Can regional anaesthesia for lymph-node dissection improve the prognosis in malignant melanoma?

Transcription

1 British Journal of Anaesthesia 109 (2): (2012) Advance Access publication 15 June doi: /bja/aes176 REGIONAL ANAESTHESIA Can regional for lymph-node dissection improve the prognosis in malignant melanoma? A. Gottschalk 1 *, G. Brodner 2, H. K. Van Aken 1, B. Ellger 1, S. Althaus 1 and H.-J. Schulze 3 1 Department of Anaesthesiology, Intensive Care and Pain Medicine, University Hospital Münster, Münster, Germany 2 Department of Anaesthesiology, Intensive Care Medicine and Pain Therapy and 3 Department of Dermatology, Hornheide Specialist Hospital, Münster, Germany * Corresponding author: Department of Anaesthesiology, Intensive Care and Pain Medicine, University of Münster, Albert-Schweitzer- Campus 1, Building A1, Münster, Germany. antjegottschalk@gmx.net Editor s key points Understanding the impact of anaesthetic technique on cancer survival is important. There is some evidence that regional techniques may improve long-term outcomes after cancer surgery. This retrospective study of malignant melanoma patients found a non-significant trend for improved survival after spinal. While there are limitations in a retrospective study, this important area clearly warrants further research. Background. Optimized anaesthetic management might improve the outcome after cancer surgery. A retrospective analysis was performed to assess the association between spinal (SpA) or general (GA) and survival in patients undergoing surgery for malignant melanoma (MM). Methods. Records for 275 patients who required SpA or GA for inguinal lymph-node dissection after primary MM in the lower extremity between 1998 and 2005 were reviewed. The follow-up ended in Survival was calculated as days from surgery to the date of death or last patient contact. The primary endpoint was mortality during a 10 yr observation period. Results. Of 273 patients included, 52 received SpA and 221 GA, either as balanced (sevoflurane/sufentanil, n¼118) or as total i.v. (propofol/ remifentanil, n¼103). The mean follow-up period was 52.2 (SD 35.69) months after operation. Significant effects on cumulative survival were observed for gender, ASA status, tumour size, and type of surgery (P¼0.000). After matched-pairs adjustment, no differences in these variables were found between patients with SpA and GA. A trend towards a better cumulative survival rate for patients with SpA was demonstrated [mean survival (months), SpA: 95.9, 95% confidence interval (CI), ; GA: 70.4, 95% CI, ; P¼0.087]. Further analysis comparing SpA with the subgroup of balanced volatile GA confirmed this trend [mean survival (months), SpA: 95.9, 95% CI, ; volatile balanced : 68.5, 95% CI, , P¼0.081]. Conclusions. These data suggest an association between anaesthetic technique and cancer outcome in MM patients after lymph-node dissection. Prospective controlled trials on this topic are warranted. Keywords:, general;, spinal; melanoma; surgery; survival Accepted for publication: 5 March 2012 Cancer is a heterogeneous disease. In most cases, surgeons can successfully remove the primary tumour, but a large proportion of cancer-related deaths are due to the development of metastases rather than directly related to the primary cancer. Potentially, curative surgery may contribute to metastatic spread: it suppresses the immune system, facilitates the growth of pre-existing micrometastases, and allows malignant cells to disseminate during tumour manipulation. 12 Preventing immunosuppression during the immediate postoperative period might be extremely important, as there is a high risk of tumour cells spreading during this period. 3 Both cell-mediated and humoural immune responses are adversely affected by general (GA) and surgical trauma, and this may result in tumour progression. In vitro studies have shown dose-dependent alterations in neutrophil, monocyte, and lymphocyte functions after exposure to different anaesthetics. 3 (SpA), which for most operations in the lower body provides very good analgesia, attenuates the surgical stress response and prevents inhibition of the immune These authors contributed equally to this paper. This article is accompanied by Editorial III. & The Author [2012]. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved. For Permissions, please journals.permissions@oup.com

2 BJA Gottschalk et al. system. In an animal model, it has been shown that innate tumour immunity is impaired through inhibition of the cytotoxic Th1 response of hepatic mononuclear cells after lower abdominal surgery. block attenuates this impairment, thereby inhibiting the promotion of metastatic spread after surgery. 45 A survival analysis of patients with malignant melanoma (MM) showed an improved survival rate after primary excision of the malignant tumour under local in comparison with GA. 6 Sole use of local is, however, often not possible in various types of surgery. In addition, there have been no prospective clinical trials so far that have been adequately powered to assess long-term survival relative to different anaesthetic techniques, and no definitive conclusions can therefore be drawn. The prognostic impact of the type of used for excision of malignant tumours is thus currently a matter of controversy. The present retrospective study was therefore carried out to allow improved hypothesis formation regarding the long-term effects of. The aim of the study was to assess the association between SpA or GA and cancer survival after lymph-node dissection in patients with MM. It was hypothesized that the use of SpA would improve the survival rate in these patients. Methods After ethical approval for the retrospective study had been received (from the ethical committee of the Medical Council of Westphalia-Lippe and from the Medical School of the University of Münster, Germany), records were reviewed for all patients who had required for inguinal lymph-node dissection after primary MM in the lower extremity between February 1998 and April 2005 at the Department of Anaesthesia, Intensive Care and Pain Medicine, Hornheide Hospital, Münster, Germany. The follow-up period ended in May A total of 353 surgical records were reviewed. The procedures included 247 sentinel lymph-node biopsies (SLNBs) and 101 complete lymph-node dissections (CLNDs), and the type of surgery could not be determined in five records. These five operations were excluded from further analyses. Seventy-three operations were identified as duplicates that is, with SLNB followed by CLND in the same patient. In these duplicates, the SLNB was excluded and only the CLND was used for analysis. Records for a total of 275 cases were thus analysed: 174 SLNBs (247 73) and 101 CLNDs. Two patients were excluded due to incomplete data sheets lacking important information, and 273 patients remained for the final analyses. Fifty-two patients received SpA and 221 underwent surgery with GA. GA was administered either as balanced GA (n¼118) or as total i.v. (n¼103) (Fig. 1). Hornheide Specialist Hospital is the main national cancer institute for the treatment of skin malignancies in Germany and is one of the major institutes for this type of cancer in Europe. The hospital incidence of MM amounted to more than 1000 patients in The centre has been certified in accordance with DIN EN ISO 9001/2000 as a centre for the treatment of skin cancer. The melanoma team consists of a combination of specialists, including surgeons, dermatologists, pathologists, medical oncologists, and also anaesthesiologists. The hospital s Department of Medical Documentation is responsible for maintaining the database of primary and followup information on patients tumour status. After the primary operation, patients receive annual follow-up examinations, which include an interview, physical examination, ultrasound of the lymph nodes, and if necessary a computed tomography scan. If patients do not appear for the annual follow-up appointment, their family or family practitioner is contacted to obtain information about the patient s health status. An electronic database (megamanager , 2008 version; megapharm Ltd, St Augustin, Germany) was used to determine baseline variables (for patient characteristic factors, tumour, physical condition, risk factors for survival), surgical variables, and follow-up variables in all patients treated for MM. The patients charts were also screened for additional information if necessary. The type of was recorded from the original charts. The patients underwent surgery either under SpA with ml hyperbaric bupivacaine 0.5% or GA. GA was administered as balanced volatile using isoflurane and sufentanil, or as total i.v. with continuous infusions of propofol and remifentanil. All data were transferred to the SPSS program for statistical analysis (IBM SPSS Statistics, PASW Predictive Analytics Software, version 18). The following potential confounders were tested in the statistical analyses to identify any significant differences between SpA and GA: Baseline variables: age, gender, BMI, smoking status, alcohol abuse, ASA status. Medical history: myocardial infarction, coronary artery disease, arrhythmias, hypertension, chronic obstructive pulmonary disease (COPD), diabetes, renal insufficiency, liver diseases, coagulopathies, and immunosuppression. Tumour stage: tumour thickness, tumour ulcerations. American Joint Committee on Cancer (AJCC) classification: Stage 0 melanoma involves the epidermis, but has not reached the underlying dermis melanoma in situ. Stage I melanoma is characterized by tumour thickness, presence and number of mitoses, and ulceration status. There is no evidence of regional lymph-node metastasis or distant metastasis. Stage II melanoma is characterized by tumour thickness and ulceration status. There is no evidence of regional lymph-node metastasis or distant metastasis. Stage III melanoma is characterized by the level of lymph-node metastasis. There is no evidence of distant metastasis. Stage IV melanoma is characterized by the location of distant metastases and the level of serum lactate dehydrogenase. Postoperative variables: signs of infection, use of antibiotics, seromas, repeat surgery. Long-term variables: additional malignancies. 254

3 Regional and cancer prognosis BJA Surgeries n = 353 Kind of surgery SLNB n = 247 CLND n = 101 No information n = 5 excluded Duplicates: SLNB followed by CLND n = 73 Primary cases: Only SLNB n = 174 excluded selected Number of cases selected n = 275 Complete data set n = 273 missing data n = 2 excluded Type of n = 52 Balanced volatile n = 118 n = 221 Total intravenous n = 103 Fig 1 Flow diagram presenting the enrolment, intervention allocation, and data analysis with numbers of patients in each group. SLNB, sentinel lymph-node biopsy; CLND, complete lymph-node dissection. Statistics The primary endpoint of the study was mortality during a postoperative observation period of 10 yr. Survival was calculated as the number of days from the date of surgery to the date of death or last contact with the patient. Nominal scale variables were described using relative and absolute frequencies. Variables with interval or higher-scale levels were described as means and standard deviation. Patient characteristic and physiological basic data were compared using Student s t-test, the x 2 test, or Fisher s test, as appropriate. Survival was analysed in two steps using the Kaplan Meier curves and subsequent log-rank tests. In the first step, the overall group was analysed for variables that might confound the effects of on survival. The following variables were included: type of surgery, tumour thickness, gender, and ASA status. In the second step using these variables, matched pairs of patients with SpA and GA were built for further statistical analyses. This was due to the small study group of only 52 patients in the SpA subgroup. The Kaplan Meier analysis and subsequent log-rank tests were done to compare the survival rate of patients in the two groups. The following alternative hypothesis was tested in the matched-pairs analysis: the long-term survival rate after inguinal lymph-node dissection in MM is different for patients undergoing surgery in SpA in comparison with GA (H 0 : m general ¼m spinal ; H 1 : m general =m spinal ). To reduce the risk of type I error, no direction for the difference was defined, statistical tests were two-tailed, and a of 0.05 was regarded as significant. Results A total of 273 patients who required for inguinal lymph-node dissection after primary MM in the lower extremity between February 1998 and April 2005 at the Department of Anaesthesia, Intensive Care and Pain Medicine, Hornheide Specialist Hospital, Münster, Germany, were included in this analysis. The follow-up ended in May The mean follow-up 255

4 BJA Gottschalk et al. time for all patients was 52.2 (SD 35.69) months (with a minimum of 0 months for patients who died within the first month after operation, and a maximum of 122 months for patients who survived for the whole study period). After testing for confounders of the effects of the type of on long-term survival, x 2 analyses comparing the SpA group (n¼52) and the GA group (n¼221) did show a significant difference with regard to gender (P¼0.021), ASA status (P¼0.004), arrhythmias in the medical history (P¼0.033), or coagulopathies (P¼0.008). There were no significant differences with regard to tumour stage, postoperative surgical complications, or long-term variables (Tables 1 3). Patients in the GA group were significantly younger than those in the SpA group (P¼0.003). In addition, the duration of surgery (P¼0.026) and hospital stay (P¼0.007) were significantly longer for patients in the GA group in comparison with the SpA group (Tables 1 and 3). The Kaplan Meier analysis showed that patients with SLNBs had a significantly longer survival in comparison with patients with CLNDs (P¼0.000). Patients with a higher ASA class had a significantly higher risk for a shorter survival after surgery, as did male patients and patients with a greater tumour thickness (all P¼0.000). Using these variables (type of surgery, ASA status, tumour thickness, gender), matched pairs of the 52 patients with SpA and 52 corresponding patients with GA were built and tested once again for the following variables: age, gender, BMI, smoking, alcohol abuse, ASA status, medical history (myocardial infarction, coronary artery disease, arrhythmias, hypertension, diabetes, COPD, renal insufficiency, liver diseases, coagulopathies, immunosuppression), tumour stage (tumour thickness, tumour ulcerations, AJCC classification), postoperative variables such as signs of infection, use of antibiotics, seromas, and repeat surgery. No significant differences between the groups now remained. However, differences in the duration of surgery [SpA min (SD 20.59), GA min (SD 35.77) (P¼0.039)] and length of hospital stay [SpA days (SD 9.91), GA (SD 12.09) (P¼0.013)] remained significant. A trend towards longer survival in patients in the SpA group was now detected [mean survival time for SpA: 95.9 months; 95% confidence interval (CI), months; GA: 70.4 months; 95% CI, months; P¼0.087] (Fig. 2). Discussion The results of this retrospective study of patients undergoing lymph-node dissection after MM are in accordance with the hypothesis that surgery with SpA might improve the longterm prognosis in patients with MM. Animal studies have indicated that immune-response control over the circulation of tumour cells and micrometastases takes place mainly through cell-mediated immunity. The main cell types involved are cytotoxic T lymphocytes, natural killer (NK) cells, NK T cells, dendritic cells, and macrophages. 7 NK cells are particularly important, as they can spontaneously recognize and kill malignant cells. NK cell suppression is also associated with increased rates of metastasis. Studies in humans have also shown that low perioperative levels of NK cell activity are associated with increased cancer-related morbidity and mortality. 89 The specific approach used is likely to be of relevance, as animal studies have shown that the choice of anaesthetic drugs and techniques has a profound influence on the immune response and, as a result, on cancer metastasis. 3 In particular, regional, including SpA, reduces the stress response caused by surgery, which is believed to be a mediator of postoperative immunosuppression This prevents noxious afferent input from reaching the central nervous system. In addition, SpA avoids inhaled anaesthetics and reduces the opioid requirements, both of which have been shown to decrease the activity of NK cells. SpA also effectively blunts the neuroendocrine response and thus decreases the production of catecholamines, which reduce NK cell activity. 3 In a mouse model, for example, it has been shown that laparotomy during sevoflurane significantly increased the number of liver metastases in comparison with sevoflurane plus SpA. 4 The addition of intrathecal local anaesthetics attenuated the suppression of tumoricidal function in hepatic mononuclear cells and thereby reduced tumour metastasis. 4 Data from retrospective studies in humans have demonstrated that the long-term outcome for patients undergoing cancer surgery is improved if they receive a neuraxial or regional block. Exadaktylos and colleagues 13 reported that the use of paravertebral nerve block in combination with GA was associated with a longer cancer-free interval and a lower incidence of recurrence in patients with breast cancer. Another retrospective trial was able to show that the epidural technique was associated with a 65% reduction in biochemical recurrences of prostate cancer, as defined by increased prostate-specific antigen after operation. 14 In contrast to these results, the use of epidural analgesia for perioperative pain control during colorectal cancer surgery was not found to be associated with a decreased rate of cancer recurrence; a potential benefit was observed in older patients. 15 These findings suggest that the benefit of regional relative to cancer recurrence, if it exists, may depend critically on the specific cancer type. However, no clinical trials have been conducted to investigate the effects of SpA on the long-term outcome after cancer surgery. The data obtained in the present study may suggest that SpA perioperatively may improve the long-term prognosis for patients with MM. However, only prospective randomized trials can fully address the relationship between regional analgesia and cancer recurrence. In principle, two different models of cancer are at issue here. On the one hand, the existence of a metastatic cascade has been proposed, with a number of sequential events being required for disseminated cancer to develop. 16 The idea behind this concept is that tumour cells migrate through the lymphatic vessels to the lymph nodes and then disseminate to distant sites. On the other hand, 256

5 Regional and cancer prognosis BJA Table 1 Patient characteristic data, type of used, and preoperative physical condition Gender Male 24 (46.2) 64 (29.0) Female 28 (53.8) 157 (71.0) Age (yr) mean (range (32/86) (4/93) min/max) ASA classification I 7 (13.5) 57 (25.8) II 34 (65.4) 148 (67.0) III 10 (19.2) 16 (7.2) IV 1 (1.9) 0 (0) Coronary artery disease No history 45 (86.5) 211 (95.5) Minimal risk 3 (5.8) 5 (2.3) High risk 4 (7.7) 5 (2.3) History of myocardial infarction No 51 (98.1) 215 (97.3).6 months 1 (1.9) 6 (2.7),6 months 0 (0) 0 (0) Arrhythmia No history 49 (94.2) 217 (98.2) Bradycardia 0 (0) 1 (0.5) Pacemaker 0 (0) 2 (0.9) Ventricular 1 (1.9) 1 (0.5) arrhythmia Atrial fibrillation 2 (3.8) 0 (0) Hypertension No 37 (71.2) 169 (76.5) Yes 15 (28.8) 52 (23.5) Coagulopathy No history 50 (96.2) 216 (97.7) Over 6 months 0 (0) 5 (2.3) previously Within the last 6 2 (3.8) 0 (0) months COPD No 48 (92.3) 208 (94.1) Dyspnoea during 4 (7.7) 11 (5.0) stress Dyspnoea during 0 (0) 2 (0.9) rest Renal failure No 49 (94.2) 218 (98.6) Compensated 3 (5.8) 3 (1.4) retention Liver failure 1.00 No 52 (100) 220 (99.5) Yes 0 (0) 1 (0.5) Diabetes mellitus No 50 (96.2) 210 (95.0) Continued Table 1 Continued.6 months 2 (3.8) 8 (3.6),6 months 0 (0) 3 (1.4) Immunosuppression No 52 (100) 218 (98.6).6 months 0 (0) 3 (1.4),6 months 0 (0) 0 (0) Smoking No 44 (84.6) 172 (77.8) Yes 8 (15.4) 49 (22.2) Alcohol consumption 0.179,1 week 45 (86.5) 204 (92.3) Weekend 1 (1.9) 8 (3.6).1 week 1 (1.9) 2 (0.9) Daily 5 (9.6) 7 (3.2) Table 2 Oncological data Tumour thickness,2 mm 25 (48.1) 105 (47.5) 2 mm 27 (51.9) 116 (52.5) Tumour ulcerations No 33 (63.5) 156 (70.6) Yes 19 (36.5) 65 (29.4) AJCC classification No data 0 (0) 1 (0.4) IA 1 (1.9) 6 (2.7) IB 19 (36.5) 70 (31.7) IIA 8 (15.4) 39 (17.6) IIB 11 (21.2) 22 (10.0) IIC 2 (3.8) 8 (3.6) IIIA 5 (9.6) 27 (12.2) IIIB 3 (5.8) 26 (11.8) IIIC 3 (5.8) 22 (10.0) IV 0 (0) 0 (0) Additional malignancies No 52 (100) 217 (98.2) Yes 0 (0) 4 (1.8) Paget 17 proposed the seed and soil hypothesis in order to explain the unusual organ-specific metastatic pattern. According to the seed and soil theory, which appears to reflect clinical reality better, metastases even from small 257

6 BJA Gottschalk et al. Table 3 Surgical information Duration of (20.59) (29.57) surgery (min) (SD) Postoperative status Wound infection No 44 (84.6) 182 (82.4) Yes 8 (15.4) 39 (17.6) Use of antibiotics No 42 (80.8) 156 (70.6) Yes 10 (19.2) 65 (29.4) Seromas No 47 (90.4) 184 (83.3) Yes 5 (9.6) 37 (16.7) Repeat surgery No 51 (98.1) 217 (98.2) Yes 1 (1.9) 4 (1.8) Hospital stay (days) (SD) (9.91) (11.58) tumours may be initiated before the primary tumour is diagnosed. The growth of the primary tumour and the process of metastasis may thus be two autonomous processes. In that case, perioperative compromise of the immune system might be fatal, as it might promote the growth of micrometastases in distant organs. Limitations of the study Potential weaknesses of this study, which is retrospective in nature, might be inaccuracies in the written records, with important data missing, and the difficulty of controlling bias and confounders. For example, differences in surgical time and hospital stay might be caused by confounders that were not recorded. The results of this study have therefore been used to generate hypotheses and will require confirmation with further prospective studies. To minimize the risk for a-error, an alternative hypothesis for a two-sided test was selected for statistical analyses. To compensate for this, a matched case control study design was chosen to adjust for potential confounders and to increase the precision of the comparison. The poor prognosis in patients with MM, particularly with positive lymph nodes, may also have masked true benefits of regional. However, this is the largest study to date providing support for the view that the use of SpA in patients Type of anasthesia 1: spinal 2: general 1: censored 2: censored Cumulative survival Months Fig 2 Kaplan Meier survival curve after adjustment of matched-pairs samples: SpA vs GA. 258

7 Regional and cancer prognosis BJA presenting with MM and undergoing inguinal lymph-node dissection may improve the long-term outcome. The study is not capable of revealing any potential mechanism of SpA relative to the tumour outcome in patients with MM. Long-term prospective studies will be required before it can be determined whether the choice of technique for cancer surgery has a significant impact on patient safety in the long run. Since 90% of cancer-related deaths are due to metastatic development rather than directly related to the primary cancer, there is a considerable potential for improving the patient outcome. 20 A prospective clinical study in patients with MMs is therefore now in progress to help to clarify a number of questions, including whether SpA alters patients perioperative immune status, which cells are affected, and how SpA impacts on the immune tumour interaction. Additionally, the impact of this on longterm survival and risk of cancer recurrence will be studied. The data presented here suggest that after lymph-node dissection, there may be an association between the anaesthetic technique used and the cancer outcome in patients with MM. A prospective randomized controlled trial on this topic is therefore warranted. Declaration of interest None declared. Funding The study was supported solely by departmental funding. References 1 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: Ben-Eliyahu S. The promotion of tumor metastasis by surgery and stress: immunological basis and implications for psychoneuroimmunology. Brain Behav Immun 2003; 17 (Suppl. 1): S Gottschalk A, Sharma S, Ford J, Durieux ME, Tiouririne M. Review article: the role of the perioperative period in recurrence after cancer surgery. Anesth Analg 2010; 110: Wada H, Seki S, Takahashi T, et al. Combined spinal and general anesthesia attenuates liver metastasis by preserving TH1/TH2 cytokine balance. Anesthesiology 2007; 106: Bar-Yosef S, Melamed R, Page GG, Shakhar G, Shakhar K, Ben-Eliyahu S. Attenuation of the tumor-promoting effect of surgery by spinal blockade in rats. Anesthesiology 2001; 94: Schlagenhauff B, Ellwanger U, Breuninger H, Stroebel W, Rassner G, Garbe C. Prognostic impact of the type of used during the excision of primary cutaneous melanoma. Melanoma Res 2000; 10: Smyth MJ, Godfrey DI, Trapani JA. A fresh look at tumor immunosurveillance and immunotherapy. Nat Immunol 2001; 2: Takeuchi H, Maehara Y, Tokunaga E, Koga T, Kakeji Y, Sugimachi K. Prognostic significance of natural killer cell activity in patients with gastric carcinoma: a multivariate analysis. Am J Gastroenterol 2001; 96: Brittenden J, Heys SD, Ross J, Eremin O. Natural killer cells and cancer. Cancer 1996; 77: Freise H, Van Aken HK. Risks and benefits of thoracic epidural. Br J Anaesth 2011; 107: Liu SS, Block BM, Wu CL. Effects of perioperative central neuraxial analgesia on outcome after coronary artery bypass surgery: a meta-analysis. Anesthesiology 2004; 101: Tonnesen E, Wahlgreen C. Influence of extradural and general on natural killer cell activity and lymphocyte subpopulations in patients undergoing hysterectomy. Br J Anaesth 1988; 60: 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: 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: Gottschalk A, Ford JG, Regelin CC, et al. Association between epidural analgesia and cancer recurrence after colorectal cancer surgery. Anesthesiology 2010; 113: Bross ID, Viadana E, Pickren JW. The metastatic spread of myeloma and leukemias in men. Virchows Arch A Pathol Anat Histol 1975; 365: Paget S. The distribution of secondary growths in cancer of the breast. Cancer Metastasis Rev 1889; 8: Sessler DI. Does regional analgesia reduce the risk of cancer recurrence? A hypothesis. Eur J Cancer Prev 2008; 17: Balch CM, Soong SJ, Gershenwald JE, et al. Prognostic factors analysis of melanoma patients: validation of the American Joint Committee on Cancer melanoma staging system. J Clin Oncol 2001; 19: Gupta GP, Massague J. Cancer metastasis: building a framework. Cell 2006; 127: