Improved patient-reported outcomes after autologous fat transplantation and corrective surgery after breast surgery

Abstract

Background: Autologous fat transplantation (AFT) is being increasingly used to improve the results after breast-conserving surgery and breast reconstruction. However, studies on patient-reported outcomes (PROs) and health-related quality of life (HRQoL) after AFT are scarce. The aim of this prospective longitudinal case-series study was to assess PRO in women who had undergone AFT after surgery for breast cancer or risk-reducing mastectomy.

Methods: Fifty women, who had undergone breast-conserving surgery or breast reconstruction, needing corrective surgery, were consecutively included between 2008 and 2013. A 20-item study-specific questionnaire (SSQ) and the Short Form Health Survey (SF-36) were used pre-operatively and 6 months, 1 year and 2 years post-operatively, to evaluate PRO and HRQoL.

Results: The patients underwent three (1–4) AFT procedures, with the injection of 164 ml (median) (range 40–516) fat. Thirty-eight and 34 patients completed the study-specific questionnaire and the SF-36, respectively, both pre-operatively and after 2 years. Sixteen of the 20 items in the SSQ were improved after 2 years, including breast size (p < 0.0001), shape (p < 0.0001), appearance (p < 0.0001), softness of the breast (p = 0.001), pain in the region (p = 0.005), scarring from previous breast surgery (p < 0.001) and willingness to participate in public physical activities (p < 0.001). HRQoL did not largely differ before and after AFT, or between the study group and a reference population.

Conclusions: AFT alone or in combination with other corrective surgical procedures, improved PRO after breast-conserving surgery and breast reconstruction in both irradiated and non-irradiated women.

Keywords:

• Quality of life
• breast reconstruction
• breast surgery
• autologous fat transplantation
• lipofilling
• fat grafting
• patient-reported outcome

Introduction

Breast reconstruction in connection with, or after, breast cancer surgery is well known to improve quality of life [1–3]. The surgical methods employed range from an oncoplastic approach in breast cancer surgery to extensive free-flap reconstructions. All have in common the possible need for surgical correction due to complications or less favourable aesthetic outcome. Patients who have received radiotherapy of the breast or the chest wall have a higher frequency of complications and reoperation [4]. Common indications for reoperation due to unfavourable aesthetic results are indurations, scar contractures, tissue deficit and breast asymmetry. Autologous fat transplantation (AFT) has become a well-established technique for improving these conditions [5]. Clinical results and surgeon’s opinions about the outcome after AFT are generally reported as good [6–9]. The results of some studies indicate that surgeon-reported outcome and patient-reported outcome may diverge after breast reconstruction [10–13]. The results of these evaluations can be valuable for patients facing quality-of-life-improving surgery in their decision-making process. There is also an increasing interest in patient-reported outcome measures (PROMs) for the evaluation of the efficacy of health care [14–16].

In most studies on patients’ opinions of the outcome of reconstructive surgery, only a single question has been asked [8,9,17]. We have previously reported patients’ satisfaction after AFT in a questionnaire-based, descriptive cross-sectional study [18]. The aim of the present study was to assess patient-reported outcome after AFT, prospectively and longitudinally, as part of reconstructive breast surgery. We also wanted to assess the patients’ health-related quality of life (HRQoL) before and after surgery and learn if it is consistent with HRQoL of a Swedish reference population.

Materials and methods

Patients and design

The study design is a prospective case series including patients treated with AFT as part of corrective surgery at the Department of Reconstructive Plastic Surgery at Karolinska University Hospital in Stockholm, Sweden. The patients were consecutively enrolled from October 2008 until June 2013. All procedures were performed under the national health service. Those included were women who had undergone breast-conserving surgery with less favourable aesthetic results, and women who had undergone mastectomy and subsequent breast reconstruction, needing corrective surgery. The inclusion criteria were time after previous breast surgery or radiotherapy longer than 1 year and mammography and ultrasound examinations performed a maximum of 3 months prior to AFT, to exclude any local recurrences. Exclusion criteria were known breast cancer recurrence, contraindications for anaesthesia or surgery due to comorbidities, other serious systemic diseases that may affect the outcome of the procedure and inability to understand verbal and written information about the study. The patients’ characteristics are given in Table 1. The study was approved by the Regional Ethics Review Board in Stockholm (2008/484–31/2).

Table 1. Patient characteristics.

Age at baseline (y)	52.7* (31.1–71.3)
BMI at baseline (kg/m^2)	24.7* (17–32.4)
BMI at follow-up (kg/m^2)	25.3* (18.8–34.3)
Previous breast surgery (n)	48
Breast-conserving surgery	25
Mastectomy + Immediate reconstruction with prosthesis	10
Mastectomy + Delayed reconstruction with prosthesis	2
Mastectomy + DIEP	1
Mastectomy + LD	7
Prophylactic mastectomy + Immediate reconstruction with prosthesis	2
Prophylactic mastectomy + Immediate reconstruction with prosthesis^†	1
Radiotherapy (n)	38
Time from radiotherapy to AFT (y)	3.6* (1.1–22.1)

n = 48, DIEP: deep inferior epigastric perforator flap; LD: latissimus dorsi flap; *Median; †Previous cancer surgery.

Autologous fat transplantation

Fat aspiration was performed with a dry technique using a Coleman cannula and a 10-ml syringe at the donor site [19]. Donor sites were the abdomen, knees, thighs, hips, flank area, in one case arms, and in another the case sub-axillary region. The fat was centrifuged at 3000 rpm for 3 min. Liquid fat and blood were poured off, and the remaining lipocytes were injected into the subcutaneous tissues at the recipient site in multiple layers and directions using a blunt cannula, 1.29 mm in diameter. Patients with implants received one dose of prophylactic antibiotics (cloxacillin provided they were not allergic, otherwise clindamycin). The patients underwent a median of three (1–4) AFT procedures unilaterally or bilaterally until satisfactory results were obtained. Other corrective procedures were performed concurrently in eleven patients: seven scheduled nipple reconstructions, one mastopexy, two scar corrections and one capsolutomy with implant change, abdominal advancement and liposuction of the axilla. The median time from the first AFT to last AFT was 12.3 months (range 4.2–27.5). All operations were performed under general anaesthesia and carried out by one of the authors.

Questionnaires and data collection

Demographic data were collected from electronic patient medical files that cover the whole catchment area. Hence, additional operations performed outside the unit and any complications treated at other hospitals would come to our knowledge. A study-specific questionnaire (SSQ) modified for this specific study population was used. The original questionnaire that SSQ was based on was developed by a psychologist [20], who specialised in oncological patients, in collaboration with experienced plastic surgeons. It has been consequently modified for different studies focusing on other breast reconstructive methods and their outcome[21,22], This SSQ contains 16 items in which answers are given on a seven-point Likert scale (Q1–Q4 & Q9–Q20), and four items (Q5–Q8) where the responses are ‘Not at all’, ‘Little’, ‘Pretty much’ or ‘Much’. The BREAST-Q questionnaire was not available at the start of this study.

The Short Form Health Survey [23] (SF-36) measures the patients’ long-term (4 weeks) HRQoL. It covers eight domains: physical functioning, physical role functioning, bodily pain, general health, vitality, social functioning, emotional role functioning and mental health in 36 items. The mean scores for each domain were transformed into a value on a 0–100 scale, where a higher number represents higher functioning. The eight domains of the SF-36 were used to compare the patients’ HRQoL with that of a Swedish reference population, change over time within the group. The outcome of the reference population is provided in the Swedish translation of SF-36 and based on answers from 3260 randomly picked women aged 30–74 in seven Swedish regions during 1991–92.

The questionnaires were answered at baseline (before AFT) and 6 months, 1 year and 2 years after the final AFT, either by mail or at the outpatient clinic. Median follow-up (time from final AFT procedure to last completed questionnaire) was 24.2 months (range 11.8–32.3). The completed questionnaires were handled by the first author and not by the surgeon who performed the operations.

Statistical methods

The items on an interval scale in the SSQ (Q1–Q4 & Q9–Q20) were visualized using descriptive statistics and mean line graphs were plotted using the results from all SSQs from all patients (n = 48), to give an overview of the change over time for each question. The items were also analysed using a statistical model, Mixed-effect model repeated measures (MMRM), to assess the longitudinal data. This model incorporates the dependence of the measurements within a patient and allows missing data (i.e. SSQ not completed at 6 months and/or 1 year) to analyse only the answers from the patients who completed both the baseline and the 2-year SSQ (n = 38). The model generates an estimate of the mean change. The outcome variable was the change relative to the baseline, and the baseline value was included as a covariate in the model. Sensitivity analysis was performed for irradiated and non-irradiated patients. This was carried out to examine if AFT especially benefits the irradiated patients. It is not inconceivable since AFT is known to have a positive impact on irradiation-damaged tissue [24]. Moreover, sensitivity analysis was performed for patients (92%) that had not undergone major additional surgery. The limit for the confidence interval was 95% for the difference between each time point and the baseline.

In the ordinal items (Q5–Q8), a quasi-symmetry model [25] was used to estimate the chance of scoring one category more positive than one category more negative at follow-up. Since the responses were given on an ordinal scale, they cannot be treated as continuous variables in the analyses. The model acknowledges the ordinal scale and tests for an ordinal trend.

The level of significance was set to 5% (p ≤ 0.05) in all tests. The statistical analyses were performed using SAS^® (Version 9.4, SAS Institute Inc., Cary, NC, USA). No imputations were performed and no last observation carried forward was used.

No statistical calculations were performed for SF-36, since the number of patients only allow for detection of 10–20 point differences both over time within the group and between the patients and the reference population. Mean scores for the general Swedish population are given by sex and age categories in the Swedish translation of SF-36. The scores for women in the nine age categories between 30 and 74 years were used to calculate a mean score for all the women aged 30–74 (n = 3260). The means were weighted with respect of number of patients in each age category. In that way, the reference population and patients are matched on sex and age. Confidence interval for the computed mean for the reference population was not calculated.

Results

Patients

Fifty patients were recruited to the study. One patient died and one dropped out before completing any questionnaires. Among the remaining 48 patients, eight dropped out during the study period, but their completed questionnaires were analysed. One dropped out because of breast cancer recurrence, one underwent bilateral breast augmentation at a private clinic, one bilateral risk-reducing mastectomy and five patients dropped out for other reasons (Figure 1). Two other patients had breast cancer recurrence, but they did not drop out of the study.

Figure 1. Flow diagram of patients’ participation in the study. Boxes on the left show the number of patients completing the study-specific questionnaire (SSQ) at each time point. Boxes in the middle show the number of patients participating in the study. Boxes on the right show the number of patients excluded and drop-outs.

Adverse events occurred in six patients: one patient had a wound rupture, one a haematoma at the donor site and four patients needed further investigations; three because of palpable lumps and one because of discomfort and swelling of the breast. Thirty-eight patients completed the SSQ and 34 patients completed the SF-36 both at baseline and after 2 years.

The total median volume of injected fat was 164 ml (40–516).

Study-specific questionnaire

Q1–Q4 and Q9–Q20

From the mean line graphs, it can be seen that the best results were obtained after 6 months, followed by a decrease after 1 year, and then an increase after 2 years (Figure 2). Fourteen of these 16 items (Q1, Q4 and Q9–Q20) were improved after 2 years. The mean score and mean change between baseline and 2 years are given for each item in Table 2 and Figure 2.

Figure 2. Mean line plots for Q1–Q4 and Q9–Q20. The mean line plots for each item at each time point are shown for all patients who completed the SSQ.

Table 2. Results for Q1–Q4 and Q9–Q20 on sensitivity, pain, appearance, softness, trouble finding a good-fitting bra and willingness to participate in physical activities.

		1     …    7		BL	2 years	Min	Max	SD
Q1	How is the sensation in the nipple area?	Normal	Reduced	4.10	4.55	−3.0	5.0	1.5
Q2	How is the sensation in the breast skin?	Normal	Reduced	4.16	3.95	−4.0	3.0	1.6
Q3	How is your ability to feel sexual sensation in the breasts?	Normal	Reduced	4.91	4.70	−5.0	5.0	1.8
Q4	Do you have pain in the breasts/breast region?	Every day	Never	5.11	5.53	−3.0	4.0	1.7
Q9	What do you think about the size of your breasts?	Too small	Too big	2.45	3.21	−2.0	4.0	1.4
Q10	Are the breasts of equal size?	Very dissimilar	Very similar	2.71	3.60	−3.0	5.0	2.0
Q11	What do you think of the shape of your breasts?	Very ugly	Very nice	2.70	4.00	−3.0	5.0	2.0
Q12	Are the breasts of equal shape?	Very dissimilar	Very similar	2.25	3.70	−2.0	5.0	1.7
Q13	What do you think about the softness of your breasts?	Very hard	Very soft	3.55	4.28	−2.0	4.0	1.5
Q14	Are you satisfied with the appearance of your breasts?	Not at all	Very satisfied	2.37	4.15	−1.0	6.0	1.6
Q15	What do you think about the appearance of the breast when you wear clothes?	Very ugly	Very nice	4.30	5.63	−3.0	5.0	1.5
Q16	What do you think of the breast's appearance when you are wearing swimwear / a bra?	Very ugly	Very nice	3.25	5.20	−1.0	6.0	1.6
Q17	What do you think of the breast's appearance when you are naked?	Very ugly	Very nice	2.31	3.72	−2.0	5.0	1.7
Q18	What do you think of the scars from previous cancer surgery?	Very ugly	Very nice	2.98	3.85	−3.0	5.0	1.8
Q19	Do you have trouble finding a bra that fits your breasts?	Very hard	Not hard at all	3.96	5.11	−4.0	6.0	2.2
Q20	Do your breasts affect your willingness to swim in public environments (swimming pool, beach, etc.) or participate in gymnastics and similar activities?	Negative	Positive	3.39	5.03	−2.0	5.0	1.8

The questions were answered on a 7-point Likert scale. Mean score at baseline (BL) and a 2-year follow-up are shown. Min: minimum change in score between baseline and 2 years. Max: maximum change in score between baseline and 2 years. Negative numbers indicate lower score at the 2-year follow-up compared to baseline. SD: Standard deviation in the change between baseline and 2 years.

In the MMRM analysis, including only patients who completed both baseline and the 2-year questionnaire, thirteen of the 16 items were significantly improved after 2 years, for the entire group and in the sensitivity analyses for the irradiated patients and the patients who had not undergone additional surgery (Q4 and Q9–Q20). Q13 (softness of the breast) was not significantly improved after 1 year (Table 3). The non-significant items (Q1–Q3) concern the sensitivity of the breasts. The non-irradiated subgroup was too small to make any inferences from the data.

Table 3. Q1–Q4 and Q9–Q20.

	1 year follow-up				2 years follow-up				Radiotherapy 2 years follow-up				No additional surgery 2 years follow-up
	ECh	SE	p	CI	ECh	SE	p	CI	ECh	SE	p	CI	ECh	SE	p	CI
Q1	0.3	0.25	0.190	−0.17–0,83	0.3	0.23	0.160	−0.13–0.80	0.4	0.27	0.190	−0.18–0.89	0.4	0.24	0.152	−0.13–0.84
Q2	−0.2	0.24	0.405	−0.68–0,28	−0.2	0.22	0.295	−0.68–0.21	−0.4	0.23	0.077	−0.86–0.04	−0.3	0.23	0.269	−0.71–0.20
Q3	0.2	0.29	0.479	−0.37–0,78	−0.2	0.26	0.451	−0.71–0.32	0.0	0.29	0.999	−0.57–0.57	−0.1	0.27	0.621	−0.66–0.40
Q4	0.5	0.22	0.021	0.08–0,97	0.6	0.22	0.005	0.19–1.05	0.6	0.24	0.014	0.13–1.09	0.6	0.23	0.015	0.12–1.02
Q9	0.5	0.17	0.003	0.18–0,85	0.7	0.16	<0.001	0.34–0.98	0.7	0.18	0.001	0.29–1.01	0.8	0.17	<0.001	0.41–1.10
Q10	1.0	0.26	<0.001	0.53–1,56	0.8	0.25	0.002	0.29–1.27	0.8	0.26	0.005	0.23–1.28	0.9	0.26	0.001	0.37–1.40
Q11	0.9	0.24	<0.001	0.43–1,40	1.1	0.23	<0.001	0.66–1.59	1.2	0.26	<0.001	0.73–1.76	1.0	0.25	<0.001	0.52–1.51
Q12	1.0	0.26	<0.001	0.48–1,52	1.3	0.25	<0.001	0.77–1.76	1.4	0.27	<0.001	0.82–1.92	1.3	0.27	<0.001	0.80–1.86
Q13	0.4	0.24	0.114	−0.09–0,86	0.8	0.23	0.001	0.32–1.22	0.9	0.24	<0.001	0.45–1.40	0.8	0.24	0.001	0.31–1.26
Q14	1.3	0.22	<0.001	0.87–1,77	1.7	0.22	<0.001	1.24–2.10	1.6	0.23	<0.001	1.14–2.07	1.6	0.20	<0.001	1.15–1.97
Q15	1.5	0.21	<0.001	1.10–1,93	1.3	0.20	<0.001	0.95–1.75	1.5	0.19	<0.001	1.14–1.91	1.2	0.20	<0.001	0.82–1.62
Q16	1.9	0.23	<0.001	1.40–2,32	2.0	0.22	<0.001	1.57–2.45	2.0	0.22	<0.001	1.55–2.45	2.0	0.23	<0.001	1.53–2.43
Q17	1.5	0.23	<0.001	1.04–1,95	1.4	0.22	<0.001	0.94–1.81	1.4	0.24	<0.001	0.96–1.91	1.3	0.22	<0.001	0.84–1.74
Q18	0.9	0.23	<0.001	0.46–1,39	1.0	0.23	<0.001	0.54–1.45	1.1	0.25	<0.001	0.63–1.62	0.9	0.23	<0.001	0.46–1.38
Q19	1.2	0.28	<0.001	0.59–1,72	1.1	0.27	<0.001	0.59–1.68	1.3	0.29	<0.001	0.77–1.92	1.0	0.28	0.001	0.45–1.59
Q20	1.3	0.27	<0.001	0.74–1,81	1.5	0.26	<0.001	1.03–2.07	1.5	0.28	<0.001	0.97–2.09	1.5	0.27	<0.001	0.95–2.03

Statistical analysis of questions in table 2 with MMRM for patients who completed the study-specific questionnaire at baseline and after 2 years (n = 38). Estimated mean change (ECh) from baseline to follow-up. Significant level was set on p ≤ 0.05 with a confidence interval (CI) on 95%. Significant items are bold. SE: Standard error. Radiotherapy 2 years follow-up = irradiated patients only. No additional surgery 2 years follow-up = patients that had not undergone major additional ipsilateral surgical procedures.

Q5–Q8

The ordinal trend, i.e., the probability of more positive scores after 2 years, was 22.47 (p = 0.002) in Q6 (sexually unattractive because of breasts), 7.9 (p = 0.003) in Q7 (seeing oneself naked) and 10.88 (p = 0.017) in Q8 (being naked in front of partner) (Figure 3). The 95% CI was wide in all three items. The ordinal trend for Q5 (importance of breasts for sexuality) was not significant.

Figure 3. Q5–Q8 focusing on sexuality and willingness to be naked. The graphs show the percentage of answers in each category, for each item, at baseline and after 2 years. The ordinal trend is the chance of scoring one category more positive than one category more negative at follow-up. Significant items are given in boldface. CI: confidence interval.

SF-36

The mean scores for the eight domains for the reference population were compared with the mean score for the study group at both baseline and after 2 years. No point differences were large enough to be able to detect any significant differences in HRQoL between the study group and the reference population. Neither were there any large differences between the mean score at baseline and at the 2-year follow-up within the study group. The descriptive statistics are shown in Figure 4.

Figure 4. Results for the eight domains of SF-36 of the study population and the age matched reference population of Swedish women (n = 3260). PF: physical functioning; RP: physical role functioning; BP: bodily pain; GH: general health; VT: vitality; SF: social role functioning; RE: emotional role functioning; MH: mental health.

Discussion

Patient-reported outcome measures are important tools in evidence-based medicine [26,27]. The results are especially important when evaluating surgery intended to improve quality of life, rather than treating severe health conditions. A debate on oncological safety has been going on over the last three decades. It has been proposed that stem cells injected in a previous cancer site could induce recurrence [28]. Consequently, women offered AFT as part of breast reconstruction or after breast-conserving surgery should be given evidence-based information about the outcome of the procedure before accepting further surgery. Information on the outcome of different procedures is also important for health-care providers and funding bodies. Despite the common use of AFT, few reports have been published on the change of patient-reported outcome over time and the quality of the reports is rather low [26]. When this study started, the validated BREAST-Q form was not yet available, and could therefore not be used as recently recommended in the VOGUE study [27]. Instead, a study-specific questionnaire was used to evaluate patient-reported outcome. The patients answered the questionnaire before AFT and at three-time points during the 2-year follow-up after completed AFT sessions. A research nurse handled the questionnaires. The aim was to assess possible changes in the patients’ opinion about the outcome over time. Resorption of the transplanted fat may lead to deterioration of the aesthetic result over time. On the other hand, as time passes, the patients may accept the cosmetic result to a larger extent and softening of the breast tissue [29] could lead to patients expressing a more positive opinion over time.

In the present study, a more positive outcome was reported in 80% of the items 2 years after AFT than pre-operatively. The patients were more satisfied with the appearance of their breast with and without clothes, found the breast symmetry better, were more comfortable about being naked with their partner and felt more sexually attractive. They were also more willing to participate in physical activities and found it easier to find a good-fitting bra. The relatively long period between radiotherapy and AFT (median >3.5 years, extending to over 22 years) indicates that AFT could the main reason for improved scoring post-operatively. If that period had been shorter, the initial negative impact of radiotherapy would decrease during the study period, making time the main reason why the patients have higher scores after 2 years. Eleven of the patients had other ipsilateral surgical procedures performed. Seven of these were elective nipple operations leaving four patients who underwent surgery that could have positive or negative impact on the outcome. Only three of them were included in the MMRM analysis. Since these three patients had to undergo additional surgery one can speculate if they have scored more negatively. On the other hand, the additional surgery may have generated even better outcome. The patients had increased in BMI from 24.7 to 25.3 kg/m² in median at the last follow-up. It is less likely that the small increase has an impact on the outcome.

Various aspects of AFT have been studied, but the number of studies using PROMs is modest. Most publications report only a one-item evaluation, asking the patient if she is satisfied with the outcome, often using a 3- to 5-point Likert scale [8,9,17,30]. In one study, 200 patients were asked personally about their degree of satisfaction with the surgical outcome at their post-operative follow-up. All of them claimed that they were satisfied or very satisfied, which may imply a significant systematic bias as the evaluation was not anonymous [31]. In the present study, care was taken to prevent the surgeons involved from handling the post-operative questionnaires, so as not to influence the outcome measures.

In the only multi-centre cohort study, on this subject, Bennett et al. [32] reported that AFT patients and ‘control patients’ scored similarly at baseline regarding physical well-being, satisfaction with the breast, sexual well-being and psychosocial well-being. Moreover, the control patients scored a little higher than the AFT patients (cases) at the 1-year follow-up. No difference between the two groups was seen after 2 years. However, almost 20% of the control patients had undergone AFT before baseline, which may lead to problems in the interpretation of the results. Moreover, no comparison was made of pre- and post-operative results. Nevertheless, the authors concluded that AFT provided measurable improvements in satisfaction, which is in line with our results.

In the present study, the nine items that can be regarded as indicators of aesthetic outcome and symmetry were all improved. The patients’ opinion of the aesthetic outcome could, therefore, be considered improved, which was one of the main reasons for performing AFT (Figure 5 and 6). It is difficult to assess the impact of breast surgery on sexuality, but this is important for the patient’s future well-being. Many breast cancer survivors suffer from impaired sexuality due to a change in body image [33]. The results of this study indicate an improvement over time after AFT regarding feeling sexual attractive and comfortable when being naked. Whether this is due to AFT or the passage of time after cancer surgery and reconstruction is unclear. Our findings are in contrast to those in the study by Bennett et al., which showed a decrease in sexual well-being over time [32]. However, we believe that the use of AFT as a complement to other breast surgical techniques may play a role in regaining a positive body image after surgery.

Figure 5. A 31-year-old women that presented with a scar retraction and volume deficiency 3.6 years after breast conserving surgery and radiotherapy (left). She underwent three AFT procedures with totally 150 ml fat. 2.2 years post-operative the scar is not visible and the symmetry is better (right).

Figure 6. A 66-year-old woman with bilateral defects after BCS and RT (left). Time from RT to AFT was 2.6 years for the left breast and 1 year for the right breast. She underwent two AFT procedures to each breast with totally 69 ml to the left side and 100 ml to the right. In addition, she underwent mastopexy on the right side. Post-operative images from the 2-year follow-up are shown to the right.

Improvement in softness of the breast is most evident at 6 months and 2 years in the mean line plots (Q13). A non-significant change in softness was seen after 1 year, in contrast to 2 years, indicating a slightly harder breast after 1 year. This could be explained by a reduction in the initial filling effect over the first year due to resorption of the transplanted fat. The improvement 2 years post-operatively could be a biological effect in radiation-damaged tissue, rather than primarily the effect of the filler. AFT has been suggested as a treatment for radiation-induced fibrosis, as clinical findings and experimental studies indicate reduced fibrosis when AFT is carried out after radiation therapy [24,34–37].

The patients, in the present study, reported a low level of pain before AFT. Despite this, reduced pain in the breast region was reported at the 2-year assessment. This finding, as well as increased breast softness, is in line with the results of a case-control study carried out by Panettiere et al. [38] and implies that AFT not only improves the aesthetic outcome of a breast reconstruction but may also have a positive impact on morbidity after treatment for breast cancer.

No large difference was found in HRQoL between the study group and the reference population, either at baseline or after 2 years. There is, therefore, no reason to believe that the patients in the study group differ from an age-matched Swedish women in general when they decide to undergo AFT. Furthermore, it appears that AFT does not lead to a large decrease in HRQoL post-operatively.

The strengths of this study include the prospective design, with baseline data collected pre-operatively. However, the study has some limitations. First, the study group is relatively small. Furthermore, additional ipsilateral surgical procedures were performed at the same time in a minority of the patients, and this could have influenced their opinions concerning the result. However, the study group represents typical patients who undergo complementary AFT. Another limitation is that the heterogeneity of the patient group may have effect on the results. A sensitivity analysis was carried out and no difference in significance was seen comparing results from the entire patient group with the sub-group (92% of the patients) that had not undergone major additional ipsilateral surgery. No measurements of remaining fat volume were made post-operatively to compare with PRO.

A validated PROM would probably have made the evidence stronger but as mentioned before, no such was available at the start of the study. Lastly, several hypotheses tests were performed; hence, type I errors may exist.

In conclusion, the patient reported outcome improved up to 2 years after AFT, when the method was used as a complement in breast reconstruction or after breast-conserving surgery. Improved outcomes included better breast symmetry, size, shape and appearance. Autologous fat transplantation also increased the softness of the breast and reduced pain in the breast area. AFT did not affect HRQoL, and the study population did not differ from the reference population in this respect. The results of this study are of clinical importance for women considering AFT, to help them to make an informed decision. Nevertheless, further studies using PROMs after AFT and breast reconstructive surgery would be valuable to confirm our findings.

Disclosure statement

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Additional information

Funding

This work was supported by the Stockholm County Council (ALF project) under Grant number 522038 and 20170121; Percy Falks stiftelse and Capio Research Foundation.