Complications and Mortality Following Surgery for Oral Cavity Cancer: Analysis of 408 Cases

Transcription

1 The Laryngoscope VC 2015 The American Laryngological, Rhinological and Otological Society, Inc. Complications and Mortality Following Surgery for Oral Cavity Cancer: Analysis of 408 Cases Zachary G. Schwam, BA; Julie A. Sosa, MD, MA; Sanziana Roman, MD; Benjamin L. Judson, MD Objectives: To analyze the postoperative complications and mortality for oral cavity cancers, their time course, and to identify modifiable risk factors associated with their occurrence. Study Design: Retrospective cohort study. Methods: Patients undergoing surgery for oral cavity cancer were identified in the American College of Surgeons National Surgical Quality Improvement Program Participant Use Data File ( ). Overall and disease-specific complication and mortality data were analyzed using chi-square and multivariate regression analysis. Results: There were 408 cases identified. The overall 30-day complication and mortality rates were 20.3% and 1.0%, respectively. The most common adverse events were reoperation (9.6%), infectious (6.6%), and respiratory (5.1%) complications. Twenty patients (4.9%) experienced postdischarge complications. Fifty-two percent of postdischarge wound dehiscences and 67% of postdischarge surgical-site infections occurred by postdischarge day 7, and 91% of all postdischarge complications occurred by postdischarge day 14. Smoking was independently associated with respiratory (odds ratio [OR] 3.59, P 5.008) and surgical site complications (OR 5.13, P 5.004). Neck dissection was independently associated with respiratory (OR 6.17, P 5.001), surgical site (OR 6.30, P 5.003), and infectious (OR 3.83, P 5.003) complications. Conclusion: Current smokers and those undergoing neck dissection are at high risk of postoperative complications after oral cavity cancer surgery. Less than 5% of patients experienced postdischarge complications, nearly all of which occurred by postdischarge day 14. Most early postdischarge complications occurred at the surgical site. In order to mitigate postdischarge complications and their sequelae, early clinical follow-up should be sought for high-risk patients. Key Words: Head and neck cancer, oral cavity cancer, NSQIP. Level of Evidence: 4. Laryngoscope, 125: , 2015 INTRODUCTION Oral cavity cancers (OCCs) account for roughly 3% of all cancers in the United States, with an estimated 27,000 new cases and 5,500 estimated deaths. 1 The most common histologic subtypes are squamous cell carcinoma (86.3%) and adenocarcinoma (5.9%). 2 Risk factors for developing OCC are smoking and excessive alcohol consumption, 3 which are estimated to cause 75% of cases. 4 Surgery is the primary treatment for OCC, with adjuvant radiation or chemoradiation reserved for certain high-risk pathologic features. 5 From the Section of Otolaryngology, Department of Surgery, Yale University School of Medicine (Z.G.S., B.L.J.), New Haven, Connecticut; the Section of Endocrine Surgery, Department of Surgery, Duke University School of Medicine (J.A.S., S.R.); the Duke Cancer Institute (J.A.S.); and the Duke Clinical Research Institute (J.A.S.), Durham, North Carolina, U.S.A. Editor s Note: This Manuscript was accepted for publication March 23, Abstract presented at the Triological Society Combined Sectional Meeting, San Diego, California, USA, January 22 24, Support: National Center for Advancing Translational Sciences Clinical and Translational Service Awards TL1 Medical Student Research Fellowship. The authors have no other funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Benjamin L. Judson, MD, Yale Otolaryngology, 333 Cedar St., PO Box New Haven, CT benjamin.judson@yale.edu DOI: /lary Readmission and mortality rates for OCC surgery have been reported, 6 but there remains a scarcity of data outside of single-institution case series describing postoperative and postdischarge complications (PDCs), despite many patients having short hospital courses. PDCs have been shown to comprise up to 42% of complications in the general surgery literature, 7 9 and postoperative complications are associated with reoperation and increased mortality in otolaryngology patients. 10 Complications may critically delay receipt of adjuvant therapy. 11,12 Therefore, characterizing these events is important in patients with OCC. Using a large administrative database, we hypothesized that patient demographics and clinical factors would influence morbidity and mortality in the 30-day postoperative period. We also sought to identify modifiable factors that could be used to predict postoperative and postdischarge complications. MATERIALS AND METHODS Patient Population Data were extracted from the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) Participant Use Data File (PUF) from 2005 to 2010, which includes more than 400 participating sites. The NSQIP has been found to be accurate in recording major postoperative complications when compared to claims data. 13 NSQIP is a prospective, risk-adjusted, validated, multicenter 1869

2 program maintained by the ACS. 14 Information is collected for 135 variables that include patient characteristics, preoperative comorbidities, intraoperative variables, and morbidity and mortality outcomes up to 30 days postoperatively. The ACS-NSQIP has been validated and is continuously audited. 15 Each patient is prospectively followed until postoperative day (POD) 30 by a trained surgical clinical nurse reviewer at each of the participating sites Cases were identified by codes from the International Classification of Diseases, 9th edition, selecting for oral cavity malignancies, namely , 144.0, 144.1, 144.8, 144.9, , 145.6, 145.8, and Cases were then screened by current procedural terminology (CPT) codes. The following codes were included in the analysis either as the principal procedure or as a concurrent, or other procedure: 21025, 21026, 21034, 21040, 21044, 21045, 21050, 21198, (bony excision), (radical tumor resection), (maxillectomy), (excision of lesion from floor of mouth), 40810, 40812, 40814, 40816, (excision of lesion from vestibule of mouth), 41110, 41112, 41114, 41120, 41130, 41135, 41140, 41145, (excision of tongue lesion, glossectomy with or without neck dissection [ND]), 41150, 41153, (glossectomy as part of composite procedure, with or without ND), 41825, 41827, (excision of dento-alveolar structures), 42104, 42106, 42107, 42120, 42140, 42160, (excision or destruction of lesion from the palate or uvula), (excision of parotid tumor), 42842, 42844, and (radical resection of retromolar trigone with or without flap reconstruction). Preoperative and Intraoperative Variables Definitions of variables included in the database are described in the ACS-NSQIP user guide. 18 Several modifications were made for analysis. Pulmonary comorbidity was defined as having dyspnea on exertion or at rest, a history of chronic obstructive pulmonary disease, or currently being ventilatordependent. Cardiovascular comorbidity was defined as a history of congestive heart failure, angina, myocardial infarction (MI), peripheral vascular disease, percutaneous coronary intervention, or cardiac surgery. Poor functional status prior to surgery was defined as being either partially or totally dependent on others for activities of daily living. Neck dissection (as a separate code or folded into another procedural code) was determined by the following CPT codes: (dissection of deep jugular nodes), 38700, (suprahyoid ND), 38720, 38724, 41135, 41155, and (includes complete lymphadenectomy, modified radical ND, radical ND, or ND in addition to another procedure). Cases in which ND was the only procedure performed were excluded from our analysis. Postoperative Outcomes Adverse events analyzed included surgical site, infectious, venous thromboembolism, respiratory, renal, and cardiovascular complications, as well as flap failure, reoperation, and death. With the exception of cardiovascular complications, classification of postoperative morbidity is defined as in Kazaure et al. 7 Cardiovascular complications are defined as either MI, cardiac arrest, or stroke. The NSQIP PUF records postoperative events for 30 days and is not restricted to in-hospital events. 18 Statistical Analysis Statistical analysis was performed using SPSS 22 (Chicago, Illinois) Cases with missing data for variables of interest (< 5%) were removed from analysis. For the calculation of simple summary statistics such as mean and median, extreme outliers were removed if they fell three times the interquartile distance away from the 75th percentile value. Univariate analysis was performed using analysis of variance for continuous variables and chi-square tests for categorical variables. Tests were 2-sided, with significance set at P.05. All variables with a significance value of P.10 from univariate analysis were entered into a multivariate logistic regression model, which was used to identify variables associated with postoperative morbidity and mortality. Because the NSQIP is a deidentified dataset, our study was granted exemption from our institutional review board. RESULTS A total of 408 cases were identified, with the most common primary sites being the tongue (40.7%), floor of mouth (21.8%), cheek mucosa (7.1%), retromolar area (6.1%), and hard palate (4.9%). A significant proportion of the study population was male, between the ages of 45 to 79, and current smokers within the last year (Table I). Neck dissection was performed in 38.2% of cases. The overall adverse event rate was 20.3%; the common complications were reoperation, infectious, and respiratory (Table II). Undergoing ND was associated with experiencing an adverse event on univariate analysis (32.7% vs. 12.7%, P <.001). In multivariate analysis, being a current smoker, significant recent weight loss, and undergoing ND were associated with multiple causes of postoperative morbidity (Table III). The only risk factor identified for postoperative mortality was steroid use for a chronic condition (odds ratio [OR] 56.44, 95% confidence interval [CI] , P 5.017). Flap failure was independently associated with ND (OR 16.97, 95% CI , P 5.014). No independent risk factors were identified for postoperative cardiovascular or renal complications. The overall median and mean length of stay were 3.0 and 4.8 days, respectively. Twenty patients (4.9%) experienced PDCs, and 17.3% of all complications occurred in the postdischarge (PD) period. A significant proportion of surgical site complications (42.1%), severe sepsis (20.0%), flap failure (22.2%), DVTs (33.3%), and death (25.0%) occurred PD. Twenty-one of 23 (91.3%) PDCs occurred by postdischarge day (PDD) 14. However, 52% of PD wound dehiscences and 67% of PD surgicalsite infections occurred by PDD 7. Most complications clustered between POD 2 10, although the majority of DVTs and progressive renal insufficiency occurred considerably later (median 22.0 and 22.5 days, respectively). Patients experiencing complications postoperatively spent a significantly longer time in the hospital on admission (mean 14.8 vs. 3.1 days, P <.001). DISCUSSION There is a paucity of multi-institutional data regarding risk factors for postoperative complications and PDCs following OCC surgery. This is the first study to characterize the time course of major complications following surgery for OCC. 1870

3 TABLE I. Patient Demographics and Clinical Characteristics. % (n)* Sex Male 60.0 (245) Female 40.0 (163) Age (years) (32) (177) (141) (58) Race Caucasian 83.8 (295) Black 10.2 (36) Hispanic 2.6 (9) Asian 3.4 (12) BMI < (90) (156) (136) (22) ASA classification (10) (121) (260) (17) Diabetes 11.8 (48) Current smoker 32.8 (134) Heavy alcohol use 9.1 (37) Pulmonary comorbidity 15.7 (64) Cardiovascular comorbidity 11.5 (47) Hypertension 52.5 (214) Prior stroke/tia 8.3 (34) Neoadjuvant chemotherapy 1.0 (4) Neoadjuvant XRT 2.0 (8) Chronic steroids 3.2 (13) Coagulopathy 2.7 (11) Poor functional status 4.2 (17) Significant weight loss prior to surgery 6.6 (27) Operation within prior 30 days 3.0 (12) *Numbers may not add to total due to missing data. ASA 5 American Society of Anesthesiologists; BMI 5 body mass index; TIA 5 transient ischemic attack; XRT 5radiotherapy. Awad et al. compared administrative and clinical registry data from 355 OCC patients from Memorial Sloan-Kettering, New York, New York. Their retrospective chart review stopped at discharge or up to 45 days, whichever occurred later. 13 They next compared their own records to what was recorded in the NSQIP for the same cohort, identifying 27% of patients, 33% of all complications, and 100% of all major complications. Our overall adverse event rate was 20.3%, which is significantly lower than their rate of 62%. This is likely due to their tracking several specialty-specific complications not found in the NSQIP, such as chyle leak, trismus, and orocutaneous fistula. In our study, ND emerged as an independent risk factor for several complications, and 33% of those undergoing ND experienced an adverse event as compared to Awad et al. s 34%. Neck dissection is typically performed for more advanced disease, 5 involves large incisions, prolongs operating time, and puts several critical structures at risk of injury, which may explain the high complication rates in these patients. In using the National Cancer Data Base to examine only 30-day readmission and mortality following surgery for oral cavity cancer in 21,681 patients, Luryi et al. found a readmission rate of 3.5% and mortality rate of 1.0%. 6 The 30-day mortality rate in our sample was 1.0%, which is in line with both Luryi et al. s finding and Awad et al. s finding of 0.8%. 13 Additionally, Luryi et al. linked male gender, ND, and stage T3 to unplanned readmission within 30 days postdischarge. 6 Although the NSQIP does not record clinical staging or readmission data, we found current smokers, recent significant weight loss, and history of stroke/transient ischemic attack to be at high risk of reoperation. Because these factors were also associated with surgical site and respiratory complications, it is possible that a post-hoc tracheostomy or stoma revision was required. Smoking has also been associated with flap failure, 19 which may necessitate reoperation. It is not known whether a return to the operating room led to readmission, which is increasingly being used as a marker of quality of care 20 and is also known to influence Medicare reimbursements. 21 This study showed that being a current smoker and recent significant weight loss were associated with surgical site complications. Cigarette smoking is a wellknown risk factor for impaired wound healing and surgical site infection. 25,26 In Schwartz et al. s NSQIP study of 2,063 laryngectomy patients, predictors of wound complications included prior radiation therapy, diabetes, and hypoalbuminemia. 27 In our cohort, significant weight loss is likely a marker of cancer-induced inflammation, which may be compounded by the lesion s TABLE II. Complications and Adverse Events Within 30 Days of Surgery. % (n)* Overall 20.3 (83) Reoperation 9.6 (39) Infectious 6.6 (27) Respiratory 5.1 (21) Surgical site 4.7 (19) Flap failure 2.2 (9) VTE 1.7 (7) Cardiovascular 1.5 (6) Renal 1.2 (5) Death 1.0 (4) *Numbers do not add to total due to some patients experiencing multiple complications. VTE 5 venous thromboembolism. 1871

4 TABLE III. Independent Risk Factors for Postoperative Events. Surgical Site Infectious VTE Respiratory Reoperation Risk Factor OR (95% CI) P OR (95% CI) P OR (95% CI) P OR (95% CI) P OR (95% CI) P Black 3.89 ( ).034 NS NS NS NS Current smoker 5.13 ( ).004 NS NS 3.59 ( ) ( ).002 Significant weight loss 4.59 ( ).043 NS NS 3.19 ( ).021 Poor functional status NS 6.73 ( ).013 NS 5.57 ( ).027 NS Chronic steroid use NS NS ( ).007 NS NS Operation in prior 30 days NS NS ( ).006 NS NS Prior stroke/tia NS NS NS 3.41 ( ) ( ).006 Neck Dissection 6.30 ( ) ( ).003 NS 6.17 ( ).001 NS CI 5 confidence interval; Hx 5 history of; NS 5 not significant; OR 5 odds ratio; TIA 5 transient ischemic attack; VTE 5 venous thromboembolism. location in the oral cavity. Elevated levels of inflammatory response markers such as C-reactive protein and interleukin-6 have been associated with weight loss in cancer patients, 28 and there is evidence that systemic inflammation 29 and significant weight loss 30 determine albumin concentration. Weight loss with physiologic impairment has been found to increase septic complications in surgical patients. 31 Our data reveal the 14 days after discharge to be a high-risk time for patients because nearly all PDCs occurred in that time frame. Many PD surgical-site complications such as superficial wound infections and wound dehiscences occurred by PDD 7. Chen et al. s analysis of NSQIP patients who underwent a range of otolaryngologic procedures found that 73% of PDCs occurred by PDD 14, and that the majority of PDCs were surgical site and infectious in nature. 10 The literature shows that early clinical follow-up is a potential method of reducing unintended PD encounters such as hospital readmission or trips to the emergency room; Jenks et al. reported that 50% of those readmitted for PDCs had not seen a physician in the 30-day period after discharge. 9 Tuggle et al. showed that a postdischarge visit with a physician could decrease rehospitalization rates in elderly patients undergoing thyroidectomy by 55%, 32 and Robles et al. demonstrated that early telephone communication with a nurse practitioner could decrease PD emergency room visits by 48%. 33 In interpreting the findings of this study, it is important to take into consideration several limitations. Many complications are procedure-specific, and patients in this cohort underwent a diverse range of primary and concurrent procedures. Similarly, important complications for head and neck cancer, such as specific cranial nerve palsies, chyle leak, trismus, orocutaneous fistula, hoarseness, stridor, and tracheostomy-specific complications are not reported by the NSQIP. Unfortunately, NSQIP does not record data on time from diagnosis to treatment or hospital or surgeon volume, which have been correlated with postoperative complication rates in the literature Additionally, NSQIP does not provide information regarding the rationale for reoperation or its timing. CONCLUSION Current smokers and those undergoing ND are at high risk of postoperative complications after OCC surgery. Less than 5% of patients experienced PDCs, nearly all of which occurred by PDD 14. Most early PDCs occurred at the surgical site. In order to mitigate PDCs and possible sequelae such as reoperation, readmission, or death, early PD clinical follow-up should be sought for high-risk patients. Acknowledgment This research was made possible by the National Center for Advancing Translational Sciences Clinical and Translational Service Awards TL1 Medical Student Research Fellowship. 1872

5 BIBLIOGRAPHY 1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013; CA Cancer J Clin 2013;63: Funk GF, Karnell LH, Robinson RA, Zhen WK, Trask DK, Hoffman HT. Presentation, treatment, and outcome of oral cavity cancer: a National Cancer Data Base report. Head Neck 2002;24: Blot WJ, McLaughlin JK, Winn DM. Smoking and drinking in relation to oral and pharyngeal cancer. Cancer Res 1988;48: Chaturvedi AK, Engels EA, Anderson WF, Gillison ML. Incidence trends for human papillomavirus related and unrelated oral squamous cell carcinomas in the United States. J Clin Oncol 2008;26: National Comprehensive Cancer Network: NCCN Clinical Practice Guidelines in Oncology (NCCN GuidelinesVR ) Head and Neck Cancers (Version ). May 30, Available at: physician_gls/pdf/head-and-neck.pdf. Accessed June 15, Luryi AL, Chen MM, Mehra S, Roman SA, Sosa JA, Judson BL. Hospital re-admission and 30-day mortality following surgery for oral cavity cancer: analysis of 21,681 cases. Head Neck doi: /hed Kazaure HS, Roman SA, Sosa JA. Association of postdischarge complications with reoperation and mortality in general surgery. Arch Surg 2012;147: Bilimoria KY, Cohen ME, Ingraham AM, et al. Effect of postdischarge morbidity and mortality on comparisons of hospital surgical quality. Ann Surg 2010;252: Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med 2009;360: Chen MM, Roman SA, Sosa JA, Judson BL. Postdischarge complications predict reoperation and mortality after otolaryngologic surgery. Otolaryngol Head Neck Surg 2013;149: Aloia TA, Lee JE, Vauthey JN, et al. Delayed recovery after pancreaticoduodenectomy: a major factor impairing the delivery of adjuvant therapy? J Am Coll Surg 2007;204: Donat SM, Shabsigh A, Savage C, et al. Potential impact of postoperative early complications on the timing of adjuvant chemotherapy in patients undergoing radical cystectomy: a high-volume tertiary cancer center experience. Eur Urol 2009;55: Awad MI, Shuman AG, Montero PH, Palmer FL, Shah JP, Patel SG. Accuracy of administrative and clinical registry data in reporting postoperative complications following surgery for oral cavity squamous cell carcinoma. Head Neck doi: /hed Sullivan MC, Roman SA, Sosa JA. Does chemotherapy prior to emergency surgery affect patient outcomes? Examination of 1912 patients. Ann Surg Oncol 2012;19: Shiloach M, Frencher SK Jr, Steeger JE, et al. Toward robust information: data quality and inter-rater reliability in the American College of Surgeons National Surgical Quality Improvement Program. J Am Coll Surg 2010;210: Gupta PK, Smith RB, Gupta H, Forse RA, Fang X, Lydiatt WM. Outcomes after thyroidectomy and parathyroidectomy. Head Neck 2012;34: Khuri SF, Daley J, Henderson W, et al. The Department of Veterans Affairs NSQIP: the first national, validated, outcome-based, riskadjusted, and peer-controlled program for the measurement and enhancement of the quality of surgical care. National VA Surgical Quality Improvement Program. Ann Surg 1998;228: American College of Surgeons National Surgical Quality Improvement Program. User Guide for the 2010 Participant Use Data File. June 24, Available at: User-Guide_FINAL.pdf. Accessed June 1, Haughey BH, Wilson E, Kluwe L, et al. Free flap reconstruction of the head and neck: analysis of 241 cases. Otolaryngol Head Neck Surg 2001; 125: Benbassat J, Taragin M. Hospital readmissions as a measure of quality of health care: advantages and limitations. Arch Intern Med 2000;160: Centers for Medicare & Medicaid Services. Readmissions Reduction Program. August 4, Available at: Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/Readmissions- Reduction-Program.html. Accessed February 20, Mosely LH, Finseth F, Goody M. Nicotine and its effect on wound healing. Plast Reconstr Surg 1978;61: Silverstein P. Smoking and wound healing. Am J Med 1992;93:22S 24S. 24. Moller AM, Villebro N, Pederson Tonneson, H. Effect of preoperative smoking intervention on postoperative complications: a randomised clinical trial. Lancet 2002;359: Lotfi CJ, Cavalcanti Rde C, Costa e Silva AM, et al. Risk factors for surgical-site infections in head and neck cancer surgery. Otolaryngol Head Neck Surg 2008;138: Sorensen LT, Horby J, Friis E, Pilsgard B, Jorgensen T. Smoking as a risk factor for wound healing and infection in breast cancer surgery. Eur J Surg Oncol 2002;28: Schwartz SR, Yueh B, Maynard C, Daley J, Henderson W, Khuri SF. Predictors of wound complications after laryngectomy: a study of over 2000 patients. Otolaryngol Head Neck Surg 2004;131: Scott HR, McMillan DC, Crilly A, McArdle CS, Milroy R. The relationship between weight loss and interleukin 6 in non-small-cell lung cancer. Br J Cancer 1996; 73: Scott HR, McMillan DC, Forrest LM, Brown DJF, McArdle CS, Milroy R. The systemic inflammatory response, weight loss, performance status and survival in patients with inoperable non-small cell lung cancer. Br J Cancer 2002;87: McMillan DC, Watson WS, O Gorman P, Preston T, Scott HR, McArdle CS. Albumin concentrations are primarily determined by the body cell mass and the systemic inflammatory response in cancer patients with weight loss. Nutr Cancer 2001;39: Windsor JA, Hill GL. Weight loss with physiologic impairment. A basic indicator of surgical risk. Ann Surg 1988;207: Tuggle CT, Park LS, Roman S, Udelsman R, Sosa JA. Rehospitalization among elderly patients with thyroid cancer after thyroidectomy are prevalent and costly. Ann Surg Oncol 2010;17: Robles L, Slogoff M, Ladwig-Scott E, Zank D. The addition of a nurse practitioner to an inpatient surgical team results in improved use of resources. Surgery 2011;150: Zuckerman JD, Skovron ML, Koval KJ, Aharonoff G, Frankel VH. Postoperative complications and mortality associated with operative delay in older patients who have a fracture of the hip. J Bone Joint Surg Am 1995;77: Gouma DJ, van Geenen RC, van Gulik TM, et al. Rates of complications and death after pancreaticoduodenectomy: risk factors and the impact of hospital volume. Ann Surg 2000;232: Bach PB, Cramer LD, Schrag D, Downey RJ, Gelfand SE, Begg CB. The influence of hospital volume on survival after resection for lung cancer. N Engl J Med 2001;345: