Survival Benets After Surgery of Renal Cell Carcinoma Patients with Inferior Vena Cava Thrombus

OBJECTIVE: To evaluate surgical and oncological outcomes after surgery in renal cell carcinoma (RCC) patients with inferior vena cava (IVC) tumor thrombus METHODS: A total of 58 patients from 2002 to 2019 underwent radical nephrectomy and IVC thrombectomy at Siriraj Hospital, Bangkok, Thailand, were retrospectively reviewed. Kaplan-Meier analysis was utilized to compare survival benets between cohorts and Cox regression to evaluate predictors of patient survival. RESULTS: There were 5 (8.6%), 21 (36.2%), 23 (39.7%) and 9 (15.5%) patients with tumor thrombus level I, II, III and IV respectively. The major complications (Clavien 3-5) were observed in 15 patients (25.8%) and 80% were patients with high thrombus level (III-IV). There was 9% mortality (5 patients): 2 intraoperatively and 3 postoperatively. Median follow-up was 15 months (IQR:5-41). Two-year overall survival (OS) was 80% and 75% in all patients and pN0M0 cohort, respectively. There was signicant difference in OS among each IVC thrombus level cohort (p<0.02). Two-year OS of metastatic RCC patients was 67% and not signicantly different when compared to non-metastatic cohort (p=0.12). On multivariate analysis, only sarcomatoid dedifferentiation was associated with OS (p=0.04). Disease-free survival was not signicantly different among thrombus-level cohorts (p=0.65). CONCLUSION: Our study suggested that surgical treatment for RCC with IVC thrombus provided acceptable OS outcomes, even in a small volume experience. Although the survival was signicantly reduced with higher IVC thrombus level cohort, the level of thrombus itself was not an independent factor. Only sarcomatoid dedifferentiation was a predictor for OS after radical nephrectomy and tumor thrombectomy. were observed in patients with high thrombus level (III-IV). The perioperative mortality rate was 8.6%; 2(3.4%) intraoperatively and 3(5.2%) postoperatively. Two intraoperative deaths were oating massive pulmonary emboli during intra-atrial thrombus surgery using CPB and unexplained cardiac arrest post-uneventful IVC thrombus level II removal. All of 3 postoperative mortality occurred in thrombus level IV patients, due to coagulopathy, pulmonary embolism and sepsis.


Introduction
Renal cell carcinoma (RCC) has a unique propensity to develop tumor thrombus into the inferior vena cava (IVC) about 4-10% 1,2 . Patients can be asymptomatic or present with symptoms based upon extension of tumor thrombus such as leg edema, varicocele, or dyspnea due to pulmonary embolism 3 .
Data from Reese and colleagues demonstrated that untreated patients had median survival of 5 months 4 . Radical nephrectomy and IVC thrombectomy remain the only potential treatment for these patients. Patients with and without metastatic disease who have undergone surgical removal of kidney and IVC thrombi have survival of 60% and 90% at 1 year, respectively 2 . However, the perioperative morbidity and mortality are high 1,2
Nonetheless, the impact of thrombus level on survival is still controversial 1 . Among untreated patients, supradiaphragmatic thrombi and distant metastases were associated with reduced disease-speci c Page 3/16 survival (DSS) 4 . The 5-year DSS was 40-60% after surgery 2,3 . Several prognostic factors were found associated with survival including tumor size and grade, tumor necrosis, positive lymph nodes, sarcomatoid differentiation, IVC thrombus, perinephric fat and adrenal gland invasion and distant metastases 5 . Data from IRCC-VTC, also demonstrated that tumor-thrombus level was an independent factor of survival 6 . However, another study from a Chinese institution found that thrombus level was not associated with DSS while IVC wall and nodal invasion and metastatic disease were 7 .
Our center has initiated this operation since 2002 using liver mobilization technique and reported our initial experience in 2008 8 . We evaluate surgical and survival outcomes and also assess predictive factors of survival among patients with RCC and IVC thrombi. To our knowledge, this is the largest study of RCC patients with IVC thrombi in Southeast Asian population. control. Liver mobilization technique without any forms of venous bypass procedure was performed in patients with thrombus level II-III. In patients with supradiaphragmatic IVC thrombus, additional diaphragm cutting for cephalic IVC control to avoid cardiopulmonary bypass (CPB) were applied. CPB was only used in intra-atrial thrombus removal procedure.

Methods
Patients were evaluated for tumor recurrence at 3-6 months interval postoperatively. The data included patient demographic, perioperative parameters, tumor characteristic, tumor recurrence and follow-up time.
Disease free survival (DFS) was calculated from the date of surgery to radiologic evidence of tumor recurrence, the last follow-up or death. Overall survival (OS) was de ned from date of surgery to the date of last follow-up or death.

Statistical analysis
Continuous variables were showed as median [interquartile range(IQR)] while categorical variables were in numbers (percentages). Comparing between cohorts were performed using Kruskal-Wallis and Chisquare tests. The survival graphs were created by Kaplan-Meier curve and compared using log-rank test.
Univariate and multivariate analyses to de ne the predictors of survival were performed using Coxproportional-hazards model. All statistical signi cance were de ned at p<0.05. SPSS version18.0 was utilized for all analyses.
On univariate analysis, tumor size, papillary and sarcomatoid histology, supradiaphragmatic tumor thrombus and positive-tumor margin were associated with OS (Table 2). However, only sarcomatoid dedifferentiation was associated with reduced OS (HR 31.3;1.27-767.8). There was no factor associated with DFS on univariate and multivariate analyses. Mortality and morbidity of radical nephrectomy and IVC thrombectomy were considered high; 1.5-10% and 15-80% for mortality and morbidity, respectively 2,3,12 . Previous study from IRCC-VTC, a total of 2147 patients from 22 institutions in the US and Europe, demonstrated 34% postoperative complications; 13% were classi ed as Clavien grade 3-5. Mortality rate was 2% within 30 days and high level of tumor thrombus was associated with increased complication rate particularly high-grade complications (p=0.03) 1 . In our study, perioperative morbidity and mortality were unsurprisingly high especially in supradiaphragmatic thrombus patients. As a result, operative time, blood loss and transfusion, and ICU stay were signi cantly greater in these patients (p<0.01). Furthermore, 80% of major complications were observed in patients with thrombus level III-IV. Recent study of 62 patients with tumor-thrombus level III-IV from four tertiary centers in the US reported that preoperative limited performance status and reduced serum albumin were associated with increased postoperative ninety-day mortality 12 .
The role of preoperative renal arterial embolization (RAE) to reduce intraoperative blood loss and perioperative complications has been debating 1-3 . RAE, generally performed on the operation day, could cut off blood supply to renal mass and thus reduce blood loss and complications. Previous study from Cleveland Clinic demonstrated signi cant risk of perioperative mortality in patients with preoperative embolization (OR 5.5, p=0.03) as well as postoperative complications 13 . Recent study from Tang and colleagues, however, supported reduced blood loss and transfusion in patients receiving perioperative embolization(p<0.03) 14 . Notably, patients from Cleveland Clinic cohort had higher thrombus level when compared to Tang's study. No preoperative RAE was performed in our series.
Level of IVC thrombus on patients' survival has been controversial 5,6,11,15 . Our study showed that tumor size and histology, supradiaphragmatic tumor thrombus and positive surgical margin were associated with OS on univariate analysis. However, tumor thrombus level was not an independent factor for OS on multivariate analysis. A study of 1192 pT3b and pT3c patients from 13 European institutions also demonstrated similar results 15 . IVC thrombus level was not a predictor for OS after radical surgery. Nevertheless, subsequent study based on a pooled analysis of 11 international centers (IRCC-VTC), the level of IVC thrombus was shown to be an independent factor of survival; HR 2.  7 . The advantages of surgery in metastatic patients were found in our study with 2-year OS rate of 67% and no signi cant difference when compared to non-metastatic group(p=0.12). Additionally, one patient after thrombectomy and lung metastatectomy has been alive with stable disease at 110 months follow-up time.
Retrospective design with single center is the limitation of this study. Regarding to tertiary center, some patients were loss to follow-up and OS was utilized as our primary outcome instead of CSS. Nevertheless, our study provides medium to long term experience with the number of patients with RCC and IVC thrombi in Thailand and represents Southeast Asian population. Limited number of the patients in this study is another limitation regarding to low incidence of disease in our area and some were not t for surgery. No prior study has ever been addressed this clinical information of such patients in the Southeast Asian region. Even though, multicenter study from this area would be required.

Conclusion
Our experience over 18 years suggested that surgical treatment for RCC with IVC thrombus was feasible and provided acceptable surgical and oncological outcomes. However, major complications and perioperative mortality in high-thrombus level(III-IV) were greater when compared to low-thrombus level(I-II). Although our experience demonstrated that tumor-thrombus level was associated with patient survival, the level of tumor thrombus itself was not an independent factor. Cytoreductive surgery in mRCC patients provided acceptable outcomes in selected patients. Only sarcomatoid variant was a signi cant adverse factor associated with reduced OS in this study.