De Oliveira, Gildasio S. Jr MD, MSCI*†; McCarthy, Robert PharmD*; Turan, Alparsalan MD†; Schink, Julian C. MD‡; Fitzgerald, Paul C. MS*; Sessler, Daniel I. MD†
At the time of diagnosis, even early diagnosis, cancer cells are usually circulating in the bloodstream.1 Many steps are required for a circulating cancer cell to become a clinical metastasis,2 but immune defenses are certainly a critical protection.3 The most important of these is probably tumor cytotoxicity by natural killer cells.4 The perioperative period may be decisive since cancer surgery may augment release of cancer cells into the systemic circulation.5 Furthermore, the stress of surgery, volatile anesthetics, and opioids all impair natural killer cell function.6–8 Laboratory evidence suggests that perioperative impairment of immune function increases the risk of cancer recurrence in patients having oncologic surgery.9,10 In contrast, it is not defined whether perioperative clinical interventions can have an effect on cancer outcomes.
Perioperative dexamethasone administration decreases postoperative nausea and vomiting, reduces postoperative pain, and improves patients’ emotions and physical independence after surgery.11,12 However, dexamethasone, similar to other corticosteroids, causes generalized immunosuppression.13 In addition, dexamethasone specifically depresses natural killer cell function.14,15 Perioperative administration of dexamethasone may therefore augment tumor recurrence after curative cancer surgery. However, it remains unknown whether typical perioperative dexamethasone doses are associated with cancer recurrence.
Our primary objective was thus to examine the association between systemic dexamethasone administration and tumor recurrence in patients having primary cytoreductive surgery for ovarian cancer. Specifically, we tested the hypothesis that women given perioperative dexamethasone have earlier propensity-matched adjusted time to cancer recurrence than women who were not given the steroid.
The study protocol was approved by the Robert L. Lurie Cancer Center of Northwestern University and the Northwestern University IRBs. Informed consent was waived by the IRBs. Women who had primary ovarian cytoreductive surgery between January 1997 and October 2007 were identified using a database maintained by the division of Gynecologic Oncology at Northwestern University. Inpatient medical records and records from surgeons’ offices were evaluated, and women who had optimal cytoreductive or “tumor debulking” surgery (residual tumor <1 cm) were included. Those with a benign pathology or peritoneal tumor were excluded. After the inclusion and exclusion criteria, all subjects in the database who qualified were included in the study.
The following data were extracted from patients’ medical records: demographic characteristics (age, height, and weight), medical history, ASA physical status, surgical duration, surgeon performing the procedure, estimated blood loss, tumor staging using the International Federation of Gynecologists & Obstetricians (FIGO) staging classification (stage I, limited to 1 or both ovaries; stage II, presence of pelvic extensions or implants; stage III, microscopic peritoneal implants outside the pelvis or limited to the pelvis with extension to the small bowel or omentum; stage IV, distant metastases to the liver or outside the peritoneal cavity) tumor grading (well or poorly differentiated), tumor’s cell type, presence or absence of transfusion, number of units transfused, time to recurrence and survival. We also recorded whether dexamethasone was given in the perioperative period, and the dose.
All women were given postoperative chemotherapy, usually starting the second or third postoperative week. Most had 6 to 8 cycles of chemotherapy with carboplatin and paclitaxel. Radiation therapy was rare and reserved for salvage therapy in patients who had persistent or progressive disease after surgery and chemotherapy. The numbers of cycles of carboplatin/paclitaxel and any other chemotherapy received were recorded.
Patients were generally seen every 3 months for the first 2 postoperative years, every 6 months for the next 2 years, and then yearly. At each visit, patients had a complete history and physical examination and blood analysis for carcinoantigen 125 Computerized tomography scans of the abdomen and pelvis were obtained yearly. The follow-up period for the study ended April 2011 with the interval of follow-up varying from 4 to 10 years.
Our primary outcome was propensity-matched time to cancer recurrence.Time to recurrence was defined as the time in months from the surgery to a carcinoantigen 125 >21 U/mL and/or a computerized tomography scan with evidence of the disease followed by tissue confirmation. Data were compared between women who were and were not given perioperative systemic dexamethasone. Dexamethasone was commonly administered to prevent postoperative nausea and vomiting. Since no hospital guidelines to prevent postoperative nausea and vomiting were in place and subjects were high risk for postoperative nausea and vomiting, provider preference was the driving force for administration of the drug.
The Shapiro-Wilk test was used to evaluate deviation from a normal distribution. Subject and surgical characteristics were evaluated using the Mann-Whitney U tests. Categorical variables were evaluated using Fisher exact test. We used propensity scores to reduce the effect of covariate bias on the primary outcome that had been previously demonstrated to affect ovarian cancer recurrence times including tumor staging, histological grade, intraoperative neuraxial anesthesia, and perioperative transfusion.16,17 The year of surgery and the use of postoperative dexamethasone were also included in the analysis. The propensity score was the conditional probability for women who were and were not given perioperative systemic dexamethasone as a function of the above predetermined covariates, added into a multiple logistic regression. Continuous variables were dichotomized by analyzing their discriminant properties after constructing receiver operator curves for cancer recurrence. Individualized propensity scores derived from the logistic regression were determined.
Using the estimated propensity scores, a 1-to-1 matched analysis (nearest neighbor with caliber matching) was followed by a random selection of a woman given dexamethasone with one who was not among the closest estimated propensity score matches. The pair was eligible for matching if the caliper width of the pair was within 0.6 standard deviation (SD) of all selected pairs. Cochran and Rubin suggested that a caliper width of 0.6 SD will remove approximately 90% of the bias in observed confounders.18 Women who did not have an acceptable range of match were excluded. Kaplan-Meier curves were then constructed, and the log-rank test was used to evaluate tumor recurrence between propensity-matched groups. Women with no recurrence of ovarian cancer were right censored at the last day they were known to be recurrence free or alive at the end of the follow-up period for the study. Women lost to follow-up or who died of unrelated causes were also right censored at the last completed appointment. Median difference and 95% confidence interval between the propensity-matched groups were calculated using a 10,000 sample bootstrap method.
All tests were 2-sided; P < 0.05 was statistically significant. Data were analyzed using STATA version 11 (College Station, TX).
Among 346 women with a diagnosis of ovarian cancer who had primary cytoreductive surgery between January 1997 and December 2007, 260 met our inclusion criteria. Eighty-six women were excluded for the following reasons: 45 had benign or inconclusive pathology, 24 had tumor histology other than ovarian cancer (primary peritoneal, uterine, fallopian tube, and germ cell), 12 were secondary surgical procedures, and 5 were suboptimally debulked. The median (IQR) follow-up time for the sample was 48 (24–75) months. Three patients received radiation therapy after surgery, 1 in the dexamethasone and 2 in the control group.
One hundred two women were given intraoperative IV dexamethasone. Seventy-two were given 4 to 5 mg, and 30 were given 8 to 10 mg. Demographic, morphometric, surgical, and tumor characteristics were not statistically different in women who were and were not given dexamethasone (Table 1). Ovarian cancer recurrence was documented in 178 subjects. The unadjusted median (IQR) time to recurrence in all subjects was 18 (7–50) months.
Eighty-seven women given dexamethasone were propensity matched to 87 who did not. Demographic, morphometric, surgical, and tumor characteristics were not statistically different in the groups (Table 2). In the propensity-matched groups, the median (IQR) time to recurrence in the dexamethasone group was 23 (6–46) compared with 18 (8–53) months in the control group (P = 0.63, Fig. 1). The median (95% confidence interval) difference of time to recurrence between the propensity adjusted control and dexamethasone group was 5 (−8 to 17) months.
In addition, there was also no significant difference in tumor recurrence in women given 8 to 10 mg dexamethasone compared with women not given the steroid, with median times to recurrence of 25 (8–49) and 17 (6–48) months; (P = 0.79), respectively.
The importance of perioperative host defense in limiting cancer recurrence is clear, and the immunosuppressive effects of steroids are beyond question. In distinct contrast to our hypothesis, there was nonetheless no association between intraoperative administration of 4 to 10 mg of systemic dexamethasone and ovarian cancer recurrence after primary cytoreductive surgery. Recurrence was not statistically different in the subgroup of women given 8 to 10 mg of dexamethasone. Our results thus suggest that doses of dexamethasone commonly used for prevention of postoperative nausea and vomiting do not increase the risk of recurrence in women having surgery for ovarian cancer and are probably safe in that patient population.
Several factors may contribute to lack of association between dexamethasone and cancer recurrence. First, dexamethasone has been shown to decrease the perioperative stress response, and the surgical stress response has a detrimental effect on perioperative immune function.19,20 Direct immune suppression by dexamethasone may thus be countered by its beneficial effect on the normal stress response to surgical tissue injury. Second, dexamethasone has proapoptotic properties which may have prosurvival effects in certain types of cancer.21–23 Last, it is possible that larger doses of dexamethasone than the commonly administered perioperative doses may be required to offset the aforementioned positive effects of dexamethasone on immune function to result in an overall detrimental effect.
We have previously demonstrated that intraoperative neuraxial anesthesia, but not isolated postoperative neuraxial analgesia, is associated with fewer recurrences in ovarian cancer patients.17 Similarly, Lin et al.24 found that neuraxial analgesia reduces recurrence in ovarian cancer patients. In contrast, other investigators have not identified a beneficial effect of regional anesthesia on cancer outcomes for colon cancer.25,26 The aforementioned studies suggested that the type of tumor may be responsible for different outcomes of the same perioperative intervention. The effects of dexamethasone on cancer recurrence may also depend on the type of cancer.
Our study was relatively small, with just 87 propensity-matched patients who were given perioperative dexamethasone and 87 who were not. The recurrence rate of ovarian cancer is high; we thus had 80% power to detect a 6-month difference between the groups. While our results do not even hint at a deleterious effect of dexamethasone on recurrence, even very small differences (1 month) would be clinically important since recurrences are frequently fatal.
The question of dexamethasone-related immunosuppresion affecting cancer outcomes has been examined by other investigators. Münstedt et al.27 demonstrated that dexamethasone given concurrently with chemotherapy did not have adverse effects on ovarian cancer outcomes but may, in fact, have protective effects on the bone marrow. In the current study, we adjusted for the use of dexamethasone as part of the postsurgical chemotherapy regimen by including it as a variable in our propensity-matched analysis. Recently, the use of perioperative dexamethasone was not associated with more frequent and severe wound infection in patients undergoing endometrial cancer surgery.28 Nevertheless, the lack of association between dexamethasone and poor cancer outcomes, observed in the current and previous studies, may be a function of the relatively small sample size.
Our study is only valid when interpreted within the context of its limitations. There may be inaccuracy in the estimated times to cancer recurrence because we did not collect information on missed appointments before the time when cancer reoccurrence was detected but have no reason to suggest that it was not homogenous between subjects that received dexamethasone and those who did not. There was a lack of standardization of the intraoperative and postoperative analgesic management of patients which precludes us from making a statement regarding the role of the opioid-sparing effects on angiogenesis and cancer outcomes. We cannot exclude the possibility of selection bias in our retrospective analysis. For example, clinicians may have withheld dexamethasone from “sicker” patients, and this practice could potentially be responsible for our inability to detect an association between perioperative dexamethasone and worse cancer outcomes. We attempted to minimize group imbalances and covariate bias by propensity-matching study groups. However, we could only compensate for known factors, and many potential confounders were unavailable to us. Only an adequately powered randomized trial will confirm or refute an effect of perioperative dexamethasone on recurrence of ovarian cancer.
In summary, considerable previous work indicates that dexamethasone is effective for prevention of postoperative nausea, vomiting, and pain. We could not find evidence for an association between perioperative administration of dexamethasone and ovarian cancer recurrence. Our results thus do not support avoiding low-dose perioperative dexamethasone in women having primary cytoreductive surgery for ovarian cancer.
Name: Gildasio S. De Oliveira Jr, MD, MSCI.
Contribution: This author participated in the design and conduct of the study, statistical analysis, and manuscript preparation.
Attestation: Gildasio De oliveira approved the final manuscript and attests to the integrity of the original data and the analysis reported in this manuscript.
Name: Robert McCarthy, PharmD.
Contribution: This author participated in the statistical analysis and manuscript preparation.
Attestation: Robert McCarthy attests the integrity of the data and the analysis reported. Robert McCarthy approved the final manuscript and is the archival author who is responsible for maintaining the study records.
Name: Alparsalan Turan, MD.
Contribution: This author helped to design the study and prepare the manuscript.
Attestation: Alparsalan Turan approved the final manuscript.
Name: Julian C. Schink, MD.
Contribution: This author prepared the manuscript and helped to conduct the study.
Attestation: Julian Schink approved the final manuscript.
Name: Paul C. Fitzgerald, MS.
Contribution: This author helped to conduct the study and prepare the manuscript.
Attestation: Paul Fitzgerald approved the final manuscript.
Name: Daniel I. Sessler, MD.
Contribution: This author helped to design the study and prepare the manuscript.
Attestation: Daniel I. Sessler approved the final manuscript.
This manuscript was handled by: Peter S. A. Glass, MB, ChB.
1. Lange PH, Hekmat K, Bosl G, Kennedy BJ, Fraley EE. Accelerated growth of testicular cancer after cytoreductive surgery. Cancer. 1980;45:1498–506
2. Baum M, Demicheli R, Hrushesky W, Retsky M. Does surgery unfavourably perturb the “natural history” of early breast cancer by accelerating the appearance of distant metastases? Eur J Cancer. 2005;41:508–15
3. Colacchio TA, Yeager MP, Hildebrandt LW. Perioperative immunomodulation in cancer surgery. Am J Surg. 1994;167:174–9
4. Vivier E, Ugolini S, Blaise D, Chabannon C, Brossay L. Targeting natural killer cells and natural killer T cells in cancer. Nat Rev Immunol. 2012;12:239–52
5. Lundy J. Anesthesia and surgery: a double-edged sword for the cancer patient. J Surg Oncol. 1980;14:61–5
6. Lai R, Peng Z, Chen D, Wang X, Xing W, Zeng W, Chen M. The effects of anesthetic technique on cancer recurrence in percutaneous radiofrequency ablation of small hepatocellular carcinoma. Anesth Analg. 2012;114:290–6
7. Colvin LA, Fallon MT, Buggy DJ. Cancer biology, analgesics, and anaesthetics: is there a link? Br J Anaesth. 2012;109:140–3
8. Fujioka N, Nguyen J, Chen C, Li Y, Pasrija T, Niehans G, Johnson KN, Gupta V, Kratzke RA, Gupta K. Morphine-induced epidermal growth factor pathway activation in non-small cell lung cancer. Anesth Analg. 2011;113:1353–64
9. Shakhar G, Abudarham N, Melamed R, Schwartz Y, Rosenne E, Ben-Eliyahu S. Amelioration of operation-induced suppression of marginating pulmonary NK activity using poly IC: a potential approach to reduce postoperative metastasis. Ann Surg Oncol. 2007;14:841–52
10. Heaney A, Buggy DJ. Can anaesthetic and analgesic techniques affect cancer recurrence or metastasis? Br J Anaesth. 2012;109:i17–28
11. De Oliveira GS Jr, Castro-Alves LJ, Ahmad S, Kendall MC, McCarthy RJ. Dexamethasone to prevent postoperative nausea and vomiting: an updated meta-analysis of randomized controlled trials. Anesth Analg. 2013;116:58–74
12. De Oliveira GS Jr, Ahmad S, Fitzgerald PC, Marcus RJ, Altman CS, Panjwani AS, McCarthy RJ. Dose ranging study on the effect of preoperative dexamethasone on postoperative quality of recovery and opioid consumption after ambulatory gynaecological surgery. Br J Anaesth. 2011;107:362–71
13. Kunicka JE, Talle MA, Denhardt GH, Brown M, Prince LA, Goldstein G. Immunosuppression by glucocorticoids: inhibition of production of multiple lymphokines by in vivo
administration of dexamethasone. Cell Immunol. 1993;149:39–49
14. Holbrook NJ, Cox WI, Horner HC. Direct suppression of natural killer activity in human peripheral blood leukocyte cultures by glucocorticoids and its modulation by interferon. Cancer Res. 1983;43:4019–25
15. Bush KA, Krukowski K, Eddy JL, Janusek LW, Mathews HL. Glucocorticoid receptor mediated suppression of natural killer cell activity: identification of associated deacetylase and corepressor molecules. Cell Immunol. 2012;275:80–9
16. De Oliveira GS Jr, Schink JC, Buoy C, Ahmad S, Fitzgerald PC, McCarthy RJ. The association between allogeneic perioperative blood transfusion on tumour recurrence and survival in patients with advanced ovarian cancer. Transfus Med. 2012;22:97–103
17. de Oliveira GS Jr, Ahmad S, Schink JC, Singh DK, Fitzgerald PC, McCarthy RJ. Intraoperative neuraxial anesthesia but not postoperative neuraxial analgesia is associated with increased relapse-free survival in ovarian cancer patients after primary cytoreductive surgery. Reg Anesth Pain Med. 2011;36:271–7
18. Gu XS, Rosenbaum PR. Comparison of multivariate matching methods: structures, distances, and algorithms. J Comput Graph Stat. 1993;2:405–20
19. Karaman K, Bostanci EB, Aksoy E, Ulas M, Yigit T, Erdemli MO, Ercin U, Bilgihan A, Saydam G, Akoglu M. Effects of dexamethasone and pheniramine hydrogen maleate on stress response in patients undergoing elective laparoscopic cholecystectomy. Am J Surg. 2013;205:213–9
20. Cardinale F, Mastrototaro MF, Cappiello A, Caffarelli C, Caimmi S, Barberi S, Bernardini R. Immunological modifications induced from products used during the perioperative period. Int J Immunopathol Pharmacol. 2011;24:S13–20
21. Liu S, Wang B, Li S, Zhou Y, An L, Wang Y, Lv H, Zhang G, Fang F, Liu Z, Han R, Jiang T, Kang X. Immune cell populations decrease during craniotomy under general anesthesia. Anesth Analg. 2011;113:572–7
22. Webb MS, Miller AL, Johnson BH, Fofanov Y, Li T, Wood TG, Thompson EB. Gene networks in glucocorticoid-evoked apoptosis of leukemic cells. J Steroid Biochem Mol Biol. 2003;85:183–93
23. Khan S, Lopez-Dee Z, Kumar R, Ling J. Activation of NFkB is a novel mechanism of pro-survival activity of glucocorticoids in breast cancer cells. Cancer Lett. 2013;337:90–5
24. Lin L, Liu C, Tan H, Ouyang H, Zhang Y, Zeng W. Anaesthetic technique may affect prognosis for ovarian serous adenocarcinoma: a retrospective analysis. Br J Anaesth. 2011;106:814–22
25. Day A, Smith R, Jourdan I, Fawcett W, Scott M, Rockall T. Retrospective analysis of the effect of postoperative analgesia on survival in patients after laparoscopic resection of colorectal cancer. Br J Anaesth. 2012;109:185–90
26. Gottschalk A, Ford JG, Regelin CC, You J, Mascha EJ, Sessler DI, Durieux ME, Nemergut EC. Association between epidural analgesia and cancer recurrence after colorectal cancer surgery. Anesthesiology. 2010;113:27–34
27. Münstedt K, Borces D, Bohlmann MK, Zygmunt M, von Georgi R. Glucocorticoid administration in antiemetic therapy: is it safe? Cancer. 2004;101:1696–702
28. Bolac CS, Wallace AH, Broadwater G, Havrilesky LJ, Habib AS. The impact of postoperative nausea and vomiting prophylaxis with dexamethasone on postoperative wound complications in patients undergoing laparotomy for endometrial cancer. Anesth Analg. 2013;116:1041–7