Introduction

Over the past decades, the multidisciplinary treatment of breast cancer has changed, and improved 5-year survival of breast cancer up to 88% in the Netherlands, especially for early stage breast cancer [1,2,3]. The improved survival has led to more focus on health related quality of life (HRQoL) with emphasis on shared-decision making and aesthetic outcomes [4, 5].

In the Netherlands, approximately 40% of the patients underwent a mastectomy in the management of invasive breast cancer and ductal carcinoma in situ (DCIS) [6]. According to the Dutch breast cancer guidelines, every patient scheduled for a mastectomy must be informed about the possibility of breast reconstruction [7]. Nevertheless, some treatment and patient-specific contraindications may limit the possibilities for breast reconstruction [8]. Despite the known positive effects on body-image, self-esteem, physiological well-being, sexuality, cosmetic outcomes and quality of life [9,10,11,12,13], some women do not opt for (immediate or delayed) breast reconstruction [14, 15].

In recent years, the rate of immediate breast reconstruction in the Netherlands has increased. Although less documented, delayed reconstruction rates seem to stay behind [6]. A large variation in the rate of immediate post-mastectomy breast reconstruction is observed between Dutch hospitals, with rates ranging from 0 to 64% [12, 14, 16]. Tumor and patient characteristics could not fully explain these differences [6]. Some studies evaluated patients’ motives to opt for or reject breast reconstruction [14, 17,18,19,20,21]. However, often the focus is on tumor and patient characteristics as predictors of breast reconstruction, and not on patients’ preferences. Nevertheless, knowledge of patients’ preferences and decision-making is of utmost importance in the counseling of breast cancer patients in the outpatient clinic and tailoring treatment to patients’ needs. As one of the first, this study combines the Breast-Q questionnaire in combination with a short survey to gain insight into the motives of women to opt for or reject post-mastectomy breast reconstruction. Additionally, this study established the breast reconstruction rate in a large Dutch teaching hospital.

Methods

Study design

The present study was a retrospective, cross-sectional study of all consecutive patients who underwent mastectomy in the management of breast cancer between January 2015 and December 2017 at a large teaching hospital in the Netherlands. The study was approved by the hospital’s institutional review board.

Patient selection

Patients were included if they underwent mastectomy in the treatment of invasive breast cancer and/or DCIS. The study included patient with primary breast cancer as well as patient with a recurrence of breast cancer. Patient who received prophylactic mastectomy or mastectomy for another reason without a history of breast cancer were excluded. Males were also excluded from the study.

According to the Dutch breast cancer guideline, the management of patients was discussed in a multidisciplinary consultation. During this consultation, a radiologist, a medical oncologist, an oncologic surgeon, a plastic surgeon, a pathologist, and a radiation therapist were present. All patients were treated following the national guidelines for breast cancer [7].

Questionnaires were sent by mail to all patients of 18 years or older, who were mentally competent, and capable of understanding and responding in the Dutch language. All questionnaires were sent at the same time in the spring of 2019, regardless of the follow-up time after treatment. Evaluation of patient-reported outcomes measures (PROMs) was performed at one moment in time, resulting in different periods of follow-up after mastectomy per patient. When patients did not respond, a phone call was made followed by a postal reminder. Written informed consent was obtained from all participating patients.

Medical data

Patient-related variables were collected from electronic medical records of our institutional database. These data included patient demographics, details of oncologic treatment, and postoperative variables. Surgical complications were defined as breast cancer surgery-related events requiring intervention.

Post-mastectomy breast reconstruction can be performed in an immediate or delayed setting. Immediate breast reconstruction is defined as breast reconstruction during the same operation as the primary oncologic surgery. Delayed breast reconstruction is breast reconstruction at any time after the primary oncologic surgery. Patients were categorized into two groups based on the presence of breast reconstruction: (1) patients who did not undergo post-mastectomy breast reconstruction (non-reconstruction group), and (2) patients who did undergo immediate or delayed post-mastectomy reconstruction (reconstruction group).

Questionnaires

Patient-reported outcomes were assessed using two questionnaires. We used the Breast-Q, which is validated in breast cancer patients [22,23,24]. The Breast-Q is a PROM developed to evaluate outcomes among women undergoing different types of breast surgery. For every type of breast surgery (breast conserving surgery, mastectomy and breast reconstruction), a specific module is available. Each module covers two domains: (1) HRQoL, and (2) patient satisfaction. The domain QoL consists of subthemes psychosocial, physical, and sexual well-being, and the domain patient satisfaction consists of the subthemes satisfaction with breasts, satisfaction with outcome, and satisfaction with care. Questions are scored on a 4- or 5-point Likert scale. Raw scores can be transformed into scores ranging from 0 to 100, and higher scores indicate better QoL or higher satisfaction [22,23,24]. In the current study, the Breast-Q for breast cancer was used. The non-reconstruction group received the mastectomy module, and the reconstruction group received the mastectomy with reconstruction module.

No validated questionnaires were available that specifically address decision-making on post-mastectomy breast reconstruction. Previous studies showed different variables to be associated with post-mastectomy breast reconstruction (i.e. age, appearance, more surgical procedures, and fear of complications or recurrence) [14, 17,18,19, 21]. These variables were incorporated into a short survey to evaluate the impact of these factors on decision making (Fig. 1). Via the option ‘Other’, a free text field appeared and patients had the possibility to elaborate on their motive. The responses to the free text field of the short survey were categorized based on meaning of these responses. The categories were: ‘Choosing for the easiest solution or prosthesis is (un)practical’, ‘provided information by the doctor’, ‘appearance or confrontation with breast cancer’, ‘pressure or limited time to make a decision’, ‘type of available reconstruction’, ‘age’, ‘number of procedures’, ‘medical indication’, ‘trust in their own breasts’, ‘no necessity’, and doubt about the decision’. Medical indication meant presence of comorbidities or need of adjuvant (radio)therapy that make patients less ideal candidates for immediate breast reconstruction. No necessity meant that patients stated that they did not feel the need for breast reconstruction.

Fig. 1
figure 1

Short-list with questions. Patients received a digital link to complete the questionnaire. All questionnaires were anonymous. All patients received a Dutch questionnaire. All questions needed to have an answer to continue to the Breast-Q questionnaire and complete the survey. The questionnaire about decision-making was the same for the non-reconstruction group as the reconstruction group

Full size image

Statistical analyses

Patient-related variables and patient-reported outcomes were analyzed using descriptive statistics presented as means with standard deviations, or numbers with percentages. In univariable analyses, differences in patient-related variables and response outcomes between the non-reconstruction group and the reconstruction group and between respondents and non-respondents were assessed using Chi-square tests, Fisher exact tests and Mann–Whitney tests. Multivariable analyses were performed to adjust differences in response outcomes between the non-reconstruction group and reconstruction group for relevant patient characteristics. Outcomes of the short questionnaire were analyzed using multivariable logistic regression, and Breast-Q scores were analyzed using multiple linear regression. Potential confounders were patient characteristics (BMI, medical history), tumor characteristics (stage, nodal management), and treatment characteristics (adjuvant chemotherapy, radiotherapy). In both multivariable analyses, we adjusted for age, ASA classification, and adjuvant radiotherapy. Taking into account the available sample size, we selected these three variables as a consensus of known characteristics that often differed between reconstruction and non-reconstruction patients (based on existing literature [16, 21, 25, 26]) and significant differences found in our study population.

The Breast-Q scores were compared to normative values reported by Oemrawsingh, and Clarijs et al. [27] In this study Dutch women were invited to complete a web-based survey that was disseminated through social media platforms of the Erasmus Medical Center between January and July 2020. Data were collected in 9037 women with a mean age ± std of 44 ± 13 years. Patients who reported breast cancer in their medical history were excluded from analysis. These authors kindly provided us with the individual-level (raw) data of their study, which enabled a comparison of our Breast-Q scores with the normative values based on statistical tests for independent samples. The comparison of the Breast-Q scores with the normative values were performed separately for the reconstruction and the non-reconstruction group, using independent samples t-tests.

In case of missing data, complete case analyses were performed. Two sided p-values of p < 0.05 were considered statistically significant. All statistical analyses were performed using the statistical software IBM SPSS version 24.

Results

Patient characteristics

In total, 319 patients were identified who had undergone mastectomy in the management of breast cancer between January 2015 and December 2017, and were eligible for PROMs (Fig. 2).

Fig. 2
figure 2

Flowchart. Patient selection and flowchart of the responders to the patient-reported outcome measures

Full size image

Of these patients, 217 (68%) patients underwent mastectomy without post-mastectomy reconstruction and the reconstruction group consisted of 102 (32%) patients. Patients in the reconstruction group were significantly younger than those in the non-reconstruction group. Also, a significantly lower BMI was found in the reconstruction group, whereas comorbidities and higher ASA classification were found in the non-reconstruction group. Although no significant differences in surgical treatment and postoperative complications were found between the two groups, nodal management was found to be more extensive in the non-reconstruction group. Also, significant differences were found in adjuvant radiotherapy and chemotherapy (Table 1).

Table 1 Patient characteristics of 319 patients with mastectomy in the management of breast cancer
Full size table

Reconstructive characteristics

The reconstruction group consisted of 102 patients of whom the majority (n = 95) opted for immediate breast reconstruction. The majority of patients opted for an implant-based breast reconstruction (88%). In case of an implant-based breast reconstruction, anatomical shaped implants were most often used (78 patients). In case of an autologous breast reconstruction, a deep inferior epigastric perforator (DIEP) flap was used most often (n = 14) (see Table 2 for more details).

Table 2 Reconstructive details of 102 patients who opted for breast reconstruction
Full size table

Five patients lost the tissue expander due to postoperative complications of whom three did not receive any other form of breast reconstruction. In one case, loss of the tissue expander was caused by development of a large postoperative hematoma, and in the other four cases, it was due to infection.

Five patients who opted for immediate breast reconstruction required adjuvant radiotherapy. In two of these cases, the tissue expander was preliminarily removed and no further reconstruction was performed. In one case, the tissue expander was preliminarily removed and eventually delayed reconstruction was performed. In two other cases, the definitive implant was placed before the start of the radiation therapy.

Patient-reported outcomes

Of the 319 patients eligible for PROMs, 155 (49%) patients completed the questionnaire (Fig. 2). In the responder group, post-mastectomy breast reconstruction was significantly more often performed than in the non-responder group (p = 0.012). Responders were significantly younger than non-responders (p = 0.001), had a lower ASA classification (p = 0.015), and received more often adjuvant chemotherapy (p = 0.007). Also, fewer smokers were counted in the responder group (p = 0.006; data not shown).

The non-reconstruction group had 100 (46%) responders. One patient was excluded from PROM analysis because she received delayed breast reconstruction within six months after completion of the survey (no breast reconstruction module). The reconstruction group had 55 (53%) responders of whom two had a delayed breast reconstruction, and 53 had an immediate breast reconstruction. Between the responders of the non-reconstruction and reconstruction group, comparable significant differences in patient and treatment characteristics were found as in the entire study population (Table 3).

Table 3 Patient characteristics of 99 responders without breast and 55 responders with breast reconstruction
Full size table

In the non-reconstruction group, the majority of patients (n = 82; 83%) confirmed that they did not want a post-mastectomy breast reconstruction. Seventeen patients (17%) stated that they had a desire for breast reconstruction. These patients had tumor and/or patient characteristics that count as a contra-indication for immediate breast reconstruction: nine out of these seventeen patients had an indication for adjuvant radiotherapy, one patient had an indication for adjuvant radiotherapy and a high BMI (> 30), three patients had a high BMI, and three patients did smoke at the time of their oncologic surgery. Four patients received a tissue-expander during the primary operation but due to infection or adjuvant radiation therapy, the tissue-expander was removed. In three cases no other reconstructive surgery was performed. In the reconstruction group, the majority of patients (n = 53; 96%) confirmed that they had the desire for a breast reconstruction.

Forty (40%) patients of the non-reconstruction group stated that age had an impact on their decision not to opt for breast reconstruction. However, the majority of patients (60%) stated that age did not influence the decision about breast reconstruction, and the stated influence of age was not significantly different between the two groups. In both groups, the majority stated that they received sufficient information about breast reconstruction. Nineteen patients of the non-reconstruction group stated that they had not received sufficient information, and four of these patients did have a desire for breast reconstruction. Correction for confounders in the multivariable analysis did not considerably change the results (Fig. 3).

Fig. 3
figure 3

Patient-reported outcomes of 154 patients after mastectomy in the management of breast cancer. (A1) Outcomes of short-survey about decision on breast reconstruction. Every question of the survey is showed in the figure. The outcomes of the non-reconstruction group were compared with the outcomes of the reconstruction group (non-adjusted p-values and adjusted p-values for age, ASA classification, and adjuvant radiotherapy); OR = Odds ratio. (A2) Outcomes of the free text field responses of the short-survey in the reconstruction group and non-reconstruction group. Categorized factors that contributed to the decision on breast reconstruction: (1) decision based on easiest outcome (prosthesis is (un)practical); (2) provided information during consultation; (3) appearance or confrontation with breast cancer; (4) pressure or time limit to make their decision; (5) age of patient at the time of decision; (6) type of reconstruction; (7) more (surgical) procedures; (8) medical indication; (9) trust in their own breasts; (10) no necessity of breast reconstruction; (11) doubt about decision. (B) Outcomes of the Breast-Q questionnaire. The figure includes the four domains of the Breast-Q questionnaire: (1) satisfaction with breasts, (2) psychosocial well-being, (3) physical well-being, and (4) sexual well-being. The outcomes of the non-reconstruction group [(1) 60.5 ± 19.3, (2) 63.0 ± 16.0, (3) 72.4 ± 15.3, and (4) 51.0 ± 23.2)] were compared with the outcomes of the reconstruction group [(1) 63.4 ± 16.8, (2) 67.5 ± 14.4, (3) 70.4 ± 16.8, and (4) 55.3 ± 22.6], (non-adjusted p-values and adjusted p-values for age, ASA classification, and adjuvant radiotherapy); B = coefficient.

Full size image

In the non-reconstruction group, the option ‘other’ was used thirty-six times and resulted in 45 motives. These motives could be classified into eleven different categories. The categories ‘medical indication’, ‘no necessity’ and ‘type of reconstruction’ were mentioned most often (12, 6, and 5 times, respectively). Medical indication included the need for adjuvant radiation therapy, but also a higher BMI, the presence of comorbidities or smoking (for more details see Fig. 3b).

In the reconstruction group, the option ‘other’ was used ten times and resulted in fourteen motives. These motives could be classified into six different categories. Two patients selected the option ‘other’ but did not specify their motives in the free text field. The categories ‘appearance’, ‘easy/(un)practical’, and ‘provided information’ were mentioned most often (4, 4, and 3 times, respectively) (for more details see Fig. 3B).

The mean Breast-Q ‘psychosocial well-being’ score significantly differed between the non-reconstruction group (63.0 ± 19.3) and the reconstruction group (67.5 ± 14.4; p = 0.042). No significant differences were found between the two groups with respect to the other Breast-Q scores. Correction for confounders in the multivariable analysis did not considerably change the results (Fig. 3c).

The normative Dutch population consisted of 9037 women. The mean patient-reported outcomes ± standard deviation per Breast-Q domain were for ‘satisfaction with breasts’ 64.2 ± 18.6, ‘psychosocial well-being’ 72.0 ± 15.9, ‘physical well-being’ 89.6 ± 12.4, and’sexual well-being’ 60.4 ± 15.3 [27]. Compared to Dutch normative Breast-Q data, responders of the non-reconstruction group, on average reported poorer Breast-Q ‘satisfaction with breasts’ (60.5 ± 19.3; p = 0.025), ‘psychosocial well-being’ (63.0 ± 16.0; p < 0.001), ‘physical well-being’ (72.4 ± 15.3; p < 0.001), and ‘sexual well-being’ (51.0 ± 23.2; p < 0.001) scores [27]. Responders of the reconstruction group from the current study reported poorer average Breast-Q ‘psychosocial well-being’ (67.5 ± 14.4; p = 0.046), and ‘physical well-being’ (70.4 ± 16.8; p < 0.001) scores compared to Dutch normative Breast-Q data [27]. For the domains ‘satisfaction with breasts’ (63.4 ± 16.8; p = 0.552) and ‘sexual well-being’ (55.3 ± 22.6; p = 0.081) no significant differences were found (Fig. 4).

Fig. 4
figure 4

Patient-reported outcomes of the four Breast-Q questionnaire. Mean patient-reported outcomes ± standard deviation per Breast-Q domains: (1) satisfaction with breasts, (2) psychosocial well-being, (3) physical well-being, and (4) sexual well-being. Comparing the non-reconstruction group [(1) n = 98, (2) n = 99, (3) n = 99, and (4) n = 64], the reconstruction group [(1) n = 55, (2) n = 55, (3) n = 55, and (4) n = 46], and normative Dutch data [(1) n = 9037, (2) n = 9037, (3) n = 9037, and (4) n = 8957]

Full size image

Discussion

Post-mastectomy breast reconstruction restores the breast contour and positively affects body-image and well-being [8, 9, 28, 29]. Despite these benefits, not all women opt for post-mastectomy breast reconstruction. The current study showed a reconstruction rate of 32%. In most cases, it concerned an immediate breast reconstruction (93%). This study is one of the first to combine the Breast-Q questionnaire with a short survey to gain insight into the motives of women to opt for or reject post-mastectomy breast reconstruction. In addition, the Breast-Q scores were compared to normative values.

In line with previous studies [6, 15], this study showed several significant differences in patient characteristics between the non-reconstruction group and reconstruction group. A higher BMI, presence of comorbidities such as hypertension or diabetes mellitus, and a high ASA classification make patients less ideal candidates for immediate breast reconstruction. Adjuvant treatment also seems to have impact on the possibility to opt for immediate breast reconstruction. In our hospital, delayed breast reconstruction is preferred and immediate breast reconstruction is discouraged when radiotherapy is indicated due to a higher risk of complications and non-optimal aesthetic results. This issue mainly applies to implant-based reconstructions [30, 31].

Differences between the non-reconstruction group and the reconstruction group arise in factors that affected patients’ decision making. The majority (73%) of the reconstruction group stated that the importance of appearance affected their decision. In the non-reconstruction group, there were considerable differences between patients to reject breast reconstruction. As expected, age differed between the non-reconstruction group and the reconstruction group. In line with previous studies, patients who opt for breast reconstruction are younger than those who do not choose for reconstruction [15, 16]. Interestingly, more than half of our respondents without breast reconstruction (60%) stated that their age did not affect their decision. Although it is expected that an older age at time of mastectomy is correlated with the rejection of breast reconstruction, this study could not find a significant correlation.

The responses of the free text field of the short survey included a wide range of factors that influenced the decision regarding breast reconstruction. Besides, in the responses to the survey of the current study, some arguments are used in both groups with different perspectives. In the non-reconstruction group multiple patients stated that they wanted to be practical, and the external prosthesis is seen as an easy solution to restore breast contour while being dressed. Interestingly, ‘being practical’ is also used as an argument in favor of breast reconstruction, since external breast prosthesis can also be experienced as inconvenient. The meaning of breast reconstruction seems to differ amongst women. Some stated it as a way to forget breast cancer, while others do not want to think about breast cancer ever again and see breast reconstruction as a constant reminder of this period. So, in daily practice, it seems that two types of women may be identified. The women who desire breast reconstruction, and those who do not seem to need nor desire breast reconstruction. It is important to acknowledge these differences, but these are difficult to identify based on patient characteristics and Breast-Q scores only. Besides, in line with other studies, some patients were not ready to decide about breast reconstruction at the time of their oncologic surgery. Although the free text field was not a mandatory field, and our results should be interpreted with car, these results do highlight the importance of personalized guidance and consultation about breast reconstruction, before and after oncologic treatment, to provide women with the needed support.

In the current study, the Breast-Q was used in addition to our non-validated short survey. We hoped that a validated questionnaire could provide extra information on patients’ choice on breast reconstruction and the relation to HRQoL. Based on the short list survey, a majority of patients (83%) of the non-reconstruction group did not have a desire for breast reconstruction. Nevertheless, compared to the Dutch normative population, patients without breast reconstruction had poorer results on all four domains of the Breast-Q questionnaire. These results give the impression that all these women have a lower quality of life than the Dutch normative population. Unfortunately, we did not have a baseline HRQoL score. Patients in the non-reconstruction group had a higher BMI, and a more extensive medical history. These factors may have resulted in a lower baseline HRQoL even before the diagnosis of breast cancer. Consequently, also the HRQoL after treatment will be lower than the HRQoL of the Dutch normative population. Therefore, differences in the patient-reported outcomes may not solely be explained by the breast cancer treatment. The reconstruction group had poorer results on two domains of the Breast-Q questionnaire (psychosocial well-being, and physical well-being). The majority of patients (93%) of the reconstruction group had an immediate implant-based breast reconstruction. Yoon et al. (2018) showed that immediate breast reconstruction is still the preferred choice, but delayed breast reconstruction does not appear to compromise clinical and patient reported outcomes [29]. Nevertheless, we can imagine that patients who opt for delayed breast reconstruction have more time for their decision and they had a period without a breast resulting possibly in a higher satisfaction with the reconstructed breast.

This study was limited by its retrospective design, which resulted in missing data, possible recall bias, and missing baseline HRQoL scores. This makes interpretation of differences in HRQoL difficult. Also the use of NAC in the responders cohort significantly differed (p = 0.006). Unfortunately, despite multiple attempts to increase the response rate, the response rate remained rather low. Some patients clearly stated not to be interested in answering the questions, with as main reason that they did not want to recall those memories. The poor response rate may result in selection bias, and perhaps neglecting unsatisfied patients. Consequently, this neglect may have resulted in an underestimation of the impact of mastectomy and post-mastectomy breast reconstruction. On the other hand, the poor response may also have neglected the satisfied patient. The retrospective design of the study may have also led to possible recall bias. Nevertheless, we still think that the majority of the responses clearly show the factors that patients used to make their decision. Another limitation may be the confounding due to differences in patient characteristics between the non-reconstruction group and the reconstruction group. This was partially corrected with the multivariable analysis. However, outcomes should be interpreted with care.

Our study resulted from the question on the uptake rate of breast reconstructions after mastectomy in our hospital. To address this question, it was designed as descriptive retrospective inventory. Our study showed that there seem to be a difference in patients who opt for and who reject post-mastectomy breast reconstruction. To make more solid recommendations in the pre- and postoperative consultation, a prospective longitudinal study is the next step.

Conclusion

This study showed that patients are well-informed when making their decisions regarding breast reconstruction. Although the positive effects of breast reconstruction are known, some women had well-informed motives not to choose for post-mastectomy breast reconstruction. Patients have personal motives to opt for or reject breast reconstruction. Patients differ in the rating of values that affect their decision since the same arguments were used to opt for or to reject reconstruction. Therefore, our study highlights the importance of personalized guidance and consultation.