SCIENTIFIC ARTICLE

Reconstruction of Thumb Pulp Defects Using a

Modified Kite Flap
Xu Zhang, MD, Xinzhong Shao, MD, Chunzhen Ren, MD, Zhijie Zhang, MD, Sumin Wen, MD,
Jianxin Sun, MD

Purpose We report on a modified kite flap for the reconstruction of thumb pulp defects. We
performed nerve repair to improve thumb pulp sensation.
Methods From May 2005 to December 2008, 42 thumbs in 42 patients were treated. The
average size of the thumb pulp defects was 2.1 ⫻ 2.6 cm (range, 1.6 ⫻ 1.8 cm to 2.8 ⫻ 3.1
cm). The mean flap size was 2.5 ⫻ 2.9 cm (range, 1.8 ⫻ 2.2 cm to 3.2 ⫻ 3.5 cm). The radial
branch of the second dorsal digital nerve was coapted to one of the proper digital nerves of
the thumb. The required average length of the nerve branch was 1.2 cm (range, 0.7 to 1.6
cm). At follow-up, flap sensation was assessed using a static 2-point discrimination (2PD)
test. For comparison, we also included 32 patients without nerve repair from April 2003 to
April 2005. Outcomes were rated using the modified American Society for Surgery of the
Hand Guidelines for Stratification of 2PD.
Results In the study group, full flap survival was achieved in 40 thumbs, and partial distal flap
necrosis was noted in 2 thumbs. At final follow-up (mean, 26 mo; range, 24 to 27 mo), we
obtained a fair result, with a mean 2PD of 7.9 mm (range, 7 to 10 mm) on all flaps. In the
comparison group without nerve repair, there were 26 fair and 6 poor results, with a mean
2PD of 12 mm (range, 8 to 18 mm) at final follow-up (mean, 24 mo; range, 22 to 26 mo).
There was a highly significant difference between the 2 groups.
Conclusions We suggest performing nerve repair to improve the sensation of the kite flap
when reconstructing a thumb pulp defect. (J Hand Surg 2011;36A:1597–1603. Copyright
© 2011 by the American Society for Surgery of the Hand. All rights reserved.)
Type of study/level of evidence Therapeutic II.
Key words Dorsal digital nerve, dorsum of index finger, first dorsal metacarpal artery,
modified kite flap, thumb pulp defect.

HE THUMB PULP plays an important role in tactile fied surgical technique to improve thumb pulp sensa-

T sensation. A kite flap based on the first dorsal

metacarpal artery (FDMA) is commonly used
for thumb pulp reconstruction.1 We report on a modi-
From the Hand Surgery Department, The Second Hospital of Qinhuangdao, Changli, Qinhuangdao,
Hebei, China; Hand Surgery Department, Third Hospital of Hebei Medical University, Shijiazhuang,
Hebei, China.
Received for publication March 2, 2011; accepted in revised form June 30, 2011.
No benefits in any form have been received or will be received related directly or indirectly to the
subject of this article.
Corresponding author: Xu Zhang, MD, Hand Surgery Department, The Second Hospital of Qin-
huangdao, Changli, Qinhuangdao, Hebei 066600, P.R. China; e-mail: ahand@sina.com.
0363-5023/11/36A10-0004$36.00/0
doi:10.1016/j.jhsa.2011.06.033
tion.
Foucher et al2 described the kite flap, a sensate skin
island flap created from the dorsum of the index finger.
Sherif3,4 showed that the superficial branch of the radial
nerve divides into the dorsal digital nerves (DDNs). The
second DDN further divides into the radial and ulnar
branches that innervate the skin over the ulnodorsal
portion of the index finger proximal phalanx and the
radiodorsal skin over the middle finger proximal pha-
lanx (Fig. 1). Transection of the radial nerve branch at
the base of the flap is a conventional procedure. Ege et
al5 reconstructed thumb pulp defects using the flap in
21 cases that recovered Semmes-Weinstein (SW)6 sen-

©  ASSH 䉬 Published by Elsevier, Inc. All rights reserved. 䉬 1597

1598 MODIFIED KITE FLAP
FIGURE 1: The second dorsal digital nerve (DDN) gives off a
radial branch that innervates the ulnodorsal skin over the
proximal phalanx of the index finger.
sitivity of 3.61 to 4.31 and a mean static 2-point dis-
crimination (2PD)7 of 10.8 mm (range, 8 mm to 20
mm) on the flap. Muyldermans et al8 used this tech-
nique to obtain an average SW of 3.31 and static 2PD
of 10.6 mm.
Based on the conventional technique and anatomical
studies, we performed a nerve repair to improve flap
sensation.
MATERIALS AND METHODS
From May 2005 to December 2008, we performed the
modified kite flap with nerve repair on 42 thumbs in 42
patients (33 men and 9 women). The mean age at
surgery was 36 years (range, 17 to 58 y). The mecha-
nisms of injury included avulsion (n⫽21), crushing
(n⫽10), and third-degree burn (n⫽1).
Patients were selected on the basis of the following
criteria: (1) loss of thumb pulp of at least 1.5 cm, (2)
necessity of thumb pulp reconstruction for sensation,
(3) single-stage procedure, and (4) exposed tendon or
bone. Patients were excluded when they had any of the
following: (1) injury to the dorsum of the proximal
portion of the index finger, (2) injury to the course of
the FDMA, (3) a thumb pulp defect less than 1.5 cm in
length, or (4) no Doppler sound at the location of the
FDMA. In this study period, only 2 patients without a
signal at the FDMA were excluded.
The average size of the thumb pulp defects was
2.1 ⫻ 2.6 cm (range, 1.6 ⫻ 1.8 cm to 2.8 ⫻ 3.1 cm).
The mean flap size was 2.5 ⫻ 2.9 cm (range, 1.8 ⫻
2.2 cm to 3.2 ⫻ 3.5 cm). For comparison, we also
collected a consecutive series of 32 patients who
JHS 䉬 Vol A, October 
received the conventional kite flap in our unit from
April 2003 to April 2005. The mean age at surgery
was 31 years (range, 20 to 54 y). The mechanisms of
injury were avulsion (n⫽21) and crush (n⫽10). The
average size of the thumb pulp defects was 2.3 ⫻ 2.4
cm (range, 1.3 ⫻ 1.6 cm to 2.5 ⫻ 3.3 cm). The mean
flap size was 2.6 ⫻ 2.8 cm (range, 1.6 ⫻ 1.9 cm to
2.9 ⫻ 3.6 cm). All patient demographic and surgical
details were collected by the same surgical team who
performed the operation.
Surgical technique
We used a Doppler probe to detect the FDMA before
surgery. After wound debridement (Fig. 2A), a kite flap
was designed on the dorsum of the proximal phalanx of
the index finger to match the size and shape of the
defect. The pivot point was designed at the radial border
of the second metacarpal base (Fig. 2B).
We made a lazy S-shaped skin incision on the dor-
sum of the second metacarpal. The pedicle was dis-
sected distally to the base of the flap and proximally to
the pivot point. At least one subcutaneous vein or a strip
of subcutaneous tissue 5 mm wide was dissected with
the pedicle to provide venous return. The first DDN
innervating the radiodorsal skin over the proximal pha-
lanx was harvested with the pedicle. A strip of skin over
the pedicle was harvested with the pedicle to avoid
compression after flap transfer. At the base of the flap,
we dissected the radial nerve branch of the second DDN
innervating the dorsum of the proximal portion of the
index finger proximally until the required length was
achieved (Fig. 3). If necessary, we obtained additional
branch length by using a combination of sharp and
blunt division of the DDN trunk. After transection of
the nerve branch, the flap was raised along its borders,
which left the paratenon undisturbed to ensure a free
tendon gliding (Fig. 4). Thereafter, the flap was trans-
ferred to the defect through a subcutaneous tunnel and
was sutured in place (Fig. 5A). With the aid of a
microscope, the nerve branch in the flap was co-
apted to the end of the radial or ulnar proper digital
nerve of the thumb, depending primarily on which
one was left after injury. If both nerves were
available, we performed radial coaptation because
the first DDN already in the flap covered the ulnar
territory. We then covered the donor site with a
full-thickness skin graft that was harvested from
the inner aspect of the upper arm. We secured the
skin graft with a tie-over bolster dressing.
In this series, the mean pedicle length was 7.8 cm
(range, 6.5 to 8.5 cm). The required average length of
 MODIFIED KITE FLAP 1599

FIGURE 2: A A thumb pulp defect. B The FDMA flap, 3.0 ⫻ 2.0 cm, is designed on the dorsum of the index finger.

FIGURE 4: The radial branch of the second DDN (arrow) is

harvested with the flap.

the radial nerve branch of the second DDN was 1.2 cm

(range, 0.7 to 1.6 cm). Nerve coaptation was performed
to the radial and ulnar proper digital nerves in 40
thumbs and 2 thumbs, respectively. In our series, injury
associated with partial nail bed loss occurred in 6
thumbs. We performed bone shortening (mean, 4 mm;
FIGURE 3: The asterisk shows the second DDN, and the
range, 2 to 7 mm) of the distal phalanx to the level of
arrow shows its radial branch.
nail bed.

JHS 䉬 Vol A, October 

1600 MODIFIED KITE FLAP
FIGURE 5: A After suturing the nerve branch to the radial proper digital nerve of the thumb, the defect is covered with the flap.
B Result 25 months later.
Postoperative treatment
After surgery, the hand was placed above heart level to
reduce venous congestion of the flap. Flap circulation
was monitored by visual inspection of the tissue color
and capillary refilling for at least 2 days. To protect the
skin graft, we splinted the hand for 7 to 10 days. Active
range-of-motion exercises were initiated thereafter. All
patients attended formal hand therapy for sensory re-
education and range of motion exercises.
Evaluation of outcomes
Ten days after surgery, the surviving flap displayed
warmth, good capillary refill, and pink color.
At final follow-up evaluation, the senior hand sur-
geon visually assessed the color, texture, and thickness
of the donor site and the flap.
Active motion of the joints was measured using a
goniometer. The motion arcs of the repaired thumbs
were compared with those of the opposite side. Scar
appearance and pain at the donor and recipient sites and
hypersensitivity of the thumb pulp were evaluated using
the patients’ self reported assessments. Cold intolerance
of the flap was measured using the self-administered
Cold Intolerance Severity Score questionnaire.10 The
maximum score is 100 and is grouped into 4 ranges
(0 –25, 26 –50, 51–75, and 76 –100), corresponding to
JHS 䉬 Vol A, October 
mild, moderate, severe, and extremely severe, respec-
tively.
The sensitivity of the flap and the donor site were
measured using the SW monofilament test and static
2PD test. The test point was at the center of the flap.
The 2PD measurements were rated using the Modified
American Society for Surgery of the Hand Guidelines
for Stratification of 2PD (excellent, ⬍6 mm; good,
6 –10 mm; fair, 11–15 mm; poor, ⬎15 mm).9 Sensory
results were compared between groups with and with-
out nerve repair. A chi-square test was used to deter-
mine whether there was a difference between the 2
groups. A difference was considered statistically signif-
icant or highly significant if the P value was less than
.05 or .01, respectively.
RESULTS
In the study group, full flap survival was achieved in 40
thumbs. Partial distal flap necrosis was noted in 2
thumbs, which healed without surgical intervention. We
encountered some mild congestion that did not require
surgical intervention. Wound infection was not ob-
served in this series.
At final follow-up (mean, 26 mo; range, 24 to 27
mo), skin color, texture, and the thickness of the
donor site and flap were acceptable in all cases
 MODIFIED KITE FLAP
FIGURE 6: A Donor site. B Flexion of the donor finger.
(Figs. 5B, 6A). Slight discoloration at the donor
site was reported in 38 digits.
The mean active motion arcs of the metacarpopha-
langeal (MCP) and interphalangeal joints of the thumbs
were 80° (range, 72° to 83°) and 82° (range, 75° to 90°),
respectively (Fig. 6B). The mean active motion arcs of
the same joints of the opposite thumbs were 83° (range,
75° to 92°) and 85° (range, 80° to 89°), respectively.
The mean active motion arcs of the MCP and proximal
interphalangeal joints of the donor index digits were 84°
(range, 80° to 90°) and 102° (range, 88° to 105°),
respectively. The measurements of the contralateral
side were 87° (range, 82° to 90°) and 102° (range, 96°
to 112°), respectively.
All scars on the donor and recipient sites were pain-
less. Three patients reported hypersensitivity on the
thumb pulp. According to the Cold Intolerance Severity
Score, all flaps scored less than 25 and experienced
mild cold intolerance.
In the comparison group, the mean active motion
arcs of the MCP and interphalangeal joints of the
thumbs were 75° (range, 68° to 81°) and 80° (range, 72°
to 88°), respectively. The range of motion of the joints
was similar in both groups. All scars on the donor and
recipient sites were painless. No patient reported hyper-
sensitivity on the thumb pulp. The mean SW sensitivity
score was 3.30 (range, 3.22 to 4.31) on the flap and 4.59
(range, 3.22 to 6.65) on the donor site. The mean static
JHS 䉬 Vol A, October 
1601
2PD was 7.9 mm (range, 7 to 10 mm) on the flap and
10.2 mm (range, 8 to 12 mm) on the donor site. In the
comparison group, the sensation tests were performed
by the same examiner. At final follow-up (mean 24 mo;
range, 22 to 26 mo), the mean SW sensitivity score was
4.38 (range, 3.62 to 5.07) on the flap and 4.05 (range,
3.62 to 4.56) on the donor site. The mean static 2PD
was 12 mm (range, 8 to 18 mm) on the flap and 10 mm
(range, 7 to 12 mm) on the donor site. We did not
re-examine these patients because some of the contact
information for the patients was invalid after a long
period of time.
The flaps with nerve repair were superior to the flaps
without nerve repair (25% and 34% improvement in
mean SW sensitivity and static 2PD scores, respec-
tively). Donor site sensation was similar in both groups.
There were no significant differences between the 2
groups in patient age, defect and flap size, and pedicle
length. In the study group, 39 of 42 flaps achieved SW
sensitivity scores of ⱕ3.84. In the comparison group, 3
of 32 flaps achieved scores of ⱕ3.84. We obtained a P
value ⬍.01, which showed that there was a significant
difference between the 2 groups. Based on the Modified
American Society for Surgery of the Hand Guidelines
for Stratification of 2PD, we observed fair 2PD results
in all cases in the study group. In the comparison group,
we observed fair and poor 2PD results in 26 and 6
cases, respectively. We obtained a P value of .003,
1602 MODIFIED KITE FLAP
which showed that there was a highly significant dif-
ference between the 2 groups in the fair result.
DISCUSSION
Clinical observation shows that both the radial and
ulnar sides of the thumb pulp play a fundamental role in
pinching by virtue of their specialized covering.11 Var-
ious techniques have been reported for thumb pulp
reconstruction.
A V-Y advancement flap is used for defects less than
1.5 cm long.12 A Moberg flap can cover a relatively
large defect and return glabrous skin and sensation to
the thumb tip.13 However, this technique carries the risk
of dorsal skin and nail bed necrosis because the original
injury might involve the dorsal arteries or risk an injury
to the arteries during dissection. The abdominal, cross-
finger, and cross-arm flaps are reliable but insen-
sate.14 –17 Reverse dorsal metacarpal flaps based on the
dorsal metacarpal arteries have the drawback of leaving
a long scar, which runs almost from the wrist to the
middle phalanx.18,19 A reverse digital artery flap har-
vested from over the proximal phalanx is reliable but
could result in the loss of one of the proper digital
arteries.20 The transfer of a free flap, such as a partial
toe transfer, requires prolonged surgery and has a risk of
anastomotic failure.21
A conventional kite flap has a pedicle length of up to
7 cm, which allows a wide arc of rotation, and has
proven useful in the restoration of the contour and
sensation of the thumb pulp.3 However, sensation on
the radial side of the thumb pulp is usually poor because
the first DDN mainly innervates the radial side of the
flap, and when it is transposed, the DDN lies on the
ulnar side of the thumb. Our technique provides im-
proved thumb pulp sensation. Although normal sensa-
tion is not restored, our results were better than previ-
ously reported studies and our comparison group. We
assumed that the flap could be cross-innervated by both
the radial branch of the second DDN and the first DDN,
which might be the reason for the improved sensitivity.
Therefore, we suggest that nerve repair should be per-
formed, if possible, when using a kite flap.
Only a few patients without nerve repair obtain 2PD
scores of 4 to 6 mm on the flap.22,23 We realize that a
drawback of the 2PD measurement instruments is the
uncertainty of how much pressure to apply while test-
ing, and a disadvantage of SW monofilaments is that
the tip of the filaments is not uniform in size.24 In
addition, scores can vary based on the test points and
testers. Therefore, we conducted a comparison group
without nerve repair to minimize variability.
JHS 䉬 Vol A, October 
According to our intraoperative observation, the ul-
nar branch of the second DDN has a mean diameter of
1.2 mm (range, 0.8 to 1.8 mm) at its bifurcation in all
digits. It is suitable for coaptation to either the radial or
ulnar proper digital nerves. During the operation, the
radial branch of the second DDN can be dissected from
the base of the flap to the bifurcation where the second
DDN arises. The branch length is usually sufficient for
nerve repair. If necessary, additional length can be
obtained by splitting the second DDN.
Hypersensitivity can occur after nerve injury and
repair. In our series, this complication was rare. Thumb
stiffness is more likely related to the original injury than
to the flap transfer. In the study group, MCP and inter-
phalangeal joint motions were generally satisfactory,
but less motion would be predictable in a patient with a
badly damaged thumb. Our study showed that the sen-
sory and motor functions of the donor finger were
similar to those in previously reported studies.2,5,8
Therefore, transection of the nerve branch, either at the
base of the flap or at the point proximal to the flap, can
result in similar results. Stiffness related to the dissec-
tion of the ulnar aspect of the MCP joint is also possi-
ble, but we noted minimal donor finger morbidity.
A flap based on the FDMA might not always easily
reach or go over the tip of the thumb. When it does not
reach far enough distally, a splint can be applied to
maintain the thumb in less abduction and pronation to
obtain a tension-free pedicle.11 In a patient with partial
nail bed loss, the dorsal defect can be repaired with a
flap, skin graft, or toe nail bed graft. These techniques
help maintain normal thumb length. Nevertheless,
shortening less than 1 cm in length does not lead to loss
of hand function.25 In our series, the maximum nail bed
loss was 7 mm. Bone shortening is a simpler and more
reliable procedure.
Based on our results, we suggest nerve repair for kite
flap transfer, if possible. It does require a microsurgical
nerve repair and does not restore the specialized gla-
brous skin on the thumb pulp. A contraindication would
be that the flap is so small that the nerve branch is not
included in the flap.
In future studies, which ideally will be prospective,
randomized, and blinded, a local anesthetic should be
placed on the intact nerve to better ascertain the effect
of the nerve repair.
REFERENCES
1. Saalabian A, Rab M, van Schoonhoven J, Prommersberger KJ.
Foucher’s first dorsal metacarpal artery island flap [in German]. Oper
Orthop Traumatol 2009;21:614 – 619.
2. Foucher G, Braun JB. A new island flap transfer from the dorsum of
the index to the thumb. Plast Reconstr Surg 1979;63:344 –349.
 MODIFIED KITE FLAP
3. Sherif MM. First dorsal metacarpal artery flap in hand reconstruc-
tion. II. Clinical application. J Hand Surg 1994;19A:32–38.
4. Sherif MM. First dorsal metacarpal artery flap in hand reconstruc-
tion. I. Anatomical study. J Hand Surg 1994;19A:26 –31.
5. Ege A, Tuncay I, Ercetin O. Foucher’s first dorsal metacarpal artery
flap for thumb reconstruction: evaluation of 21 cases. Isr Med Assoc
J 2002;4:421– 423.
6. Berquin AD, Lijesevic V, Blond S, Plaghki L. An adaptive proce-
dure for routine measurement of light-touch sensitivity threshold.
Muscle Nerve 2010;42:328 –338.
7. Ozaksar K, Toros T, Sugun TS, Bal E, Ademoglu Y, Kaplan I.
Reconstruction of finger pulp defects using homodigital dorsal mid-
dle phalangeal neurovascular advancement flap. J Hand Surg 2010;
35B:125–129.
8. Muyldermans T, Hierner R. First dorsal metacarpal artery flap for
thumb reconstruction: a retrospective clinical study. Strategies
Trauma Limb Reconstr 2009;4:27–33.
9. Dellon AL, Kallman CH. Evaluation of functional sensation in the
hand. J Hand Surg 1983;8:865– 870.
10. Irwin MS, Gilbert SE, Terenghi G, Smith RW, Green CJ. Cold intol-
erance following peripheral nerve injury. Natural history and factors
predicting severity of symptoms. J Hand Surg 1997;22B:308 –316.
11. Zhang X, He Y, Shao X, Li Y, Wen S, Zhu H. Second dorsal
metacarpal artery flap from the dorsum of the middle finger for
coverage of volar thumb defect. J Hand Surg 2009;34A:1467–1473.
12. Moberg E. Aspects of sensation in reconstructive surgery of the
upper extremity. J Bone Joint Surg 1964;46A:817– 825.
13. Baumeister S, Menke H, Wittemann M, Germann G. Functional
outcome after the Moberg advancement flap in the thumb. J Hand
Surg 2002;27A:105–114.
14. Urushidate S, Yotsuyanagi T, Yamauchi M, Mikami M, Ezoe K,
Saito T, et al. Modified thin abdominal wall flap (glove flap) for the
JHS 䉬 Vol A, October 
1603
treatment of acute burns to the hands and fingers. J Plast Reconstr
Aesthet Surg 2010;63:693– 699.
15. Cohen BE, Cronin ED. An innervated cross-finger flap for fingertip
reconstruction. Plast Reconstr Surg 1983;72:688 – 697.
16. Fejjal N, Belmir R, El Mazouz S, Gharib N, Abbassi A, Belmahi A.
Reversed cross finger subcutaneous flap: A rapid way to cover finger
defects. Indian J Plast Surg 2008;41:55–57.
17. Megerle K, Palm-Bröking K, Germann G. The cross-finger flap.
Oper Orthop Traumatol 2008;20:97–102.
18. Lai-jin L, Xu G. The reverse dorsal metacarpal flap: experience with
153 cases. Ann Plast Surg 2006;56:614 – 617.
19. Koch H, Bruckmann L, Hubmer M, Scharnagl E. Extended reverse
dorsal metacarpal artery flap: clinical experience and donor site
morbidity. J Plast Reconstr Aesthet Surg 2007;60:349 –355.
20. Takeishi M, Shinoda A, Sugiyama A, Ui K. Innervated reverse
dorsal digital island flap for fingertip reconstruction. J Hand Surg
2006;31A:1094 –1099.
21. Rybka FJ, Pratt FE. Thumb reconstruction with a sensory flap from
the dorsum of the index finger. Plast Reconstr Surg 1979;64:141–
144.
22. Satish C, Nema S. First dorsal metacarpal artery islanded flap: A
useful flap for reconstruction of thumb pulp defects. Indian J Plast
Surg 2009;42:32–35.
23. Tränkle M, Germann G, Heitmann C, Sauerbier M. Defect coverage
and reconstruction of thumb sensibility with the first dorsal meta-
carpal artery flap [in German]. Chirurg 2004;75:996 –1002.
24. Wong KH, Coert JH, Robinson PH, Meek MF. Comparison of
assessment tools to score recovery of function after repair of trau-
matic lesions of the median nerve. Scand J Plast Reconstr Surg Hand
Surg 2006;40:219 –224.
25. Chris EW. Upper extremity. In: Wiggins CE, ed. A concise guide to
orthopaedic and musculoskeletal impairment ratings. Philadelphia,
PA: Lippincott Williams & Wilkins, 2007:19 –33.