Our search strategy identified 40 citations. Searching additional sources yielded an additional 9 potentially eligible citations. Of these, 30 were excluded based on title and abstract screening, because of duplicated papers (n = 5), missing outcomes (n = 7), or the lack of a standardized pain treatment in the control group (n = 18). Thus a total of 19 randomized controlled trials were included in this systematic review. The flow diagram (see Figure 1) shows the results from the literature search and the study selection process.
The characteristics of included studies and outcomes assessed in this review are presented in Table 2A, 2B and 2C.
The surgical procedures performed in the reviewed trials included lumpectomy or mastectomy in 14 of 19 trials,10-23 with additional Sentinel Lymph Node Biopsy (SLNB) or Axillary Lymph Node Dissection (ALND).10, 12, 14-18, 20-21, 23 Three trials included patients undergoing breast augmentation,24-26 and two trials also included patients undergoing breast reduction.27-28
The 19 randomized controlled trials involved a total of 1307 patients, of which 749 received nerve blocks, 330 received local infiltration, 115 received IV lidocaine, 73 received infusion of local anesthetic via pump, and 40 received tumescent anesthesia.
According to the COSMIN checklist, all studies included in this review showed an excellent-to-good quality. The majority of clinical trials had a low risk of bias.
Figure 1. Flow diagram study selection process.
Table 2A, 2B, 2C. Studies characteristics.
LUMPECTOMY/MASTECTOMY. In the included studies 1086 patients underwent lumpectomy/mastectomy. There were no demographic differences between the two groups as shown in Table 3.
Table 3. Personal and clinical characteristics of patients undergoing lumpectomy/mastectomy.
The largest studies involved 140 patients (Mohamed et al., 2013; Versyck et al., 2017), whereas the smallest consisted of 44 patients (Couceiro et al., 2014). All analysed studies were conducted in inpatient settings.
Patients underwent mastectomy, while lumpectomy was performed in 4 studies (Campbell et al., 2014; Cros et al., 2018; Gürkan et al., 2018; Versyck et al., 2017). Lanier et al 2018 and Wang et al 2018 reported immediate tissue expander or implant based breast reconstruction. Surgical treatment for breast cancer was associated with a sentinel lymph node biopsy (SLNB) or a axillary lymph node dissection (ALND) in 10 papers (Campbell et al., 2014; Cros et al., 2018; Gürkan et al., 2018; Ilfeld et al., 2014; Lanier et al., 2018; Mohamed et al., 2013; Neethu et al., 2018; Versyck et al. 2017;Wang et al., 2018; Yao et al., 2019).
Regional anesthetic techniques were performed ten times: the most common procedure was pectoral nerve (Pecs) block type I and II (5 studies: Cros et al., 2018; Neethu et al., 2018; Versyck et al. 2017; Wang et al., 2018; Wang et al., 2019), followed by serratus plane block (SPB) (2 studies: Wang et al., 2019; Yao et al., 2019), erector spinae plane (ESP) block (Gürkan et al., 2018), and paravertebral block (PVB) (Ilfeld et al., 2014). In Lanier et al 2018, intraoperative nerve blocks of intercostal and pectoral nerves were performed under direct visualization by the attending plastic surgeon at the completion of the mastectomy.
Local infiltration was reported in 3 studies. In Campbell et al 2014, patients received 20 mL of bupivacaine 0.25% with or without adrenaline to be infiltrated into the breast wound and a further 20 mL of bupivacianee 0.25% with adrenaline to be infiltrated into the axilla wound when applicable. In Mohamed et al 2013, 5 ml of bupivacaine 0.5% with or without clonidine were diluted with saline 0.9% to 15 mL volume and irrigated into the surgical field before skin closure. An infusion pump of levobupivacaine 0.50% for approximately 48 hours was evaluated in Ferreira Laso et al 2014.
Couceiro et al 2014 and Terkawi et al 2014 investigated i.v. lidocaine infusion. In the first paper, a bolus dose of lidocaine was not administered and, after incision, a lidocaine infusion at 3 mg/kg was started. In the other study, lidocaine was administered as a bolus to all patients before anesthetic induction, at a dose of up to 1.5 mg/kg, followed by a lidocaine infusion at 2 mg/kg/h until 2 hours after arrival in PACU.
Almost all studies were conducted under general anesthesia, except for Ilfeld et al 2014 conducted under sedation. General anesthesia was induced and then maintained with opioids, such as alfentanil (Ferreira Laso et al., 2014), fentanyl (Couceiro et al., 2014; Gürkan et al., 2018; Ilfeld et al., 2014; Mohamed et al., 2013; Neethu et al., 2018; Terkawi et al., 2014; Wang et al., 2018), remifentanil (Wang et al., 2019) and sufentanil (Versyck et al., 2017; Wang et al., 2019; Yao et al., 2019). For the postoperative pain management, four studies (Couceiro et al., 2014; Cros et al., 2018; Ferreira Laso et al., 2014; Ilfeld et al., 2014) provided infiltration of the chest wall ipsilateral to the mastectomy with local anesthetic; acetaminophen and other NSAIDs were systematically administered. Two studies did not report the anesthesia protocol (Campbell et al., 2014; Lanier et al., 2018).
Pain Intensity. Different investigators recorded this outcome on different scales and at different intervals. We normalized all NRS to a zero to 10 range (see Table 4). The majority of authors reported pain intensity at 1, 6, 12, 24 and 48 hours after surgery.
Table 4. NRS at 1, 6, 12, 24 and 48 hours after lumpectomy/mastectomy.
As shown in Figure 2, the average NRS score was 1.83 ± 0.81 at 1 h, 2.02 ± 1.03 at 6 h, 1.92 ± 1.57 at 12 h, 1.74 ± 1.29 at 24 h and 1.51 ± 1.03 at 48 h in the intervention group; in the placebo group NRS scores were recorded of 3.36 ± 1.65 at 1 h, 3.10 ± 1.29 at 6 h, 2.35 ± 1.14 at 12 h, 2.37 ± 1.27 at 24 h and 2.77 ± 1.12 at 48 h.
Figure 2. NRS at 1, 6, 12, 24 and 48 hours after lumpectomy/mastectomy.
One study, Couceiro et al 2014, did not report NRS at any interval: only 2 (9.09%) vs 3 (13.6%) patients in the lidocaine and placebo groups, respectively, experienced severe to very severe pain 24 hours after surgery.
Opioid consumption. The most frequently used opioids were codeine (Campbell et al., 2014; Couceiro et al., 2014), fentanyl (Neethu et al., 2018), oxycodone (Campbell et al., 2014), piritramide (Versyck et al. 2017), tramadol (Campbell et al., 2014; Mohamed et al., 2013; Versyck et al. 2017), and sufentanil (Yao et al., 2019). Morphine was used in the other studies.
At discharge from PACU, the overall mean amounts of morphine consumed in the intervention and placebo groups were 3.0 ± 3.63 mg and 4.87 ± 5.76 mg, respectively, with the difference being not statistically significant (p = 0.51).
After 48 hours, the overall mean amounts of morphine consumed in the intervention and placebo groups were, respectively, 2.68 ± 0.88 mg and 4.94 ± 4.61 mg (p = 0.18). Among regional anesthetic techniques, postoperative opioid consumption for the first 48 hours was respectively 2.14 ± 1.52 mg and 4.84 ± 4.63 mg; no statistically significant difference was observed (p = 0.16). The average per patient opioid consumption up to 48 h after surgery was 1.33 ± 1.49 mg vs 3.01 ± 3.05 mg (p = 0.52) among patients treated with local infiltration.
Table 5 shows the mean amounts of morphine consumed in the intervention and placebo groups.
Table 5. Average per patient opioid consumption in PACU and up to 48 h after lumpectomy/mastectomy.
In Couceiro et al 2014, opioid consumption in the first 24 hours after surgery was similar in the lidocaine and placebo groups.
Adverse Events (AEs). An adverse event is defined as any undesirable experience associated with the use of a medical product in a patient. A total of 379 AEs were recorded. Three studies (Couceiro et al., 2014; Ilfeld et al., 2014; Versyck et al., 2017) did not report the number of AEs. The most frequently reported AEs were nausea, vomiting and postoperative nausea and vomiting (PONV), pruritus, dizziness, haematoma/bleeding, seroma and bruising (see Table 6 and Figure 3). Some studies did not specify the timing of adverse events.
Table 6. Number of adverse events (AEs) after lumpectomy/mastectomy.
Figure 3. Distribution of adverse events (AEs) among intervention groups (lumpectomy/mastectomy).
Nausea, vomiting and PONV. Nausea, vomiting and PONV were the most frequent AEs (235 events, 62% of AEs). 96 vs 139 episodes were respectively recorded in intervention vs placebo group (p = 0.25). In 9 studies (Cros et al., 2018; Ferreira Laso et al., 2014; Gürkan et al., 2018; Lanier et al., 2018; Neethu et al., 2018; Terkawi et al., 2014; Versyck et al. 2017; Wang et al., 2018; Yao et al., 2019) prophylaxis was administered; in the others studies prescription of antiemetic drugs was missed or not reported.
Pruritus. A total of 23 cases of pruritus was reported in two studies (Ferreira Laso et al., 2014; Wang et al., 2019). No statistically significant difference was observed (7 vs 16, p = 0.47).
Dizziness. Fifteen episodes were recorded among patients not receiving treatment; only two patients reported dizziness in the intervention group. No statistically significant difference (p = 0.06) was observed in these 3 studies (Ferreira Laso et al., 2014; Wang et al., 2019; Yao et al., 2019).
Haematoma/Bleeding. Three studies (Campbell et al., 2014; Cros et al., 2018; Ferreira Laso et al.) reported a total of 21 episodes (10 vs 11, p = 0.85).
Seroma. Seroma was found in 10 cases in the intervention group vs 11 cases in the placebo group (Campbell et al., 2014; Ferreira Laso et al., 2014) with no statistically significant difference (p = 0.91).
Bruising. Campbell et al 2014 reported 36 episodes of bruising (20 vs 16, p = 0.70).
Others. Haemodynamic changes were rarely reported. Hypotension was reported in 3 studies (Cros et al., 2018; Ferreira Laso et al., 2014; Mohamed et al., 2013) for a total of 7 AEs (5 vs 2, p = 0.25). Ferreira Laso et al 2014 reported a case of hypertension. Two episodes of bradycardia were reported from Mohamed et al 2013.
Infection was observed, respectively, 3 vs 2 times in intervention and placebo groups in two studies (Campbell et al., 2014; Ferreira Laso et al.); no statistically significant difference was noticed (p = 0.69). Deep vein thrombosis (DVT), pulmonary thromboembolism (PTE) and acute respiratory infection were equally distributed (1 vs 1, p = 1.0) in Ferreira Laso et al 2014.
Patient satisfaction. Patient satisfaction results were presented as different degrees of subjective satisfaction levels. We normalized all of them to "satisfied / not satisfied".
Only 5 studies (Cros et al., 2018; Ferreira Laso et al., 2014; Lanier et al., 2018; Neethu et al., 2018; Wang et al., 2019;) were available for analysis of satisfaction (see Table 7 and Figure 4).
Table 7. Patient satisfaction after lumpectomy/mastectomy.
Figure 4. Patient satisfaction after lumpectomy/mastectomy.
Among the intervention group, there were 164 satisfied patients vs 14 not satisfied patients; in the placebo group 149 patients were satisfied and 40 not satisfied. No statistically significant difference was observed between the two groups (satisfied, p = 0.28; not satisfied, p = 0.14).
In Versyck et al 2017, both patient-groups were very satisfied with their management; while in Yao et al 2019, patient satisfaction scores were higher in the SPB group.
BREAST AUGMENTATION. 142 patients underwent breast augmentation. As shown in Table 8, the groups were similar in age, weight, height and body mass index.
Table 8. Personal and clinical characteristics of patients undergoing breast augmentation.
All patients underwent subpectoral bilateral cosmetic breast augmentation.
Regional anesthetic techniques were performed in two studies: PVB in Gardiner et al 2012; association of Pecs type II and PSB in Schuitemaker et al 2019. In the other study, Picard et al 2017, a local infiltration was performed.
Patients received general anesthesia in Schuitemaker et al 2019, and sedation in Gardiner et al 2012. In both studies patients received fentanyl. Picard et al 2017 did not report the anesthesia protocol.
Pain Intensity. Different investigators recorded this outcome on different scales and at different intervals. We normalized all NRS to a zero to 10 range (see Table 9). The majority of authors reported pain intensity at 1, 6, 24 and 72 hours after surgery.
Table 9. NRS at 1, 6, 24 and 72 hours after breast augmentation.
As shown in Figure 5, the average NRS scores were 3.4 ± 0.5 at 1 h, 3.0 at 6 h, 3.65 ± 1.15 at 24 h and 3.05 ± 0.25 at 72 h in the intervention group; in the placebo group NRS scores were recorded of 5.25 ± 0.05 at 1 h, 3.0 at 6 h, 3.65 ± 1.15 at 24 h and 4.2 ± 0.5 at 72 h.
Figure 5. NRS at 1, 6, 24 and 72 hours after breast augmentation.
Opioid consumption. In Gardiner et al 2012, 6 patients in the placebo group required supplementary opioid use. No patient received opioids in the intervention group.
After 24 hours in Schuitemaker et al 2019, the overall mean amounts of morphine consumed in the intervention and placebo groups were 0.2 ± 0.8 mg and 0.6 ± 1.2 mg, respectively, with the difference being not statistically significant (p = 0.29).
No data on opioid consumption was available in Picard et al 2017.
Adverse Events (AEs). A total of 19 (9 vs 10) AEs were recorded.
Gardiner et al 2012 reported 12 episodes of nausea and vomiting (5 vs 7, p = 0.36). In this study, hypotension occurred 3 times among patients in the intervention group and once in the placebo group. A single case of bradycardia occurred in both groups.
No differences were observed between groups concerning the appearance of AEs in Schuitemaker et al 2019.
No adverse effects were reported in Picard et al 2017.
Patient satisfaction. Only Schuitemaker et al 2019 reported data about patient satisfaction: after 24 hours, 80% vs 53% of patients in intervention and placebo groups were satisfied.
BREAST REDUCTION. 79 patients underwent breast reduction. The mean age was 38.28 ± 2.71years vs 38.78 ± 3.21 years (see Table 10).
Table 10. Personal and clinical characteristics of patients undergoing breast reduction.
All patients underwent breast reduction surgery.
Tumescent anesthesia was performed in Christie et al 2017; in Valente et al 2014, patients received local infiltration. All patients in these studies underwent general anesthesia.
Pain Intensity. Different investigators recorded this outcome on different scales and at different intervals. We normalized all NRS to a zero to 10 range (see Table 11). The authors reported pain intensity at 24 hours after surgery.
Table 11. NRS at 24 hours after breast reduction.
The average NRS score was 2.55 ± 1.72 in the intervention group; in the placebo group a NRS score was recorded of 2.85 ± 1.14 (p = 0.89).
Opioid consumption. After 24 hours in Christie et al 2017, the overall mean amounts of morphine consumed in the intervention and placebo groups were 0.58 mg and 0.64 mg, respectively, with the difference being not statistically significant (p = 0.71).
No data on opioid consumption is available in Valente et al 2014.
Adverse Events (AEs). There was no significant difference in occurrence of nausea or vomiting in the first 24 hours between the two groups in Christie et al 2017. No adverse effects were reported in Valente et al 2014.
Patient satisfaction. The level of satisfaction was not reported in these studies.