Introduction

Breast cancer is the leading cancer in women worldwide affecting 1 in every 9 women [1, 2].

Though many patients are treated by breast conserving therapy [3], some patients will require mastectomy possibly combined with a breast reconstruction which may be performed as an immediate procedure [4].

Implant-based immediate breast reconstruction (IBR) requires a healthy well-vascularized flap accomplished by careful planning and close collaboration with breast surgeons. However, the overall risk of complications can be as high as 60% [4,5,6] with rates of surgical site infections and flap necrosis of 20–30% [7].

Prevention of surgical site complications in oncological patients is crucial, not only to avoid a negative effect on patients’ quality of life (QoL), but also risk of postponing adjuvant treatment.

affecting both recurrence risk and overall survival—constituting a significant burden for society [8,9,10].

Negative pressure wound therapy (NPWT), known as vacuum assisted closure (VAC), has been widely used treating different types of wounds by generating negative pressure leading to removal of excess fluid, reduction of edema, increased dermal perfusion, stimulating the formation of granulation tissue and reducing bacterial colonization [10]. Studies have reported NPWT to be associated with a reduced risk of complications, improved scar quality and enhanced recovery [11,12,13]. In breast surgery, use of NPWT is reported to lower the risk of infections, wound dehiscence, necrosis and seroma and cost-efficient [8,9,10, 14, 15].

Incisional negative pressure wound therapy (iNPWT), is a newer approach for surgical site closure based on the principles behind NPWT. Recently, Prevena™ has been approved as an iNPWT-system applied at completion of the breast reconstructive procedure.

Prevena™ consists of a sponge foam with a bactericidal silver linning, dressed with an adherent film, connected to a vacuum device [16]. The Prevena™ foam and dressing is placed over the suture line and kept in situ for 7 days (https://www.3m.com/3M/en_US/p/d/b5005265118/).

Indocyanine green angiography (ICG-A) is an intraoperative imaging modality developed to assist in quantification of tissue perfusion and perforators in breast reconstructive procedures. Use of ICG-A is associated with a lower risk of per- and postoperative complications within breast reconstructive procedures [17,18,19,20].

We hypothesize that iNPWT, administrated by Prevena™ can potentially reduce the postoperative complications in patients undergoing IBR, compared to conventional postoperative dressing (3 M™ Micropore™ tape).

Furthermore, that intraoperative results of ICG-A are associated with mastectomy skin flap necrosis and postoperative outcomes.

The aim of this study is to evaluate whether iNPWT provides superior wound treatment for patients undergoing an IBR, compared to treatment with conventional postoperative dressing. Furthermore, to investigate if intraoperative use of ICG-A can predict and prevent postoperative complications.

Secondly, to evaluate scars, QoL, timing of adjuvant therapy and incidence of postoperative lymphedema of the affected limb/limbs after IBR.

Accordingly, we conducted a randomized controlled study (RCT) applying iNPWT in IBR using implants or tissue expanders (TE). Intraoperative ICG-A was applied in all procedures.

We present this paper in accordance with the CONsolidated Standards Of Reporting Trials (Consort) guidelines.

Methods

Patients

Forty-one patients undergoing IBR using implants or TE were included.

Patients were enrolled consecutively at the Department of Plastic Surgery and Burns Treatment, Copenhagen University hospital, Rigshospitalet, Denmark between February and December 2020.

The inclusion criteria were: Patients older than 18 years of age deemed suitable for IBR by the consultant plastic- and breast surgeon. Patients who were pregnant, breastfeeding or not able to understand enough Danish to comprehend the given information and to complete the study questionnaires were excluded. Patients allergic to iodine were also excluded [21].

Data collection

Authors adhered to the Consort guidelines. The study was approved by The Capital Denmark Region Committees on Health Research Ethics (H-19074545, 70,429).

Ethical considerations.

The study was approved and registered by the Capital Region of Denmark joint application for the Danish Data Protection Agency, the Scientific Ethical Committee of the Capital Region of Denmark, and the research project was carried out in accordance with the Helsinki Declaration II.

Randomization

The included patients were randomized to one of two arms: Prevena™ iNPWT treatment for 7 days (intervention group) or conventional postoperative dressing (control group), in an allocation ratio of 1:1, using a block randomization according to the SNOSE principles [22]. Regardless of which type of post-operative dressing the patient was randomized to receive, the surgical procedure was performed in the same way (standard of care). The study was not blinded, since the Prevena™ device makes a sound, when generating negative pressure, and the dressing looks a lot different compared to the conventional post-operative dressing.

Intraoperative ICG-A application and perfusion assessment

Twenty-five milligrams of ICG were diluted in 10 mL sterile water, an intravenous bolus administration of ICG (Verdye® 5 mg/mL, Diagnostic Green GmbH Feldkirchener Str. 7c 85,551 Kirchheim b. Munich, Germany) of 2.5 mg/mL is followed by a 10 mL flush with normal saline (2.5 mL of ICG solution for each administration)[23].

Peroperative tissue perfusion is visualized within 20 s using the SPY-Elite Fluorescence Imaging system® (Stryker AB, Malmö, Sweden, www.stryker.com) (Video 1). Perfusion was scored and quantified 45 s after injection, and recording concluded after 60 s [24]. The relative perfusion values were quantified using a cut-off value of 33% [25]. Each ICG injection was performed with a minimum of 20 min, and the same investigator conducted the angiographies and quantification of perfusion [26]. Tissue perfusion was consecutively assessed by the ICG-A and compared to the surgeons’ clinical assessment (capillary refill and bleeding from skin edges).

After mastectomy, skin flaps were assessed clinically to identify areas in risk of hypoperfusion. A sizer of appropriate size was inserted, and the dermis temporarily sutured. The first ICG-A was then performed assessing mastectomy skin flaps, with healthy skin outside the surgical field (usually the skin covering the area over the xiphoied process) serving as reference point (100%).

If the angiography was sufficient, breast reconstruction proceeded with either a prepectoral implant wrapped in acellular matrix (Tutomesh®), or a subpectoral (i.e., dual plane) implant/TE with similar matrix.

If the angiography showed values < 33%, the procedure was re-evaluated clinical assessment and a repeated ICG-A. Perfusion < 33% on the second angiography resulted in excision of the hypoperfused area (if located near incision area), a smaller implant or reconstruction with subpectoral placement of a TE. Upon completing the breast reconstruction, ICG-A was performed again to confirm and ensure sufficient perfusion. Two drains were used for each patient, and removed when producing less than 30 ml. for 2 consecutive days.

Complications

Complications were reported and analyzed both per patient and per breast including minor- and major complications, as well as loss of reconstruction.

Major complications were defined as complications requiring surgery/debridement in local- or general anesthesia. Minor complications were defined as complications treated conservatively: seroma, infections treated by antibiotics, small hematoma etc. Loss of reconstruction was defined as implant removal without replacement with TE.

Follow-up and postoperative clinical evaluation

Clinical evaluations were conducted 4 times during the trial: before surgery and at 4 weeks, 4–6 months, and 12 months postoperatively. Outpatient visits included assessment of patient well-being and a clinical examination, involving inspecting the surgical fields for signs of infection, necrosis, seroma, hematoma etc. Scars were assessed using the POSAS, and BREAST-Q questionnaires to evaluate QoL. Lymphedema measurements were also performed.

Patient reported satisfaction and aesthetic outcome (BREAST-Q)

Before surgery (at baseline), patients completed the Danish version of the BREAST-Q 1.0 pre-reconstruction module. At each clinical follow-up, the post-reconstruction module of BREAST-Q 1.0 was used to assess patients' QoL using the BREAST-Q score [27].

Patient and Observer Scar Assessment Score (POSAS 2.0)

Scars were assessed using the POSAS (Patient and Observer Scar Assessment Scale) at each post-operative clinical visit [28].

Assessment of postoperative lymphedema

To assess the incidence of postoperative lymphedema measurements of the affected limb/limbs were performed before surgery (baseline) and at each postoperative visit [29]. A specialized physiotherapist performed circumferential measurements (figure-of-eight), and bioimpedance spectroscopy (BIS) using the SOZO system (https://www.impedimed.com/products/sozo/). Presence of lymphedema was categorized based on the International Society of Lymphology staging system [30].

Endpoints

Our primary endpoint was the postoperative time to removal of the surgical drains, as we hypothesize that Prevena™ leads to faster and better regeneration, with fewer complications, thus leading to lover formation of seroma and thereby shorter drain time. Drains were removed when producing less than 30 ml. for 2 consecutive days.

Secondly, proportion of flaps where the ICG-A assessment showed sufficient/insufficient perfusion, and proportion where the intraoperative results of the ICG-A changed surgical decision making. Furthermore, analysis of the association between intraoperative ICG-A and postoperative complications: infection, hematoma, necrosis, epidermolysis and partial- or full flap loss was evaluated [31]. Development of lymphedema, changes in QoL and differences in POSAS were all considered as secondary endpoints, as well as timely initiation of adjuvant treatment. Timely initiation was defined as the date set by the oncologists at the first postoperative multidisciplinary conference.

Statistical analysis

Statistical analysis was performed using R for Mac OS X GUI R 4.2.2 GUI 1.79 High Sierra build (8160). R: Copyright © 2004–2021. The R Foundation for Statistical Computing. http://www.R-project.org.

Continuous variables were analyzed using Wilcox-signed rank test assuming non-distributed data.

Categorical variables were analyzed using Fisher’s exact test due to small sample size.

Analysis of the POSAS were made using Wilcox-signed rank test assuming non-distributed data.

The BREAST-Q was analyzed comparing pre- and postoperative scores within and between the patients allocated to either Prevena™ or conventional postoperative dressing. Two-way repeated measures ANOVA tests were used to analyze repetitive measurements comparing the 2 patient groups based on the assumption of normative data [24]. Statistical significance was set by p-values < 0.05.

The minimal clinically relevant difference the study was looking for, was a decrease of 3 days from surgery until removal of drains. The standard deviation for days until drain removal after immediate breast reconstruction was found by Kim et al. to be 3.66 days [26].

To achieve a statistical significance level of 5% and a study power of 80%, a minimum of 24 patients were required in each group. To raise the power to 88%, we planned to include 30 patients in each group.

Results

Of 41 included patients, 2 dropped out leaving 39 patients for inclusion. Twenty-two patients received unilateral procedures and 17 bilateral, comprising a total of 56 breasts with 30 breasts receiving Prevena™ and 26 standard dressing. Of the 39 patients, 25 received a prepectoral implant-based IBR and 14 were subpectoral.

Mean age was 46.1 years (range 31–65) and mean body mass index (BMI) 23.3 kg/m2 (range 18.7–32.5). Baseline demographics are depicted in Table 1.

Table 1 Patient characteristics
Full size table

Patients were randomly allocated according to the randomization protocol, with 21 patients (30 breasts) in the Prevena™ group and 18 patients (26 breasts) treated with conventional post-operative dressing (Fig. 1).

Fig. 1
figure 1

On the left a patient randomized to postoperative iNPWT Prevena™. On the right a patient randomized to conventional postoperative dressing

Full size image

Statistical analysis on smoking status, alcohol consumption, indication for surgery and adjuvant treatment showed no difference in baseline characteristics between the 2 groups. Furthermore, there were no significant difference in baseline characteristics or complication rates between patients receiving a pre- or subpectoral IBR.

Time to removal of the surgical drains: Prevena™ or conventional dressing

Mean time from breast reconstruction to removal of surgical drains was 10.2 days in the Prevena™ group and 9.3 days in the group treated with conventional postoperative dressing. Analysis showed no significant difference (p = 0.30, W = 202) Table 2.

Table 2 Complications per patient
Full size table

Hospitalization: Prevena™ or conventional dressing

Mean hospitalization time did not differ in the groups (p = 0.28, W = 151) Table 2. Patients receiving conventional postoperative dressing had a mean hospitalization time of 4.9 days and 4.3 days in the Prevena™ group.

Postoperative dressing and complications: Prevena™ or conventional dressing

Complications including epidermolysis, necrosis, infection, seroma and hematoma were analyzed both per breast and per patient according to the randomization arm: Prevena™ or conventional dressing Tables 2 and 3. Also, complications were grouped in minor, major and loss of reconstruction according to postoperative dressing. Analysis showed no significant differences in incidence of postoperative complications Tables 2 and 3.

Table 3 Complications per breast
Full size table

Peroperative ICG-A results and postoperative complications

Intraoperative ICG-A found sufficient perfusion in 30/39 patients (77%) Table 4.

Table 4 Intraoperative ICG-A results and surgical complications (n = patients)
Full size table

Overall, 9/30 patients (23%) demonstrated insufficient perfusion by the ICG-A, whereof 8 of the 9 (88.9%) affected/influenced the intraoperative surgical decision making. Of these 8 patients, 7 (87.5%) experienced postoperative complications consisting of 4 minor complications, 3 major complications and 1 loss of reconstruction. In total, 33.3% with insufficient ICG-A developed major complications.

In 1 patient where the ICG-A showed insufficient perfusion, the surgeons chose to wait and see. This patient healed without complications.

Of the 30 patients where the ICG-A assessed the perfusion to be sufficient, 12 (40%) developed postoperative complications. Of these 12, 10 had minor complications, 2 major complications and 1 loss of reconstruction. Of the patients with sufficient perfusion on the ICG-A 6.67% developed major complications.

Analysis showed no significant association between ICG-A-result and incidence of overall postoperative complications. Also, there were no significant association between ICG-A result and major complications Table 5.

Table 5 Fisher’s Exact Test. Association between ICG-A and complications
Full size table

Postoperative follow-up, BREAST-Q and POSAS

BREAST-Q was completed by > 95% at each follow-up Table 1.

Analysis of the BREAST-Q showed a significant improvement in psychosocial- (p = 0.023) and physical well-being (p < 0.001) from the 4-weeks to 12 months follow-up Table 6.

Table 6 Postoperative follow-up. BREAST-Q and POSAS
Full size table

Analyzing the 2 treatment groups, patients receiving the conventional postoperative dressing showed significant better physical- (p = 0.023) and sexual well-being (p = 0.017) after 12 months compared to the Prevena™ group. The were no significant difference between the 2 groups in satisfaction with outcome Table 6.

The POSAS 4-weeks assessment were completed > 85% at each follow-up Table 1.

For both groups, there were no significant difference between the patients’ and observers’ scars scoring after 12 months Table 6, Fig. 2 + 3.

Fig. 2
figure 2

Bilateral subpectoral implant-based IBR. Patient randomized to conventional postoperative dressing. On the left a clinical photo before breast reconstruction, on the right a clinical photo 12 months after breast reconstruction

Full size image

Assessment of postoperative lymphedema and administration of adjuvant therapy

Four patients had preoperatively been diagnosed with lymphedema of the upper extremity Table 1. All 4 had previously undergone axillary lymph node dissection (ALND) and underwent a delayed breast reconstructive procedure. One had been treated with neo-adjuvant chemotherapy (NACT), 2 had undergone adjuvant chemo- and radiation therapy (RT) and 1 had NACT and adjuvant RT.

After 12 months follow-up, 1 additional patient presented with lymphedema of the upper extremity. She had undergone immediate breast reconstruction and ALND in the same procedure and received NACT and adjuvant RT Table 1.

Adjuvant treatment was administered on time in 32 patients (82%). Two patients (5%) experienced delayed adjuvant treatment due to surgical complications, delaying treatment in 7 and 10 days, respectively. Five patients (13%) had their adjuvant treatment delayed due to non-surgical reasons Table 1.

Due to few patients, we did not investigate associations between surgical treatment and lymphoedema. Nor between delayed adjuvant treatment and postoperative complications in relation to postoperative dressings.

Discussion

This single-center RCT investigated if iNPWT administrated by Prevena™ leads to faster recovery and reduces the rate of postoperative complications in patients undergoing IBR. Patients were randomized to receive either Prevena™ or conventional postoperative dressing (3 M™ Micropore™ tape) on the closed incisions.

Time to removal of surgical drains was set as an indicator for faster recovery. Also, we investigated if the use Prevena™ was associated with an increased QoL and satisfaction with scars.

Secondly, intraoperative ICG-A was applied in all procedures analyzing association between ICG-A results and rate of postoperative complications.

Patients were followed for 1 year, including a clinical evaluation before the breast reconstructive procedure and 3 clinical evaluations postoperatively.

Use of iNPWT for closed incisions have been described by several authors, and there is growing evidence that it is associated with reduced complication rates following surgery [8, 10, 11, 15].

A multicenter RCT by Pieszko et al. demonstrated a significant decrease in surgical-site wound complications in patients treated with iNPWT for closed incisions in IBR [14]. iNPWT has also been useful for reduction of complications in mammoplasties and prepectoral breast reconstruction [28, 32], but inconclusive in oncoplastic breast surgery [29, 33].

This study did not find any significant association between the use of iNPWT by Prevena™ and reduced time to removal of drains. Also, there were no significant difference in mean hospitalization time. The rate of postoperative complications was analyzed both per breast and per patient, with no significant differences between the 2 groups in both minor- and major complications, and loss of reconstruction.

The insignificant results may be caused by a substantial lower sample size (39 patients were included) than anticipated when planning the study (a total of 60 patients, with 30 in each group).

Of the total patient population, 48.7% developed postoperative complications with major complications constituting 12.8% and 5.1% experiencing loss of reconstruction. These rates are comparable to previous published studies [14, 34, 35].

Intraoperative change according to the ICG-A was performed in 88.9% with insufficient ICG-A, yet the majority (87.5%) experienced postoperative complications with 33.3% being major complications.

Of the cases with sufficient perfusion by the ICG-A, 40% experienced postoperative complications and 6.7% of these were major complications.

Previous studies have demonstrated ICG-A to be associated with a reduced risk of postoperative complications in breast reconstructive procedures [17, 18, 20, 25, 36,37,38,39,40]. However, we did not find a significant association between ICG-A and postoperative complications,

Though the association between ICG-A and complications was not significant, complication rate was much higher in the group demonstrating insufficient ICG-A. This suggests that an inadequate angiography should result in thorough reevaluation of the surgical procedure, with particular emphasis on enhancing postoperative monitoring and care. This is essential to preempt potential complications.

Administration of adjuvant RT may be the most important risk factor for developing surgical-site complications [41, 42]. Analysis of association between complications and RT was not accomplished due to few patients receiving RT with no difference between groups.

There was no difference in overall postoperative QoL or satisfaction with scars, which is consistent to previous studies [14].

Adjuvant therapy was administered on time in > 80%. The majority of patients receiving delayed adjuvant treatment were caused by non-surgical reasons. During follow-up 1 patient developed lymphedema of the upper extremity which could be related the postoperative adjuvant treatment consisting of RT [43].

Limitations

This study is a RCT with 1 year follow-up. All but 1 patient completed the postoperative follow-up eliminating the risk of missing data.

The study is limited by a relatively small sample size of 39 patients. Unfortunately, due to structural changes, it was not possible to include the number of patients initially planned making the study prone to bias, and conclusions is to be assessed considering this.

Conclusions

This randomized controlled study investigated the use of iNPWT by Prevena™ compared to conventional postoperative dressing in immediate breast reconstructions using implants or tissue expanders. Furthermore, ICG-A was applied to investigate intraoperative tissue perfusion and postoperative complications.

No significant difference in time to recovery and time of hospitalization were found.

The 2 groups did not differ in rate of postoperative complications, QoL and satisfaction with scars within 1 year follow-up.

Though the rate of major complications were higher in the group with insufficient intraoperative ICG-A, there was no significant association between ICG-A and postoperative complications.

Larger randomized studies applying intraoperative ICG-A are needed to obtain higher quality data.