Critical care therapy and in-hospital mortality after radical nephroureterectomy for nonmetastatic upper urinary tract carcinoma

Bladder cuff excision at radical nephroureterectomy improved survival in upper tract urothelial carcinoma

OBJECTIVE

To test for survival differences after radical nephroureterectomy (RNU) in pT1-T3N0 upper tract urothelial carcinoma (UTUC) patients according to bladder cuff excision (BCE) status.

METHODS

Within the Surveillance, Epidemiology, and End Results database (2004-2020), we identified RNU patients and documented BCE status. Prior and after propensity score matching (ratio 1:1), cumulative incidence plots and competing risk regression (CRR) models addressed cancer specific mortality (CSM) and other-cause mortality (OCM).

RESULTS

Of 4,426 RNU-treated patients, 3,186 (72%) patients underwent BCE and 1,240 (28%) did not. BCE rates significantly increased over time from 65.2 to 77.0% (EAPC: +1.0%, 95% CI +0.4 to +1.6%, P = 0.004) between 2004 and 2020. After 1:1 propensity score matching for T stage, 1,240 of 1,240 (100%) RNU patients without BCE and 1,240 of 3,186 (39%) RNU patients with BCE were included in subsequent analyses. Five-year CSM rates were 30% in RNU with BCE vs. 36% in RNU without BCE patients (Δ = 6%). In multivariable CRR, RNU with BCE independently predicted lower CSM (HR: 0.81, 95% CI: 0.70-0.93; P < 0.001). Conversely, BCE status did not affect OCM (P = 0.4).

CONCLUSION

The rate of guideline recommended BCE at RNU increased over time. Additionally, RNU with BCE is associated with significantly lower CSM than RNU without BCE in pT1-T3N0 UTUC patients. In consequence, BCE should represent an integral part of a comprehensive RNU.

Costs of robot-assisted vs. open approaches for 5 major cancers

To test for differences in total hospital cost (THC) between robot-assisted vs. open major cancer surgeries, colectomy, esophagectomy, radical hysterectomy, lung resection and pancreatectomy. Within the National Inpatient Sample (2016-2019), we identified all robot-assisted vs. open procedures for the above stated surgeries. Multivariable Poisson regression models were fitted. Of all surgeries, 6830 (14%) were robot-assisted colectomies, 333 (7%) esophagectomies, 5985 (24%) radical hysterectomies, 6500 (21%) lung resections and 449 (4%) pancreatectomies. Relative to open surgery, robot-assisted esophagectomy (181,462 vs. 96,195 $, Δ = 85,267 $), pancreatectomy (123,872 vs. 95,707 $, Δ = 28,168 $), lung resection (93,910 vs. 80,770 $, Δ = 13,140 $) and colectomy (82,898 vs. 71,279 $, Δ = 11,619 $) were associated with higher THC (all p < 0.001), except for radical hysterectomy (63,793 vs. 62,558 $, p = 0.8). After multivariable adjustment for patient and hospital characteristics, robot-assisted esophagectomy (risk ratio [RR]: 1.40), robot-assisted pancreatectomy (RR: 1.24), robot-assisted colectomy (RR: 1.20), robot-assisted lung resection (RR: 1.11) as well as robot-assisted radical hysterectomy (RR: 1.10) independently predicted higher THC (all p < 0.001). For the five examined procedures, THC are invariably higher when the robot-assisted approach is used. This THC disadvantage of the robot-assisted approach requires a careful consideration to the other benefits of robotic-assisted surgery, such as shorter convalescence and earlier return to regular activities, that could not be addressed in the current analysis.

Total hospital cost of robot-assisted approach in major urological cancer surgeries

BACKGROUND

The aim of this study was to test for differences in total hospital cost (THC) between robot-assisted vs. open partial nephrectomy (PN), radical cystectomy (RC), radical prostatectomy (RP) and radical nephroureterectomy (NU).

METHODS

Within the National Inpatient Sample (2010-2019), we identified all robot-assisted vs. open PN, RC, RP and NU patients. Multivariable Poisson regression models were fitted.

RESULTS

Of all surgeries, 22,572 (56%) were robot-assisted PN (RPN), 5114 (24%) were robot-assisted RC (RARC), 99,134 (70%) were robot-assisted RP (RARP), and 1138 (24%) patients were robot-assisted NU (RNU). Relative to open surgery, RARC (115,511 vs. 103,531$), RNU (64,761 vs. 54,768$), RARP (49,629 vs. 40,850$) and RPN (56,288 vs. 50,875$) were associated with higher THC (all P<0.001). After multivariable adjustment, RARP (risk ratio [RR]: 1.25), RNU (RR: 1.13), RPN (RR: 1.11) as well as RARC (RR: 1.10) independently predicted higher THC (all P<0.001). Additionally, Charlson Comorbidity Index ≥2 (RR: from 1.07 to 1.08), large bed size hospitals (RR: from 1.03 to 1.08), length of stay (RR: from 1.02 to 1.06), and overall complications (RR: from 1.09 to 1.19) invariably predicted higher THC.

CONCLUSIONS

THC is invariably higher when robot-assisted approach is applied instead of open approach in PN, RC, RP and NU patients. This THC disadvantage of robot-assisted approach requires consideration in the light of other benefits of robot-assisted surgery that could not be addressed in the current analyses.

Effect of Race/Ethnicity on In-hospital Outcomes After Radical Cystectomy

OBJECTIVE

Previously, African American race/ethnicity predisposed to higher rate of adverse in-hospital outcomes after radical cystectomy (RC). We tested whether this association applies to contemporary RC patients.

METHODS

Patients were identified within the National Inpatient Sample (NIS 2000-2019). Multivariable logistic and Poisson regression models were fitted.

RESULTS

Of 19,370 RC patients, 1,089 (5.6%) were African American, while 18,281 (94.4%) were Caucasian. Relative to Caucasians, African Americans were younger (median age 66 vs. 70 years; p < 0.001), more frequently female (33.8 vs. 18.5%; p < 0.001) and more frequently in the lowest income quartile (46.8 vs. 18.6%; p < 0.001). Relative to Caucasians, after RC, African Americans exhibited higher rates of postoperative complications (61.3 vs. 58.3%; multivariable odds ratio [MOR] 1.2; p = 0.009). Specifically, African Americans exhibited higher rates of blood transfusions (30.2 vs. 24.1%; MOR 1.3; p < 0.001), gastrointestinal (26.7 vs. 24.1%; MOR 1.2; p = 0.003), and infectious (6.2 vs. 4.2%; MOR 1.5; p = 0.001) complications, as well as deep vein thrombosis (3.1 vs. 1.7%; MOR 1.9; p < 0.001). Additionally, after RC, African Americans exhibited higher rates of critical care therapy use (CCT; 13.9 vs. 12.2%; MOR 1.3; p = 0.002) and in-hospital mortality (2.8 vs. 1.7%; MOR 1.8; p = 0.002). Finally, African Americans exhibited higher rates of length of stay ≥ 75th percentile (40.9 vs. 31.2%; MOR 1.6; p < 0.001).

CONCLUSIONS

In contemporary RC patients, African American race/ethnicity predisposes to less favorable in-hospital outcomes, including higher in-hospital mortality and longer hospital stay. Unfortunately, these race/ethnicity disadvantages have not been improved upon relative to the previous report.

Perioperative Complications and In-Hospital Mortality in Radical Nephroureterectomy Patients with Heart Valve Replacement

BACKGROUND

Perioperative complication rates and in-hospital mortality after radical nephroureterectomy (RNU) in patients with a history of heart valve replacement are unknown.

METHODS

Within the National Inpatient Sample (2000-2019), we identified non-metastatic upper urinary tract carcinoma patients treated with RNU according to the presence or absence of a history of heart valve replacement. Propensity score matching (1:10 ratio) and multivariable logistic regression as well as Poisson regression models were used.

RESULTS

Of 15,423 RNU patients, 151 (1.0%) harbored heart valve replacement. This proportion increased over the study span from 0.5% to 1.4% (p = 0.02). Heart valve replacement patients were older (median 78 vs. 72 years; p < 0.001), more frequently male (73 vs. 60%; p = 0.002), and more frequently harbored a Charlson Comorbidity Index (CCI) ≥3 (36 vs. 27%; p = 0.002). After 1:10 propensity score matching for age, sex, and CCI, the patients exhibited higher rates of perioperative bleeding (7.4 vs. 2.4%; odds ratio [OR] 3.2; p < 0.001), blood transfusions (24.0 vs. 17.0%; OR 1.6; p = 0.02), and cardiac complications (24.0 vs. 14.0%; OR 2.1; p < 0.001). Conversely, heart valve replacement patients did not exhibit higher rates of critical care therapy (OR 1.06; p = 0.8), higher rates of in-hospital mortality (OR 0.8; p = 0.8), and longer length of stay (OR 1.4; p = 0.052) than their RNU counterparts without a history of heart valve replacement.

CONCLUSION

RNU patients with a history of heart valve replacement are at significantly increased risk of perioperative bleeding, blood transfusions, and cardiac complications; however, despite these increased risks, they neither exhibited higher critical care therapy rates or higher in-hospital mortality rates, nor did they require significantly longer hospital stay.

Trends and Disparities in Inpatient Palliative Care Use in Metastatic Renal Cell Carcinoma Patients Receiving Critical Care Therapy

PURPOSE

Temporal trends in and predictors of inpatient palliative care use in patients with metastatic renal cell carcinoma (mRCC) undergoing critical care therapy are unknown.

METHODS

Relying on the National Inpatient Sample (2008-2019), we identified mRCC patients undergoing critical care therapy, namely invasive mechanical ventilation, percutaneous endoscopic gastrostomy tube insertion, dialysis for acute kidney failure, total parenteral nutrition, or tracheostomy. Estimated annual percentage changes (EAPC) analyses and multivariable logistic regression models addressed inpatient palliative care use.

RESULTS

Of 3802 mRCC patients undergoing critical care therapy, 817 (21.5%) received inpatient palliative care. Overall, inpatient palliative care use increased from 4.9% to 31.5% between 2008 and 2019 (EAPC +9.2%). In subgroup analyses, the highest increase in inpatient palliative care use was observed in the Midwest (EAPC: +11.9%), in the South (EAPC +10.4%), and in teaching hospitals (EAPC +9.0%; all P ≤ .004). In logistic regression models, teaching hospital status (odds ratio [OR] 1.41) and contemporary year interval (OR 2.12; all P < .001) independently predicted higher inpatient palliative care rates. Conversely, hospital admission in the Northeast (OR 0.53) or in the South (OR 0.79; all P ≤ .03) was associated with lower inpatient palliative care rates than in the West.

CONCLUSION

In mRCC patients, inpatient palliative care rates have improved over time, with the highest increase in hospitals in the Midwest and in the South. Moreover, admission to teaching hospitals or in the West is associated with higher inpatient palliative care rates. In consequence, regional disparities, as well as differences according to teaching hospital status represent targets to achieve comprehensive inpatient palliative care coverage in mRCC patients receiving critical care therapy.

In-Hospital Outcomes after Robotic Vs Open Radical Nephroureterectomy

Objective: To test whether the contemporary robot-assisted nephroureterectomy (RNU) is associated with more favorable in-hospital outcomes than historical RNU, relative to the same endpoints in open NU (ONU). Methods: Within the National Inpatient Sample (2008-2019), we identified RNU and ONU patients. Multivariable logistic and Poisson regression models were fitted. Results: Of 8032 NU patients, historical (2008-2013) vs contemporary (2014-2019) proportions were 776 (41%) vs 1104 (59%) for RNU and 3719 (60%) vs 2433 (40%) for ONU. The rates of RNU have increased over time (2008-2019; Δ absolute: +18%; p < 0.001). Contemporary RNU patients exhibited significantly better in-hospital outcomes in 6 of 12 comparisons vs historical that ranged from -54% for genitourinary complications to -12% for median length of stay (LOS). Contemporary ONU patients also exhibited significantly better in-hospital outcomes in 11 of 12 comparisons vs historical that ranged from -67% for blood transfusions to -26% for gastrointestinal complications. When historical RNU was compared with historical ONU, RNU in-hospital outcomes were better in 7 of 12 comparisons that ranged from -61% for median LOS to -16% for postoperative complications. Conversely, when contemporary RNU was compared with contemporary ONU, RNU in-hospital outcomes were only better in 2 of 12 comparisons: -25% cardiac complications and -13% for median LOS. Conclusion: The magnitude of in-hospital outcomes categories improvement between historical vs contemporary was two-fold more pronounced in ONU (11 improved categories) than in RNU (6 improved categories). Few outcome benefits remained (two categories only) when contemporary RNU was compared with contemporary ONU.

Non-Caucasian Race/Ethnicity Predisposes to Unfavorable Stage and Grade at Upper Tract Urothelial Carcinoma Diagnosis

OBJECTIVE

To test whether race/ethnicity affects stage or grade distribution at upper tract urothelial carcinoma (UTUC) diagnosis.

METHODS

Within the Surveillance, Epidemiology, and End Results (SEER) database 2004-2020, UTUC patients were identified. Multivariable logistic regression models tested for the association between race/ethnicity and stage as well as grade at diagnosis according to renal pelvis vs. ureteral origin. Stage at presentation was defined as (i) metastatic stage (TanyNanyM1) and (ii) advanced stage (locoregional or metastatic (T3-4N0-2M0-1)).

RESULTS

Of 14,384 UTUC patients, 8926 (62%) were renal pelvis. Of renal pelvis UTUC patients, 7064 (79%) were Caucasian, 797 (9%) Hispanic, 623 (7%) Asian Pacific Islander (API), and 442 (5%) African American (AA). Relative to Caucasians, API (odd ratio (OR) 1.85, 95% confidence interval (CI) 1.44-2.37, p < 0.001) and AA (OR 1.37, 95% CI 1.02-1.83, p = 0.03) patients more frequently harbored metastatic stage. APIs also more frequently harbored advanced (OR 1.33, 95% CI 1.11-1.60, p = 0.002) stage and high (OR 1.33, 95% CI 1.07-1.67, p = 0.01) grade. Of 5458 (38%) ureteral UTUC patients, 4360 (80%) were Caucasian, 362 (7%) Hispanic, 509 (9%) API, and 227 (4%) AA. Relative to Caucasians, Hispanic (OR 1.45, 95% CI 1.01-2.05, p = 0.03) patients more frequently harbored metastatic stage. APIs also more frequently harbored advanced stage (OR 1.22, 95% CI 1.02-1.47, p = 0.03) and high (OR 1.94, 95% CI 1.49-2.55, p < 0.001) grade.

CONCLUSIONS

Race/ethnicity other than Caucasian, such as API and Hispanic, may predispose to higher odds of metastatic (TanyNanyM1) or advanced (T3-4N0-2M0-1) stages as well as to higher grade at initial diagnosis.