237 L N Pontius et al. Lymphovascular invasion for 23:7 555–562 Research PTC Lymphovascular invasion is associated with survival for papillary thyroid cancer Lauren N Pontius, Linda M Youngwirth, Samantha M Thomas, Randall P Scheri, Correspondence should be addressed Sanziana A Roman and Julie A Sosa to J A Sosa Duke University Medical Center, Durham, North Carolina, USA Email [email protected] Abstract Data are limited regarding the association between tumor lymphovascular invasion and Key Words survival for patients with papillary thyroid cancer (PTC). This study sought to examine f lymphovascular invasion lymphovascular invasion as an independent prognostic factor for patients with PTC f papillary thyroid cancer undergoing thyroid resection. The National Cancer Data Base (2010–2011) was queried f thyroidectomy for patients with PTC who underwent total thyroidectomy or lobectomy. Patients were f lymph nodes classified into two groups based on the presence/absence of lymphovascular invasion. f survival Demographic, clinical and pathological features were evaluated for all patients. A Cox proportional hazards model was utilized to identify factors associated with survival. Results show that 45,415 patients met inclusion criteria; 11.6% had lymphovascular invasion. Patients with lymphovascular invasion were more likely to have larger tumors (2.8 cm vs Endocrine-Related Cancer Endocrine-Related 1.5 cm, P < 0.01), metastatic lymph nodes (74.1% vs 32.5%, P < 0.01), and distant metastases (3.0% vs 0.5%, P < 0.01). They were also more likely to receive radioactive iodine (69.3% vs 44.9%, P < 0.01). Unadjusted overall 5-year survival was lower for patients who had tumors with lymphovascular invasion (86.6% vs 94.5%) (log-rank P < 0.01). After adjustment, increasing patient age (HR = 1.06, P < 0.01), male gender (HR = 1.68, P < 0.01), presence of metastatic lymph nodes (HR = 1.77, P < 0.01), distant metastases (HR = 3.49, P < 0.01), and lymphovascular invasion (HR = 1.88, P < 0.01) were associated with compromised survival. For patients with lymphovascular invasion, treatment with RAI was associated with reduced mortality (HR = 0.43, P < 0.01). The presence of lymphovascular invasion among patients with PTC is independently associated with compromised survival. Patients who have PTC with lymphovascular invasion should be considered higher risk, and adjuvant RAI should be more strongly considered. Endocrine-Related Cancer (2016) 23, 555–562 Introduction Thyroid cancer is the fastest increasing cancer in the Jillard et al. 2015). PTC has an excellent prognosis, United States among both men and women, largely with a cause­specific 10­year survival rate of over as a result of an increase in the incidence of papillary 90% (Biersack & Grünwald 2005). In general, optimal thyroid cancer (PTC), which now represents approxi­ treatment includes thyroid resection, with or without mately 90% of all new cases (Sosa & Udelsman 2006, administration of postoperative radioactive iodine (RAI). http://erc.endocrinology-journals.org © 2016 Society for Endocrinology Published by Bioscientifica Ltd. DOI: 10.1530/ERC-16-0123 Printed in Great Britain Downloaded from Bioscientifica.com at 10/01/2021 02:32:24PM via free access 10.1530/ERC-16-0123 Research L N Pontius et al. Lymphovascular invasion for 23:7 556 PTC The impact of lymphovascular invasion on survival found that there was a significant association between LVI is important for several cancers; it is an independent and both lateral cervical lymph node metastases and an risk factor for survival in colorectal, urothelial and increased risk of disease recurrence. breast cancers (Lotan et al. 2005, Meguerditchian et al. The primary aim of this study was to look for a 2005, Rovera et al. 2013). Data regarding the impact of potential association between tumor LVI and survival lymphovascular invasion (LVI) on patient outcomes for patients with PTC undergoing thyroid resection. in PTC have been limited, based primarily on small, Such an association is not recognized in current retrospective institutional studies (Kim et al. 2006, Girardi differentiated thyroid cancer staging systems, including et al. 2013). One single institution study by Kim et al., UICC/AJCC TNM (Union for International Cancer including 662 PTC patients, of whom 33 (5%) had LVI, Control/American Joint Commission on Cancer tumor, Table 1 Patient demographic, clinical and pathological features by the presence and absence of lymphovascular invasion. LVI present n = 5284 (%) LVI absent n = 40,131 (%) All patients n = 45,415 P value Age (years) (mean ± S.D.) 47.6 ± 16.4 50.2 ± 14.5 49.9 ± 14.7 <0.01 Gender (female) 3645 (69.0) 31,459 (78.4) 35,104 (77.3) <0.01 Race 0.01 White 4434 (85.8) 33,941 (86.0) 38,375 (86.0) Black 343 (6.6) 3043 (7.7) 3386 (7.6) Asian 316 (6.2) 1964 (5.0) 2280 (5.1) Unknown/other 73 (1.4) 496 (1.3) 569 (1.3) Insurance status <0.01 None 208 (4.0) 1146 (2.9) 1354 (3.0) Private 3610 (69.3) 28,185 (71.1) 31,795 (70.9) Government 1390 (26.7) 10,292 (26.0) 11,682 (26.1) Annual income 0.04 ≤$35,999 1329 (26.1) 9462 (24.5) 10,791 (24.7) ≥$36,000 3764 (73.9) 29,099 (75.5) 32,863 (75.3) Distance traveled to treatment facility (miles) 28.1 ± 94.7 25.8 ± 97.3 26.1 ± 97.0 0.05 (mean ± S.D.) Charlson/Deyo score 0.543 0 4402 (83.3) 33,338 (83.1) 37,740 (83.1) Endocrine-Related Cancer Endocrine-Related 1 724 (13.7) 5669 (14.1) 6393 (14.1) ≥2 158 (3.0) 1124 (2.8) 1282 (2.8) Facility type <0.01 Academic 2573 (48.7) 17,112 (42.6) 19,685 (43.4) Comprehensive 2371 (44.9) 20,255 (50.5) 22,626 (49.8) Community 337 (6.4) 2757 (6.9) 3094 (6.8) Geographic region <0.01 Northeast 1466 (27.7) 10,641 (26.5) 12,107 (26.7) South 1822 (34.5) 13,468 (33.6) 15,290 (33.7) Midwest 1042 (19.7) 9116 (22.7) 10,158 (22.3) West 954 (18.1) 6906 (17.2) 7860 (17.3) Tumor size (cm) (mean ± S.D.) 2.8 ± 2.4 1.5 ± 1.5 1.7 ± 1.7 <0.01 Metastatic cervical lymph nodes 2991 (74.1) 6853 (32.5) 9844 (39.1) <0.01 Distant metastases 142 (3.0) 179 (0.5) 321 (0.8) <0.01 Stage-NCDB <0.01 I 2592 (51.5) 28,719 (76.0) 31,311 (73.1) II 429 (8.5) 3174 (8.4) 3603 (8.4) III 1235 (24.5) 4378 (11.6) 5613 (13.2) IV 775 (15.5) 1502 (4.0) 2277 (5.3) Lobectomy 263 (5.0) 4828 (12.0) 5091 (11.2) <0.01 Length of stay (days) (mean ± S.D.) 2.1 ± 6.0 1.4 ± 4.9 1.5 ± 5.0 <0.01 Readmission 266 (5.1) 1533 (3.9) 1799 (4.0) <0.01 Positive surgical margins 1427 (27.8) 3658 (9.2) 5085 (11.3) <0.01 Extrathyroidal extension 405 (7.6) 547 (1.3) 952 (2.1) <0.01 RAI 3660 (69.3) 18,023 (44.9) 21,683 (47.7) <0.01 Data presented as n (%) unless otherwise specified. Percentages were calculated from patients with non-missing data, which may be less than the total number of patients in each category. http://erc.endocrinology-journals.org © 2016 Society for Endocrinology Published by Bioscientifica Ltd. DOI: 10.1530/ERC-16-0123 Printed in Great Britain Downloaded from Bioscientifica.com at 10/01/2021 02:32:24PM via free access Research L N Pontius et al. Lymphovascular invasion for 23:7 557 PTC nodes, metastases) (Edge et al. 2010), the National stage according to the AJCC/TNM system. All variables Thyroid Cancer Treatment Cooperative Study (NTCTCS) were summarized based on the presence or absence of LVI. (Sherman et al. 1998), the European Organization for Research and Treatment of Cancer (EORTC) (Byar et al. Statistical analysis 1979), AGES (age, grade, extent, size) (Hay et al. 1987), AMES (age, metastases, extent, size) (Cady & Rossi 1988), Baseline characteristics were reported using frequencies and MACIS (metastases, age, completeness of resection, and proportions for categorical variables and mean and invasion locally, size) (Hay et al. 1993, Dean & Hay 2000, standard deviation for continuous variables. Data were Adam et al. 2015). Our secondary aim was to describe compared using χ2 and Student’s t-tests. Overall survival the treatment patterns employed in the United States was defined as the time from diagnosis to death or the last for patients with PTC whose tumors have LVI. We follow­up visit. Survival time was censored for patients hypothesized that LVI would be associated with patient alive at the end of the study period. Patients with zero outcomes and that more aggressive treatment for these months of follow­up evaluation were excluded from the tumors would be warranted. study. Estimates and 95% confidence intervals (95% CI) of overall survival proportions were computed using the Kaplan–Meier method, and survival distributions were Methods compared between study groups using the log­rank test. The National Cancer Data Base (NCDB) is a joint A multivariate Cox proportional hazards model was collaboration of the Commission on Cancer of the used to examine the adjusted association of LVI with American College of Surgeons and the American Cancer overall survival. The model adjusted for the effects of Society. It is a national database that contains more than patient age, gender, race, annual income, insurance status, 29 million cancer cases from more than 1500 Commission facility type, tumor size, lymph node involvement, and on Cancer accredited cancer programs.