Patient characteristics
Among 147 patients with evaluable baseline eGFR, only 49 (33.3%) had normal eGFR values. CKD grade 3a was the most common (39.5%), and end-stage kidney disease (ESKD; kidney disease requiring hemodialysis [HD]) was also present (5.4%) (Fig. 1). The median baseline eGFR (interquartile range [IQR]) was 53.0 (43.6–69.4) ml/minute/1.73 m2. The baseline eGFR was statistically correlated with nephrectomy (66.3 ml/minute/1.73 m2 in patients with nephrectomy vs. 50.6 ml/minute/1.73 m2 in patients without nephrectomy, P < 0.001, Fig. 2A), proteinuria (45.6 ml/minute/1.73 m2 in patients with proteinuria vs. 54.4 ml/minute/1.73 m2 in those without proteinuria, P = 0.015, Fig. 2B), and HT (48.2 ml/minute/1.73 m2 in patients with HT vs. 55.9 ml/minute/1.73 m2 in those without HT, P < 0.001, Fig. 2C). The median baseline eGFR in patients with DM was relatively lower than that in those without DM, although it did not show statistical difference (46.8 vs. 53.4 ml/minute/1.73 m2, P = 0.053, Fig. 2D). The median OS (95% confidential interval [CI]) was 35.4 (26.2–42.8) months. There was no statistically significant correlation between OS and baseline eGFR (hazard ratio; 1.005, 95% CI; 0.9978–1.013, P = 0.169).
The characteristics of Cohort 1 (patients with ≥ 15 ml/minute/1.73 m2 eGFR, n = 139) and Cohort 2 (patients with ≥ 30 ml/minute/1.73 m2 eGFR, n = 134) are shown in Table 1. The median follow-up time (IQR) was 2.81 (1.24–5.20) and 2.89 (1.23–5.40) years in cohorts 1 and 2, respectively. More than half of patients were censored due to RCC death. Approximately 70% of the patients underwent nephrectomy in both cohorts. Most of the patients (86.3% in Cohort 1 and 85.8% in Cohort 2) received anti-VEGFR therapy (Table 2).
Probability Of Progress To Ckd Grades 4 And 5 For Each Baseline Ckd Grade
No patients with normal baseline eGFR progressed to CKD grade 4 or grade 5 (Fig. 3A and 3B). Among 85 patients with baseline CKD grade 3, 2 progressed to grade 5 after 8 years. Among five patients with baseline CKD grade 4, three progressed to grade 5 within 2 years (Fig. 3A). The estimated probabilities of progress to CKD grade 4 in the patients with baseline CKD grade 3a in 1, 2, and 5 years after the start of systemic therapy were 1.8%, 3.9%, and 15.6%, respectively. The patients with baseline CKD grade 3b had a higher rate of progress to CKD grade 4, and the estimated probabilities in 1, 2, and 5 years were 7.4%, 24.8%, and 55.9%, respectively (P < 0.001, Fig. 3B).
Clinical Course Of Two Cases Introduced To Hemodialysis Despite Baseline Ckd Grade 3a
In two cases, baseline CKD grade 3a progressed to ESKD. Figure 4 shows the progression to hemodialysis.
The first case was that of a man who had undergone radical nephrectomy. Two years later, he developed multiple lung metastases at the age of 71. He had HT and his baseline eGFR was 48.3 ml/minute/1.73 m2. He was initially treated using interferon-α + interleukin-2 + tegafur/uracil. Twenty months later, the disease progressed, and the treatment was switched to sunitinib. Although sunitinib treatment showed partial response, proteinuria appeared immediately after the introduction. CKD progressed to grade 4 3 years after the start of sunitinib and then to grade 5 7 years later, and hemodialysis was introduced 9 years later (Fig. 4A).
The second case was that of a 55-year-old woman who underwent radical nephrectomy. Three months later, she developed multiple lung and muscle metastases. She had no comorbidity other than CKD grade 3a with eGFR 47.8 ml/minute/1.73 m2. She was treated with sorafenib. She had hypertension grade 3 and proteinuria shortly after the introduction of sorafenib, while she showed complete response. Two years later, although a complete response was maintained, CKD progressed to grade 4, and sorafenib treatment was discontinued. Ten years later, although a complete response was still maintained, CKD progressed to grade 5, and hemodialysis was introduced (Fig. 4B).
Factors Affecting Progression To Ckd Grade 4
The patients with proteinuria at baseline had a high probability of early progression to CKD grade 4 (P = 0.0195, Fig. 5A). Nephrectomy had no statistical correlation with the probability of CKD progression (P = 0.898, Fig. 5B). The patients without anti-VEGFR agents did not progress to CKD grade 4 (P = 0.032, Fig. 5C). The use of mammalian targets of rapamycin inhibitors had no statistically significant correlation with the probability of CKD progression (P = 0.194, Fig. 5D).
Finally, we conducted multivariate analysis using six variables (baseline eGFR, follow-up duration, duration of treatment with anti-VEGF agents, proteinuria, HT, and DM). Multivariate analysis showed that the duration of anti-VEGFR treatment (P = 0.014), baseline eGFR (P = 0.002), and DM (P = 0.040) were independently correlated with progression to CKD grade 4 (Table 3).