J Periodontol • February 2016

Clinical Comparison of Platelet-Rich

Fibrin and a Gelatin Sponge in the
Management of Palatal Wounds After
Epithelialized Free Gingival Graft
Harvest: A Randomized Clinical Trial
Beatrice Femminella,* Maria Chiara Iaconi,* Marcella Di Tullio,* Luigi Romano,* Bruna Sinjari,*
Camillo D’Arcangelo,* Paolo De Ninis,† and Michele Paolantonio*

Background: Platelet-rich fibrin (PRF) promotes tissue regen-

eration by releasing various growth factors. The palatal donor site
of the epithelialized connective tissue (CT) graft significantly in-
fluences the patient’s morbidity. The aim of this study is to
compare the effects of PRF and gelatin sponge on the healing
of palatal donor sites and the patient’s morbidity.
Methods: Forty patients with at least one site of Miller Class I

M
any surgical techniques have been
or II gingival recession were treated by a coronally advanced proposed for the correction of
flap with CT graft resulting from the de-epithelialization of a gingival recession.1 Although the
free gingival graft. In the test group (20 patients), a PRF membrane amount of root coverage yielded by re-
was placed over the palatal wounds; conversely, the 20 control generative techniques2,3 and pedicle grafts4
group patients were treated with an absorbable gelatin sponge. is similar to that of free graft procedures,
Patients were monitored at 1, 2, 3, and 4 weeks after surgery these techniques produce less of an in-
for the complete re-epithelialization of the palatal wound (CWE), crease in gingival thickness.5 Conversely,
the alteration of sensitivity around the wound area, postoperative bilaminar techniques (BTs) consisting
discomfort, and changes in feeding habits (CFH). Furthermore, of the association of a connective tissue
the consumption of analgesics during the postoperative week 1 graft (CTG) with a pedicle graft are pres-
was assessed. ently considered the most predictable
Results: The test group showed a significantly faster CWE treatment choices to achieve exposed root
(P <0.001); 35% of the test patients showed CWE at the end of coverage6 while providing the greatest in-
week 2 (controls, 10%), whereas at the end of week 3, all palatal crease in gingival thickness.2
wounds in the test patients epithelialized completely (controls, Gingival thickness is very important
25%). Similarly, test patients reported significantly less discom- for long-term stability; in fact, gingival
fort and CFH (P £ 0.02) and took a significantly lower dose of thickness plays an important role in pre-
analgesics (P = 0.02). venting the recurrence of recession.7 The
Conclusion: The PRF-enriched palatal bandage significantly main disadvantage of BTs is the need for
accelerates palatal wound healing and reduces the patient’s a palatal wound, which often produces pain
morbidity. J Periodontol 2016;87:103-113. and discomfort.4
Many tissue-harvesting procedures have
KEY WORDS
been described previously.8,9 The epi-
Autografts; biocompatible materials; clinical trial; pain; thelialized free gingival graft (EFGG)9 is
palate; wound healing. easy to perform and enables the harvest
of large quantities of high-quality connec-
* Department of Medical, Oral, and Biotechnological Sciences; G. D’Annunzio University; tive tissue (CT). Conversely, it produces
Chieti-Pescara, Italy.
† Luisa D’Annunzio Institute for High Culture, Pescara, Italy. a site of secondary-intention wound heal-
ing with discomfort and pain.10,11
To overcome this problem, different
procedures with primary-intention healing

doi: 10.1902/jop.2015.150198

103
PRF and Gelatin Sponge in the Management of Palatal Wounds Volume 87 • Number 2

have been proposed.8 However, they are more dif- root coverage.12 Conversely, the more superficial
ficult to perform and require the presence of thick tissue from the EFGG is stable and dense tissue,
palatal tissues to obtain a sufficiently thick graft and suitable for use in root coverage.
to avoid necrosis of the epithelial–connective layer To reduce the postoperative problems in the EFGG
left at the donor site. Furthermore, this graft usually donor site, Rossmann and Rees13 suggest the use of
contains a considerable amount of fatty and glan- a hemostatic dressing. Platelet-rich fibrin (PRF) is
dular tissues with less CT; it may be inadequate for a platelet concentrate obtained by a simple and in-
expensive procedure that does
not require biochemical blood
handling; its three-dimensional
fibrin network promotes effective
neovascularization, accelerated
wound closing, and fast cicatri-
cial tissue remodeling.14 There-
fore, PRF is used in many fields
of regenerative medicine,15 in-
cluding orthopedics,16 oral and
maxillofacial surgery,17 and sports
medicine;16 it has also been used
with interesting results for the
treatment of skin wound ulcers.18
Recently, Aravindaksha et al.19
presented results from four pa-
tients whose palatal donor sites
were covered by PRF membranes
as palatal bandages, which showed
very fast healing.
To the best of the authors’
knowledge, to date, no random-
ized trials have evaluated the
usefulness of PRF in the man-
agement of soft tissue donor
sites by testing whether it could
accelerate tissue healing and
reduce the patient’s morbidity
compared with a conventional
hemostatic material. This is the
aim of the present study.

MATERIALS AND
METHODS
Experimental Design
This was a prospective, controlled
randomized clinical trial (RCT)
with a parallel design, performed
to evaluate the healing time and
the patients’ morbidity produced
by the harvest of an EFGG from
the palate. The wounds were
treated with PRF (test group) and
with an absorbable gelatin sponge
(control group) (Fig. 1).20

Study Population
Figure 1. Forty patients (15 males and 25
20
CONSORT (Consolidated Standards of Reporting Trials) diagram showing the study layout.
females, aged 18 to 47 years

104
J Periodontol • February 2016 Femminella, Iaconi, Di Tullio, et al.

[mean age, 32.4 – 5.0 years]) who sought treatment Surgeon Training
at the Unit of Periodontology of G. D’Annunzio Uni- To minimize differences related to the surgical tech-
versity for at least one site of Miller Class I or II re- nique, all procedures were performed by one experi-
cession21 (‡3 mm in depth) were selected for this enced clinician (MP).
study. To obtain grafts of similar size from all patients,
The inclusion criteria for this research were as fol- producing wounds of similar characteristics and di-
lows: 1) systemic factors (no systemic diseases; no mensions, the surgeon underwent preclinical training on
coagulation disorders; no medications affecting peri- animal tissues with the objective of withdrawing a 15 ·
odontal status in the previous 6 months; no pregnancy 8–mm graft of an even 2-mm thickness, as measured in
or lactation); 2) behavioral factors (no smoking habit); its central part by means of a caliper. The training was
and 3) dental and periodontal factors (a full-mouth continued until the size of the graft (height, width, and
plaque score [FMPS]22 and a full-mouth bleeding score thickness) differed by no more than 5% in five con-
[FMBS]23 lower than 20% at the time of surgery; no secutive samples.
periodontal surgery on the experimental sites; no in-
Presurgical Treatment
adequate endodontic treatment or tooth mobility at the
All selected patients underwent a session of prophylaxis
site of surgery).
with instruction in proper oral hygiene measures and
The participants volunteered for the study after they
professional tooth cleaning. The use of an electric
received verbal and written information and signed
toothbrush with an extrasoft head‡ with controlled
a consent form. The protocol was approved by the
pressure§ was recommended. The patients were in-
Ethical Committee of G. D’Annunzio University for
structed concerning the optimal use of the electric
human participants. The study protocol was in accor-
toothbrush, dental floss, and/or interdental brush.
dance with the Declaration of Helsinki of 1975, revised
in Tokyo in 2004. This study was performed from May Intrasurgical Measurement
2013 to January 2014. After local anesthesia, the thickness of the palatal soft
This study is registered at ClinicalTrials.gov as tissues in the harvesting area was measured according
NCT02438046. to Paolantonio.2 The measurement was made at the
midpalatal location, 5 mm apical to the gingival
Sample Size and Randomization margin of the first premolar, by means of a no. 15
The primary outcome was to assess the time needed endodontic reamer. The reamer was inserted perpen-
to obtain complete re-epithelialization of the palatal dicular to the mucosal surface through the soft tissue
wound (CWE). Secondary outcomes were to evalu- with light pressure until a hard surface was felt. The
ate the following: 1) the alteration of sensitivity (AS) silicone disk stop was then placed in tight contact with
around the palatal wound; 2) the postoperative dis- the soft tissue surface and fixed with a drop of cyano-
comfort (D); 3) changes in patients’ feeding habits acrylate; after careful removal of the reamer, the pen-
(CFH); 4) the consumption of analgesics; and 5) the etration depth was measured with a caliper accurate to
existence of delayed bleeding from the palatal wound the nearest 0.1 mm. The thickness of the grafts was
(DWB) during postoperative week 1. measured in both the test and control groups using
The sample size was calculated to provide a power a caliper positioned at the central part of the graft.
1-b = 90% to detect the difference in the proportion of The mesial–distal dimensions and the apical-coronal
patients who exhibited epithelialization after 3 weeks dimensions of the grafts were measured with a manual
among patients undergoing FGG and single-incision probei and rounded up to the nearest millimeter. Graft
procedures, as reported in the study by Del Pizzo measurements were performed by a different examiner
et al.,11 with a = 0.05. At a minimum, 17.3 patients (BF).
per treatment arm would have been required. Twenty
Surgical Technique
patients per group were recruited to avoid a loss of
The 40 patients had their gingival recession treated
statistical power as a consequence of patient drop-
by a coronally advanced flap (CAF) + CTG surgical
out. The balance of experimental groups by age and
technique.
sex was tested by Student t test for unpaired samples
After local anesthesia, the EFGG was harvested as
and x2 analysis, respectively.
follows: the donor site extended from the distal line
Each patient was allocated randomly to one of the
angle of the canine to the mesial line angle of the
experimental groups. Assignment was performed by
maxillary first molar. The most coronal horizontal
a custom-made computer-generated table. To conceal
incision, 15 mm long, was made 2 mm apical from
allocation, opaque envelopes containing the treatment
of the specific patients were assigned to the specific
‡ Oral B Sensitive EBS17, Procter & Gamble, Gattatico, Italy.
patient and were opened during surgery, immediately § Oral-B Pro 6000 CrossAction; Procter & Gamble.
before fabricating the palatal bandage. i XP 23/UNC 15, Hu-Friedy, Chicago, IL.

105
PRF and Gelatin Sponge in the Management of Palatal Wounds Volume 87 • Number 2

the gingival margin; a second horizontal incision of Table 1.

the same length was drawn 8 mm away from the first,
Palatal Thickness and Graft Dimensions in
in a more apical position. Two vertical incisions were
made to join the ends of the horizontal incisions and the Experimental Groups
to delimitate the graft area. A rectangular-shaped
partial-thickness incision was drawn to obtain a graft Test Group Control Group Difference
2 mm in thickness. The EFGG was then measured; Palatal thickness 3.62 – 1.22 3.90 – 1.30 NS
afterward it was trimmed, and the fatty tissue was
eliminated. Then, the graft was de-epithelialized with Graft thickness 2.11 – 0.81 1.89 – 0.76 NS
a 15c blade and adapted to the tooth to be grafted. Graft width 14.89 – 2.22 15.02 – 3.01 NS
In the test group (n = 20 patients), the palatal wound
was protected by a quadruple layer of PRF, obtained by Graft height 8.11 – 01.55 7.93 – 1.67 NS
folding two PRF membranes on themselves; conversely, NS = not significant.
the control group patients (n = 20) had their wound
medicated by an absorbable gelatin sponge.¶ Both cally treated area was maintained by chlorhexidine
bandages were maintained in situ by compressive sling rinsing for an additional 1 week after suture removal.
3-0 silk sutures.# Patients were instructed again in mechanical tooth
To prevent immediate and/or delayed bleeding from cleaning of the grafted area using an ultrasoft manual
the donor site, in both experimental groups, two vertical toothbrush for 1 month. Patients were recalled once
mattress sutures were made mesial and distal to the a week for the first 4 weeks after the surgery, when
tissue harvesting site. The ligatures were made with they underwent gentle supragingival professional
a 2-0 silk suture and a semicircular needle.** The tooth cleaning and oral hygiene reinforcement. Then,
needle was inserted 0.5 mm coronal to the coronal the patients were enrolled into a 3-month mainte-
horizontal incision and 2 mm distal or mesial to the nance program.
vertical distal or mesial incision. The needle was left to
slide on the bony surface of the alveolar process and Wound Healing and Patient Morbidity
was more apically resurfaced beyond the apical hori- Patients were monitored at 1, 2, 3, and 4 weeks after
zontal incision of the donor site. These sutures were surgery. The following parameters were recorded: 1)
made with the aim of choking the blood vessels in the CWE; 2) AS; 3) D; and 4) CFH. Furthermore, post-
submucosa, thereby reducing the bleeding tendency. operative pain during the first week was evaluated by
recording the patients’ mean analgesic (ketoprofen)
PRF Preparation consumption (in milligrams).25 In particular, patients
The PRF was prepared according to Choukroun et al.24 were instructed to take ketoprofen if they felt signifi-
Immediately before surgery, a 40-mL blood sample was cant pain. Patients who did not fulfill this condition
taken by venipuncture of the antecubital vein without were excluded from the per-protocol analysis.
anticoagulant, and it was divided into four tubes of The number of episodes of DWB during the first
10 mL each. The tubes were centrifuged immediately postoperative week was also recorded.
by a dedicated centrifuge†† at 3,000 rpm for 10 min- CWE was evaluated clinically by the peroxide test.26
utes. Such a preparation protocol produces a structured This test is based on the principle that if the epithelium
fibrin clot in the middle of the tube, between the is discontinuous, then H2O2 diffuses into the CT; the
erythrocytes at the bottom and acellular plasma at the enzyme catalase acts on H2O2 to release water and
top. After removal of acellular plasma, the PRF was oxygen. This is shown clinically by the production of
separated from the erythrocytes using sterile scissors; bubbles on the wound. The area to be evaluated was
fibrin membranes were obtained by squeezing the se- dried, and 3% H2O2 was sprinkled on the wound with
rum from the clot with a specific mechanical press.‡‡ a syringe, waiting for the appearance of bubbles, which
Each membrane thus obtained was turned in on it- suggested that the surgical site was not completely
self, and two membranes placed one over the other epithelialized. CWE was recorded as a dichotomous
(quadruple PRF layer) represented the palatal bandage variable (yes/no). For each observation week, the
of the test group. number of new patients experiencing CWE was
Postoperative Care
All patients were administered 2 g/d amoxicillin plus ¶ Surgifoam, Ethicon, Johnson & Johnson, Pomezia, Italy.
clavulanic acid for 6 days;§§ pain was controlled by oral # Ethicon, Johnson & Johnson.
ketoprofen if needed.ii Patients were advised to rinse **
††
PC-02, Process, Nice, France.
IntraSpin Centrifuge, Intra-Lock International, Boca Raton, FL.
twice a day with 0.12% chlorhexidine digluconate sol- ‡‡ IntraSpin Xpression fabrication kit, Intra-Lock International.
§§ Augmentin, SmithKline Beecham, Milan, Italy.
ution¶¶ for 3 weeks after surgery. Sutures were removed ii OKi 80, Dompé, L’Aquila, Italy.
14 days after the surgery. Plaque control in the surgi- ¶¶ Dentosan 0.12 monthly treatment, Johnson & Johnson.

106
 Table 2.
Clinical Parameter Scores for Each Week and Pooled Data, Respectively, Using Univariate Tests
Group · Time Effect, 2 · 4 Contingency
Tables, Log-Likelihood Ratio

Week 1 Week 2 Week 3 Week 4 Weeks 1 to 4

Dependent
variable Test Difference Control Test Difference Control Test Difference Control Test Difference Control Test Difference Control
J Periodontol • February 2016

2 · 2 Contingency Tables (Bonferroni Adjusted Fisher Exact Tests)

Distribution free
CWE in the
week (%)
ITT 0 NS 0 35 NS 10 65 P = 0.003 15 0 P <0.001 70 P <0.001
PP 0 NS 0 35 NS 11 65 P = 0.003 11 0 P <0.001 72 P <0.001

Univariate Analyses (Mann–Whitney U test)

AS (VAS) PMe
ITT 4 NS 4 2.25 NS 3 2 NS 2 1.50 NS 1.50 10 NS 10.5
PP 4 NS 4 2.25 NS 3 2 NS 2 1.50 NS 1.50 10 NS 10.5

D (VAS) PMe
ITT 2 P <0.001 4.50 2 P <0.001 4 1 P <0.001 2.50 0 P = 0.02 1 5.5 P <0.001 12
PP 2 P <0.001 4.50 2 P <0.001 3.50 1 P <0.001 2 0 P = 0.02 1 5.5 P <0.001 11.5

CFH (VAS) PMe

ITT 3.50 P = 0.01 4.50 1.75 P <0.001 2.50 1 P <0.001 2.50 0 NS 0.50 7 P <0.001 9.50
PP 3.50 P = 0.02 4.50 1.75 P <0.001 3 1 P <0.001 2.50 0 NS 0.50 7 P <0.001 9.50

Analgesic usage
(mg), mean
ITT 204 NS 296
PP 204 P = 0.04 328

Group Effect,
Bonferroni-Adjusted
Univariate Tests in
Group · Time Effect, Bonferroni-Adjusted Univariate Tests in RM-MANOVA RM-MANOVA

Parametric
AS (VAS),
mean – SE
ITT 4.15 – 0.18 NS 4.3 – 0.18 2.6 – 0.2 NS 2.85 – 0.2 1.85 – 0.11 NS 1.95 – 0.11 1.5 – 0.11 NS 2.52 – 0.1 NS 2.67 – 0.1
PP 4.15 – 0.18 NS 4.33 – 0.19 2.6 – 0.2 NS 2.88 – 0.22 1.85 – 0.11 NS 1.94 – 0.12 1.5 – 0.11 NS 2.52 – 0.1 NS 2.68 – 0.11
D (VAS),
mean – SE
ITT 2.4 – 0.2 P <0.001 4.6 – 0.2 1.75 – 0.22 P <0.001 3.75 – 0.22 1.1 – 0.18 P <0.001 2.6 – 0.18 0.15 – 0.17 P = 0.004 1.35 – 0.12 P <0.001 2.96 – 0.12
PP 2.4 – 0.2 P <0.001 4.5 – 0.21 1.75 – 0.21 P <0.001 3.55 – 0.22 1.1 – 0.17 P <0.001 2.44 – 0.18 0.15 – 0.17 P = 0.02 1.35 – 0.11 P <0.001 2.86 – 0.12
Femminella, Iaconi, Di Tullio, et al.

107
PRF and Gelatin Sponge in the Management of Palatal Wounds Volume 87 • Number 2

calculated; patients whose palatal wounds were pre-

2.46 – 0.08
2.45 – 0.08
viously completely healed were not considered again.
AS, D, and CFH were evaluated by showing the in-
Bonferroni-Adjusted
Univariate Tests in
RM-MANOVA tensity of the given event on a 100-mm visual analog
Group Effect,

scale (VAS)27 divided into 10 segments of 10 mm each

P <0.001
P <0.001
and numbered from 0 to 10. AS was scored (0 to 10) by
asking the patient to compare the sensitivity he/she had
when the surrounding areas of the wound were touched
1.68 – 0.08
1.68 – 0.08

by the tip of a periodontal probe and comparing this

feeling to the feeling that he/she felt on the other side of
Clinical Parameter Scores for Each Week and Pooled Data, Respectively, Using Univariate Tests

the palate. D was assessed as the level of pain (0 to 10)

experienced by the patients during the postoperative
NS
NS

experimental weeks as a result of the palatal wound.

CFH was described as the degree of change in the
patient’s eating habits (0 to 10) resulting from the
0.25 – 0.1
0.25 – 0.1

presence of the palatal wound.

DWB was considered to represent the occurrence of
prolonged hemorrhaging during the first post-surgical
2.35 – 0.13
2.27 – 0.14

week. This parameter was recorded as the number of

observed episodes.
Group · Time Effect, Bonferroni-Adjusted Univariate Tests in RM-MANOVA

ITT = intention-to-treat analysis; PP = per-protocol analysis; PMe: Hodges-Lehmann pseudo-median score; NS = not significant.

Data Analyses
P <0.001
P <0.001

A double parallel analysis was performed: 1) an ‘‘in-

tention-to-treat’’ analysis on data from all the patients;
and 2) a ‘‘per-protocol’’ analysis after excluding patients
1.2 – 0.13
1.2 – 0.13

who did not respect the experimental protocol, i.e., they

did not consume analgesics although they experienced
significant pain.
The main outcome analysis was made by comparing
2.75 – 0.14
2.77 – 0.15

experimental groups on the frequency distribution of

patients who experienced CWE in each of the 4 post-
operative weeks. The Williams-adjusted log-likelihood
ratio was performed, followed by post hoc tests using
P <0.001
P <0.001

Bonferroni-corrected Fisher exact test (two-tailed).

The other parameters were analyzed using a distri-
bution-free test (Mann–Whitney U test). Because in
1.7 – 0.14
1.7 – 0.14

a previous study12 the dependent variables were ana-

lyzed in univariate contexts, multiple univariate ana-
lyses were performed,28 reporting the correlation matrix
4.4 – 0.19

296 – 43.2
328 – 40.8

among dependent variables according to Huberty and

4.38 – 0.2

Morris.28
Similarly, for the VAS score analyses, the distribution-
free test was compared with a repeated-measures
P = 0.005
P = 0.02

P = 0.02

multivariate analysis of variance (RM-MANOVA),29 es-

NS

timating the time · treatment effect and the confidence

intervals (CIs).
3.6 – 0.19
3.6 – 0.19

Finally, for the first week, differences in painkiller

204 – 32
204 – 32

consumption were evaluated by the Mann–Whitney U

Table 2. (continued )

test, followed by a correlation matrix for D, CFH, AS, and

painkiller consumption, assessing the association level
(mg), mean – SE

between the outcome variables. The D, AS, and CFH

mean – SE

Analgesic usage

scores were analyzed using the 4 weeks of pooled data

CFH (VAS),

and then the data from each single week. Student t test
was used to analyze the differences between the control
ITT

ITT
PP

PP

and test groups in the thickness of the palatal mucosa

and the graft size. Significance was set at a = 0.05.

108
J Periodontol • February 2016 Femminella, Iaconi, Di Tullio, et al.

RESULTS
The experimental groups were bal-
anced by age and sex (P >0.05). No
patients dropped out of the study,
and no postoperative complications
were reported by the patients.
The FMPS and FMBS remained
<20% throughout the entire study
without significant differences be-
tween the groups.
At the first week of examination,
two patients in the control group
reported that they did not take
analgesics although they had felt
significant pain. Therefore, two
separate data analyses were per-
formed: the first according to the
intention-to-treat principle and the
second according to the per-protocol
analysis.
The thickness of the palatal tis-
sues and the dimensional charac-
teristics of the grafts from test and
control patients did not show sig-
nificant differences (Table 1).
Table 2 summarizes the results
obtained in this study. The test
group showed a significantly faster
CWE (P <0.001); more than one
third of test patients showed a CWE
at the end of postoperative week 2,
and at the end of week 3, the palatal
wounds of all patients treated with
PRF had completely epithelialized;
in contrast, one patient from the
control group did not show com-
plete healing at the end of post-
operative week 4 (Fig. 2).
A similar trend was shown by
the assessments of D and CFH: in
fact, as early as the postoperative
week 1, patients in the test group
reported a more favorable evolu-
tion of these parameters; these
differences remained significant
until the end of week 3.
A difference in the AS level be-
tween the experimental groups was
Figure 2. never detected.
Healing process at test and control donor sites.
The per-protocol analysis showed
that, in week 1, patients from the test
group took a significantly lower dose
of analgesics compared with the
control group patients. This differ-
ence was not statistically significant

109
PRF and Gelatin Sponge in the Management of Palatal Wounds Volume 87 • Number 2

when analyzing the data according to the

RM-MANOVA, Test-Control Univariate Differences Between the Means and 95% CIs for Each Week on Clinical Data

0.225
0.223

-0.259
-0.285

0.199
0.172
Upper
Limit
intention-to-treat principle.

95% CI
No episodes of DWB were reported by any

-0.425
-0.446

-1.241
-1.304

-0.399
-0.450
Lower
Limit
test or control patient during week 1.
The similarity of the results from distri-

Significance P
Bonferroni-
Adjusted bution-free and parametric tests allowed for

0.004
0.003
NS
NS

NS
NS
Week 4

the calculation of the CIs (Table 3).28 The

magnitude of the differences between the two
0.16 experimental groups for the dependent pa-
0.16

0.24
0.25

0.14
0.15
SEd

rameters D and CFH suggests the existence of

Test – Control

a significant difference from both the statis-

the Means
Difference
Between

-0.100
-0.111

-0.750
-0.794

-0.100
-0.139
tical and clinical points of view.
According to Huberty and Morris,28 Table 4
shows the correlations among the dependent
0.220
0.244

-0.964
-0.826

-0.765
-0.685
Upper
Limit

variables; a strong correlation among all

95% CI

variables was shown, with the only exception

-0.420
-0.433

-2.036
-1.862

-1.535
-1.471
Lower
Limit

being AS.
RM-MANOVA: Group · Time Effect, Bonferroni-Adjusted Univariate Tests and CIs

Significance P
Bonferroni-

DISCUSSION
Adjusted

<0.001
<0.001

<0.001
<0.001
NS
NS
Week 3

Although CAF alone produces excellent root-

ITT = intention-to-treat analysis; PP = per-protocol analysis; SEd = standard error of differences between means; NS = not significant.
coverage results in the short term,30 it pro-
0.15
0.16

0.26
0.25

0.19
0.19
SEd

duces a considerable recurrence of recession

in long-term follow-up.31 However, the addi-
Test – Control

the Means
Difference
Between

tion of a CTG to CAF, while improving the

-0.100
-0.094

-1.500
-1.344

-1.150
-1.078

long-term results,32 requires a palatal surgical

site. The use of heterologous materials has
been reported to yield fewer benefits than
0.350
0.381

-1.347
-1.182

-0.635
-0.648
Upper
Limit

autogenous grafts.33 The need for a palatal

95% CI

donor site makes periodontal plastic surgery

-0.850
-0.848

-2.653
-2.429

-1.465
-1.507
Lower
Limit

an often painful procedure.25,34

Although many techniques with primary-
Significance P
Bonferroni-
Adjusted

<0.001
<0.001

<0.001
<0.001

intention healing have been described for CT

NS
NS
Week 2

harvesting,8 it may sometimes be necessary

to take a conventional EFGG, for example,
0.29
0.30

0.32
0.30

0.20
0.21
SEd

when the palatal tissue is very thin. Moreover,

EFGG is easier to obtain and requires less
Test – Control

the Means
Difference
Between

-0.250
-0.233

-2.000
-1.806

-1.050
-1.078

operative time. Furthermore, the high quality

of the dense subepithelial tissue is very dif-
ferent from that of the deep palatal tissue,
0.367
0.361

-1.616
-1.506

-0.254
-0.217
Upper

which is rich in adipose and glandular mate-

Limit
95% CI

rial, is less consistent and not suitable for root

-0.667
-0.727

-2.784
-2.694

-1.346
-1.360
Lower

coverage, and produces a lesser increase in

Limit

buccal soft tissue thickness.35,36

In a recent case report, Aravindaksha et al.19
Significance P
Bonferroni-
Adjusted

<0.001
<0.001

0.005
0.008
NS
NS

showed that the use of a PRF membrane as

Week 1

a palatal bandage is effective in accelerating

soft tissue healing. The present results from this
0.25
0.26

0.28
0.29

0.27
0.28
SEd

RCT on 40 patients confirm this observation by

Test – Control

comparing the use of a PRF membrane with the

the Means
Difference

-0.150
-0.183

-2.200
-2.100

-0.800
-0.789

P <0.001
P <0.001
Between

use of a commonly used hemostatic agent.

When comparing the present data to those
from other similar studies, it should be con-
Multivariate
time effect:
RM-MANOVA

sidered that great heterogeneity exists in the

group ·
Table 3.

CFH (VAS)
Dependent

treatment of palatal wounds. In fact, Del Pizzo

AS (VAS)

test
D (VAS)
Variable

ITT

ITT

ITT

ITT

et al.11 did not use any bandage, Wessel and

PP

PP

PP

PP

Tatakis25 protected the donor site with a stent,

110
J Periodontol • February 2016 Femminella, Iaconi, Di Tullio, et al.

and Zucchelli et al.12 treated the wound with equine-

<0.001
<0.001
0.24
0.27
derived collagen. Conversely, Rossmann and Rees13

Spearmann

P
evaluated three treatment modalities for the donor site:

0.148
0.185

0.809
0.792
CFH (VAS)
1) oxidized cellulose; 2) an absorbable gelatin sponge;

r
and 3) sterile gauze with external pressure.

<0.001
<0.001
0.21
0.22
P In the present study, 35% of the test group showed
Pearson

CWE at the week-2 visit, and 100% showed CWE at

0.201
0.206

0.794
0.785
week 3, with a significant difference relative to the
R

control group (P = 0.003).

Pooled Data Correlation Matrix

0.051

<0.001
<0.001
0.07
These data agree with results from Aravindaksha
Spearman

0.286 et al.,19 who reported complete healing at week 3 in

0.318

0.809
0.792
100% of PRF-treated patients.
r
D (VAS)

A similar outcome was reported by Del Pizzo et al.11

<0.001
<0.001
0.07
0.04

in patients treated with the single-incision technique.37

P
Pearson

This method is currently considered to be less trau-

0.288
0.333

0.794
0.785
matic, leading to primary-intention healing.11
R

Although this study does not include patients treated

0.051

0.27
0.07

0.24
Spearman

with the single-incision technique,37 a comparison of

the present data to those from Del Pizzo et al.11 sug-
0.318

0.185
0.286

0.148
AS (VAS)

gests that the EFGG technique, along with a PRF

0.04

0.22
0.07

0.21

bandage, may lead to similar postoperative morbidity

P
Pearson

and offer greater ease and speed in the procedure and

0.333

0.206
0.288

0.201

a better graft quality.35

R

In this study, at week 4, complete palatal wound

0.001
<0.001
0.03
NS

NS
NS
Spearman

healing is observed in 95% of the control patients.

Week 1 and Pooled Data Correlation Matrices From Clinical Data

This observation was different from that of Del Pizzo

0.267
0.360

0.085
0.097

0.531
0.521
CFH (VAS)

et al.,11 who reported complete healing in 16% and

r

50% of patients at weeks 2 and 3, respectively, in

0.001
0.001
NS

NS
NS
0.04
P
Pearson

EFGG donor sites without any bandage treatment.

This can be explained by the difference in methods
0.275
0.340

0.137
0.148

0.518
0.504
R

for evaluating healing. In fact, Del Pizzo et al.11 as-

0.001
<0.001

sessed the complete wound epithelialization on the

0.03
NS

NS
NS
Spearman

basis of a clinical evaluation; conversely, in the present

study, the epithelial barrier was considered to be com-
0.196
0.360

0.073
0.104

0.531
0.521
D (VAS)

ITT = intention-to-treat analysis; PP = per-protocol analysis; NS = not significant.

pletely formed when no bubble formation occurred after

Week 1 Correlation Matrix

0.001
0.001

H2O2 irrigation. The latter may be considered to be

NS

NS
NS
0.04
P
Pearson

a more objective and reliable evaluation method.

0.191
0.336

0.101
0.130

0.518
0.504

It was hypothesized that the important biologic me-

R

diators within PRF are responsible for the shorter amount

NS
NS

NS
NS

NS
NS
Spearman

of time needed for CWE in the test group. The fibrin clot
0.203
0.200

0.073
0.104

0.085
0.097

is enough in itself to account for the significant cicatricial

r
AS (VAS)

capacity of the PRF: in fact, the physiologic three-

NS
NS

NS
NS

NS
NS
P

dimensional fibrin network leads to more effective cell

Pearson

migration and proliferation.14 PRF represents a combi-

0.231
0.223

0.101
0.130

0.137
0.148
R

nation of cytokines, structural glycoproteins, and gly-

0.03

0.03
NS
NS

NS

NS

canic chains that play a synergetic role in healing and

Spearman

P
Analgesic Usage (mg)

stimulating angiogenesis, immunity, and epithelializa-

0.203
0.200

0.196
0.360

0.267
0.360

tion.14 Angiogenesis is enhanced by soluble factors such

r

as vascular endothelial growth factor, platelet-derived

0.04

0.04
NS
NS

NS

NS
P
Pearson

growth factor (PDGF), and fibroblast growth factor.14

0.231
0.223

0.191
0.336

0.275
0.340

Fibroblast proliferation and epithelial cell migration are

R

influenced positively by the presence of fibrin, fibro-

usage (mg)

nectin, PDGF, and transforming growth factors.14

Table 4.

CFH (VAS)
Dependent

Another interesting finding from the present study is

AS (VAS)
Analgesic

D (VAS)
Variable

ITT

ITT

ITT

ITT

the lower degree of D and the minor CFH observed in

PP

PP

PP

PP

the test group. This observation cannot be compared

111
PRF and Gelatin Sponge in the Management of Palatal Wounds Volume 87 • Number 2

with others in the literature because the only other clinical advantages in accelerating palatal wound healing
study of PRF palatal bandages19 did not evaluate the and reducing the patient’s morbidity.
patients’ morbidity.
It could be hypothesized that the presence of mul- ACKNOWLEDGMENTS
tilayer PRF exerts mechanical protection, covering the This study was supported by a grant (Ex 60%-014)
injured sensitive structures exposed by surgery. This from the Italian Ministry of University and Scientific
may explain why the PRF-treated patients showed Research. The authors report no conflicts of interest
significantly less D and CFH as early as week 1. related to this study.
Zucchelli et al.12 suggested that the thickness of the
remaining soft tissues covering the palatal bone plays REFERENCES
a pivotal role in reducing the patients’ postoperative 1. Chambrone L, Sukekava F, Araújo MG, Pustiglioni FE,
discomfort and that ‡2 mm of soft tissue should be left Chambrone LA, Lima LA. Root-coverage procedures
for the treatment of localized recession-type defects: A
to cover the bone. In the present study, the mean soft Cochrane systematic review. J Periodontol 2010;81:
tissue thickness remaining on the palatal bone in the 452-478.
test group (1.51 mm) (Table 1) is inadequate to pre- 2. Paolantonio M. Treatment of gingival recessions by
vent significant discomfort. It could be speculated that combined periodontal regenerative technique, guided
the multilayered PRF placed on the wound may have tissue regeneration, and subpedicle connective tissue
graft. A comparative clinical study. J Periodontol 2002;
produced the same protective effects as a thick residual 73:53-62.
layer of CT. 3. Pini Prato G, Clauser C, Cortellini P, Tinti C, Vincenzi G,
The accelerated healing process induced by PRF Pagliaro U. Guided tissue regeneration versus muco-
may explain the test patients’ more favorable condi- gingival surgery in the treatment of human buccal
tions during the following weeks. The reduced pain recessions. A 4-year follow-up study. J Periodontol
1996;67:1216-1223.
experienced by test patients was also demonstrated by 4. Wennström JL. Mucogingival surgery. In: Lang NP,
their lower use of analgesics compared with control Karring T, eds. Proceedings of the 1st European Work-
patients (P <0.05) (Table 2). When the lower mean shop on Periodontology. London: Quintessence Pub-
discomfort VAS score from the test group of this study lishing; 1994:193-209.
(2.4 – 0.88) was compared to that from the study by 5. Ahmedbeyli C, Ipc xi SxD, Cakar G, Kuru BE, Yılmaz S.
Clinical evaluation of coronally advanced flap with or
Zucchelli et al.12 (EFGG group; 3.2 – 1.99) and bearing without acellular dermal matrix graft on complete
in mind that, in the above-mentioned study,12 the defect coverage for the treatment of multiple gingival
collected gingival grafts were thinner than the present recessions with thin tissue biotype. J Clin Periodontol
ones, it can speculated that, in adopting a PRF ban- 2014;41:303-310.
dage, thicker EFGG can be collected while producing 6. Chambrone L, Pannuti CM, Tu YK, Chambrone LA.
Evidence-based periodontal plastic surgery. II. An in-
limited postoperative discomfort.
dividual data meta-analysis for evaluating factors in
Although Zucchelli et al.35 showed that thin grafts achieving complete root coverage. J Periodontol 2012;
associated with CAFs obtain excellent root-coverage 83:477-490.
results with less patient morbidity and better esthetics, 7. Müller HP, Eger T, Schorb A. Gingival dimensions after
there are situations in which a thick graft may be root coverage with free connective tissue grafts. J Clin
Periodontol 1998;25:424-430.
preferable for the clinician (e.g., soft tissue ridge aug-
8. Liu CL, Weisgold AS. Connective tissue graft: A
mentation procedures; Miller FGG technique for root classification for incision design from the palatal site
coverage).38 and clinical case reports. Int J Periodontics Restorative
In the present study, no difference is observed be- Dent 2002;22:373-379.
tween the experimental groups concerning the recovery 9. Edel A. Clinical evaluation of free connective tissue
grafts used to increase the width of keratinised gingiva.
of sensitivity at the donor site. Indeed, the presence of
J Clin Periodontol 1974;1:185-196.
specific biologic agents able to influence the growth of 10. Farnoush A. Techniques for the protection and cover-
nervous fibers in PRF was not reported. When in- age of the donor sites in free soft tissue grafts. J
vestigating the nervous regeneration in a rat model, Periodontol 1978;49:403-405.
Lichtenfels et al.39 reported that PRF was unable to 11. Del Pizzo M, Modica F, Bethaz N, Priotto P, Romagnoli
significantly enhance peripheral nerve repair. R. The connective tissue graft: A comparative clinical
evaluation of wound healing at the palatal donor site. A
Finally, the lack of episodes of DWB in both ex- preliminary study. J Clin Periodontol 2002;29:848-
perimental groups during week 1 may be explained by 854.
the palatal ligatures placed to constrict the blood ves- 12. Zucchelli G, Mele M, Stefanini M, et al. Patient morbidity
sels within the submucosa. and root coverage outcome after subepithelial connec-
tive tissue and de-epithelialized grafts: A comparative
CONCLUSION randomized-controlled clinical trial. J Clin Periodontol
2010;37:728-738.
On the basis of the present results, it can be concluded 13. Rossmann JA, Rees TD. A comparative evaluation of
that a PRF palatal bandage produces significant hemostatic agents in the management of soft tissue

112
J Periodontol • February 2016 Femminella, Iaconi, Di Tullio, et al.

graft donor site bleeding. J Periodontol 1999;70:1369- 28. Huberty CJ, Morris JD. Multivariate analysis versus mul-
1375. tiple univariate analysis. Psychol Bull 1989;105:302-308.
14. Dohan DM, Choukroun J, Diss A, et al. Platelet-rich 29. Koch GG. The use of non-parametric methods in the
fibrin (PRF): A second-generation platelet concentrate. statistical analysis of a complex split plot experiment.
Part I: Technological concepts and evolution. Oral Surg Biometrics 1970;26:105-128.
Oral Med Oral Pathol Oral Radiol Endod 2006;101: 30. Zucchelli G, Mele M, Mazzotti C, Marzadori M, Montebugnoli
e37-e44. L, De Sanctis M. Coronally advanced flap with and without
15. Bielecki T, Dohan Ehrenfest DM. Platelet-rich plasma vertical releasing incisions for the treatment of multiple
(PRP) and platelet-rich fibrin (PRF): Surgical adjuvants, gingival recessions: A comparative controlled random-
preparations for in situ regenerative medicine and tools ized clinical trial. J Periodontol 2009;80:1083-1094.
for tissue engineering. Curr Pharm Biotechnol 2012;13: 31. Pini Prato G, Rotundo R, Franceschi D, Cairo F, Cortellini
1121-1130. P, Nieri M. Fourteen-year outcomes of coronally ad-
16. Dohan Ehrenfest DM, Andia I, Zumstein MA, Zhang CQ, vanced flap for root coverage: Follow-up from a ran-
Pinto NR, Bielecki T. Classification of platelet concen- domized trial. J Clin Periodontol 2011;38:715-720.
trates (platelet-rich plasma-PRP, platelet-rich fibrin-PRF) 32. Pini-Prato GP, Cairo F, Nieri M, Franceschi D, Rotundo
for topical and infiltrative use in orthopedic and sports R, Cortellini P. Coronally advanced flap versus connec-
medicine: Current consensus, clinical implications and tive tissue graft in the treatment of multiple gingival
perspectives. Muscles Ligaments Tendons J 2014;4:3-9. recessions: A split-mouth study with a 5-year follow-
17. Del Corso M, Vervelle A, Simonpieri A, et al. Current up. J Clin Periodontol 2010;37:644-650.
knowledge and perspectives for the use of platelet-rich 33. Cairo F, Nieri M, Pagliaro U. Efficacy of periodontal
plasma (PRP) and platelet-rich fibrin (PRF) in oral and plastic surgery procedures in the treatment of localized
maxillofacial surgery part 1: Periodontal and dentoalveo- facial gingival recessions. A systematic review. J Clin
lar surgery. Curr Pharm Biotechnol 2012;13:1207-1230. Periodontol 2014;41(Suppl. 15):S44-S62.
18. Cieslik-Bielecka A, Choukroun J, Odin G, Dohan 34. Curtis JW Jr., McLain JB, Hutchinson RA. The incidence
Ehrenfest DM. L-PRP/L-PRF in esthetic plastic surgery, and severity of complications and pain following peri-
regenerative medicine of the skin and chronic wounds. odontal surgery. J Periodontol 1985;56:597-601.
Curr Pharm Biotechnol 2012;13:1266-1277. 35. Zucchelli G, Mounssif I, Mazzotti C, et al. Does the
19. Aravindaksha SP, Batra P, Sood V, Kumar A, Gupta G. dimension of the graft influence patient morbidity and
Use of platelet-rich fibrin membrane as a palatal ban- root coverage outcomes? A randomized controlled
dage. Clin Adv Periodontics 2014;4:246-250. clinical trial. J Clin Periodontol 2014;41:708-716.
20. The CONSORT Group. CONSORT. Available at http:// 36. Harris RJ. Histologic evaluation of connective tissue
www.consort-statement.org/. Accessed May 2, 2013. grafts in humans. Int J Periodontics Restorative Dent
21. Miller PD Jr. A classification of marginal tissue reces- 2003;23:575-583.
sion. Int J Periodontics Restorative Dent 1985;5:8-13. 37. Hürzeler MB, Weng D. A single-incision technique to
22. O’Leary TJ, Drake RB, Naylor JE. The plaque control harvest subepithelial connective tissue grafts from the
record. J Periodontol 1972;43:38. palate. Int J Periodontics Restorative Dent 1999;19:
23. Ainamo J, Bay I. Problems and proposals for recording 279-287.
gingivitis and plaque. Int Dent J 1975;25:229-235. 38. Miller PD Jr. Root coverage using a free soft tissue
24. Choukroun J, Adda F, Schoeffler C, Vervelle A. An autograft following citric acid application. Part 1: Tech-
opportunity in perio-implantology: The PRF (in French). nique. Int J Periodontics Restorative Dent 1982;2:65-70.
Implantodontie 2000;42:55-62. 39. Lichtenfels M, Colomé L, Sebben AD, Braga-Silva J.
25. Wessel JR, Tatakis DN. Patient outcomes following Effect of platelet rich plasma and platelet rich fibrin on
subepithelial connective tissue graft and free gingival sciatic nerve regeneration in a rat model. Microsurgery
graft procedures. J Periodontol 2008;79:425-430. 2013;33:383-390.
26. Silva CO, Ribeiro Edel P, Sallum AW, Tatakis DN. Free
gingival grafts: Graft shrinkage and donor-site healing Correspondence: Dr. Michele Paolantonio, Via Trilussa no.
in smokers and non-smokers. J Periodontol 2010;81: 21, I-65122 Pescara, Italy. E-mail: [email protected].
692-701.
27. Aitken RC. Measurement of feelings using visual ana- Submitted May 13, 2015; accepted for publication August
logue scales. Proc R Soc Med 1969;62:989-993. 23, 2015.

113