Standardized Musculocutaneous Flap
for the Coverage of Deep Sternal Wounds
After Cardiac Surgery
Olimpiu Bota, MD, Christoph Josten, MD, Michael A. Borger, MD, PhD,
Nick Spindler, MD, and Stefan Langer, MD
Department of Orthopedics, Trauma and Plastic Surgery, University Hospital Leipzig, Leipzig; and Department of Cardiac Surgery,
Leipzig Heart Center, University of Leipzig, Leipzig, Germany
Background. Deep sternal wound infection remains a
serious complication after cardiac surgery, leading to
increased morbidity, mortality, and cost. The goal of our
study was to develop a standardized, reproducible
method to safely cover deep sternal wounds and ensure
improved healing rates.
Methods. The study was developed as a retrospective
cohort study. We included 58 patients who received
standardized latissimus dorsi flap coverage of a sternum
defect wound after poststernotomy mediastinitis at our
institution between September 2015 and June 2017.
Results. The average age of the cohort was 66.75 years,
and 51.72% of patients were men. The mean hospital
stay was 26.83 days. Eight patients (14.75%) died during
the hospital stay due to sepsis or heart failure. The
average flap size was 137.13 cm2. The mean operative
time was 155 minutes. Seventy-four percent of patients
D eep sternal wound infection (DSWI) remains a
serious complication after cardiac surgery, leading
to increased morbidity, mortality, and cost. The reported
prevalence varies between 1% and 3% [1, 2]. Several risk
factors have been associated with increased rates of
DSWI, related to the patient (age, sex, diabetes, obesity,
respiratory insufficiency), to surgical factors (context of
emergency, operative time, type of surgical procedure,
early surgical revision for bleeding, harvesting of both
internal mammary arteries), and hospitalization (duration
of preoperative stay, patient preparation) [1, 3]. The
development of flap coverage of these wounds as well as
the emergence of negative pressure wound therapy has
led to a decrease in the mortality associated with this
complication [4]. Nevertheless, there is currently no
consensus regarding a treatment strategy for DSWI [5].
The goal of our work was to develop a standardized,
reproducible method to safely cover deep sternal wounds
(DSWs) and to ensure improved healing rates. Of the
Accepted for publication Sept 7, 2018.
Address correspondence to Dr Bota, Klinik und Poliklinik f€ur Orthop€adie,
Unfallchirurgie und Plastische Chirurgie, Operatives Zentrum,
Liebigstraße 20, 04103, Leipzig, Germany; email: olimpiu.bota@gmail.com.
Ó 2019 by The Society of Thoracic Surgeons
Published by Elsevier Inc.
developed a seroma at the donor site, which was treated
conservatively with compression garments and taps, and
7% of patients developed a wound dehiscence of the
donor site, which was treated conservatively with
dressings. The mean follow-up time was 15 weeks. All
50 surviving patients showed complete wound healing
on follow-up.
Conclusions. The latissimus dorsi pedicled flap is a
safe, reproducible technique for coverage of deep sternal
wounds, with few relevant perioperative complications.
By setting definite parameters for the flap dissection and
by ensuring a reliable blood supply, our method enables
the coverage of these complex wounds by an interdisci-
plinary team in any cardiovascular surgical setting.
(Ann Thorac Surg 2019;-:-–-)
Ó 2019 by The Society of Thoracic Surgeons
plastic coverage methods available for sternal coverage
(omentum, fasciocutaneous, and muscle flaps) the latis-
simus dorsi musculocutaneous flap (LDMF) provides
sufficient tissue to fill the dead space after sternectomy
with reduced donor site morbidity [6, 7].
Material and Methods
The study has been approved by the Ethical Committee
our University’s Medical Faculty. The study was devel-
oped as a retrospective cohort study. The patients with
DSWI were treated in cooperation with the heart center,
which performs around 3,500 open heart surgeries a year.
After diagnosing the sternal osteomyelitis and stating the
impossibility of a secondary wound closure, the cardio-
vascular surgeon contacted a plastic surgeon from our
institution to engage in the interdisciplinary treatment of
these patients. This was done based on a preexisting
protocol between the heart center and the plastic surgery
department. We included all patients who received flap
coverage of a sternum defect wound after poststernotomy
mediastinitis between September 2015 and June 2017. We
identified 58 patients, all of whom received standardized
flap coverage as described below. The LDMF is the only
flap used for the coverage of DSWI in our institutions.
0003-4975/$36.00
https://doi.org/10.1016/j.athoracsur.2018.09.017
2 BOTA ET AL Ann Thorac Surg
STANDARDIZED DEEP STERNAL WOUND COVERAGE 2019;-:-–-

Fig 2. Flap design and measurements.

island is drawn. Depending on the skin and soft tissue

defect size the skin island may be chosen up to 23 to 25
cm in length and 7 to 9 cm in width. The pinch test
ensures the defect can still be closed primarily. The skin
island is drawn starting 5 cm anterior to the posterior
axillary line. If an extralong skin island is needed, care
must be taken to draw the skin island at least 5 cm away
from the spine to preserve skin viability (Fig 2).
The incision begins with the 15 cm line along the
posterior axillary line. The anterior edge of the muscle is
identified and dissected (Fig 3). Depending on the course
of the muscle the skin island may now be slightly
adjusted such that the tip of the flap lies on the inter-
digitation between the latissimus dorsi, the serratus
anterior, and the oblicus externus muscles. The upper
border of the flap is now incised, and the latissimus dorsi
Fig 1. Sternal wound after radical debridement.
muscle is freed from the subcutaneous tissue up to its
upper border and its spine origin. The subcutaneous
After diagnosing the mediastinitis the standard tissue around the flap is incised in an oblique manner
procedure in our institutions is a radical debridement of down to the muscle, such that 1 to 2 cm of subcutaneous
the presternal soft tissues, sternum, rib cartilages, fat tissue surrounds the flap (Fig 4). Subsequently the
sternoclavicular joint, and mediastinal dead tissue (Fig 1) inferior border of the skin island is incised in a similar
followed by simultaneous or subsequent flap coverage of
the defect with a pedicled LDMF as described below. The
preferred side for the muscle harvest is the nondominant
hand side. In cases where there is important scarring or
an implanted pacemaker on the dominant hand side, the
contralateral side is chosen.
The patient, under general anesthesia, is positioned
laterally with the arm in 90 degrees of abduction, the
elbow flexed 90 degrees, cushioned, and fixed to the
anesthesia bar. If simultaneous mediastinal debridement
is being performed, one surgical team operates on the
front of the patient and one operates on the back. If the
radical debridement has already been performed in a
previous operation, there is only one surgical team
operating on the back of the patient.
The design of the flap begins by finding the intersection
of the axillary fold with the posterior axillary line. From
this landmark a 15-cm line is drawn along the posterior
axillary line. At the bottom of this line the cutaneous Fig 3. Dissection of the anterior border of the latissimus dorsi muscle.
Ann Thorac Surg BOTA ET AL 3
2019;-:-–- STANDARDIZED DEEP STERNAL WOUND COVERAGE

Fig 4. Flap before pedicle dissection. Notice the 15-cm distance from
the axillary fold.

matter, and the muscle is dissected toward the iliac crest

over 3 to 5 cm.
At this point the anterior border of the flap is elevated
to identify the right dissection plane. The muscle with the
skin island is now elevated from lateral to medial. The
inferior border of the muscle is incised. The latissimus
dorsi muscle and the erector spinae muscle are separated
Fig 5. The flap elevated from the thoracic wall, before pedicle
from each other dorsally. The perforator vessels under-
dissection. Notice the branch to the serratus anterior muscle, which
neath the muscle are carefully clipped. After separating
needs to be clipped.
the flap from the serratus anterior and the teres major
muscles (Fig 5) the thoracodorsalis pedicle is micro-
compressive bra is fitted starting the second post-
surgically dissected while clipping the collateral branches
operative day to avoid shear forces across the wound. The
(Fig 6). The thoracodorsalis nerve is identified, and 1 cm is
patient undergoes mobilization by a physical therapist on
excised to denervate the flap.
the first postoperative day, if possible.
At this moment a subcutaneous tunnel toward the
sternum defect is dissected. The flap is transected at
its humeral origin and passed through the tunnel so that
the skin island covers the mediastinal defect (Fig 7).
The tendinous muscle origin is fixed on the thorax fascia
to avoid pedicle tensioning or kinking. If necessary,
further releasing of the pedicle toward the axilla can be
performed to achieve a completely tension-free position
of the pedicle.
The excess muscle is folded into itself anteriorly into
the mediastinum and fixed to the chest musculature with
absorbable sutures. The wound is closed airtight in layers
over suction drainages (Fig 8). The donor site is primarily
closed over suction drainages. The patient is instructed to
keep the arm in 45 degrees of abduction for the first 5
postoperative days so there is no pressure on the pedicle.
The patient’s arm is placed postoperatively on a pillow in
a light abducted position. With ambulation the patients
are instructed to keep their hand on their thigh to keep
the arm abducted. After 5 days a compressive bandage is
fitted for 6 weeks. In female patients with macromastia a Fig 6. Dissected flap, before passing through the tunnel.
4 BOTA ET AL Ann Thorac Surg
STANDARDIZED DEEP STERNAL WOUND COVERAGE 2019;-:-–-

Fig 7. Flap lying in the tunnel.

Results
The average age of the cohort was 66.75 years (range, 48 to
85), and 51.72% were men. The duration of hospital stay
ranged from 7 to 98 days, with a median of 19.5 and a
mean of 26.83. Eight patients (14.75%) died during the
hospital stay because of sepsis or heart failure.
In 30 patients the sternum debridement was performed
simultaneously with the flap, and in 28 patients the
sternum was debrided before the flap coverage. The
average flap size was 137.13 cm2, ranging from 72 to 230
cm2 (ellipse area ¼ short radius  long radius  3.14). The
mean operative time was 155 minutes. Four patients had
to be reoperated immediately postoperatively because of
bleeding. There were two partial flap necroses in the distal
Fig 9. Flap healed 3 months postoperatively.
part, which eventually needed revision operation, and one
of which received a similar flap from the contralateral
side. There was no complete flap failure. Twenty patients chronic fistula due to persistent rib cartilage infection and
(11 men, 9 women) showed wound-healing problems received revision operation and a second similar flap from
in the caudal pole of the flap because of persistent the contralateral side. One female patient received a
secretion and infection and required revision operation. pedicled omentum flap on revision operation for the
Of these, 2 female patients received a second, similar flap closure of the dead space around the infected aortic
from the contralateral side. One male patient showed a prosthesis.
Seventy-four percent of patients developed a seroma at
the donor site, which was treated conservatively with
compression garments and punctures. Seven percent of
patients developed a wound dehiscence of the donor site,
which was treated conservatively with dressings.
The mean follow-up time was 15 weeks, ranging from
12 to 25 weeks. All 50 surviving patients showed complete
wound healing on follow-up (Figs 9 and 10).

Fig 8. Anterior view of the flap fixed into place, before skin suture.
Comment
Radically debriding the DSW usually results in a long,
wide, and especially deep wound that may stretch around
the ascendant aorta and the myocardia. Fasciocutaneous
flaps (internal mammary artery perforator flap) [8, 9],
muscle flaps (pectoralis major flap, rectus abdominis
flap), the omentum flap [5, 7, 10], and free vascularized
Ann Thorac Surg
2019;-:-–-
Fig 10. Donor site healed 3 months postoperatively.
flaps (anterolateral thigh flap, gracilis muscle flap) [11, 12]
have been used for covering this type of wound, with the
aid of negative pressure wound therapy as a “bridging
procedure” [13]. Nevertheless there is currently no
consensus regarding the optimal therapeutic option and a
complete lack of prospective randomized trials
comparing these techniques [7, 10, 14].
The internal mammary artery perforator flap relying on
the second intercostal perforator has been reported in
small patient series as an alternative for sternal coverage
[8, 9]. The reduced tissue bulk and the reported flap
necrosis at the lower edge of the flap make it a solution
for smaller, superficial sternal wounds. The pectoralis
major flap is considered by many surgeons as a gold
standard for sternal coverage [10]. It usually requires the
harvest of both pectoral major muscles with or without a
skin island, producing significant donor site morbidity
[7]. An important drawback of this flap is its reduced
availability in the lower part of the sternum [15]. The
turnover pectoralis major flap based on the intercostal
arteries may be used for the coverage of the lower part of
the sternum. Nevertheless these vessels are often lost
during the radical sternal and rib cartilage debridement.
The rectus abdominis musculocutaneous flap has also
been used for sternal coverage, especially in the lower
part of the sternum, even after the harvest of both internal
mammary arteries [15, 16]. However, this method leads to
the weakening of the abdominal wall and provides
relatively reduced tissue bulk compared with other flaps.
The omentum flap is a well-known source of soft tissue
for the coverage of thoracic wounds and may be
harvested through laparotomy or laparoscopy. It provides
immunologic tissue with a reliable blood supply and a
high reachability for most sternal wounds, used alone or
BOTA ET AL 5
STANDARDIZED DEEP STERNAL WOUND COVERAGE
in combination with the pectoralis flap [7, 10, 17] or free
flaps [11]. Its place in the sternal reconstruction still needs
to be established, because of the donor site morbidity and
the need for an intraabdominal procedure. Free flaps
have also been used in the treatment of DSWI, using
either an arteriovenous loop or the right gastroepiploic
vessels for the blood supply [11, 12]. The need for
microsurgical skills, the prolonged operative time, and
the risk of total flap loss limits the use of such flaps to
specialized centers in selected cases.
The LDMF represents a reliable, voluminous, well-
perfused source of tissue for covering DSW. The redun-
dant muscle bulk is capable of filling out most of the dead
space while providing healthy, remote, well-vascularized
tissue. The large skin island can close even the largest
soft tissue defects while allowing for the donor site to be
closed primarily [6]. The vascularization of the muscle is a
type 5 under the Mathes and Nahai classification [18] and
is provided by the thoracodorsal vascular bundle as the
main blood supply and the intercostal perforator vessels
as a segmental blood supply. The main pedicle splits in
the upper part of the muscle into the transverse and the
vertical branches. Watanabe and coworkers [19]
researched the angiosome distribution of the skin over-
lying the latissimus dorsi muscle and found a separation
in two vascular territories. The first vascular territory
consists of perforator vessels originating from the two
branches of the thoracodorsal artery, which form a
vascular network with the perforator vessels from the 9th,
10th, and 11th intercostal arteries and the scapular
circumflex artery. As the latter vessels are being sectioned
during flap dissection, the skin vascularity in this territory
is ensured by the direct anastomosis between the two
thoracodorsalis branches and the intercostal skin perfo-
rators. The second vascular territory is formed by a
vascular network between the perforating vessels from
the subcostal and the first two lumbar arteries. This
second territory is connected to the first one through
choke vessels. The distance from the axillary origin of the
main pedicle to the skin island is a key factor influencing
the blood supply of the skin. The 15-cm landmark ensures
that the skin island relies on the safe vascularity of the first
vascular territory, avoiding the dependence on the choke
vessels and preventing necrosis in the critical lower pole
of the flap.
Known as a standard flap for breast reconstruction, the
LDMF provides low donor site morbidity with incon-
spicuous scars. The most common donor site complication
with this flap, seroma [7], is easily treated in an outpatient
setting with periodic punctures and compressive gar-
ments. Wound dehiscence, not a rare occurrence in these
multimorbid patients, can also be safely treated conser-
vatively without added risk to the patient. Temporary
functional weakness of the arm compared with the
contralateral side may be seen in the first months after
operation, although it should completely resolve within
1 year [20].
Here we present our experience with the use of LDMF
to treat DSWI and base our discussion on a relatively
small series of 58 cases. These patients were operated by
6 BOTA ET AL
STANDARDIZED DEEP STERNAL WOUND COVERAGE
an interdisciplinary team in the context of an interinsti-
tutional collaboration protocol between a large university
heart center (which performs around 3,500 heart
interventions per year) and a plastic surgery department,
resulting in an incidence of less than 1% DSWI per year,
The reported incidence in the literature for DSWs lies
between 0.8% and 1.5% [7], which situates our report in
the lower part of the interval. Because the decision to
refer to the interdisciplinary team is made at the heart
center, the patients treated in this series are a selection of
the most severe cases, which could not be treated by
secondary wound closure.
The four cases of postoperative bleeding that required
a surgical revision can be attributed to the continuous
administration of anticoagulant and antiplatelet agents
after cardiovascular surgical procedures. Because of the
robustness of the latissimus dorsi flap, revisional opera-
tion can be performed without jeopardizing the flap
viability.
In a large population-based study Sears and colleagues
[21] found the 1-year mortality to be 10.7% in patients
with DSWI. In another large population-based study
Kubota and coworkers [22] found a higher incidence of
DSWI with more complex heart interventions and the
30-day mortality to be significantly higher with reexplo-
ration for bleeding (23.0% vs 8.1%). The 14.75% mortality
in our study reflects the difficulty and complexity of the
treated cases.
The mean hospital stay is reported to be 33 days in
patients with DSWI. The lower mean hospital stay of
26.83 days in our study reflects the fact that the primary
treatment of the sternal infection was performed in the
heart institution. However it might also be reduced by the
simultaneous debridement and flap coverage of 30
patients (51.72%).
The shortcoming of this flap is its limited capacity to
absorb secretions after infection, which sometimes leads
to prolonged retention of drainages and to wound
dehiscence in the caudal pole [8]. The nature of this
patient population, with multiple comorbidities and
decreased cardiac output, also predisposes them to
wound-healing problems [9, 23, 24]. Nevertheless, when
secondary flap coverage is necessary, the latissimus dorsi
from the contralateral side can be used as a salvage
procedure for wound closure without significant addi-
tional morbidity.
Although the LDMF is considered to be a workhorse
flap for chest reconstruction [25] and the account of the
first 5 patients with DSW treated with LDMF were
published in 1985 by Tizian and associates [26], most
authors report it as a salvage procedure [5–7, 10, 26, 27],
probably because of the lack of studies establishing a
precise operative technique. By developing a standard-
ized, reproducible procedure for performing the
LDMF we provide a reliable alternative for the surgeon
to treat this kind of wound. Respecting the exact mea-
surements provided above ensures that the flap will
reach the defect in any type of patient without jeop-
ardizing the viability of the skin island, especially on its
dorsal side.
Ann Thorac Surg
2019;-:-–-
We conclude that the latissimus dorsi pedicled flap is a
safe, reproducible technique for covering DSW, with few
relevant perioperative complications. By setting definite
parameters for flap dissection and by ensuring a reliable
blood supply, our method enables the coverage of these
complex wounds in any cardiovascular surgical setting,
without requiring advanced microsurgical training. The
complications associated with these troublesome wounds
in multimorbid patients can be treated by revisional
operation and backed up by the availability of the
contralateral latissimus dorsi flap for secondary coverage.
Further randomized prospective studies have to be
developed to establish the exact position of each flap in
the coverage of DSW.
References
1. Buttar SN, Yan TD, Taggart DP, Tian DH. Long-term and
short-term outcomes of using bilateral internal mammary
artery grafting versus left internal mammary artery grafting:
a meta-analysis. Heart 2017;103:1419–26.
2. Gudbjartsson T, Jeppsson A, Sjogren J, et al. Sternal wound
infections following open heart surgery—a review. Scand
Cardiovasc J 2016;50:341–8.
3. Lepelletier D, Bourigault C, Roussel JC, et al. Epidemiology
and prevention of surgical site infections after cardiac sur-
gery. Med Mal Infect 2013;43:403–9.
4. Yu AW, Rippel RA, Smock E, Jarral OA. In patients with
post-sternotomy mediastinitis is vacuum-assisted closure
superior to conventional therapy? Interact Cardiovasc
Thorac Surg 2013;17:861–5.
5. Cotogni P, Barbero C, Rinaldi M. Deep sternal wound
infection after cardiac surgery: evidences and controversies.
World J Crit Care Med 2015;4:265–73.
6. Spindler N, Lehmann S, Steinau HU, Mohr FW, Langer S.
[Complication management after interventions on thoracic
organs: deep sternal wound infections]. Chirurg 2015;86:
228–33.
7. Kaul P. Sternal reconstruction after post-sternotomy media-
stinitis. J Cardiothorac Surg 2017;12:94.
8. Kannan RY. The internal mammary artery perforator flap
and its subtypes in the reconstruction of median sternotomy
wounds. J Thorac Cardiovasc Surg 2016;152:264–8.
9. Koulaxouzidis G, Orhun A, Stavrakis T, Witzel C. Second
intercostal internal mammary artery perforator (IMAP) fas-
ciocutaneous flap as an alternative choice for the treatment of
deep sternal wound infections (DSWI). J Plast Reconstr
Aesthet Surg 2015;68:1262–7.
10. van Wingerden JJ, Lapid O, Boonstra PW, Mol BA de. Muscle
flaps or omental flap in the management of deep sternal
wound infection. Interact Cardiovasc Thorac Surg 2011;13:
179–87.
11. Dornseifer U, Kleeberger C, Ehrl D, Herter F, Ninkovic M,
Iesalnieks I. Arteriovenous loop-independent free flap
reconstruction of sternal defects after cardiac surgery.
J Reconstr Microsurg 2016;32:506–12.
12. Taeger CD, Horch RE, Arkudas A, et al. Combined free flaps
with arteriovenous loops for reconstruction of extensive
thoracic defects after sternal osteomyelitis. Microsurgery
2016;36:121–7.
13. Ennker IC, Ennker JC. Management of sterno-mediastinitis.
HSR Proc Intensive Care Cardiovasc Anesth 2012;4:233–41.
14. Rupprecht L, Schmid C. Deep sternal wound complications:
an overview of old and new therapeutic options. Open J
Cardiovasc Surg 2013;6:9–19.
15. Davison SP, Clemens MW, Armstrong D, Newton ED,
Swartz W. Sternotomy wounds: rectus flap versus modified
pectoral reconstruction. Plast Reconstr Surg 2007;120:929–34.
16. Netscher DT, Eladoumikdachi F, Goodman CM. Rectus
abdominis muscle flaps used successfully for median
Ann Thorac Surg
2019;-:-–-
sternotomy wounds after ipsilateral internal mammary ar-
tery ligation. Ann Plast Surg 2001;47:223–8.
17. Spindler N, Etz CD, Misfeld M, Josten C, Mohr FW,
Langer S. Omentum flap as a salvage procedure in deep
sternal wound infection. Ther Clin Risk Manag 2017;13:
1077–83.
18. Mathes SJ, Nahai F. Classification of the vascular anatomy of
muscles: experimental and clinical correlation. Plast Reconstr
Surg 1981;67:177–87.
19. Watanabe K, Kiyokawa K, Rikimaru H, Koga N, Yamaki K,
Saga T. Anatomical study of latissimus dorsi musculocuta-
neous flap vascular distribution. J Plast Reconstr Aesthet
Surg 2010;63:1091–8.
20. Glassey N, Perks GB, McCulley SJ. A prospective assessment
of shoulder morbidity and recovery time scales following
latissimus dorsi breast reconstruction. Plast Reconstr Surg
2008;122:1334–40.
21. Sears ED, Wu L, Waljee JF, Momoh AO, Zhong L, Chung KC.
The impact of deep sternal wound infection on mortality and
BOTA ET AL 7
STANDARDIZED DEEP STERNAL WOUND COVERAGE
resource utilization: a population-based study. World J Surg
2016;40:2673–80.
22. Kubota H, Miyata H, Motomura N, et al. Deep sternal wound
infection after cardiac surgery. J Cardiothorac Surg 2013;8:132.
23. Fu RH, Weinstein AL, Chang MM, Argenziano M,
Ascherman JA, Rohde CH. Risk factors of infected sternal
wounds versus sterile wound dehiscence. J Surg Res
2016;200:400–7.
24. Hassan M, Smith JM, Engel AM. Predictors and outcomes of
sternal wound complications in patients after coronary artery
bypass graft surgery. Am Surg 2006;72:515–20.
25. Seder CW, Rocco G. Chest wall reconstruction after extended
resection. J Thorac Dis 2016;8(Suppl 11):S863–71.
26. Tizian C, Borst HG, Berger A. Treatment of total sternal
necrosis using the latissimus dorsi muscle flap. Plast
Reconstr Surg 1985;76:703–7.
27. Dejesus RA, Paletta JD, Dabb RW. Reconstruction of the median
sternotomy wound dehiscence using the latissimus dorsi
myocutaneous flap. J Cardiovasc Surg (Torino) 2001;42:359–64.