skin graft (Capla et al., 2006; O'Ceallaigh et al., 2006) also
graft, the number and area fractions of vessels decrease. contributes to the graft's revascularization. In our current
It was a little surprising that the two-dimensional data for study, each skin graft was placed so that it was located several
number and area fraction were no different than the three- millimeters from nearby grafts and from the intact skin around
dimensional data for volume fraction in our current study. The the recipient site, which prevented skin graft revascularization
assays most often used to study tumor angiogenesis are from occurring from anywhere except the underlying muscle.
frequently thought to oversimplify microvascular network Thus, only vessels growing from the tissue under the graft
were measured, enabling quantitative analysis of the vascular
metrics for as long as the fibrin interface was present. Because
Table 4 we used conventional light microscopy at a high power, we
Comparison of area fraction or equivalent in two studies
were able to determine the landmarks that delineate the
Author of study Postoperative Mean value Proportional (times) boundaries of the total imaged area and, thus, to limit our
day increase from
analysis to the microvasculature growing in the fibrin interface
postoperative
day 3 value of the skin graft and recipient site. The use of frozen sections
avoided the shrinkage and processing artifacts of the
Wu et al. (current study)
3 0.009 mm2 microvasculature associated with paraffin and resin fixation
7 0.021 mm2 +2.3 and, at least in theory, should have supplied a more accurate
10 0.012 mm2 +1.3 reflection of the actual dimensions of the microvasculature
Brey et al. (2004) (Baker, 1960).
3 ⁎2.42 ± 3.08
In this study, we did not attempt to distinguish capillaries
7 *4.99 ± 0.71 +2.1
14 *9.68 ± 1.03 +4 from vessels of lymphatic origin because this would require
21 0 0 special molecular antibodies against endothelial cell luminal
Area fraction was not measured in the studies of Capla et al. (2006) and surface antigens that are specific for lymphatics and such
O'Ceallaigh et al. (2006). antibodies were not readily available when we started this study.
⁎ Percentage of endothelial cell area stained. Moreover, although several lymphatic markers have recently
128 X. Wu et al. / Microvascular Research 75 (2008) 119–129
been discovered, the best quantification method for assessing
lymphatic density has not been determined (Witte et al., 2006).
The histology and imaging techniques used in this study did
have some shortcomings, such as the long amount of time
required for processing, the tedious nature of the technique,
technical difficulties, and the need for extensive computer
memory. Nevertheless, the histology and imaging techniques
used allowed us to quantify the microvasculature in the fibrin
interface with a resolution and accuracy that no other currently
available imaging technique allows, to our knowledge.
The data obtained from this study are significant in that they
set a standard for analyzing and comparing microvascular
growth metrics in the fibrin interface of other types of
conventional and tissue-engineered grafts. In addition to
determining vessel density, our method of analysis could be
used to determine the morphology and morphometry of the
microvasculature in the interface of a graft and its recipient site
and to compare the metrics of rat and human skin graft
revascularization, of full-thickness skin grafts and STSGs, of
conventional skin grafts and tissue engineered skin equivalent,
and of skin to other types of tissues (grafts).
Acknowledgments
We thank Linda Vega for assistance with manuscript
preparation. This study was funded in part by grants from The
University of Texas M. D. Anderson Cancer Center (number
80094186), the Plastic Surgery Education Foundation Grant
(number 01-99-00141), the National Institutes of Health (HL
062341), and a generous gift from the Maurice Amado
Foundation.
References
Artacho-Perula, E., Roldan-Villalobos, R., Cruz-Orive, L., 1999. Application of
the fractionator and vertical slices to estimate total capillary length in
skeletal muscle. J. Anat. 195, 429–437.
Baker, J.R, 1960. Principles of Biological Microtechnique. Methuen, London,
p. 32.
Bellman, S., Velander, E., Frank, H.A., Lambert, P.B., 1964. Survival of arteries
in full thickness skin autografts. Transplantation 2, 167–172.
Bert, P., 1865. Notes sur un cas de greffe animale. Comptes Rendus de la Societe
de Biologie 61, 587.
Birch, J., Brånemark, P.I., 1969a. The vascularization of a free full thickness
skin graft. I. A vital microscopic study. Scand. J. Reconstr. Surg. 3, 1–10.
Birch, J., Brånemark, P.I., 1969b. The vascularization of a free full thickness skin
graft. II. A microangiographic study. Scand. Plast. Reconstr. Surg. 3, 11–17.
Black, A.F., Berthod, F., L'Heureux, N., Germain, L., Auger, F.A., 1998. In
vitro reconstruction of a human capillary-like network in a tissue-engineered
skin equivalent. FASEB J. 12, 1331–1340.
Brey, E.M., King, T.W., Johnston, C., McIntire, L.V., Reece, G.P., Patrick Jr.,
C.W., 2002. A technique for quantitative three-dimensional analysis of
microvascular structure. Microvasc. Res. 63, 279–294.
Brey, E., McIntire, L.V., Johnson, C., Reece, G.P., Patrick Jr., C.W., 2004. Mural
cell markers desmin and smooth muscle alpha actin are differentially
expressed during angiogenesis in vivo and are not necessarily indicators of
vessel stability. Ann. Biomed. Eng. 32, 1100–1107.
Boyce, S.T., Goretsky, M.J., Greenhalgh, D.G., Kagan, R.J., Rieman, M.T.,
Warden, G.D., 1995. Comparative assessment of cultured skin substitutes
and native skin autograft for treatment of full-thickness burns. Ann. Surg.
222, 743–752.
Bullitt, E., Liu, A., Prizer, S.M., 1997. Three-dimensional reconstruction of
curves from pairs of projection views in the presence of error. II. Analysis of
error. Med. Phys. 24, 1679–1686.
Capla, J.M., Ceradini, D.J., Tepper, O.M., Callaghan, M.J., Bhatt, K.A., Galiano,
R.D., Levine, J.P., Gurtner, G.C., 2006. Skin graft vascularization involves
precisely regulated regression and replacement of endothelial cells through
both angiogenesis and vasculogenesis. Plast. Reconstr. Surg. 117, 836–844.
Clemmensen, T., 1964. The early circulation in split-thickness skin grafts.
Restoration of blood supply to split-thickness autografts. Acta Chir. Scand.
127, 1–8.
Converse, J.M., Smahel, J., Ballantyne Jr., D.L., Harper, A.D., 1975.
Inosculation of vessels of skin graft and host bed: a fortuitous encounter.
Br. J. Plast. Surg. 28, 275–282.
Davis, J.S., Traut, H.F., 1925. Origin and development of blood supply of
whole-thickness skin grafts. Ann. Surg. 82, 871–879.
Edgerton, M.T., Peterson, H.A., Edgerton, P.J., 1957. The homograft rejection
mechanism. Arch. Surg. 74, 238–244.
Haller Jr., J.A., Billingham, R.E., 1967. Studies of the origin of the vasculature
in free skin grafts. Ann. Surg. 166, 896–901.
Henry, L., Marshall, D.C., Friedman, E.A., Goldstein, D.P., Dammin, G.J.,
1961. A histologic study of the human skin autograft. Am. J. Pathol. 39,
317–486.
Henry, L., Marshall, D.C., Friedman, E.A., Dammin, G.H., Merrill, J.P., 1962.
The rejection of skin homografts in the normal human subject: Part II.
Histological findings. J. Clin. Invest. 41, 420–446.
Hynes, W., 1954. The early circulation in skin grafts with a consideration of
methods to encourage their survival. Br. J. Plast. Surg. 6, 257–263.
Jones, I., Currie, L., Martin, R., 2002. A guide to biological skin substitutes. Br.
J. Plast. Surg. 55, 185–193.
Kamrin, B.B., 1960. Studies on the successful healing of skin homografts in
albino rats. Ann. N.Y. Acad. Sci. 87, 323–351.
Kamrin, B.B., 1961. Analysis of the union between host and graft in the albino
rat. Plast. Reconstr. Surg. 28, 221–226.
King, T.W., Brey, E.M., Youssef, A., Johnston, C., Patrick Jr., C.W., 2002.
Quantification of vascular density: an application of a semiautomated
technique for immunohistochemically stained tissue sections. Anal. Quant.
Cytol. Histol. 24, 39–48.
Kunzi-Rapp, K., Rück, A., Kaufmann, R., 1999. Characterization of the chick
chorioallantoic membrane model as a short-term in vivo system for human
skin. Arch. Dermatol. Res. 291, 290–295.
Lambert, P.B., 1971. Vascularization of skin grafts. Nature 232, 279–280.
Less, J.R., Skalak, T.C., Sevick, E.M., Jain, R.K., 1991. Microvascular
architecture in mammary carcinoma: branching patterns and vessel
dimensions. Cancer Res. 51, 265–273.
Ljungqvist, A., Almgård, L.E., 1966. The vascular reaction in the free skin allo-
and autografts. A stereomicroangiographic and histological study in the
rabbit. Acta Pathol. Microbiol. Scand. 68, 553–562.
Maniotis, A.J., Folberg, R., Hess, A., Seftor, E.A., Gardner, L.M.G., Pe'er, J.,
Trent, J.M., Maltzer, P.S., Hendrix, M.J.C., 1999. Vascular channel
formation by human melanoma cells in vivo and in vitro: vasculogenic
mimicry. Am. J. Pathol. 155, 739–752.
Mansbridge, J., 2002. Tissue-engineered skin substitutes. Expert Opin. Biol.
Ther. 2, 25–34.
Markmann, A., 1966. Autologous skin grafts in the rat: vital microscopic studies
of the microcirculation. Angiology 17, 475–482.
McDonald, D.M., Munn, L., Jain, R.K., 2000. Vasculogenic mimicry: how
convincing, how novel, and how significant? Am. J. Pathol. 156, 383–388.
Medawar, P.B., 1944. The behavior and fate of skin autografts and skin
homografts in rabbits. J. Anat. 78, 176–199.
Neumann, T., Nicholson, B.S., Sanders, J.E., 2003. Tissue engineering of
perfused microvessels. Microvas. Res. 66, 59–67.
Norrby, K., 1998. Microvascular density in terms of number and length of
microvessel segments per unit tissue volume in mammalian angiogenesis.
Microvasc. Res. 55, 43–53.
O'Ceallaigh, S., Herrick, S.E., Bluff, J.E., McGrouther, D.A., Ferguson, M.W.J.,
2006. Quantification of total and perfused blood vessels in murine skin
autografts using a fluorescent double-labeling technique. Plast. Reconstr.
Surg. 117, 140–151.
 X. Wu et al. / Microvascular Research 75 (2008) 119–129
O'Donoghue, M.N., Zarem, H.A., 1971. Stimulation of neovascularization-
comparative efficacy of fresh and preserved skin grafts. Plast. Reconstr.
Surg. 48, 474–477.
Okada, T., 1986. Revascularization of free full thickness skin grafts in rabbits: a
scanning electron microscopy of microvascular casts in full thickness skin
grafts. Br. J. Plast. Surg. 39, 183–189.
Patan, S., Munn, L.L., Tanda, S., Roberge, S., Jain, R.K., Jones, R.C., 2001a.
Vascular morphogenesis and remodeling in a model of tissue repair: blood
vessel formation and growth in the ovarian pedicle after ovariectomy. Circ.
Res. 89, 723–731.
Patan, S., Tanda, S., Roberge, S., Jones, R.C., Jain, R.K., Munn, L.L., 2001b.
Vascular morphogenesis and remodeling in a human tumor xenograft: blood
vessel formation and growth after ovariectomy and tumor implantation.
Circ. Res. 89, 732–739.
Reece, G., Patrick Jr., C., 1998. Tissue engineered construct design
principles. In: Patrick Jr., C.W., Mikos, A.G., McIntire, L.V. (Eds.),
Frontiers in Tissue Engineering. Elsevier Science (Pergamon), Oxford,
pp. 166–196.
Rolle, G.K., Taylor, A.C., Charipper, H.A., 1959. A study of vascular changes in
skin grafts in mice and their relationship to homograft breakdown. J. Cell
Comp. Physiol. 53, 215–239.
Scothorne, R.J., McGregor, I.A., 1953. The vascularization of autografts and
homografts of rabbit skin. J. Anat. 87, 379–386.
Smahel, J., 1968. The problem of revascularization of free skin autografts. Acta
Chir. Plast. 11, 78–84.
Smahel, J., Ganzoni, N., 1970. Contribution to the origin of the vasculature in
free skin autografts. Br. J. Plast. Surg. 23, 322–325.
129
Soejima, K., Negishi, N., Nozaki, M., Sasaki, K., 1998. Effect of cultured
endothelial cells on angiogenesis in vivo. Plast. Reconstr. Surg. 101,
1552–1560.
Taylor, A.C., Lehrfeld, J.W., 1953. Determination of survival time of skin
homografts in rat by observation of vascular changes in graft. Plast.
Reconstr. Surg. 12, 423–431.
Ueda, M., Sumi, Y., Kaneda, T., Eto, K., 1986. Revascularization of reversed
dermis grafts. J. Oral Maxillofac. Surg. 44, 534–540.
U.S. Department of Health and Human Services, Center for Disease Control and
Prevention/National Center for Health Statistics. Ambulatory and Inpatient
Procedures in the United States. Vital and Health Statistics. November 1998,
Series 13, No. 139, p. 28.
U.S. Department of Health and Human Services, Center for Disease Control and
Prevention/National Center for Health Statistics. National Hospital Dis-
charge Survey: 2002 Annual Summary with Detailed Diagnosis and
Procedure Data. Vital and Health Statistics. March 2005, Series 13, No.
158, pp. 178–179.
Witte, M.H., Jones, K., Wilting, J., Dictor, M., Selg, M., McHale, N.,
Gershenwald, J.E., Jackson, D.G., 2006. Structure function relationships in
the lymphatic system and implications for cancer biology. Cancer Metastasis
Rev. 25, 159–184.
Young, D.M., Greulich, K.M., Weier, H.G., 1996. Species-specific in situ
hybridization with fluorochrome-labeled DNA probes to study vasculariza-
tion of human skin grafts on athymic mice. J. Burn Care Rehabil. 17,
305–310.
Zarem, H.A., Zweifach, B.W., 1967. Development of microcirculation in full
thickness skin grafts in mice. Am. J. Physiol. 212, 1081–1085.