If it becomes clear during the preoperative multidisciplinary meeting that adjuvant radiation therapy (RT) of the breast/chest wall will likely or certainly be required, it is insufficiently clear whether immediate breast reconstruction is contraindicated.

1. Comparison of immediate implant reconstruction followed by radiation therapy versus immediate implant reconstruction without radiation therapy

2. Comparison of immediate autologous reconstruction followed by radiation therapy versus immediate implant reconstruction without radiation therapy

3. Comparison of immediate autologous reconstruction followed by radiation therapy versus immediate implant reconstruction followed by radiation therapy

1. Comparison of immediate implant reconstruction followed by radiation therapy versus immediate implant reconstruction without radiation therapy

One systematic review (Barry et al, 2011) of four comparative retrospective studies compared patients undergoing immediate breast reconstruction with a definite prosthesis or tissue expander followed by postoperative radiation therapy with patients who did not receive postoperative radiation therapy (Tallet et al, 2003; Krueger et al, 2001; Cordeiro et al, 2004; Whitfield et al, 2009). Average follow-up varied from 25 months (Tallet et al, 2003) to 51 months (Whitfield et al, 2009). As the review by Barry et al did not report the correct data from the individual studies, the data for the outcome measures aesthetic result and postoperative complications were extracted from the individual studies.

After the search date for the review by Barry, one additional study (Nava et al, 2011) was published and included in the literature analysis. This study compared three groups: a) 109 patients with an immediate reconstruction using a tissue expander and prosthesis who underwent radiation therapy after placement of the definite prosthesis, b) 50 patients who underwent immediate reconstruction with a tissue expander and prosthesis followed by radiation therapy after expansion of the tissue expander and c) 98 patients who underwent immediate reconstruction with a tissue expander and prosthesis without subsequent radiation therapy. Average follow-up was 50 months.

Local recurrence

The outcome measure local recurrence was not studied.

Capsular contracture

The pooled result (N=411) shows that the chances of capsular contracture are three times higher if immediate implant reconstruction with radiation therapy is performed (statistically significant, see figure 1). Cordeiro et al (2004) followed patients for at least one year after adjuvant radiation therapy.

Figure 1. Capsular contracture

In the implant with RT group, the total number of capsular contractures was 372/1000, in the implant without RT group this was 158/1000.

Infection

The pooled result (N=781) shows that the chance of infection is twice as high if immediate implant reconstruction with radiation therapy is performed (statistically significant, see figure 2). Cordeiro et al (2004) followed patients for less than one year after adjuvant radiation therapy (reason for different N in the outcome measures).

Figure 2. Infection

In the implant with RT group, the number of infections was 84/1000, in the implant without RT group this was 29/1000.

Re-operation

The pooled result (N=781) shows that the chance of re-operations is two and a half as high if immediate implant reconstruction with radiation therapy is performed (statistically significant, see figure 3). Cordeiro et al (2004) followed patients for less than one year after adjuvant radiation therapy (reason for different N in the outcome measures).

Figure 3. Re-operation

In the implant with RT group, the number of re-operations was 187/1000, in the implant without RT group this was 64/1000.

Fibrosis

The outcome measure fibrosis was not studied.

Aesthetic result

Four studies included aesthetic result as an outcome measure (Tallet et al, 2003; Krueger et al, 2001; Cordeiro et al, 2004; Nava et al, 2011). Aesthetic result was measured in a number of different ways.

In the study by Tallet et al (n=77), the aesthetic result was assessed by a physician based on five criteria: shape of the breast without a bra, contralateral symmetry, definition of the inframammary fold, mobility and consistency of the breast. A figure showed that 43% of patients with an implant reconstruction followed by adjuvant radiation therapy were graded as ‘good’. In the implant group without radiation therapy, this was 80% (the authors did not provide absolute numbers).

In the study by Kruger et al, the aesthetic result was assessed using a questionnaire. The questionnaire used a 5-point Likert scale, where 1 indicated high satisfaction and 5 low satisfaction. Only questionnaires with scores of 1 or 2 on all items were scored as ‘satisfied’. In the group that had received postoperative radiation therapy, 29% (4/14) of patients were satisfied, while 23% (10/43) of the group without postoperative radiation therapy was satisfied.

The study by Cordeiro et al does not describe how aesthetic result was measured. A distinction was made between excellent, good or poor to fair. Of all patients with an implant reconstruction followed by radiation therapy, 38% (25/66) were assessed as ‘excellent’. In the implant group without radiation therapy, this was 69% (52/75).

In the study by Nava et al (N=257), breast shape and symmetry were assessed by the surgeon and classified as good, moderate or poor. The outcome was generally scored as ‘good’ in the group that had not received postoperative radiation therapy (72.2% for shape and 46.1% for symmetry). In the group that had received immediate reconstruction using a tissue expander and prosthesis and had undergone radiation therapy after placement of the definite prosthesis, this was 58.7% and 28.8%, respectively. In the group that had undergone immediate reconstruction with a tissue expander and prosthesis followed by radiation therapy after expansion of the tissue expander, this was 30.8% and 15.4%, respectively (p=0.0009).

Comment: studies may have been missed for this outcome measure, because this literature analysis is an update of the review by Barry et al, and he did not include this outcome measure. The data were extracted from the individual studies.

Patient satisfaction

One study examined the outcome measure patient satisfaction (Nava et al 2011). Patients were asked to rate their reconstruction as good, moderate or poor. In the group that had not received postoperative radiation therapy, 68.1% of patients scored the result as good, while this was 52.2% in the group who had undergone immediate reconstruction using a tissue expander and prosthesis and had received RT after placement of the definite prosthesis, and 46.2% for reconstruction with a tissue expander and prosthesis followed by radiation therapy after expansion of the tissue expander (p=0.04).

Level of evidence of the literature

The level of evidence for the outcome measures postoperative complications (capsular contracture, infection, re-operation and fibrosis), aesthetic result and patient satisfaction is very low, as there are very serious limitations in study design (non-randomized studies) and due to the low number of patients (imprecision). Furthermore, statistical heterogeneity is present for the outcome measure capsular contracture (I²=56%). Furthermore, it is unclear whether the applied fractioning schedules and radiation therapy techniques are comparable to those used in The Netherlands, and whether the incidence of adverse events after radiation therapy can be extrapolated to the Dutch situation.

The level of evidence for the outcome measure local recurrence was not assessed, as there are no comparative studies with these outcome measures.

2. Comparison of immediate autologous reconstruction followed by radiation therapy versus immediate implant reconstruction without radiation therapy

One systematic review (Schaverien et al, 2013) compared patients with an immediate autologous breast reconstruction followed by postoperative radiation therapy with patients who had not received postoperative radiation therapy. Ten retrospective comparative studies were included (Berry et al, 2010; Lee et al, 2010; Leonardi et al, 2010; Chatterjee et al, 2009; Carlson et al, 2008; Thomson et al, 2008; Spear et al, 2005; Rogers and Allen, 2002; Tran et al, 2000; Williams et al, 1997). Follow-up was unknown for four studies, and ranged from 1 to 5 years for patients who had received postoperative radiation therapy in the other studies.

Local recurrence

The outcome measure local recurrence was not studied.

“Overall” complications

For the outcome measure complications, only those complications reported in all studies were included. Which complications were included specifically could not be determined based on the systematic review. The review did report that the complications ‘anastomosis revision’ and ‘abdominal weakness and bulging’ were excluded. The pooled result (N=1121) shows that the risk of complications is about the same when postoperative radiation therapy is performed compared to when no postoperative radiation therapy is performed (see figure 4).

Figure 4. “Overall” complications

Fat necrosis

The pooled result (N=2313) shows that the chance of fat necrosis is 2.8 times higher if postoperative radiation therapy is performed than if no postoperative radiation therapy is performed (statistically significant, see figure 5).

Figure 5. Fat necrosis

Re-operation

Re-operation is defined as a follow-up operation for reconstruction of the ipsilateral breast. The pooled result (N=272) shows that the chance of re-operation is lower if postoperative radiation therapy is performed than if no postoperative radiation therapy is performed (not statistically significant, see figure 6).

Figure 6. Re-operation

Aesthetic result

Seven studies included aesthetic result as an outcome measure (Lee et al, 2010; Leonardi et al, 2010; Carlson et al, 2008; Thomson et al, 2008; Spear et al, 2005; Rogers and Allen, 2002; Tran et al, 2000). Aesthetic result was measured in a number of different ways.

In the study by Lee et al (N=407), the aesthetic result was assessed based on patient satisfaction. Patient satisfaction was assessed with a questionnaire that used a 5-point Likert scale. In the group that had received postoperative radiation therapy, 66.7% of patients were satisfied, while 75.7% of the group without postoperative radiation therapy was satisfied.

In the study by Leonardi et al (N=33), 21 physicians evaluated the aesthetic result based on photographs of women who had undergone autologous breast reconstruction. Four items were scored: symmetry, volume, position of the inframammary fold and ptosis. Each item was scored between 0 (worst result) and 10 (best result) based on which the aesthetic result was assessed as ‘excellent’ (treated breast practically identical to the other breast), ‘good’ (treated breast only slightly different from the other breast), ‘fair’ (treated breast clearly different from the previous breast, but not deformed) or ‘poor’ (treated breast seriously deformed) as described by Harris et al (1979). In the group receiving postoperative radiation therapy, 47.6% was evaluated as ‘good’ or ‘excellent’. In the group that had not received postoperative radiation therapy this was 68.6% (statistically significant, p<0.0001).

In a study by Carlson et al (174 flaps), photographs of women were evaluated by four independent evaluators. They scored the breast reconstruction based on volume, breast shape, location of the breast and position of the inframammary fold. As was the case for the study by Leonardi et al, the total aesthetic result was assessed as ‘excellent, good, fair or poor’, which were given 4, 3, 2 or 1 points, respectively. On average, the postoperative radiation therapy group scored 2.56 points, and the group without postoperative radiation therapy 3.02 points.

In the study by Thompson et al (N=20), the aesthetic result was assessed twice, once based on patient satisfaction, and once based on the judgment of three medical professionals. Patient satisfaction was determined using a questionnaire covering 13 items identified based on the literature, scored with 5-point Likert scales. The medical professionals also gave their score on a 5-point Likert scale, based among other things on symmetry, shape, size, skin color, scars and nipple reconstruction. The medical professionals gave a mean score of 3.3 on the Likert scale to the postoperative radiation therapy group, and 4.5 points to the group without postoperative radiation therapy. Patients gave a mean score of 3 on the Likert scale when receiving postoperative radiation therapy, and a mean of 4 when not treated with postoperative radiation therapy.

In the study by Spear et al (N=112), photographs of women who underwent autologous breast reconstruction were assessed by 16 blinded judges. They scored the aesthetic result as ‘excellent, good, fair or poor’, which were given 4, 3, 2 or 1 points, respectively. The mean score for the postoperative radiation therapy group was 2.76, the mean score for the group without postoperative radiation therapy was 3.77 (statistically significant, p=0.0001).

In the study by Rogers and Allen (N=60), photographs of women who underwent autologous breast reconstruction were assessed by eight judges, including medical specialists and medical students. Photographs were made immediately after breast reconstruction, and new photographs were taken again later, potentially after radiation therapy. The photographs were assessed for symmetry, position and overall aesthetic proportion. Scores between 1 (worst result) and 5 (best result) were given on these three items, and an average was calculated based on the three scores. In the group treated with postoperative radiation therapy, photographs were scored 0.56 lower than photographs before radiation therapy. In the group not treated with postoperative radiation therapy, photographs were scored 0.50 higher than earlier photographs (statistically significant, p=0.006).

The study by Tran et al (N=1484) examined whether breast symmetry remained intact if radiation therapy was given after breast reconstruction. Data were collected retrospectively in a number of different ways, such as personal communications, physical examination and photographic review. Only seven of the 41 patients (17%) achieved and retained symmetry.

Patient satisfaction

The outcome measure patient satisfaction was not studied.

Level of evidence of the literature

The level of evidence for the outcome measures “overall” complications, fat necrosis and aesthetic result is very low, as there are very serious limitations in study design (non-randomized studies) and because studies were not powered for these outcome measures (imprecision). In addition, statistical heterogeneity is present for the outcome measure fat necrosis (I²=56%). Furthermore, it is unclear whether the applied fractioning schedules and radiation therapy techniques are comparable to those used in The Netherlands. Thus, the question remains whether the results from the study can be extrapolated to Dutch clinical practice.

The level of evidence for the outcome measure re-operation is very low for the same reasons listed above. A clinically relevant effect is present. However, this effect is highly uncertain due to the wide confidence intervals (imprecision).

The level of evidence for the outcome measures local recurrence and patient satisfaction was not assessed, as there are no comparative studies with these outcome measures.

3. Comparison of immediate autologous reconstruction followed by radiation therapy versus immediate implant reconstruction followed by radiation therapy

One systematic review (Barry et al, 2011) of four comparative retrospective studies compared patients undergoing immediate autologous reconstruction with an immediate implant reconstruction, both followed by radiation therapy (Anderson et al, 2004; Jhaveri et al, 2008;Wong et al, 2008; Strålman et al, 2008). After the search date for the review by Barry, one additional study (Christante et al, 2010) was published and included in the literature analysis. Follow-up varied from 10 months (Wong et al, 2008) to 34 months (Strålman et al, 2008).

Re-operation

The pooled result (N=441) shows that the chance of re-operation is more than twice as high if immediate implant reconstruction followed by radiation therapy is performed (statistically significant, see figure 7).

Figure 7. Re-operation

In the implant group + RT, the number of re-operations was 261/1000, in the autologous group +RT this was 122/1000.

Infection

Two studies examined the outcome measure infection (Anderson et al, 2004; Wong et al, 2008). Wong only looked at serious infections that required additional intervention, and found 13.3% (2/15) in the implant group versus 8.5% (4/47) in the autologous group.

In the study by Anderson et al, 4% (2/50) of patients with an implant reconstruction developed an infection, compared to 0% (0/35) in the autologous reconstruction group.

Aesthetic result

Two studies included aesthetic result as an outcome measure (Anderson et al, 2004; Jhaveri et al, 2008). Aesthetic result was measured in a number of different ways. In the study by Anderson et al, 82% (41/50) of patients with an implant reconstruction rated the result as ‘excellent/good’, compared to 90% (31.5/35) in the autologous reconstruction group. In the study by Jhaveri et al, 51% (35/69) of patients with an implant reconstruction rated the result as ‘acceptable’, compared to 83% (19/23) in the autologous reconstruction group. In summary, the two studies found better aesthetic results if, in case of postoperative radiation therapy an immediate autologous reconstruction was performed.

Patient satisfaction

The outcome measure patient satisfaction was not studied.

Level of evidence of the literature

The level of evidence for the outcome measure local recurrence, postoperative complications (re-operation, infection, fibrosis and fat necrosis) and aesthetic result is very low, as the study is not randomized (highly significant limitations in study design) and the number of patients with complications is low (imprecision). Furthermore, statistical heterogeneity is present for the outcome measure capsular contracture (I²=88%).

The level of evidence for the outcome measures local recurrence and patient satisfaction was not assessed, as they were not studied.

Three systematic literature analyses were performed in order to answer the primary question, examining the following questions:

1. What is the difference between the effects of immediate implant reconstruction followed by radiation therapy compared with immediate implant reconstruction not followed by radiation therapy on local recurrence, postoperative complications, aesthetic result or patient satisfaction?
2. What is the effect of immediate autologous reconstruction followed by radiation therapy compared with immediate autologous reconstruction not followed by radiation therapy on local recurrence, postoperative complications, aesthetic result or patient satisfaction?
3. What is the effect of immediate autologous reconstruction followed by radiation therapy compared with immediate implant reconstruction followed by radiation therapy on local recurrence, postoperative complications, aesthetic result or patient satisfaction?

Medline (OVID), Embase and Cochrane databases were searched for immediate reconstruction. The search justification is listed at the end of this chapter. The literature search yielded 136 results. Studies that met the following selection criteria were included in the literature summary: original studies; comparative studies; systematic review of comparative studies; comparison of immediate breast reconstruction followed by postoperative radiation therapy to immediate breast reconstruction not followed by radiation therapy, and at least one of the following outcome measures: local recurrence, postoperative complications, aesthetic result or patient satisfaction.

Of the 136 references identified, three studies met the selection criteria for the literature analysis (Barry et al, 2011; Nava et al, 2011; Christante et al, 2010), including one systematic review (Barry et al, 2011). The evidence tables for these studies may be found at the end of this chapter.

After the search date, Schaverien et al (2013) published an additional systematic review that also met the selection criteria and was therefore included in the literature analysis.