Key words breast cancer - implant reconstruction - influencing factors - aesthetic result
Schlüsselwörter Mammakarzinom - Implantatrekonstruktion - Einflussfaktoren - ästhetisches Ergebnis
Introduction
One of eight women will be diagnosed with breast cancer in the course of her life [1 ], [2 ]. Around 74 500 cases are registered annually in Germany, and breast cancer still is one of the ten most common causes of mortality. While therapy options for breast cancer have evolved [3 ], surgery remains mandatory for all breast cancer patients. Despite the oncological safety of breast-conserving therapy and radiation for most patients [4 ], [5 ], mastectomy is necessary for a selected group of patients with large or multicentric tumors or skin involvement, for whom breast conserving therapy is not possible. With mastectomy rates of between 20 and 50 % in some countries, breast reconstruction is an important issue for preserving the body image of affected women. Mastectomy is also associated with a number of quality-of-life issues and physical sequelae [6 ], [7 ], [8 ]. In many women who have undergone mastectomy, the wish for aesthetic improvement does not decrease even after long-term follow-up [7 ].
The decision whether to opt for heterologous or autologous reconstruction needs to be considered with regard to both patient safety and patient satisfaction. The impact of patient characteristics and comorbidities on patient satisfaction still has to be determined. With half of all breast cancer patients younger than 65 years at diagnosis and every 6th woman younger than 50 years [1 ], increasing numbers of women will desire breast reconstruction.
Fernández-Delgado evaluated the role of breast reconstruction in a cohort of patients who underwent either immediate or delayed reconstruction after mastectomy and found significantly lower levels of stress and anxiety in those women who pursued reconstruction compared to those who did not opt for reconstruction [9 ].
To facilitate a better patient selection and achieve a higher degree of patient satisfaction with reconstructive results, we analyzed surgical outcome as a function of patient characteristics and comorbidities in a large cohort of post-mastectomy patients who had had implant reconstruction. Patient self-image and the impact of self-image on partner interaction were also analyzed.
Patients and Methods
Patients
The survey was carried out in a cohort of patients treated consecutively at the Breast Center Düsseldorf, Luisenkrankenhaus. All patients who underwent either immediate or deferred implant reconstruction after mastectomy between 2000 and 2010 were eligible for inclusion in the study. Patients were identified from a prospectively maintained database which registers all patients who have reconstructive breast surgery. The database had datasets for 567 patients who underwent reconstructive surgery in the period stated above. Of those patients, 310 patients were excluded because they had autologous reconstruction (n = 310), resulting in study population of 257 implant reconstructions, 97 of which were simultaneous reconstructions and 151 of which were two-stage reconstructions after tissue expander procedures (n = 151). The study was approved by the Institutional Review Board (IRB) and complies with the Declaration of Helsinki.
Data collection
Patient characteristics at the time of diagnosis and data on the initial surgery were obtained from the patientsʼ medical charts. Prospectively documented surgery complications were retrieved from the patientsʼ charts for this analysis.
Patients were asked to complete a customized questionnaire which was sent to patients by regular mail. The questionnaire comprised a set of 60 questions addressing short- and long-term sequelae after surgery, patient satisfaction with the aesthetic outcome, and partner interaction.
Primary endpoint was the correlation between patient characteristics and surgical outcome. Secondary endpoint was the influence of patient characteristics on patient satisfaction. Data were retrieved from patientsʼ medical charts and questionnaires to evaluate aesthetic outcomes and mid- and long-term surgical outcomes.
Statistics
For explorative statistical analysis, we used standard univariate methods appropriate to the scale of the variables (e.g. Fisherʼs Exact test for n × m tables or Mann-Whitney-Wilcoxon test to compare two distributions) and applied a significance level α of 5 % to each test.
Results
Description of patient cohort
One hundred and eighty-five out of 257 patients (71.98 %) completed the questionnaire. Mean age was 49 years (range: 24–79 years). Mean body weight was 65.71 ± 10.92 kg, and height ranged from 148 to 185 cm (median: 168 cm; mean: 167.7 cm ± 6.46 cm), resulting in a median BMI of 22.44 (range: 16.33–40.09; mean: 23.29 ± 3.56). Patient characteristics are summarized in [Table 1 ].
Table 1 Patient characteristics.
Minimum
Maximum
Median
Mean
Standard deviation
Age
24 years
79 years
49 years
50.16 years
± 10.95 years
Height
148 cm
185 cm
168 cm
167.7 cm
± 6.46 cm
Weight
41 kg
120 kg
65 kg
65.71 kg
± 10.92 kg
BMI
16.33
40.09
22.44
23.29
± 3.56
One hundred and thirty patients (50.58 %) received chemotherapy, 30 (23 %) of them in a neoadjuvant, and 78 (60 %) in an adjuvant setting; 22 (17 %) patients did not specify the chemotherapy setting. Thirty-five patients (13.62 %) had radiotherapy, 2 of them after implantation of an expander system. Forty-eight of 257 patients were smokers (18.68 %); 9 (3.50 %) of them smoked less than 5 cigarettes per day, 22 (8.56 %) smoked 5–10 cigarettes per day, and 17 (6.62 %) smoked more than 10 cigarettes per day. Seven (2.72 %) women had been smokers in the past; 2 (0.78 %) of them had stopped smoking within the last 3 years, and 5 (1.95 %) had stopped smoking more than 3 years ago. Data analysis was done between 2.37 and 128.6 months (71 to 3858 days) after reconstruction. Follow-up was defined from the date of the last implant reconstruction to the date of sending the questionnaire or the patientʼs last visit to the clinic. Median follow-up was 36.8 months (mean: 41.29 ± 26.52 months). Median time from mastectomy to reconstruction was 6 months (mean: 6.29 ± 2.9 months).
Surgical procedures
The majority of patients (n = 151; 58 %) underwent two-stage reconstruction using an expander followed by subsequent replacement by an implant. One-stage immediate breast reconstruction was less common (n = 97; 38 %); one-stage reconstruction of the one side combined with two-stage reconstruction of the contralateral side was performed in 9 patients (4 %).
Ninety-seven women (37.74 %) had unilateral implant reconstruction; 7 of them had additional latissimus dorsi flap breast reconstruction with implant. Bilateral implant reconstruction was performed in 56 women (21.79 %), in 4 cases also with an additional latissimus dorsi flap.
Contralateral alignment of the breast – i.e., augmentation or reduction mammaplasty – was done in 102 patients (39.69 %). A titanized mesh was used with a dual-plane technique in 10 (3.89 %) of 257 patients; in 157 cases (61.09 %) corial flap was used for coverage of the lower breast pole. Seventy-one patients (27.63 %) with deferred breast reconstruction had secondary nipple-areola reconstruction.
Complications
Bleeding was reported as an early complication in 25 (9.73 %) of 257 women; 16 (6.23 %) had infection and fever, 13 (5.06 %) reported impaired wound healing, and 4 (1.56 %) had scar insufficiency ([Table 2 ]).
Table 2 Early complications (day 1–14).
Complication
N
%
Bleeding
mild (no revision)
5
1.95
intense (revision)
20
7.78
Fever/infection
16
6.23
Scar dehiscence
4
1.56
Impaired wound healing
13
5.06
The following late complications were recorded: dysesthesia in 101 cases (39.39 %); any type of restriction of arm movement in 49 patients (19.07 %); keloid scarring in 27 women (10.51 %); capsular fibrosis 24 cases (9.34 %); seroma in 20 cases (7.78 %); and scar insufficiency in 9 cases (3.5 %).
It was notable that impaired wound healing, scar insufficiency and dehiscence occurred more often in smokers. Of all the patients who were smokers – 9 patients who smoked less than 5 cigarettes, 22 patients who smoked 5–10 cigarettes, and 17 patients who smoked more than 10 cigarettes per day – only the last group had an increased number of complications related to wound healing, with impaired wound healing found in 11.76 % and scar insufficiency in 5.88 %. Former smokers, even those who had only stopped smoking within the last 3 years, did not experience any increase in these complications.
Implant rotation occurred in 19 cases (7.4 %) and loss of implant in 4 cases (1.56 %). The implant had to be removed because of local recurrence in 2 cases (0.78 %) and due to infection in another 2 cases (0.78 %). The infection rate in our cohort was comparatively low, with an incidence of 5.08 % (n = 3) in the group of patients where drainage remained in situ for 1–7 days, 3.18 % (n = 3) in the group where drainage remained in place for 8–14 days, and 8.75 % in the group where drainage remained for 15–21 days, and 50 % in the group were drainage remained in situ for more than 21 days (n = 2). [Table 3 ] lists the late complications after implant reconstruction.
Table 3 Late complications (> 14 days post surgery) after implant reconstruction.
Complication
n
%
Keloid
Total
27
10.51
15
5.48
12
4.67
Scar dehiscence
9
3.5
Dysesthesia
Total
101
39.39
61
23.74
25
9.73
7
2.7
3
1.17
2
0.78
1
0.39
1
0.39
1
0.39
Seroma
Total
20
7.78
7
2.72
13
5.06
Capsular fibrosis
diagnosed by physician
19
7.39
according to patient
5
1.95
Implant removal
Total
5
1.95
2
0.78
2
0.78
1
0.39
Implant rotation
3
1.17
Implant displacement
16
6.23
Implant expulsion
2
0.78
Rupture
implant
2
0.78
expander
1
0.39
not stated
1
0.39
Restriction of movement
49
19.07
Implant removal on demand
5
1.95
Exchange of implant
Total
11
4.28
10
3.89
1
0.39
The factors influencing implantation rotation or loss were analyzed. Patients with higher BMI were more prone to implant rotation and implant loss, as shown in [Fig. 1 ]. Meshes protected against implant rotation (p = 0.034). Other examined factors did not influence implant rotation or loss.
Fig. 1 Implant loss correlated to BMI.
Patient-reported outcome (PRO)
Questionnaires investigated mid- and long-term complications and patient satisfaction with the surgical outcome. A total of 185 out of 257 patients (71.98 %) responded to the questionnaires. Eighty-eight patients (47.57 %) reported a better body image after mastectomy and reconstruction; 35 (18.92 %) reported no change, and only 29 (15.68 %) reported a deterioration in body image. The patientʼs partner described the surgical outcome as improved in 26 cases (14.05 %), as no-change in 103 cases (55.7 %), and as deteriorated in 14 cases (7.57 %).
Around 73 % of women (122/167) reported no pain or almost no pain (score 1–4) on a pain scale from 1 (excellent, no pain) to 10 (worst pain).
The aesthetic result after breast reconstruction was rated by the women using a scale from 1 (excellent) to 6 (failure). One hundred and three patients (74 %) expressed their satisfaction with the surgical result, giving it a rating of 1 (excellent), 2 (good), or 3 (satisfactory). [Fig. 2 ] shows the distribution of ratings.
Fig. 2 Patient satisfaction with the aesthetic result.
Analysis of patient-reported outcome
A good aesthetic result was associated with a better self-image (p = 0.0001), and this also correlated with partner assessment (p = 0.0001) ([Fig. 3 ]). A good aesthetic result was also associated with less postoperative pain. Similarly, a positive self-image after reconstruction also correlated with a lower pain rating (p < 0.001). Patients with a higher BMI reported a worse patient-reported aesthetic outcome (p = 0.004), as shown in [Fig. 4 ]. Additional correlations between age or BMI and PRO parameters are shown in [Fig. 5 ]. An overview of other correlations is shown in [Table 4 ].
Fig. 3 Impact of self-image on partner interaction.
Fig. 4 Correlation between patient satisfaction, aesthetic result and BMI (1 = very good, 6 = very bad).
Fig. 5 Exploratory analysis of various influencing factors.
Table 4 Significant correlations between variables in implant-based breast reconstruction.
Variable 1
Variable 2
p-value
Mesh
Implant dislocation (lower)
0.034
Corial flap
Self-image (better)
0.002
Aesthetic result (worse)
BMI (higher)
0.004
Aesthetic result (better/worse)
Age (higher)
0.054
Self-image (better)
Interaction with partner (better)
0.0001
Aesthetic result (better)
Pain (less)
0.038
Aesthetic result (better)
Self-image (better)
0.0001
Interaction with partner (better)
Aesthetic result (better)
0.002
BMI
Implant loss (higher rate)
0.005
BMI
Implant dislocation (higher rate)
0.014
Adherence to decision
Finally, patients were asked whether they considered reconstruction to have been the right choice and whether they would opt for the same surgery again. The majority of patients confirmed that reconstruction had been the right decision (145 patients [78.38 %] vs. 10 patients [5.41 %]). A total of 139 patients (75.14 %) would opt for the same type of surgery again vs. 15 patients (8.11 %) who would not.
Use of mesh protected against implant rotation. Self-image with mesh was better compared to corial flap.
Discussion
We explored various patient and surgery-related characteristics influencing the surgical outcome after breast reconstruction with implant. Obesity and smoking > 10 cigarettes per day had a negative impact on implant reconstruction results. A higher BMI (> 30) made patients more prone to implant loss and dislocation, and aesthetic results were rated as less favorable in obese patients. Autologous reconstruction may be an option offering a higher patient satisfaction for this group.
Age in itself did not seem to be a contraindication for implant breast reconstruction; it did not have a negative impact on aesthetic outcome or patient satisfaction. It should be noted that the oldest patients with breast reconstruction in our study were 79 years old and they were highly satisfied with the aesthetic result.
We also found evidence that a good aesthetic result is associated with less pain after surgery and that it also corresponded with a good partner interaction. The use of corial flap or titanized mesh had a positive impact on self-image, and titanized polypropylene meshes prevented implant loss.
Use of the minimal touch technique for implant insertion and antibiotic prophylaxis as long as drainage remains in place resulted in a low infection rate of 6 % in our cohort as well as a low capsular contraction rate of < 10 % and minimal implant loss of < 2 % at 3.1 yearsʼ follow-up.
This study shows which patients have a higher risk of an unfavorable outcome after implant reconstruction and which surgical devices could reduce implant complications. Our findings could be used in pre-surgical counselling of patients desiring breast reconstruction after mastectomy.
Breast reconstruction with an expander and implant is currently being investigated in a number of studies, especially with regard to sequelae such as capsular fibrosis. In a review of 49 publications, each with a minimum follow-up time of one year, Kronowitz et al. [10 ], reported that capsular fibrosis occurs in more than 40 % of patients. Whitfield et al. [11 ], Cowen et al. [12 ] and Piroth et al. [13 ] detected significant differences in the frequency of capsular fibrosis between patients with and those without radiotherapy. A review by Lam et al. [14 ] confirmed a higher implant failure rate in patients undergoing radiotherapy. In our study, the rate of capsular fibrosis was low, at 6.25 % after a median follow-up of 3.1 years. This differs from the results reported in other studies. Late sequelae such as implant failure in patients with radiotherapy did not occur in our study. Baschnagel et al. [15 ], demonstrated that even with radiotherapy, implant-based breast reconstruction is feasible with a good cosmetic result and an acceptable rate of implant loss/failure (9.7 % after 1 year, 19.3 % after 2 years and 25.5 % after 3 years).
The correlation between infection and fever and the duration of antibiotic therapy was also analyzed. Infection occurred in 16 cases (6.23 %), a rate that is markedly lower than the rate reported by Washer et al. [16 ]. Patients who received antibiotic prophylaxis for more than two weeks did not suffer from any infection. This demonstrates the beneficial effect of antibiotic prophylaxis in implant reconstruction.
No significant differences in surgical outcome associated with the length of antibiotic prophylaxis were detected in the first two weeks after surgery. A retrospective study by Weichmann et al. [17 ] recommended the use of oral fluorquinolones as primary prophylaxis, based on their finding of an infection rate of 4.8 % after implant reconstruction which necessitated removal of the implant.
The infection rate in our cohort was similarly low at 5.08 % (n = 3) in the group of patients with drainages remaining in place for 1–7 days, 3.18 % (n = 3) in patients where drainage remained in place for 8–14 days, 8.75 % in patients where drainage remained for 15–21 days, and 50 % in patients where drainage remained in place for more than 21 days (n = 2).
Seroma occurred in 5.6 % (13) of patients in all three groups (1–7 days, 8–14 days, 15–21 days) without significant differences between groups. However, no seroma occurred in the group where drainage remained in place for > 21 days. Spear et al. [18 ], explored the occurrence of late seroma (1 to 4.7 years) after augmentation and implant reconstruction in 25 patients with 28 operations. They reported that seroma were idiopathic, without clear evidence of infection or malignancy in the majority of cases. Drainages had been placed before seroma occurrence in all cases. Complete capsulectomy was performed in 61 % of cases, and implant removal or a combination of capsulectomy procedure and implant removal was done in 64 % of cases. The median follow-up in our study was 41 months (3.4 years) after implant reconstruction – one year longer than in Spearsʼ cohort. In our cohort, implant removal was only required in one woman with seroma (due to capsular fibrosis).
Patient satisfaction was an important issue in our analysis. Case cohort studies often compare heterogeneous groups with autologous or heterologous reconstruction. In the study by Hu et al. [19 ], patient satisfaction in the first 5 years after the procedure did not differ between TRAM flap and expander-implant reconstruction. However, as Bodin et al. [20 ] demonstrated, patient satisfaction subsequently decreases in later years in the group with heterologous reconstruction. In our study, patient satisfaction reached its maximum level in the first year after surgery (30.77 % rated their satisfaction with the outcome after surgery as “very good” and 23.08 % as “good”) and decreased in subsequent years. Yueh et al. [21 ] reported an overall satisfaction with expander-implant reconstruction of 56.5 %. If we group together the ratings “very good”, “good” and “satisfactory” achieved in our cohort as overall satisfaction, then our results are even higher at 61.08 %. Koslow et al. [22 ] reported that contralateral prophylactic mastectomy combined with immediate reconstruction also had a positive impact on patient satisfaction.
Fischer et al. [23 ] examined risk factors and found them to be significantly associated with higher BMI (p < 0.0001), higher age (p < 0.001), obesity (p < 0.001), smoking (p < 0.0001), and hypertension (p < 0.001). These findings were confirmed in multivariate regression analysis (age > 55 years: OR = 2.0, p = 0.004; obesity [BMI ≥ 30 kg/m2 ]: OR 1.7, p = 0.03; smoking: OR = 4.0; p < 0.004). In our cohort, BMI was confirmed as an independent risk factor for implant loss (BMI > 30 kg/m2 ; p = 0.005). Impaired wound healing and scar insufficiency only occurred in the group of patients smoking > 10 cigarettes; however, this finding did not reach statistical significance (p = 0.069).
Ho et al. [24 ] investigated the long-term outcomes for patients with expander-implant reconstruction and post-mastectomy radiotherapy. Between 1996 and 2006, 1639 patients underwent modified radical mastectomy, 751 of whom had immediate expander placement. A total of 151 of these patients received chemotherapy, with replacement of the expander by a permanent implant and radiotherapy. After a median follow-up of 86 months, 21 implants had been exchanged (17.1 %) and 17 implants had been removed (13.3 %). Reasons for implant removal were infection, implant failure (rupture, leak), patient request or multifactorial [24 ].
In our case cohort study, the rate of implant exchange was comparably low at 4.28 % (n = 11). Implant removal was performed in 10 patients (3.90 %). Reasons for implant removal included patient request in 5 cases, infection in 2 cases, in-breast recurrence or development of another primary breast cancer in another 2 cases and capsular fibrosis in one case. Implant rupture was recorded separately in our study and only occurred in two cases (0.78 %), which was lower compared to the rate in Hoʼs study. Our rate refers to the whole cohort whereas in Hoʼs study, it refers only to post-radiotherapy patients. None of the 35 patients with implant reconstruction and radiotherapy in our study experienced implant loss. A recent study by Jagsi et al. [25 ] on US trends in breast reconstruction for women undergoing mastectomy showed that autologous techniques are used more often in patients who have both reconstruction and radiotherapy (OR 1.8; p < 0.001). However, our data suggests that a combination of radiotherapy and implant reconstruction is feasible, provided that patients are aware of potential sequelae.
Conclusion
This case cohort study of patients with expander-implant reconstruction demonstrates the safety of heterologous reconstructions, with low rates of complications and high levels of patient satisfaction. The majority of patients expressed their satisfaction with the surgical result in study questionnaires and would opt for the same operation again, demonstrating a strong adherence to their decision.
Patient characteristics such as high BMI – but not smoking less than 10 cigarettes/day or higher age – had a negative impact on patientsʼ self-image after implant reconstruction and led to higher rates of implant loss. Patients with a high consumption of cigarettes and/or obesity should be aware of the risk of an unfavorable aesthetic outcome after implant reconstruction.
The use of polypropylene mesh protected against implant displacement and the prophylactic use of antibiotics was confirmed as beneficial to avoid implant pouch infection. A good aesthetic result was found to be correlated with lower postoperative pain and a better self-image and body image. The aesthetic result did not vary with age. With a high degree of satisfaction reported across all age groups, elderly patients can also be encouraged to undergo implant reconstruction as a shorter and less time-consuming procedure.
Acknowledgements
We would like to thank André Scherag, Professor of Clinical Epidemiology, University of Jena, and Hildegard Lax, Institute of Medical Informatics, Biometry and Epidemiology, University of Duisburg-Essen, for their contribution to the statistical analysis.
