Secondary Logo

Journal Logo

Patient Safety: Research Reports

Do Intraoperative Analgesics Influence Breast Cancer Recurrence After Mastectomy? A Retrospective Analysis

Forget, Patrice MD*; Vandenhende, Julie MD*; Berliere, Martine MD, PhD; Machiels, Jean-Pascal MD, PhD; Nussbaum, Benoît MD*; Legrand, Catherine PhD§; De Kock, Marc MD, PhD*

Author Information
doi: 10.1213/ANE.0b013e3181d2ad07

Surgery remains the keystone in the treatment of solid neoplastic tumors including breast cancer. Paradoxically, the perioperative period represents a high risk of metastases development.1 Several factors may account for this phenomenon, including profound depression of antitumoral cellular immunity. Depression of this cellular immunity in the postoperative period is at least partially linked to the metabolic and hormonal changes induced by the “stress reaction” to surgery.2 Analgesics may influence immunity and tumor development either directly by interfering with cellular mechanisms (e.g., cell apoptosis) or indirectly by interacting with the endocrine and sympathetic systems.

In animal models, morphine-based analgesia reduces the number of pulmonary metastases after surgery.3 However, in humans, recent retrospective data suggest that opioid-based perioperative analgesia is associated with reduced recurrence-free survival, when compared with regional techniques for breast and prostate cancer surgery.4,5 Very few studies have evaluated the possible immune effect of the other analgesics or antihyperalgesics used for “balanced” perioperative analgesia such as the nonsteroidal antiinflammatory drugs (NSAIDs), ketamine, and clonidine.

Currently, whether perioperative analgesia has an impact on postoperative cancer recurrence remains largely unknown.1 Therefore, we performed a retrospective analysis of medical records to compare the incidence of local recurrence and metastatic disease after mastectomy in patients with breast cancer who received different intraoperative analgesics (i.e., sufentanil, ketamine, ketorolac, and clonidine).


Patients and Procedures

After approval from the Ethical Committee of the St-Luc Hospital, Brussels, Belgium, we reviewed the medical records of 327 consecutive patients who underwent mastectomy with axillary dissection between February 2003 and September 2008. The size of this convenience sample was mainly due to logistical reasons (availability of the medical records), but it also maintained the homogeneity in the oncological treatments among the patients. Patients with previous ipsilateral surgery for breast cancer were excluded.

Indications for mastectomy with axillary dissection were defined according to international recommendations and guidelines.6,7 These indications were discussed every week by the multidisciplinary board of our breast clinic and regularly updated and adjusted with new international recommendations and literature data. All mastectomies were performed by the same surgeon (MB) and jointly followed by the surgeon and the same oncologist (J-PM). Chemotherapy, radiotherapy, and endocrine therapy were performed according to international expert consensus (ninth and 10th St-Gallen consensus).811

During the first 2 postoperative years, medical consultation occurred every 3 months, then every 6 months for 3 additional years, and then once a year. For all operations, general anesthesia was induced with sufentanil (0–0.2 μg · kg−1) and a hypnotic, sodium thiopental (4 mg · kg−1) or propofol (2–3 mg · kg−1). After insertion of a laryngeal mask airway, anesthesia was maintained with continuous infusion of propofol, plus sevoflurane or desflurane in an oxygen/air mixture. The type, dosage, and combination of intraoperative analgesic therapy were left to the discretion of the 2 anesthesiologists in charge. Total doses administered were as follows: sufentanil (0–0.5 μg · kg−1); preincisional clonidine (0–6 μg · kg−1); and preincisional ketamine (0–0.5 mg · kg−1). Ketorolac, when administered before the incision, was used in IV doses of 20 mg in patients <60 kg body weight and 30 mg in patients >60 kg.

In the postanesthesia care unit, postoperative analgesia consisted of IV piritramide titrated until the visual analog scale scores were lower than 4 on a scale anchored with 0 as “no pain” and 10 as “the worst pain ever experienced.” During the first 48 postoperative hours, all patients received acetaminophen, 3 to 4 g/d. Oral diclofenac 50 mg was administered 3 times a day for 3 days as necessary, in the absence of contraindication (gastric, renal, or advanced age). No additional opioid was administered.

Primary End Point

The main objective of this study was to determine the effect, if any, of the administration of different intraoperative analgesics (sufentanil, ketamine, clonidine, and ketorolac) on cancer recurrence. Consequently, our primary end point for this analysis was the length of recurrence-free survival through February 2009. Recurrence-free survival was measured from the date of surgery to the date of first recurrence, death due to oncological cause, or to the date of last follow-up, whichever occurred first. Recurrence was defined as clinical evidence of local recurrence or development of metastases confirmed by radiological examination(s). Patients lost to follow-up or those remaining disease free at the time of analysis were censored in the statistical model at the date of the last follow-up (administrative censoring). Patients who died of nononcological causes were censored at the date of death (assuming noninformative censoring).

Data Collection

The following data were obtained from the medical records: perioperative (demographic) characteristics, tumor size, histological tumor grade, histopathological type, estrogen and progesterone receptor status, epidermal growth factor receptor type 2 (HER-2) expression, extent of axillary node disease, and administration of perioperative or postoperative adjuvant chemotherapy, radiotherapy, or endocrine therapy. The patient's current status was determined by the most recent follow-up in our breast cancer clinic through February 2009. Lost to follow-up was defined as patients who lacked follow-up for >3 months after the last clinic visit attended date. The Nottingham Prognostic Index was calculated based on the histological findings: 0.2 (tumor size) + histological grade (1 = grade 1, least aggressive tumor appearance; 2 = grade 2, intermediate appearance; and 3 = grade 3, most aggressive appearance) + axillary lymph node involvement (1 = no axillary lymph nodes involved, 2 = up to 3 axillary lymph nodes involved, and 3 = >3 axillary lymph nodes involved).4 The type and total dose of the analgesics administered intraoperatively were obtained by reviewing the electronic intraoperative and postoperative records. The duration of the surgical procedure, and type and doses of hypnotic administered were also noted.

Statistical Analysis

Patient baseline characteristics are presented as mean ± SD, median (interquartile 25–75), or numbers (percentage). Univariate Cox model and log-rank test were used to assess the potential effect of these baseline characteristics and to investigate an eventual effect of the administration of the analgesics on recurrence-free survival probability. Kaplan-Meier analyses were used to estimate recurrence-free survival probabilities. Variables considered for the univariate analyses were tumor size, histological grade and lymph node involvement,12 age, height, weight, length of surgery, Nottingham Prognostic Index, hormonal receptor status (estrogen and/or progesterone positive), HER-2 expression, hypnotics (propofol, sevoflurane, or desflurane), and postoperative analgesics (synthetic opioid piritramide and diclofenac).

After univariate analyses, the Cox regression model was used for multivariate analysis while adjusting for any baseline factors and intraoperative or oncological factors related to the outcome in the univariate analyses (P < 0.05). A single model was conducted including all the analgesics simultaneously to assess the effect of each analgesic while adjusting for taking one and not another analgesic. We then used stepwise manual backward regression, and all significant factors at a value of P < 0.05 were retained in the final model. STATISTICA (data analysis software system, 2004) version 7 (StatSoft, Tulsa, OK) was used for all analyses.


Patients and Procedure Characteristics: Histopathological Findings

The data from 327 consecutive patients who underwent mastectomy with axillary dissection were reviewed. Eight patients were excluded because of intrathoracic tumor expansion, incomplete surgical resection, and/or perioperative metastases. The data from the 319 remaining patients were analyzed. Patients, histological findings, and duration of the surgery are summarized in Table 1. Postoperative complications included wound infection in 13 patients (4%) and significant hemorrhage requiring transfusion and reoperation in 1 patient (0.3%).

Table 1:
Patient Characteristics

The median follow-up time was 27.3 months (13–44 months). Recurrence was noted in 35 patients (11%), and 17 patients (5%) died of oncological causes during the follow-up period. Deaths from nononcological causes were rare (3 patients, 0.9%), distributed equally according to the type of treatment, and not related to treatment (2 due to pneumonia and 1 due to cardiovascular disease). Sixteen patients (5%) were lost to follow-up.

Intraoperative and Postoperative Analgesics

The intraoperative analgesics included sufentanil, ketorolac, clonidine, and ketamine. Table 2 summarizes the proportion of patients receiving the various drugs. Clonidine and ketamine were often combined for their opioid-sparing effects and were associated with smaller doses of sufentanil. Half of the patients who received sufentanil also received ketorolac. All postoperative analgesics (except acetaminophen) are also listed in Table 2.

Table 2:
Number of Patients Receiving Intra- and Postoperative Analgesics and the Doses Administered

Association Between Use of Analgesics and Cancer Recurrence: Univariate and Multivariate Analyses

Potential confounders, including perioperative patient characteristics, cancer prognostic factors, and length of surgery, were equally distributed over the groups of patients treated with the different analgesics (data not shown). One exception is the younger age of the patients receiving ketorolac (56 ± 12 vs 63 ± 14 years) (P = 0.02).

In univariate analysis, intraoperative administration of ketorolac was associated with longer recurrence-free survival (P = 0.019). Cancer recurrence was more frequent in patients who did not receive intraoperative ketorolac (17%, n = 25 of 145 vs 6%, n = 10 of 174; P = 0.001). No significant change in the rate of cancer recurrence was found in the patients who received sufentanil, clonidine, ketamine, and other drugs (P > 0.05) (Table 3, Fig. 1). Univariate analysis also showed that age (P = 0.004), histological grade (P = 0.006), lymph node invasion (P < 0.001), and Nottingham Prognostic Index (P = 0.048) were significantly associated with recurrence-free survival (Table 3).

Table 3:
Univariate Analyses of Demographic Characteristics, Cancer Prognostic Factors, Length of Surgery, and Intraoperative Analgesics for Breast Cancer Surgery
Figure 1:
Kaplan-Meier recurrence-free survival estimated for 319 patients receiving (or not receiving) intraoperative analgesics (sufentanil, clonidine, ketorolac, and ketamine). Univariate analysis by log-rank tests.

The Cox regression model was used for multivariate analysis while adjusting for these factors. Using stepwise manual backward selection, age, histological grade, and lymph node invasion were retained as risk factors in our multivariate analysis model (Table 4).

Table 4:
Multivariate Association with Cancer Recurrence After Mastectomy: Cox Regression Model

When adjusting for age (P = 0.02), histological grade (P = 0.015), and lymph node involvement (P < 0.001), the risk of cancer recurrence remains significantly lower after the intraoperative administration of ketorolac (P = 0.019). No significant effect was found with sufentanil, clonidine, and ketamine (P > 0.05) (Table 4). A single model was constructed to test all the analgesics simultaneously, to assess the effect of each analgesic while adjusting for the others, and to consider the risk of potential confounder effects in the case of drug combinations.


This retrospective analysis suggests an association between a reduced risk of breast cancer recurrence after surgery and the perioperative administration of ketorolac. This reduced risk was not observed when other intraoperative analgesics were administered. In contrast with previous data suggesting a negative influence of opioids on cancer-related immunity,13 sufentanil had no deleterious effect on cancer recurrence. One reason may be the doses used in our patients, which are relatively small in comparison with those in other series.14 This is consistent with the fact that opioid-induced immunosuppression is dose dependent.13

The α2-agonist clonidine enhances hypnotics, optimizes postoperative analgesia, and stabilizes intraoperative hemodynamic variables. This drug potentially interferes with immunity via adrenergic antiinflammatory pathways and by a central analgesic effect that reduces sympathetic tone15; these effects are immunosuppressive in the perioperative period.2 The doses administered were chosen based on the patients' physical conditions and the fact that mastectomy is classically considered a minor surgery14 that induces a relatively limited “stress reaction.”1

Ketamine is a drug widely used in the perioperative period. At subanesthetic doses, it prevents hyperalgesia and enhances analgesia.16 Ketamine, similar to the other analgesics, interferes with natural killer (NK) cell activity. It possesses antiinflammatory properties in humans during the postoperative period17 or in severe sepsis.18 In our series, ketamine was not associated with improved oncological outcome.

When considering the administration of ketorolac, these results suggest that, given just before the surgery, it was associated with a lower risk of breast cancer relapse. By multivariate analysis, age was identified as a potential confounder. Indeed, advanced age was considered a relative contraindication for NSAIDs and was consequently associated with a lower use of NSAIDs. After adjustment for the significant oncological predictors (histological grade and lymph node invasion) and the potential confounder (age), the association between ketorolac administration and the lower risk of cancer recurrence remained significant. Our observation is consistent with the results obtained in previous studies that investigated the influence of NSAIDs on antitumoral immunity. These studies identified the role of prostaglandins in immunity and inflammation,19,20 and the positive effect of NSAIDs on immunity and against cancer progression in both animals and humans.2126 Indeed, cyclooxygenase-2 (COX-2) inhibitors are active in some models of breast cancer, and COX-2 could play a role in tumor development.24 In fact, overexpression of COX-2 in breast cancer leads to stimulation of epithelial cell proliferation, inhibition of apoptosis, stimulation of angiogenesis, immune suppression, and increases the production of mutagens.23 This favors breast tumor growth and increases the risk of cancer relapse. These effects are mostly mediated by prostaglandins and specifically prostaglandin E2 (PGE2).19,24 PGE2 is released by monocytes and dendritic cells to regulate the inflammatory cascade and profoundly depresses cellular antitumoral immunity, i.e., NK cell activity.24 The PGE2-induced suppression of NK cell activity is carried via membrane receptors that trigger the synthesis of cyclic adenosine monophosphate, which interferes with the cytotoxicity of NK cells. This suppression of NK cell activity dissipates quickly after removal of prostaglandins. The time course of postoperative immunosuppression extends from the first postoperative hours to a few days. As a consequence, the optimal window for administration of a drug that acts on NK cells may be very short, ideally just before surgery. This time-specific effect may explain the effect of intraoperative ketorolac and the absence of effect of postoperative diclofenac in our retrospective study.

Another explanation is a specific effect of ketorolac. Indeed, inhibition of COX-2 is necessary to counteract the deleterious effects of PGE2,25 but, although all NSAIDs act against the growth of tumors, they are probably not equivalent in their antitumor effects. Alternative targets, such as the tumor-associated reduced nicotinamide adenine dinucleotide oxydase (tNOX), are possibly involved in this anticancer effect. The existence of tNOX explains the fact that some cancer cell lines lacking COX-2 respond to certain NSAIDs but not to others, suggestive of additional COX-2–independent antitumor activities.26 Further studies are required to explore the differences between the NSAIDs and their antitumor effects in this context.

When analyzing the timing of cancer recurrence, the greatest difference between patients receiving ketorolac and those not receiving it appears between 9 and 18 months. This early cancer recurrence confirms our observation that intraoperative administration of ketorolac improves the control of residual neoplastic disease. Cancer relapse is a consequence of intraoperative spread of tumor cells in the bloodstream, local recurrence, or previously disseminated micrometastatic dormant cells.27 Inducing or maintaining dormancy in these neoplastic cells is a major goal for the adjuvant therapeutics, e.g., endocrine therapy. The control of residual disease limits the risk of early recurrence. It explains why recurrences occur gradually during the first 10 to 15 years in women treated with endocrine therapy, whereas in women not treated with endocrine therapy (i.e., with estrogen receptor-negative tumors), the majority of cancer relapses occur in the first 2 years.27

Our study is potentially limited by the retrospective design and the exploratory nature of our analyses. Uncontrolled and unrecognized biases are frequent in retrospective studies. All known variables that influence cancer outcome were analyzed, but this does not exclude unrecognized bias. Overfitting, i.e., having poor predictive performance or exaggerating minor fluctuations in the data, is also a well-known problem of prognostic factor analyses, and validation against an external database is always advisable. The collection of data was systematic and the investigators (PF, JV, BN, and MDK) used the same methodology. The quality of the follow-up (MB and J-PM) and the accessibility of data permitted us to ascertain all variables. Nevertheless, to confirm these data, prospective well-conducted and multicenter studies are required. Although never reported as immunosuppressive, we did not investigate the effect of acetaminophen, because it was prescribed systematically during the postoperative period.

In conclusion, our data support a positive effect of ketorolac administered preoperatively for breast cancer surgery. These data are consistent with previous animal and human preclinical studies that suggested an anticancer effect of preoperatively administered NSAIDs.


1. Shakhar G, Ben-Eliyahu S. Potential prophylactic measures against postoperative immunosuppression: could they reduce recurrence rates in oncological patients? Ann Surg Oncol 2003;10:972–92
2. Wrona D. Neural-immune interactions: an integrative view of the bidirectional relationship between the brain and immune systems. J Neuroimmunol 2006;172:38–58
3. Page GG. Immunologic effects of opioids in the presence or absence of pain. J Pain Symptom Manage 2005;29:S25–31
4. Exadaktylos AK, Buggy DJ, Moriarty DC, Mascha E, Sessler DI. Can anesthetic technique for primary breast cancer surgery affect recurrence or metastasis? Anesthesiology 2006;105: 660–4
5. Biki B, Mascha E, Moriarty DC, Fitzpatrick JM, Sessler DI, Buggy DJ. Anesthetic technique for radical prostatectomy surgery affects cancer recurrence: a retrospective analysis. Anesthesiology 2008;109:180–7
6. Rutgers EJ. Guidelines to assure quality in breast cancer surgery. Eur J Surg Oncol 2005;3:568–76
7. Birido N, Geraghty JG. Quality control in breast cancer surgery. Eur J Surg Oncol 2005;31:577–86
8. Goldhirsch A, Glick JH, Gelber RD, Coates AS, Thürlimann B, Senn HJ. Meeting highlights: international expert consensus on the primary therapy of early breast cancer 2005. Ann Oncol 2005;16:1569–83
9. Goldhirsch A, Wood WC, Gelber RD, Coates AS, Thürlimann B, Senn HJ. Ann Oncol 2007;18: 1133–44
10. Lee MC, Jasi R. Postmastectomy radiation therapy: indications and controversies. Surg Clin North Am 2007;87:511–26
11. Smith BD, Smith GL, Haffty BG. Postmastectomy radiation and mortality in women with T1–2 node-positive breast cancer. J Clin Oncol 2005;23:1409–19
12. Soerjomataram I, Louwman MW, Ribot JG, Roukema JA, Coebergh JW. An overview of prognostic factors for long term survivors of breast cancer. Breast Cancer Res Treat 2008; 107:309–30
13. Beilin B, Shavit Y, Hart J, Mordashov B, Cohn S, Notti I, Bessler H. Effects of anesthesia based on large versus small doses of fentanyl on natural killer cell cytotoxicity in the perioperative period. Anesth Analg 1996;82:492–7
14. Aubrun F, Valade N, Coriat P, Riou B. Predictive factors of severe postoperative pain in the postanesthesia care unit. Anesth Analg 2008;106:1535–41
15. Romero-Sandoval A, Eisenach JC. Clonidine reduces hypersensitivy and alters the balance pro-and anti-inflammatory leukocytes after local injection at the site of inflammatory neuritis. Brain Behav Immun 2007;21:569–80
16. De Kock M, Lavand'homme P, Waterloos H. ‘Balanced analgesia’; in the perioperative period: is there a place for ketamine? Pain 2001;92:373–80
17. Bentley MW, Stas JM, Johnson JM, Viet BC, Garett N. Effects of preincisional ketamine treatment in natural killer cell activity and postoperative pain management after oral maxillofacial surgery. AANA J 2005;73:427–36
18. Mazar J, Rogachev B, Shaked G, Ziv NY, Czeiger D, Chaimovitz C, Zlotnik M, Mukmenev I, Byk G, Douvdevani A. Involvement of adenosine in the antiiflammatory action of ketamine. Anesthesiology 2005;102:1174–81
19. Waymack JP, Flescher E, Venkatraman J, Fernandes G, Guzman RF, Yurt RW, Mason AD Jr, Pruitt BA Jr. Effect of prostaglandin E in multiple experimental models. VIII. Effect on host response to metastatic tumor. J Surg Oncol 1991;48:239–45
20. Okuno K, Jinnai H, Lee YS, Nakamura K, Hirohata T, Shigeoka H, Yasutomi M. A high level of prostaglandin E2 (PGE2) in the portal vein suppresses liver-associated immunity and promotes liver metastases. Surg Today 1995;25:954–8
21. Benish M, Bartal I, Goldfarb Y, Levi B, Avraham R, Raz A, Ben-Eliyahu S. Perioperative use of beta-blockers and COX-2 inhibitors may improve immune competence and reduce the risk of tumor metastasis. Ann Surg Oncol 2008;15:2042–52
22. Melamed R, Rosenne E, Shakhar K, Schwartz Y, Abudarham N, Ben-Eliyahu S. Marginating pulmonary-NK activity and resistance to experimental tumor metastasis: suppression by surgery and the prophylactic use of a beta-adrenergic antagonist and a prostaglandin synthesis inhibitor. Brain Behav Immun 2005;19:114–26
23. Singh B, Lucci A. role of cyclooxygenase-2 in breast cancer. J Surg Res 2002;108:173–9
24. Yakar I, Melamed R, Shakhar G, Rosenne E, Abudarham N, page GG, Ben-Eliyahu S. Prostaglandin e(2) suppresses NK activity in vivo and promotes postoperative tumor metastasis in rats. Ann Surg Oncol 2003;10:469–79
25. Stanojeviæ-Bakiæ N, Vuckoviæ-Dekiæ L, Radomiroviæ S, Juraniæ Z, Jovanoviæ N. The influence of surgery and anesthesia on lymphocyte functions in breast cancer patients: in vitro effects of indomethacin. Neoplasma 1999;46:54–60
26. Morré DJ, Morre DM. tNOX, an alternative target to COX-2 to explain the anticancer activities of non-steroidal anti-inflammatory drugs (NSAIDS). Mol Cell Biochem 2006;283: 159–67
27. Mansell J, Monipenny IJ, Skene AI, Abram P, Carpenter R, Gattuso JM, Wilson CR, Angerson WJ, Doughty JC. Patterns and predictors of early recurrence in postmenopausal women with estrogen receptor-positive early breast cancer. Breast cancer Res Treat 2009;117:91–8
© 2010 International Anesthesia Research Society