Eur J Cardiothorac Surg 2006;30:148-152
© 2006 Elsevier Science NL
Two-stage management of sternal wound infection using bilateral pectoralis major advancement flap
Clement H.K. Wong
a
,
*
,
Shireen Senewiratne
a
,
Bruce Garlick
a
,
Daniel Mullany
b
a Department of Cardiothoracic and Plastic Surgery, The Prince Charles Hospital, Rode Road, Chermside, Brisbane, Qld 4053, Australia
b Department of Intensive Care Services, The Prince Charles Hospital, Rode Road, Chermside, Brisbane, Qld 4053, Australia
Received 2 January 2006;
received in revised form 24 March 2006;
accepted 27 March 2006.
* Corresponding author. Address: 1/11 Sir Fred Schonell Drive, St. Lucia, Qld 4067, Australia. Tel.: +61 7 33717150; fax: +61 7 33717150. (Email: clement_wong1{at}excite.com).
 |
Abstract
|
|---|
Objective: To report our experience using two staged bilateral pectoralis major flap as the sole treatment modality for sternal wound infection. Methods: A retrospective study of 9417 open-heart surgery cases performed between 1998 and 2003 at The Prince Charles Hospital. Sixty-eight patients were referred to the plastic surgical team for consideration of bilateral pectoralis major flap as the sole treatment modality for sternal wound infection. Results: There was a trend for early referral for flap operation (median 10 days) (p
= 0.49). The median postoperative ventilation time and ICU stay were 1 and 2 days, respectively. The median hospital stay after flap operation was 15.5 days. One-year overall survival was 91%. Ninety-five per cent healed stable sternum was achieved with 100% failure in patients with chronically unstable sternum. Early referral appears to be an important factor in preventing osteomyelitis formation (p
= 0.05) with the longest recurrence at 10 months postoperatively. Conclusions: The key to the successful management of deep sternal wound infection is early referral for pectoralis major flap operation. Our approach is safe with good long-term outcomes. We recommend this approach in all severe deep sternal wound infection but not in patients with chronic unstable sternum.
Key Words: Sternum • Mediastinal infection • Wound dehiscence
|
1. Introduction
|
|---|
The management of sternal wound infection poses one of the most controversial and challenging problems for both cardiac and plastic surgeons. The incidence of deep sternal wound infection following open-heart surgery in our institution in the last five years is 1%. This is comparable to the reported worldwide incidence of 0.4–5% [1]. Although it is relatively uncommon, sternal wound infection can lead to significant morbidity, mortality and hospital cost. The mortality rate has been reported to be as high as 70% [2]. Thus, many different approaches of treating this devastating complication have been reported. These included rewiring and irrigation [3] and flap reconstruction [4]. In this study, we will report our experience with bilateral pectoralis major advancement flap as the sole treatment modality for deep sternal wound infection.
|
2. Methods
|
|---|
Of the 9417 consecutive open-heart surgery cases performed between 1998 and 2003 at The Prince Charles Hospital, Brisbane, Australia, 68 patients were referred to the plastic surgical team for consideration of bilateral pectoralis advancement major flap operation. Three were excluded from our analysis because of either (1) the operations were not performed by the same surgeon (two patients) or (2) the pectoralis major muscle was not used as the sole flap (one patient), leaving 65 patients for analysis. All patients underwent a two-stage repair process. The first stage consisted of debridement of all necrotic tissues and removal of sternal wires. The chest was then left open and packed with normal saline pack that was changed every 4 h with overnight saline irrigation. The second stage was performed only if the patients were haemodynamically stable with no temperature (>38 °C) for 24 h. Otherwise, the chest will continue to be left open with regular saline pack change. The second stage involved raising the pectoralis major flap from medial to lateral in the plane beneath the pectoralis major muscle, leaving a small medial cuff of pectoralis major medially. The dissection was performed using diathermy. Superiorly, the dissection stopped at the level of the clavicle. Laterally, the flap was raised as far as the anterior axillary line. The pectoralis minor, humeral insertion of the pectoralis major muscle and the thoracoacromial vessels and nerves were all left intact. Inferiorly, the rectus sheath was split. Sternal edges were approximated using 20–25 periosteal pulley sutures (0/0 polydioxanon monofilament absorbable sutures: PDS; Ethicon). Inferiorly, the rectus sheath was closed in two layers. The pectoralis major flaps were advanced and closed with 0/0 PDS. Two closed suction drains were put under each flap and the third one above the muscle flap but below the skin. Skin was closed using 3/0 monocryl. All patients had perioperative antibiotic cover. The choice and duration of antibiotic depended on the bacterial culture, sensitivities and clinical indications. All patients had a crop top (Tubigrip vest) for 6 weeks. Abduction of the upper arms was restricted for 2 weeks. Drains were removed when the total drainage was less than 20 ml in 24 h.
In our study, patients were purposely grouped according to the time of referral to the plastic surgical team for flap operation rather than using the classification as described by El Oakley and Wright [5]. El Oakley classified mediastinal wound infection into two broad categories: superficial and deep (mediastinitis) sternal wound infection. They defined infection confined to subcutaneous tissue as superficial whereas patients with infection involving sternum (osteomyelitis), which may or may not be associated with retrosternal space as deep sternal wound infection. They further classified deep sternal infection into four subtypes based on the time of first presentation, previous results of attempted treatment and patients’ risk factors. In our study, Group 1 patients (acute infective group, n
= 48) were referred for flap operation within 6 weeks of infection and with microbiological evidence of infection at the time of referral. Group 2 patients (acute non-infective, n
= 4) were referred within 6 weeks with unstable or discharging sternum but with no microbiological evidence of infection. Group 3 patients (chronic infective, n
= 10) had microbiological evidence of infection and were referred more than 6 weeks from the initial infection. All of these patients had failed attempts of debridement and rewiring. Group 4 patients (chronic non-infective, n
= 3) had chronically unstable sternum and were referred more than 6 weeks from the initial events with no evidence of infection at the time of referral.
2.1 Statistical analysis
Time to event curves was plotted using the product limit method of Kaplan–Meier. The number of patients at risk is shown at specific time points under each curve with 95% confidence intervals. The chi-square analysis was used to compare discrete variables and the Student's t-test for continuous variables.
|
3. Results
|
|---|
3.1 Patient characteristics (Table 1)
Most of our patients had coronary artery bypass surgery (CABG) (54%), cardiac valve replacement (13%) or combined CABG and valve replacement (22%). A minority of patients had cardiac transplant (5%) or aortic root replacement (6%). Forty-six per cent of patients had no known risk factors (Table 1
).
We observed a trend towards earlier referral for flap operation (p
= 0.49) after the diagnosis of sternal wound infection in recent years (Table 2
). The overall median length of hospital stay after flap operation was 15.5 days. In fact, 57% of patients could be discharged 11 days after flap operation. The post-flap ventilation and ICU stay was 1 and 2 days, respectively (Table 3
). Seventy-five per cent of patients had one debridement and 18% had two debridements before flap operation. Sixty per cent of patients were still ventilated before the second stage operation.
Based on El Oakley's classification, 25% of patients in Group 1 had superficial mediastinal infection and the remaining patients had deep sternal infection. All the patients in Group 3 had deep sternal wound infection. Table 4
shows further subtypes of mediastinitis. In Group 1, two patients were on glucocorticosteroid preoperatively and three patients had bilateral mammary artery grafts. Four patients in this group were initially classified as Group 2 but then reclassified as Group 1 because the sternal wound specimens sent during the flap operation had positive growth. Four patients in Group 3 were referred to the plastic surgical team with chronic sinus. The remaining six patients in Group 3 presented with repeat purulent discharge and chest pain. Five per cent of patients presented with a chronic unstable sternum (Group 4). All the patients had previous attempts of rewiring before referring for flap operation in this group. In one patient, two attempts were made to treat the chronic unstable sternum using pectoralis major flap but without success.
In this series, bilateral pectoralis major flap was used as primary treatment for sternal wound infection/dehiscence in 80% of all the referred patients. The remaining was used as a salvage method after other treatment modalities (debridement, rewiring and irrigation) failed. Fifty-eight patients had positive sternal wound culture. The most common pathogen was MRSA (35%), followed by Staphylococcus epidermidis (29%) and Staphylococcus aureus (18%). A total of 14 patients had two pathogens. Candida was cultured from two patients (one of which was posttransplant).
3.2 Outcome (Table 5)
Follow-up was incomplete for 10 patients who were lost to follow-up after 6 months. Those three patients who were excluded from our study achieved a complete healed sternum with no complication. Overall, our series achieved 95% healed and stable sternum. Five per cent of sternums remained unstable despite flap repair. There was no reportable shoulder weakness. The important risk factor for the failure of Group 4 patients appeared to be a body mass index (BMI) of greater than 32 (normal = 25–30) (p
< 0.01). Other risk factors such as age, culture results and time of referral and debridement were not statistically significant (Table 5
).
3.3 Complications
The 30 days perioperative mortality rate was zero with one-year overall survival of 91%. There were no intraoperative deaths. Two patients died from pulmonary sepsis. One died from renal failure and two died from multi-organ failure that was not directly related to sternal wound infection. Forty-eight per cent of patients had no complications after flap operation at the one-year mark. The most common complication was haematoma (Table 6
). Twenty-eight per cent of those who developed a haematoma or seroma needed further surgical drainage or exploration. Most of the seromas and haematomas occurred within 48 h after the flap operation.
Notably, patients were unlikely to develop osteomyelitis if they did not develop it within one year after the flap operation. The longest recurrence was 10 months. The most important prognostic factor in the development of chronic osteomyelitis appeared to be the referral time (p
= 0.05) (Table 7
). While 10 patients (15.4%) developed recurrent osteomyelitis following flap operation and required further debridement, all of these patients were late referral (Group 3).
|
4. Discussion
|
|---|
There has been a notable change in the management of deep sternal wound infection in the last five years in our institution. An aggressive approach has been adopted with early debridement and referral to the plastic surgical team for flap operation. This approach also decreases the amount of tissues to be removed and thus saving patients from extensive reconstruction. In addition, our series has demonstrated that by referring patients early, within 48 h after the diagnosis of sternal infection, the incidence of chronic osteomyelitis appears to be reduced.There are several reasons for using two-stage bilateral pectoralis major flap as the sole treatment modality in this series. First, we are aware of the importance of accurately diagnosing the extent of the sternal infection as this will determine the choice of treatment and subsequent outcomes. It is not always clear-cut and requires a lot of experience, which we lack in the early stage of this new approach. These concerns are highlighted in our series that four patients in Group 2 needed to be reclassified. It showed that early sternal dehiscence may lead to deep sternal wound infection as a result of migration of bacteria into mediastinum after skin separation [6]. Second, we were concerned about potential complications and recurrence associated with early closure. These concerns were shared by Lindsey [7] and Fleck et al. [8] but not by Ascherman et al. [9]. They reported good long-term results using one-stage approach (Table 8
). These conflicting results may be explained by the different subtypes of sternal wound infections. According to Francel and Kouchoukos [10], one-stage approach would be appropriate in patients with no purulent mediastinal infection and negative culture who have inadequate sternal bones. A two-stage approach should be considered if there was evidence of severe mediastinal infection or the surgeons were in doubt of the suitability of primary closure. In the future, one stage approach may be an appropriate option for patients in Group 2.
In theory, some patients in this series may be potential candidate for rewiring. It can facilitate early discharge but from our previous experience (rewiring), wires can act as nidus for further infection. Our approach may seem extreme but bear in mind that the reported failure rate of rewiring in Francel's group was 35% despite their extensive experience. With all those who failed, muscle flaps were required which resulted in prolonged hospitalisation and recovery time. Therefore, we believe flap reconstruction should be the first line treatment strategy in the majority of patients with deep sternal wound infection. Third, more than 50% of our patients had Type III or IV mediastinal infection and they were not suitable for one-stage approach. Finally, our results showed that the overall hospital stay is comparable to one-stage approach (Table 8).
It is worth pointing out that our recurrence (osteomyelitis and sinus tract) rate was approximately 20% with the longest recurrence at 10 months. This was due to many late referrals with chronic osteomyelitis in the early 2000 (Group 3). This is in contrast to the longest recurrence of 27.6 months reported by Pairolero et al. [11]. They attributed their recurrence to inadequate debridement rather than late referral and reported that extensive resection of sternum, manubrium and costocartilage was the key to prevent recurrence rather than debridement alone. From our experience, this approach may be too aggressive. Certainly, there are patients that have fragile sternum especially those who are on steroid or no sternum left after debridement. In that case, they may have inadequate bone stock for our technique. Therefore, the sternum may not be able to be re-approximated and a turn over flap or omental flap may be required. So far, all of the patients in our series had their sternum re-approximated and achieved complete wound healing with debridement alone, even in patients with chronic osteomyelitis. The key, we believe, is the two-stage approach we used and the way where we put our PDS stitches.
By re-approximating the sternum using 20–30 PDS stitches that went through the cuff of pectoralis major, tension can be distributed equally throughout and thus maximising sternal integrity and avoiding pulling through of the sutures. Moreover, big muscle flaps such as pectoralis turnover flap are not required to fill in the space. Therefore, problems such as loss of the axillary fold will not occur. Furthermore, pectoralis major advancement flap can be applied to any post-cardiac surgical patient because it does not require an intact internal mammary artery. Additionally, normal pulmonary function was restored. This was supported by Cohen et al. [12] and was evident from our series that patients can be extubated early after flap operation. None of our patients required rectus abdominis or omentum flaps to cover the lower portion of the sternotomy wound. In fact, these flaps should be avoided because they can compound the patients’ problems by opening another compartment [13]. Nevertheless, unilateral flaps may be suitable in selected cases.
Recently, Fleck et al. [14] and Hersh et al. [15] reported using vacuum-assisted closure (VAC) as an alternative to saline pack. They believed it was less labour intensive (dressing change every 48 h) and patients had a shorter ICU stay. Fuchs et al. [16] also reported that patients treated with VAC had the advantage of early rewiring and hospital discharge. However, this approach was not widely adopted in this series because the majority of CABG patients had their pleura opened. We were also concerned that the condition of the sternal wound might have changed during the period when VAC was applied. Furthermore, we supported the comment by Francel [17] that VAC might have a limited role in CABG patients using mammary artery as graft because they were at high risk of extensive bone loss and early aggressive debridement and reconstruction should be employed to facilitate early discharge and healing.
The management of chronic unstable sternum remains a real challenge. A high BMI appeared to be the major contributing factor of failure but we have to be cautious in interpreting this result due to the small number of patients in Group 4 (n
= 3). Nevertheless, the complete failure suggested that these patients may have a different pathophysiology. Since all the patients had no evidence of infection at the time of referral, we suspect that chest wall instability was due to a mechanical problem rather than an infective cause. Our technique is designed to eliminate sepsis by occluding dead space. It restores chest wall stability and to a certain extent provides mechanical support. But clearly, it is not appropriate for this group of patients. Thus, a different strategy is needed. Chase et al. [18] reported using transmanubrial compress plates for internal fixation of both infected and non-infected dehiscence sternum with 97% success rate in restoring sternal stability. However, 13% of patients required metal plate removal due to infection and screw loosening. Therefore, patients in Group 4 may benefit from this technique. In our opinion, internal fixation should only be reserved for those who have genuine chronic unstable sternum without infection though it may be hard to be certain all the time.
The management of sternal wound infection has undergone significant changes in the last few years. Despite all the controversy, our approach of using bilateral pectoralis advancement major flap as a two-stage process has significantly improved long-term outcomes and reduced financial burden to our health system. Our study has highlighted the importance of early diagnosis and referral in the management of sternal wound infection. Teamwork involving the cooperation of surgical teams, anaesthetic staffs, intensive care unit and wound care nursing staffs is paramount to the successful management of this potentially devastating complication.
|
Acknowledgments
|
|---|
I would like to thank Mr Lee Tripcony and Ms Jacqueline Keller for their assistance with the statistical analysis.
|
References
|
|---|
- Sarr MG, Gott VL, Towsend TR. Mediastinal infection after cardiac surgery. Ann Thorac Surg 1984;38:415-423.[Abstract]
- Cheung EH, Craver JM, Jones EL, Murphy DA, Hatcher CR, Guyton RA. Mediastinitis after cardiac valve operations. J Thorac Cardiovasc Surg 1985;90:517-522.[Abstract]
- Shumacker HB, Mandelbaum I. Continuous antibiotic irrigation in the treatment of infection. Arch Surg 1963;86:384-387.[Medline]
- Jurkiewicz MJ, Bostwick J, Hester TR, Bishop JB, Craver J. Infected median sternotomy wounds. Successful treatment by muscle flaps. Ann Surg 1980;191:738-743.[Medline]
- El Oakley RM, Wright JE. Postoperative mediastinitis: classification and management. Ann Thorac Surg 1996;61:1030-1036.[Abstract/Free Full Text]
- Demmy TL, Park SB, Liebler GA, Burkholder JA, Maher TD, Benckart DH, Magovern GJ, Magovern Sr. GJ. Recent experience with major sternal wound complications. Ann Thorac Surg 1990;49:458-462.[Abstract]
- Lindsey JT. A retrospective analysis of 48 infected sternal wound closures: delayed closure decreases wound complications. Plast Reconstr Surg 2002;109:1882-1885.[Medline]
- Fleck TM, Koller R, Giovanoli P, Moidl R, Czerny M, Fleck MCM, Wolner E, Grabenwoger M. Primary or delayed closure for the treatment of poststernotomy wound infections. Ann Plast Surg 2004;52:310-314.[CrossRef][Medline]
- Ascherman JA, Patel BA, Malhotra SM, Smith CR. Management of sternal wounds with bilateral pectoralis major myocutaneous advancement flaps in 114 consecutively treated patients: refinements in technique and outcomes analysis. Plast Reconstr Surg 2004;114:676-683.[Medline]
- Francel TJ, Kouchoukos NT. A rational approach to wound difficulties after sternotomy: the problem. Ann Thorac Surg 2001;72:1411-1418.[Abstract/Free Full Text]
- Pairolero PC, Arnold PG, Harris JB. Long term results of pectorlais major muscle transposition for infected sternotomy wounds. Ann Surg 1991;213:583-589.[Medline]
- Cohen M, Yaniv Y, Weiss J, Greif J, Gur E, Wertheym E, Shafir R. Median sternotomy wound complication: the effect of reconstruction on lung function. Ann Plast Surg 1997;39:36-43.[Medline]
- Oh AK, Lechtman AN, Whetzel TP, Thomas P, Stevenson TR. The infected median sternotomy wound management with the rectus abdominis musculocutaneous flap. Ann Plast Surg 2004;52:367-370.[Medline]
- Fleck Tm, Fleck M, Moidl R, Czerny M, Koller R, Giovanoli P, Hiesmayer MJ, Zimpfer D, Wolner E, Grabenwoger M. The vacuum-assisted closure system for the treatment of deep sternal wound infections after cardiac surgery. Ann Thorac Surg 2002;74:1596-1600.[Abstract/Free Full Text]
- Hersh RE, Jack JM, Dahman MI, Morgan RF, Drabe DB. The vacuum-assisted closure device as a bridge to sternal wound closure. Ann Plast Surg 2001;46:250-254.[CrossRef][Medline]
- Fuchs U, Zittermann A, Stuettgen B, Groening A, Minami K, Koerfer R. Clinical outcome of patients with deep sternal wound infection managed by vacuum-assisted closure compared to conventional therapy with open packing: a retrospective analysis. Ann Thorac Surg 2005;79:526-531.[Abstract/Free Full Text]
- Francel TJ. Invited commentary. Ann Thorac Surg 2002;74:1600.[Free Full Text]
- Chase CW, Franklin JD, Guest DP, Barker DE. Internal fixation of the sternum in median sternotomy dehiscence. Plast Reconstr Surg 1999;103:1667-1673.[Medline]
- Perkins DJ, Hunt JA, Pennington DG, Stern HS. Secondary sternal repair following median sternotomy using interosseous absorbable sutures and pectoralis major myocutaneous advancement flaps. Br J Plast Surg 1996;49:214-219.[CrossRef][Medline]
- Jeevanandam V, Smith CR, Rose EA, Malm JR, Hugo NE. Single-stage management of sternal wound infections. J Thorac Cardiovasc Surg 1990;99:256-263.[Abstract]
- Hugo NE, Sultan MR, Ascherman JA, Patsis MC, Smith CR, Rose EA. Single-stage management of 74 consecutive sternal wound complications with pectoralis major myocutaneous advancement flaps. Plast Reconstr Surg 1994;93:1433-1441.[CrossRef][Medline]
Top
Abstract
1. Introduction
