RaNeO Consortium Members

Upper urinary tract urothelial carcinoma (UTUC) accounts for 5%-10% of cases of urothelial cancers. Despite a relatively low frequency, UTUC has a high mortality rate, with more than 150,000 deaths per year. The gold standard for high-risk localized UTUC is radical nephroureterectomy (RNU), including removal of kidney, entire ureter, and bladder cuff, while low-risk disease is suitable for kidney-sparing approaches, which provide equal survival outcomes preserving renal function.^1,2 UTUC patients have an increased risk of chronic kidney disease because of age, comorbidities, smoking exposure, and potential impairment of the contralateral kidney due to diagnostic procedures or contralateral UTUC,³ and this risk further increases after RNU. Previous studies have demonstrated a benefit of adjuvant chemotherapy in prolonging survival and reducing recurrence rate in locally advanced UTUC. However, 50% of patients are not eligible for platinum-based protocols due to postoperative renal failure.^4,5

In this scenario, identification of patients at risk of significant renal function decline may allow clinicians to better assess kidney-sparing rather than extirpative surgery. Additionally, it may help in developing more appropriate protective strategies such as neoadjuvant treatments or different adjuvant approaches, and in a more adequate follow-up schedule for these patients. The aim of this study is to investigate the prevalence and predictors of renal function variation in a large multicenter cohort of patients who underwent RNU for UTUC.

We retrospectively analyzed data for consecutive patients who underwent RNU for UTUC at 17 tertiary centers between 1994 and 2020. Estimated glomerular filtration rate (eGFR) was calculated using the CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) equation. Renal function variation was evaluated at postoperative day (POD) 1 and 6 and 12 months of follow-up considering differences between time points as follows: Δ1: POD 1 eGFR – baseline eGFR; Δ2: 6-month eGFR – POD 1 eGFR; and Δ3: 12-month eGFR – 6-month eGFR. We defined POD 1 acute kidney injury (AKI) as an increase in serum creatinine by ≥0.3 mg/dL or a 1.5- to 1.9-fold increase in serum creatinine from baseline, according to the Acute Kidney Injury Network classification. Linear mixed models were used to evaluate the effect of clinical factors on eGFR variation and their interaction with follow-up. Patients who underwent previous or concurrent radical cystectomy, had a history of contralateral or metastatic UTUC, or underwent previous renal parenchymal-sparing surgery were excluded from the analysis.

A total of 576 patients were included. The median age was 72 years (IQR 64-79), median BMI was 26.2 kg/m² (IQR 24-29), 71% were male, 57.5% had a smoking history, and 46.5% had homolateral hydronephrosis. Preoperative eGFR was 62.2 (IQR 48.1-79.9) mL/min/1.73². Estimated blood loss was 170 mL (IQR 100-270), operating time was 210 min (180-255), 16.5% of patients had intraoperative complications, 50.8% had pT2 UTUC, and 4.6% had positive surgical margins. We found the following renal function variation at the considered time points: Δ1: −10.2 (−25.9; −2.2); Δ2: 1.37 (−6.2; 9.0); and Δ3: 0.9 (−3.6; 5.2). In univariable linear mixed models, elderly patients had an early positive eGFR variation after RNU compared to their younger counterpart (P = .03; Figure 1). Similarly, those with hydronephrosis showed an early positive eGFR variation after RNU (P < .001; Figure 2). Conversely, RNU had a clinically meaningful detrimental effect on those with POD 1 AKI (P < .001; Figure 3). During follow-up, eGFR recovery was more pronounced in those with preoperative hydronephrosis (ß interaction: −3.4 ± 0.9; P < .001) and in those who experienced POD 1 AKI (ß interaction: 13.9 ± 0.8; P < 0.001). The elderly’s capacity to recover was lower than their younger counterparts (ß interaction: −0.1 ± 0.04, P = .006). Multivariable regression model confirmed that age (ß 0.3 ± 0.1, P = .002), hydronephrosis (ß: 3.7 ± 1.8, P = .039), and POD 1 AKI (ß: −31.6 ± 1.8, P < .001) were independently associated with eGFR variation. In addition, all models confirmed a positive eGFR variation during follow-up (ß: 10.3 ± 2.9, P < .001).

Figure 1. Age-related estimated glomerular filtration rate (eGFR) variation according to Δ time intervals.

Figure 2. Impact of hydronephrosis on estimated glomerular filtration rate (eGFR) variation at considered Δ 1 (t1), 2 (t2), and 3 (t3) time points. t0 indicates operative day.

Figure 3. Impact of postoperative day (POD) 1 acute kidney injury (AKI) on estimated glomerular filtration rate (eGFR) variation at considered Δ 1 (t1), 2 (t2), and 3 (t3) time points. t0 indicates operative day; AKI I POD, post-operative day 1 acute kidney injury.

Our results are in line with findings coming from smaller populations involved in the most recent studies that demonstrated that low preoperative eGFR, age, and the absence of hydronephrosis were predictive factors for impaired postoperative renal function after RNU for UTUC.^6,7 These findings suggest that previously established contralateral compensatory kidney hypertrophy induced by the hydronephrosis of the ipsilateral urinary tract may play a protective role on post-surgical renal function variation. In our study, the elderly population has an early postoperative favorable eGFR variation probably because of a compensatory mechanism acting well before surgery. However, younger kidneys better compensate in the long run.

We observed that RNU has a great impact on the residual renal function, and a detrimental effect on POD 1 AKI. While extensively investigated after kidney surgeries,^8,9 the role of POD 1 AKI after RNU has not been widely studied. A small, retrospective, single cohort study recently assessed how POD 1 AKI influences eGFR decline, and for the first time showed that it was a strong predictor for eGFR decline after RNU for UTUC.¹⁰

All these results may have significant clinical implication. Knowing preoperatively which patients will more likely experience a postoperative eGFR decline may better address neoadjuvant chemotherapy regimens, resulting in an increased survival. On the other hand, patients who are ineligible for neoadjuvant regimen and who are at high risk of postoperative eGFR decline may undergo kidney sparing surgery, when oncologically indicated. Moreover, knowing the magnitude of the effect that POD 1 AKI can have on eGFR after RNU, multidisciplinary pre- and perioperative adapting strategies can be adopted, making every effort to avoid POD 1 AKI.

Acknowledgments

We thank our co-authors, Andrea Mari, MD; Riccardo Tellini, MD; Katia Odorizzi, MD; Alessandro Veccia, MD; Daniele Amparore, MD; Aliasger Shakir, MD; Umberto Carbonara, MD; Andrea Panunzio, MD; Federica Trovato, MD; Michele Catellani, MD; Letizia M. I. Janello, MD; Lorenzo Bianchi, MD; Giacomo Novara, MD; Fabrizio Dal Moro, MD; Riccardo Schiavina, MD; Elisa De Lorenzis, MD; Paolo Parma, MD; Sebastiano Cimino, MD; Ottavio De Cobelli, MD; Francesco Maiorino, MD; Pierluigi Bove, MD; Fabio Crocerossa, MD; Francesco Cantiello, MD; David D’Andrea, MD; Federica Di Cosmo, MD; Francesco Porpiglia, MD; Pasquale Ditonno, MD; Emanuele Montanari, MD; Francesco Soria, MD; Paolo Gontero, MD; Giovanni Liguori, MD, PhD; Carlo Trombetta, MD; Guglielmo Mantica, MD; Marco Borghesi, MD; Carlo Terrone, MD; Francesco Del Giudice, MD; Alessandro Sciarra, MD; Andrea Galosi, MD, PhD; Marco Moschini, MD, PhD; Shahrokh F. Shariat, MD, PhD; Marta Di Nicola, PhD; Andrea Minervini, MD, PhD; and Matteo Ferro, MD.