Higher Risk of Proteinuria with Atezolizumab plus Bevacizumab than Lenvatinib in First-Line Systemic Treatment for Hepatocellular Carcinoma

Introduction

The Barcelona Clinic Liver Cancer (BCLC) staging system recommends that patients with unresectable hepatocellular carcinoma (HCC) and preserved liver function receive systemic treatment [1]. Currently, the foremost first-line systemic treatments for unresectable HCC include the combination therapy of atezolizumab plus bevacizumab (Atezo/Bev), an amalgamation of an immune checkpoint inhibitor (ICI) and vascular endothelial growth factor (VEGF) inhibitor, as well as lenvatinib (LEN), a tyrosine kinase inhibitor (TKI) monotherapy [2, 3]. However, the increasing use of ICIs, TKIs, and targeted agents in unresectable HCC has brought about a heightened awareness of previously undocumented adverse events during treatment [4, 5]. Proteinuria, in particular, has emerged as a notable adverse event observed in patients receiving VEGF monoclonal antibodies or anti-VEGF TKIs. Clinical trials have demonstrated that Atezo/Bev and LEN exhibit a higher incidence of proteinuria as an adverse event compared with sorafenib (25% vs. 11% for LEN and 29% vs. 5% for Atezo/Bev) [2, 3].

LEN and bevacizumab exert their therapeutic effects by targeting the VEGF signaling pathway [6], which leads to increased intraglomerular injury to podocytes and has a direct effect on proximal tubular cell receptors [7]. In addition to nephrotoxicity associated with anti-VEGF treatment, the concurrent use of ICIs has led to a substantial increase in ICI-related renal adverse events [8]. In real-world clinical settings, patients often transiently or permanently discontinue treatment due to proteinuria associated with anti-VEGF therapy [9‒13]. Concerning Atezo/Bev, patients may transition to atezolizumab monotherapy by discontinuing bevacizumab. The early interruption of treatment could potentially compromise the antitumor efficacy in patients with unresectable HCC [14, 15]. Furthermore, proteinuria, unlike other adverse events, poses an extensive challenge in management and is associated with a marked decline in patient prognosis and quality of life, necessitating careful consideration.

Although risk factors for early proteinuria during anti-VEGF treatment have been investigated across various cancer types [16, 17], studies specifically focusing on HCC are limited. Notably, Ando et al. [18] identified poor baseline renal function, the use of antihypertensive drugs, and elevated systolic blood pressure (SBP) as risk factors for early-onset proteinuria during Atezo/Bev treatment for unresectable HCC. Similarly, Ikesue et al. [19] demonstrated that pre-existing proteinuria and poor baseline estimated glomerular filtration rate (eGFR) were significantly associated with a high risk of proteinuria in patients with HCC receiving LEN. However, these studies had relatively small sample sizes, and to date, no studies have systematically compared the incidence of proteinuria between Atezo/Bev and LEN in patients with unresectable HCC treated as first-line systemic therapy. Therefore, this study aimed to identify the risk factors for proteinuria in patients with unresectable HCC treated with Atezo/Bev or LEN and compare the incidence of proteinuria between Atezo/Bev and LEN.

Materials and Methods

Study Population

This was a retrospective, single-center study of adult patients with unresectable HCC undergoing first-line systemic treatment with Atezo/Bev or LEN between October 2013 and October 2022 at Asan Medical Center, Seoul, Republic of Korea. A total of 632 patients were retrospectively analyzed for study eligibility (Fig. 1). A diagnosis of HCC was established radiologically and pathologically based on international guidelines [20‒22]. Patients were excluded if they had a baseline urine dipstick test over 2+ or a urine total protein to creatinine ratio (UPCR) over 200 mg/g Cr (n = 1) and insufficient availability of medical records (n = 9). The index date was defined as the time of Atezo/Bev or LEN treatment initiation. Patients in the LEN group were administered LEN at a dosage of 12 mg/day (for body weight ≥60 kg) or 8 mg/day (for body weight <60 kg). Those in the Atezo/Bev group underwent a combination therapy comprising 1,200 mg of atezolizumab plus 15 mg per kg of body weight of bevacizumab intravenously every 3 weeks. Dosage adjustments or delays were permitted based on individual patient tolerability. Response assessments occurred every 6–8 weeks throughout the treatment period, with the option for additional evaluations as clinically warranted. Treatment in both groups was continued until disease progression, deterioration of performance status to ECOG PS 4, decline in liver function, occurrence of adverse drug reactions requiring treatment termination due to severe toxicity, or death. Ethical approval was obtained from the Institutional Review Board of Asan Medical Center (IRB No. 2023-1253), and the requirement for informed consent was waived given the retrospective nature of the study.

Fig. 1.

Flowchart of the study. Atezo/Bev, atezolizumab plus bevacizumab; LEN, lenvatinib.

Covariates and Outcomes

Patient characteristics and outcomes data were extracted from the electronic medical records of Asan Medical Center, Seoul, Republic of Korea. Clinical information encompassing demographics, body mass index, etiologies of HCC, hypertension, diabetes, type of antihypertensive agent, and ECOG PS was obtained at the index date. Laboratory parameters, including platelet counts, international normalized ratio, creatinine, eGFR, liver function tests including aspartate aminotransferase, alanine aminotransferase, albumin, total bilirubin, and tumor markers including alpha-fetoprotein and protein induced by vitamin K antagonist-II (PIVKA-II), were also collected. Liver function was further assessed by Child-Pugh class and albumin-bilirubin score. Baseline tumor characteristics included BCLC stage and the presence of extrahepatic metastasis or macrovascular invasion (MVI).

The primary outcome of interest was the incidence of proteinuria, while the secondary outcome was the occurrence of severe proteinuria in patients with unresectable HCC treated with Atezo/Bev or LEN as their first-line systemic therapy. Patients who discontinued treatment before the development of proteinuria were censored at the date of their last urinalysis during the treatment. Proteinuria was defined as a urine dipstick exceeding 2+ or the urine protein creatinine ratio (UPCR) greater than 200 mg/g Cr, as described in previous studies [23, 24]. The grade of proteinuria adhered to the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0 [10]. Urine dipstick tests and UPCR assessments were performed at baseline and during each routine visit, with UPCR serving as a quantitative alternative to the 24-h urine collection method [25]. Severe proteinuria was characterized by a UPCR exceeding 2,000 mg/g Cr.

Statistical Analysis

Continuous variables were presented as mean values ± standard deviation or median (interquartile ratio [IQR]), while categorical variables were expressed as numbers with percentages. Student’s t test or Mann-Whitney U test was used for continuous variables, and the Chi-square or Fisher’s exact test was employed for categorical variables, as appropriate.

The cumulative incidence of proteinuria was assessed using Kaplan-Meier’s method, and differences were compared using the log-rank test. To mitigate potential confounding and selection biases, a propensity score (PS)-matched analysis with a 1:1 ratio was performed. Additionally, a subgroup analysis comparing the proteinuria development between the Atezo/Bev and LEN groups was conducted using a forest plot within the PS-matched cohort. To identify the risk factors associated with proteinuria, Cox regression analysis was conducted for the entire study population and each treatment group.

Furthermore, we examined the impact of proteinuria on oncological outcomes, including overall survival (OS), progression-free survival (PFS), and objective response rate. To mitigate immortal time bias, where disease progression cannot occur before proteinuria develops in the proteinuria group, we conducted a time-dependent Cox regression analysis with proteinuria as a time-varying variable and performed a 6-month landmark analysis.

Missing values, constituting 1.1% (PIVKA-II) to 2.7% (eGFR) of the baseline laboratory data, were imputed using linear interpolation by the MICE package. Statistical analyses were performed using R software, version 4.3.2 (R Foundation Inc; http://cran.r-project.org/). A p value of <0.05 was considered statistically significant.

Results

Baseline Characteristics

Table 1 presents comprehensive information on patient demographics, tumor characteristics, and laboratory findings of the 622 study patients at baseline. The mean age was 60.5 years, with a predominance of males (82.3%). Chronic hepatitis B virus infection was identified as the leading cause of underlying liver disease, accounting for 70.9% of cases. The prevalence of diabetes was 20.7% and a majority (54.2%) had hypertension. The median SBP and diastolic blood pressure were 126 mm Hg (IQR: 115–140) and 77 (IQR: 70–84) mm Hg, respectively. Approximately 57.1% of patients were taking antihypertensive agents, with a substantial proportion using a calcium channel blocker (68.7%) and an angiotensin receptor blocker (45.9%). The median eGFR was 90.0 (IQR: 76.0–100.0) mL/min/1.73 m².

Table 1.

Baseline characteristics of the entire and PS-matched cohorts

Variables	Entire cohort				PS-matched cohort
	total (n = 622)	Atezo/Bev (n = 367)	LEN (n = 255)	p value	Atezo/Bev (n = 202)	LEN (n = 202)	SMD
Demographic findings
Age, years	60.5±11.0	61.7±11.2	58.7±10.3	0.001	60.0±10.8	60.0±9.6	0.002
Sex
Female	110 (17.7)	68 (18.5)	42 (16.5)	0.58	32 (15.8)	32 (15.8)	<0.001
Male	512 (82.3)	299 (81.5)	213 (83.5)		170 (84.2)	170 (84.2)
Body mass index, kg/m2	23.9 [21.8–25.7]	23.9 [21.9–25.7]	23.8 [21.5–25.6]	0.43	23.9 [21.9–25.8]	24.0 [21.8–25.8]	0.025
Etiologies of underlying liver disease
Hepatitis B	441 (70.9)	253 (68.9)	188 (73.7)	0.11	144 (71.3)	144 (71.3)	0.022
Hepatitis C	36 (5.8)	20 (5.4)	16 (6.3)		14 (6.9)	14 (6.9)
Alcohol	43 (6.9)	23 (6.3)	20 (7.8)		16 (7.9)	15 (7.4)
Others	102 (16.4)	71 (19.4)	31 (12.2)		28 (13.9)	29 (14.4)
Hypertension	337 (54.2)	187 (51.0)	150 (58.8)	0.06	112 (55.4)	115 (56.9)	0.030
SBP, mm Hg	126 [115–140]	125 [114–136]	128 [116–143]	0.04	128 [117–138]	126 [115–141]	0.025
DBP, mm Hg	77 [70–84]	76 [69–82]	79 [71–86]	<0.001	79 [71–84]	78 [70–85]	0.016
Diabetes	129 (20.7)	70 (19.1)	59 (23.1)	0.26	40 (19.8)	48 (23.8)	0.096
Antihypertensive agent
None	267 (42.9)	155 (42.2)	112 (43.9)	0.74	91 (45.0)	88 (43.6)	0.030
Presence	355 (57.1)	212 (57.8)	143 (56.1)		111 (55.0)	114 (56.4)
Classification of antihypertensive agent
ACE inhibitor	3 (0.8)	2 (0.9)	1 (0.7)	0.99	1 (0.9)	1 (0.9)	<0.001
ARB	163 (45.9)	98 (46.2)	65 (45.5)	0.81	59 (53.2)	53 (46.5)	0.066
Beta-blocker	134 (37.7)	77 (36.3)	57 (39.9)	0.76	36 (32.4)	44 (38.6)	0.100
Alpha-blocker	7 (2.0)	5 (2.4)	2 (1.4)	0.78	1 (0.9)	1 (0.9)	<0.001
Nonselective alpha-/beta-blocker	31 (8.7)	17 (8.0)	14 (9.8)	0.77	8 (7.2)	10 (8.8)	0.048
CCB	244 (68.7)	141 (66.5)	103 (72.0)	0.68	84 (75.7)	85 (74.6)	0.010
ECOG
PS 0	196 (31.5)	114 (31.1)	82 (32.2)	0.84	63 (31.2)	69 (34.2)	0.063
PS 1	426 (68.5)	253 (68.9)	173 (67.8)		139 (68.8)	133 (65.8)
Child-Pugh class
A	611 (98.2)	366 (99.7)	245 (96.1)	0.002	201 (99.5)	199 (98.5)	0.100
B	11 (1.8)	1 (0.3)	10 (3.9)		1 (0.5)	3 (1.5)
Ascites
None	575 (92.4)	346 (94.3)	229 (89.8)	0.05	191 (94.6)	189 (93.6)	0.102
Mild	45 (7.2)	21 (5.7)	24 (9.4)		11 (5.4)	12 (5.9)
Severe	2 (0.4)	0 (0.0)	2 (0.8)		0 (0.0)	1 (0.5)
ALBI score	−2.0 [–2.2, −1.7]	−2.0 [–2.2, −1.7]	−2.0 [–2.2, −1.6]	0.87	−1.9 [–2.2, −1.6]	−2.0 [–2.2, −1.7]	0.12
Grade of ALBI
Grade 1	14 (2.3)	6 (1.6)	8 (3.1)	0.05	5 (2.5)	6 (3.0)	0.080
Grade 2	533 (85.7)	325 (88.6)	208 (81.6)		170 (84.2)	174 (86.1)
Grade 3	75 (12.0)	36 (9.8)	39 (15.3)		27 (13.3)	22 (10.9)
Tumor characteristics
BCLC stage
B	81 (13.0)	53 (14.4)	28 (11.0)	0.25	27 (13.4)	25 (12.4)	0.030
C	541 (87.0)	314 (85.6)	227 (89.0)		175 (86.6)	177 (87.6)
Extrahepatic metastasis	442 (71.1)	248 (67.6)	194 (76.1)	0.03	152 (75.2)	150 (74.3)	0.023
MVI	261 (42.0)	163 (44.4)	98 (38.4)	0.16	74 (36.6)	76 (37.6)	0.020
Laboratory findings
AFP, ng/mL	74.2 [6.0–1,430.5]	79.2 [5.5–1,088.8]	62.0 [6.4–2,542.1]	0.54	83.8 [5.5–1,394.5]	38.0 [6.3–1,632.7]	0.011
PIVKA-II, mAU/mL	312.5 [42.0–3,987.0]	265.0 [37.5–2,538.5]	400.0 [56.0–7,221.5]	0.01	312.5 [42.0–3,404.0]	281.0 [54.0–4,449.0]	0.007
AST, IU/L	36.0 [28.0–58.0]	35.0 [28.0–52.0]	40.0 [28.0–65.5]	0.01	35.5 [29.0–55.0]	37.5 [28.0–59.0]	0.025
ALT, IU/L	23.0 [15.0–37.0]	22.0 [14.0–35.5]	24.0 [17.0–42.0]	0.01	26.0 [15.0–38.0]	23.0 [17.0–38.0]	0.041
Total bilirubin, mg/dL	0.7 [0.5–0.9]	0.7 [0.5–0.9]	0.7 [0.5–1.0]	0.40	0.7 [0.5–0.9]	0.6 [0.5–0.9]	0.041
Creatinine, mg/dL	0.9 [0.7–1.0]	0.8 [0.7–1.0]	0.9 [0.7–1.0]	0.32	0.9 [0.7–1.0]	0.9 [0.7–1.0]	0.095
eGFR, mL/min/1.73 m2	90.0 [76.0–100.0]	90.0 [78.0–100.0]	90.0 [75.0–101.0]	0.98	91.0 [78.0–100.0]	90.0 [76.0–101.0]	0.069
Albumin, g/dL	3.7 [3.4–3.9]	3.7 [3.4–3.9]	3.7 [3.3–4.0]	0.82	3.7 [3.3–3.9]	3.7 [3.4–4.0]	0.110
Platelets, ×1,0003/μL	144.0 [103.0–206.0]	140.0 [105.0–202.0]	152.0 [98.5–212.5]	0.19	146.5 [104.0–211.0]	152.0 [103.0–205.0]	0.037
INR	1.0 [1.0–1.1]	1.0 [1.0–1.1]	1.0 [1.0–1.1]	0.25	1.1 [1.0–1.1]	1.0 [1.0–1.1]	0.040

Continuous variables are described in mean ± standard deviation or median [IQR] and categorical variables are described in n (%) as appropriate.

ACE inhibitor, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; AFP, alpha-fetoprotein; ALBI, albumin-bilirubin; ALT, alanine aminotransferase; AST, aspartate aminotransferase; Atezo/Bev, atezolizumab plus bevacizumab; BCLC, Barcelona Clinic Liver Cancer; CCB, calcium channel blocker; DBP, diastolic blood pressure; ECOG PS, Eastern Cooperative Oncology Group performance status; eGFR, estimated glomerular filtration rate; INR, international normalized ratio; LEN, lenvatinib; PIVKA-II, protein induced by vitamin K absence-II; SBP, systolic blood pressure; SMD, standard mean deviation.

Of the 622 study patients, 367 patients were treated with Atezo/Bev and 255 with LEN. The Atezo/Bev group was older (61.7 vs. 58.7 years) with a lower median SBP and diastolic blood pressure (125 mm Hg vs. 128 mm Hg and 76 mm Hg vs. 79 mm Hg, respectively) compared with the LEN group. The Atezo/Bev group exhibited better underlying liver function (Child-Pugh class A: 99.7% vs. 96.1%) and had a lower level of PIVKA-II (265.0 vs. 400.0 mAU/mL). After PS matching, baseline characteristics were well balanced between the two groups, with a standardized mean difference of <0.1.

Of all patients, 269 (43.2%) experienced adverse events during treatment (Atezo/Bev: n = 137 and LEN: n = 132). In the Atezo/Bev group, 85 patients (62.0%) discontinued both drugs, while 43 (31.5%) discontinued only Bev, primarily due to proteinuria (38.6%). In the LEN group, 65 patients (49.2%) discontinued treatment, and 66 (50.0%) reduced their dose, with poor tolerance leading to 19.8% of interruptions (online suppl. Table S1; for all online suppl. material, see https://doi.org/10.1159/000541621).

Among those who developed proteinuria during Atezo/Bev treatment (n = 68), 36 patients (52.9%) discontinued both drugs, 14 (20.6%) discontinued only Bev, and 1 (1.5%) had a 30% dose reduction. In the LEN group (n = 38), 16 patients (42.1%) discontinued LEN, and 3 (7.9%) underwent dose reduction (online suppl. Table S2).

Cumulative Incidence of Proteinuria in the Entire Cohort

Over the median follow-up period of 6.3 months (95% confidence interval [CI]: 5.6–7.1), 106 patients (17.0%) developed proteinuria. The cumulative incidence rates of proteinuria at 4 months, 8 months, 12 months, and 20 months from initiation of the treatment were 9.9%, 17.3%, 27.5%, and 34.0%, respectively (Fig. 2a). Multivariable analysis revealed several significant risk factors for proteinuria: Atezo/Bev treatment (adjusted hazard ratio [aHR]: 1.57; 95% CI: 1.03–2.42; p = 0.04), hypertension (aHR: 2.27; 95% CI: 1.04–4.97; p = 0.04), diabetes (aHR: 1.64; 95% CI: 1.03–2.61; p = 0.04), Child-Pugh class B (aHR: 3.43; 95% CI: 1.34–8.78; p = 0.01), MVI (aHR: 1.58; 95% CI: 1.04–2.38; p = 0.03), and an eGFR ≤60 mL/min/1.73 m² (aHR: 3.21; 95% CI: 1.84–5.62; p < 0.001; Table 2; online suppl. Fig. S1).

Fig. 2.

Cumulative incidence of proteinuria in the entire cohort, between the Atezo/Bev and LEN group before PS matching, and between the Atezo/Bev and LEN group after PS matching. a Entire cohort. b Atezo/Bev versus LEN group before PS matching. c Atezo/Bev versus LEN group after PS matching. Atezo/Bev, atezolizumab plus bevacizumab; LEN, lenvatinib; PS, propensity score.

Table 2.

Univariate and multivariable Cox regression analysis for proteinuria in the entire cohort

Variables	Univariate analysis			Multivariable analysis
	HR	95% CI	p value	aHR	95% CI	p value
Systemic treatment
LEN	1	Reference	0.20	1	Reference	0.04
Atezo/Bev	1.31	0.88–1.95		1.57	1.03–2.42
Sex
Female	1	Reference	0.70	-	-	-
Male	1.13	0.66–1.96
Etiologies
Viral	1	Reference	0.80	-	-	-
Nonviral	1.04	0.67–1.62
Age, 1-year increments	1.02	1.00–1.04	0.02	0.99	0.97–1.01	0.50
Hypertension	1.85	1.22–2.81	0.004	2.27	1.04–4.97	0.04
Diabetes	2.22	1.48–3.33	<0.001	1.64	1.03–2.61	0.04
SBP <140 mm Hg	1	Reference	0.30	-	-	-
SBP ≥140 mm Hg	0.80	0.51–1.25
DBP <90 mm Hg	1	Reference	>0.90	-	-	-
DBP ≥90 mm Hg	0.98	0.56–1.72
Antihypertensive agent
Nonea	1	Reference		1	Reference
ACE inhibitor/ARB	1.76	1.10–2.83	0.02	0.55	0.24–1.29	0.20
Othersb	1.52	0.94–2.45	0.09	0.72	0.33–1.56	0.40
Child-Pugh class A	1	Reference	<0.001	1	Reference	0.01
Child-Pugh class B	7.83	3.38–18.1		3.43	1.34–8.78
Grade of ALBI
Grade 1/2	1	Reference	0.30	-	-	-
Grade 3	1.39	0.72–2.67
Extrahepatic metastases	0.96	0.63–1.45	0.90	-	-	-
MVI	1.42	0.97–2.10	0.07	1.58	1.04–2.38	0.03
AFP <400 ng/mL	1	Reference	0.08	1	Reference	0.20
AFP ≥400 ng/mL	1.43	0.96–2.13		1.31	0.86–2.01
ECOG PS 0	1	Reference	0.05	1	Reference	0.20
ECOG PS 1	1.52	1.00–2.32		1.31	0.85–2.03
Baseline eGFR >60 mL/min/1.73 m2	1	Reference	<0.001	1	Reference	<0.001
Baseline eGFR ≤60 mL/min/1.73 m2	3.78	2.44–5.86		3.21	1.84–5.62

aHR, adjusted hazard ratio; ACE inhibitor, angiotensin-converting enzyme inhibitor; AFP, alpha-fetoprotein; ALBI, albumin-bilirubin; ARB, angiotensin receptor blocker; Atezo/Bev, atezolizumab plus bevacizumab; CI, confidence interval; DBP, diastolic blood pressure; ECOG PS, Eastern Cooperative Oncology Group performance status; eGFR, estimated glomerular filtration rate; HR, hazard ratio; LEN, lenvatinib; SBP, systolic blood pressure.

^aOf 267 patients, 255 patients had no hypertension, and 12 patients had hypertension without medication.

^bOthers include CCBs, β-blockers, and alpha-blockers.

Cumulative Incidence of Proteinuria between the Atezo/Bev and LEN Groups

Based on the type of first-line systemic treatment, the Atezo/Bev group exhibited a higher cumulative incidence rate of proteinuria compared with the LEN group within the entire cohort (Fig. 2b). The cumulative incidence rates of proteinuria in the Atezo/Bev group at 4, 8, 12, and 20 months after treatment initiation were 9.9%, 18.2%, 32.4%, and 40.7%, respectively. In contrast, the LEN group exhibited rates of 9.8%, 15.9%, 20.5%, and 25.6% at the same time points.

Among Atezo/Bev-treated patients who developed proteinuria, the median cumulative dose of Bev was 6,540 [3,525–13,400] mg, with a median time of 3.7 [1.6–9.1] months to reach the maximal dose. For LEN, the median cumulative dose was 738 [339–1,668] mg, with a shorter median time of 2.7 [1.0–6.3] months.

In the PS-matched cohort, the Atezo/Bev group exhibited higher cumulative incidence rates of proteinuria compared with the LEN group at 4, 8, 12, and 20 months after treatment initiation (11.1% vs. 7.9%, 22.1% vs. 14.3%, 36.5% vs. 17.2%, and 39.4% vs. 23.1%, respectively; Fig. 2c). The Atezo/Bev group also had a 1.74 times higher risk of developing proteinuria compared with the LEN group (HR: 1.74; 95% CI: 1.05–2.87; p = 0.03).

Across all predefined subgroups, the Atezo/Bev group exhibited a higher risk of proteinuria compared with the LEN group (Fig. 3). Notably, the increased risk of proteinuria in the Atezo/Bev group compared with the LEN group was more pronounced in patients with MVI (HR: 2.84; 95% CI: 1.23–6.54; p = 0.01) than those without MVI (HR: 1.31; 95% CI: 0.69–2.47; p = 0.41, p for interaction = 0.14).

Fig. 3.

Forest plot of the subgroup analysis for proteinuria between the Atezo/Bev and LEN groups after PS matching. Atezo/Bev, atezolizumab plus bevacizumab; LEN, lenvatinib; PS, propensity score.

In the multivariable analysis to identify the risk factors for proteinuria in each group, Child-Pugh class B (aHR: 6.19; 95% CI: 2.16–17.80; p < 0.001) and a decreased baseline eGFR ≤60 mL/min/1.73 m² (aHR: 3.34; 95% CI: 1.60–6.97; p = 0.001) were identified as risk factors for proteinuria in the LEN group, while diabetes (aHR: 1.87; 95% CI: 1.03–3.37; p = 0.04), MVI (aHR: 2.28; 95% CI: 1.32–3.92; p = 0.003), and an eGFR of ≤60 mL/min/1.73 m² (aHR: 2.86; 95% CI: 1.32–6.17; p = 0.01) were significantly associated with a risk of proteinuria in the Atezo/Bev group (Table 3).

Table 3.

Univariate and multivariable Cox regression analysis for proteinuria in each Atezo/Bev and LEN group

Variables	Atezo/Bev						LEN
	univariate			multivariable			univariate			multivariable
	HR	95% CI	p value	aHR	95% CI	p value	HR	95% CI	p value	aHR	95% CI	p value
Sex
Female	1	Reference	0.80	-	-	-	1	Reference	0.70	-	-	-
Male	1.08	0.55–2.12					1.21	0.47–3.12
Etiologies
Viral	1	Reference	>0.90	-	-	-	1	Reference	0.90	-	-	-
Nonviral	1.03	0.60–1.77					1.08	0.51–2.28
Age, 1-year increments	1.03	1.01–1.06	0.01	1.01	0.98–1.04	0.50	1.00	0.97–1.04	0.80	-	-	-
Hypertension	1.74	1.06–2.87	0.03	2.32	0.89–6.09	0.09	2.45	1.12–5.36	0.02	1.38	0.17–11.1	0.80
Diabetes	2.18	1.26–3.79	0.01	1.87	1.03–3.37	0.04	2.76	1.46–5.24	0.002	1.83	0.90–3.69	0.09
SBP <140 mm Hg	1	Reference	0.30	-	-	-	1	Reference	0.90	-	-	-
SBP ≥140 mm Hg	0.75	0.42–1.35					0.94	0.48–1.87
DBP <90 mm Hg	1	Reference	0.70	-	-	-	1	Reference	0.60	-	-	-
DBP ≥90 mm Hg	0.82	0.33–2.05					1.25	0.59–2.64
Antihypertensive agent
Nonea	1	Reference		-	-	-	1	Reference		1	Reference
ACE inhibitor/ARB	1.89	1.08–3.31	0.03				1.79	0.74–4.32	0.20	0.59	0.07–4.80	0.60
Othersb	0.98	0.53–1.83	>0.90				2.97	1.33–6.60	0.01	1.19	0.15–9.11	0.90
Child-Pugh class A	1	Reference	0.08	0.55	0.06–4.65	0.60	1	Reference	<0.001	1	Reference	<0.001
Child-Pugh class B	5.98	0.82–43.5					7.78	2.92–20.70		6.19	2.16–17.80
Grade of ALBI
Grade 1/2	1	Reference	>0.90	-	-	-	1	Reference	0.14	-	-	-
Grade 3	0.99	0.36–2.72					1.98	0.81–4.83
Extrahepatic metastases	1.18	0.71–1.97	0.50	-	-	-	0.77	0.37–1.59	0.50	-	-	-
MVI	1.70	1.06–2.75	0.03	2.28	1.32–3.92	0.003	0.98	0.49–1.95	>0.90	-	-	-
AFP <400 ng/mL	1	Reference	0.07	1	Reference	0.40	1	Reference	0.70	-	-	-
AFP ≥400 ng/mL	1.57	0.96–2.57		1.26	0.75–2.12		1.16	0.58–2.30
ECOG PS 0	1.49	0.88–2.53	0.13	-	-	-	1	Reference	0.20	-	-	-
ECOG PS 1							1.55	0.76–3.15
Baseline eGFR >60 mL/min/1.73 m2	1	Reference	<0.001	2.86	1.32–6.17	0.01	1	Reference	<0.001	1	Reference	0.001
Baseline eGFR ≤60 mL/min/1.73 m2	3.42	1.86–6.31					4.86	2.50–9.43		3.34	1.60–6.97

aHR, adjusted hazard ratio; ACE inhibitor, angiotensin-converting enzyme inhibitor; AFP, alpha-fetoprotein; ALBI, albumin-bilirubin; ARB, angiotensin receptor blocker; Atezo/Bev, atezolizumab plus bevacizumab; DBP, diastolic blood pressure; ECOG PS, Eastern Cooperative Oncology Group performance status; eGFR, estimated glomerular filtration rate; HR, hazard ratio; SBP, systolic blood pressure.

^aIn the Atezo/Bev group, 179 patients did not take an antihypertensive agent (150 patients had no hypertension, and 29 patients had hypertension without medication). In the LEN group, 112 patients did not take an antihypertensive agent (105 patients had no hypertension, and 7 patients had hypertension without medication).

^bOthers included CCBs, β-blockers, and alpha-blockers.

Cumulative Incidence of Severe Proteinuria

Among 106 patients with proteinuria, 36 developed severe proteinuria with cumulative incidence rates of 3.5%, 6.4%, 10.4%, and 14.2% at 4, 8, 12, and 20 months, respectively (online suppl. Fig. S2a). In the multivariable analysis, Child-Pugh class B (aHR: 6.03; 95% CI: 1.60–22.70; p = 0.01) and an eGFR ≤60 mL/min/1.73 m² (aHR: 3.43; 95% CI: 1.39–8.49; p = 0.01) were identified as risk factors for severe proteinuria (online suppl. Table S3). Moreover, compared with the LEN group, the Atezo/Bev group tended to have a higher risk of severe proteinuria (HR: 1.27; 95% CI: 0.64–2.52; p = 0.50; online suppl. Table S3; online suppl. Fig. S2b), a finding consistently observed in the PS-matched cohort (HR: 1.39; 95% CI: 0.62–3.13; p = 0.42; online suppl. Fig. S2c), although it did not reach statistical insignificance. In the LEN group, multivariable analysis identified diabetes (aHR: 5.27; 95% CI: 1.46–19.00; p = 0.01), Child-Pugh class B (aHR: 19.6; 95% CI: 3.76–103.00; p < 0.001), and an eGFR of ≤60 mL/min/1.73 m² (aHR: 4.58; 95% CI: 1.36–15.40; p = 0.01) as risk factors for severe proteinuria in the LEN group, while no statistically significant risk factors for severe proteinuria were identified in the Atezo/Bev group (online suppl. Table S4).

Impact of Proteinuria on Oncological Outcomes

Next, we examined the impact of proteinuria on oncological outcomes by conducting a time-dependent Cox regression and a 6-month landmark analyses to mitigate the immortal time bias. There was no difference in the proportion of patients who received sequential treatment between the group that developed proteinuria and the group that did not (p = 0.66). Among patients who received treatment for ≥6 months, the objective response rate did not significantly differ between the group that developed proteinuria and the group that did not, both in the entire cohort (online suppl. Table S5) and within each treatment group (online suppl. Table S6).

In the 6-month landmark analysis, there was no significant difference in either OS or PFS between the two groups in the entire cohort (OS: aHR: 1.15; 95% CI: 0.83–1.59; p = 0.41 and PFS: aHR: 1.02; 95% CI: 0.77–1.36; p = 0.88; online suppl. Fig. S3), in the Atezo/Bev group (OS: aHR: 1.26; 95% CI: 0.81–1.93; p = 0.30 and PFS: aHR: 0.96; 95% CI: 0.65–1.41; p = 0.84), and in the LEN group (OS: aHR: 1.00; 95% CI: 0.60–1.68; p = 0.99 and PFS: aHR: 1.15; 95% CI: 0.73–1.81; p = 0.55; online suppl. Fig. S4). Similarly, in the time-dependent Cox analyses, where proteinuria was treated as a time-dependent variable, proteinuria had no significant impact in either OS or PFS in the entire cohort (OS: aHR: 0.96; 95% CI: 0.72–1.29; p = 0.81 and PFS: aHR: 1.16; 95% CI: 0.87–1.53; p = 0.31), in the Atezo/Bev group (OS: aHR: 0.78; 95% CI: 0.52–1.16; p = 0.22 and PFS: aHR: 1.06; 95% CI: 0.72–1.57; p = 0.76), and in the LEN group (OS: aHR: 1.18; 95% CI: 0.76–1.85; p = 0.85 and PFS: aHR: 1.18; 95% CI: 0.75–1.85; p = 0.48).

Discussion

In the present study, we elucidated the risk factors for proteinuria in patients with unresectable HCC treated with Atezo/Bev and LEN, the predominant first-line systemic treatment for HCC. We confirmed that diabetes, hypertension, Child-Pugh class B liver function, MVI, and decreased eGFR are independent risk factors for proteinuria in these patients. Notably, we observed nearly twice the incidence of proteinuria with Atezo/Bev compared with LEN, persisting even after PS matching. Among these risk factors, Child-Pugh class B and decreased eGFR emerged as potential risk factors for severe proteinuria, potentially necessitating treatment cessation. The use of Atezo/Bev, compared to LEN, tended to be associated with an increased risk of severe proteinuria. Upon individual drug analysis, Child-Pugh class B and decreased eGFR were identified as risk factors for proteinuria in the LEN group, while diabetes, MVI, and decreased eGFR were implicated in the Atezo/Bev group. Interestingly, the subgroup analysis revealed that when MVI was absent, there was no statistically significant difference in the risk of proteinuria between Atezo/Bev and LEN; however, in patients with MVI treated with Atezo/Bev, the risk increased by nearly threefold compared with LEN.

In the landscape of HCC treatment, the use of various TKIs and ICIs has significantly improved patient prognosis. However, we are witnessing the emergence of immune-related adverse events and novel complications, such as hypertension, proteinuria, thyroid dysfunction, and gastrointestinal bleeding. Among these, proteinuria poses a particularly formidable challenge in management [26]. If left untreated, proteinuria can precipitate substantial deterioration in patient quality of life and prognosis. Deterioration can manifest as hypoalbuminemia accompanied by ascites, peripheral edema, and shortness of breath, which may ultimately result in treatment discontinuation [14]. Importantly, proteinuria tends to persist even after treatment cessation, distinguishing it from other complications. In the current study, while proteinuria led to treatment modification or interruption, it did not directly affect oncological prognosis, both in the entire cohort and within each treatment group. This aligns with the post hoc analysis by Kudo et al. [27] of the IMbrave 150 study, which found no difference in treatment efficacy or safety between patients who skipped Bev due to adverse events and those who did not. Therefore, heightened vigilance for complications such as proteinuria resulting from systemic therapy, along with timely intervention, is crucial to improving patient quality of life and prognosis.

The increased rates of proteinuria observed in patients treated with Atezo/Bev and LEN versus sorafenib have been corroborated in pivotal trials such as REFLECT and IMbrave150 [2, 3]. However, in our study, the incidence rates of proteinuria for each treatment were lower than those reported in these major clinical trials (Atezo/Bev: 18.5% vs. 29% in IMbrave150 and LEN: 14.9% vs. 25% in REFLECT). This variance could be attributed to the lower proportion of non-B, non-C hepatitis (23.3% vs. 30.0% in IMbrave 150 and 29.0% in REFLECT) in our study compared with previous trials. Non-B and non-C hepatitis is associated with comorbidities such as diabetes mellitus [28, 29], which increase susceptibility to proteinuria [2, 3]. Therefore, the lower prevalence of this risk factor in our study population might have contributed to the relatively lower incidence of proteinuria observed. In a real-world setting, the incidence of proteinuria in patients with HCC receiving Atezo/Bev or LEN was significantly lower in our study compared with previous studies (Atezo/Bev: 18.5% vs. 34.4% in Ando et al.’s [18] study and LEN: 14.9% vs. 32.4% in Ikesue et al.’s [19] study). However, it is important to note that these studies included patients with prior systemic treatment history, rather than treatment-naïve patients, as in our study. The inclusion of these patients suggests a possibility that glomerular injury caused by previous TKI use may exacerbate renal function and hypertension, thereby increasing susceptibility to the development of proteinuria in subsequent treatments.

Anti-VEGF TKIs such as LEN and monoclonal antibodies against VEGF such as bevacizumab exert their therapeutic effects by inhibiting the VEGF signaling pathway. However, this inhibition can induce systemic and intraglomerular hypertension, leading to proteinuria [30]. Notably, monoclonal antibodies such as bevacizumab, which is administered at double the dose in HCC treatment compared to other malignancies (15 mg/kg vs. 7.5 mg/kg every 3 weeks) in particular, exert a more potent inhibitory effect on the VEGF signaling pathway compared with anti-VEGF TKIs such as LEN. In addition, given that thrombotic microangiopathy is the main pathology underlying proteinuria in the context of anti-VEGF therapy, the higher risk of thromboembolic events associated with Atezo/Bev compared to LEN may partly explain the increased incidence of proteinuria observed with Atezo/Bev compared to LEN [31]. Therefore, the incidence of proteinuria appears to be higher in patients treated with Atezo/Bev compared with LEN. Although the precise pathophysiology of ICI-induced proteinuria remains elusive, studies suggest that disruptions in T-cell activity may instigate an inflammatory cascade within the glomerular structure [32]. This immune dysregulation can precipitate various forms of kidney injury, including acute tubular necrosis, thrombotic microangiopathy, and multiple glomerular disease [33‒35]. Combination therapy involving ICIs and monoclonal antibodies against VEGF, such as Atezo/Bev, could potentially exacerbate proteinuria development through a synergistic mechanism. The concurrent action of these agents could amplify the inflammatory response within the glomerular microenvironment, thereby heightening the risk of proteinuria. This synergistic effect may elucidate the higher incidence of proteinuria observed in the Atezo/Bev group compared with the LEN group. While direct comparisons between the different treatment regimens are challenging due to variations in study designs, similar trends have been observed in pivotal trials. In the IMbrave 150 trial evaluating Atezo/Bev, the incidence of proteinuria was 29%, while in the REFLECT trial investigating LEN, it stood at 25%. This disparity suggests a slightly elevated incidence of proteinuria in patients treated with Atezo/Bev, aligning with the proposed mechanistic insights [2, 3].

Studies investigating the risk of proteinuria during treatment in patients with HCC are scarce. Ikesue et al. [19] reported pre-existing proteinuria and low baseline eGFR as significant predictors of higher proteinuria risk in patients with HCC receiving LEN. Similarly, Ando et al. [18] noted that poor baseline eGFR, concurrent use of antihypertensive medication, and high SBP (≥130 mm Hg) were associated with an increased risk of proteinuria in patients treated with Atezo/Bev. In particular, MVI and Child-Pugh class B were identified as risk factors for proteinuria, potentially associated with portal hypertension. Portal hypertension may lead to decreased renal perfusion through renal vasoconstriction, accelerating glomerular injury and thereby contributing to the development of proteinuria. Moreover, recent observations suggest that treatment with Atezo/Bev or LEN could lead to an increase in portal pressure due to anti-VEGF-induced impairment of neovascularization in the mesenteric vascular bed or portosystemic collaterals [36, 37]. In patients with MVI, where portal pressure elevation is already a significant concern, these effects can be amplified by Atezo/Bev, a more potent anti-VEGF agent compared with LEN, resulting in a nearly threefold higher incidence of proteinuria. However, further studies are needed to validate this finding.

Our study had several limitations. First, being a retrospective study, it was inherent to biases. However, we attempted to mitigate bias by enrolling patients consecutively with relatively large numbers and conducting rigorous statistical analyses, including multivariable-adjusted and PS-matched analyses. Second, as a single-center study conducted in Korea, where HBV is predominantly associated with HCC, our study may have limited generalizability to populations with different HCC etiologies or diverse geographic regions. Lastly, dosage modification or delays due to other adverse events were not accounted for; however, all analyses were conducted based on the intention-to-treat principle.

In conclusion, our study identified several significant risk factors associated with an increased likelihood of developing proteinuria in patients with unresectable HCC treated with Atezo/Bev or LEN as first-line systemic treatment, including the presence of diabetes, hypertension, Child-Pugh class B liver function, MVI, and decreased eGFR. Additionally, our findings underscored a higher risk of proteinuria with Atezo/Bev, particularly when accompanied by MVI, where the risk was nearly tripled compared with LEN. These observations highlight the importance of vigilant monitoring for proteinuria in patients with these identified risk factors, enabling timely intervention to mitigate potential complications and optimize treatment outcomes.

Statement of Ethics

The need for informed consent was waived by the Institutional Review Board of Asan Medical Center, where study protocol and ethical approval (IRB No. 2023-1253) were granted for this study.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Author Contributions

W.M. Choi is the guarantor of the article. All authors had full access to the data used in this study and take responsibility for its integrity and the accuracy of the analyses. J.Yang and W.M. Choi were responsible for the conception and design of the study; the acquisition, analysis, and interpretation of the data; drafting of the manuscript; and the statistical analyses. H.D. Kim, J. Choi, C. Yoo, D.B. Lee, J.H. Shim, K.M. Kim, Y.S. Lim, and H.C. Lee were responsible for the data acquisition and critical revision of the manuscript. All authors read and approved the final version of the manuscript.

References

1. Reig M, Forner A, Rimola J, Ferrer-Fàbrega J, Burrel M, Garcia-Criado Á,. BCLC strategy for prognosis prediction and treatment recommendation: the 2022 update.JHepatol. 2022;76(3): 681–93.

2. Kudo M, Finn RS, Qin S, Han K-H, Ikeda K, Piscaglia F,. Lenvatinib versus sorafenib in first-line treatment of patients with unresectable hepatocellular carcinoma:arandomised phase3non-inferiority trial. Lancet. 2018;391(10126): 1163–73.

3. Cheng AL, Qin S, Ikeda M, Galle PR, Ducreux M, Kim TY,. Updated efficacy and safety data from IMbrave150: atezolizumab plus bevacizumab vs. sorafenib for unresectable hepatocellular carcinoma.JHepatol. 2022;76(4): 862–73.

4. Sangro B, Sarobe P, Hervás-Stubbs S, Melero I. Advances in immunotherapy for hepatocellular carcinoma. Nat Rev Gastroenterol Hepatol. 2021;18(8): 525–43.

5. Ng KYY, Tan SH, Tan JJE, Tay DSH, Lee AWX, Ang AJS,. Impact of immune-related adverse events on efficacy of immune checkpoint inhibitors in patients with advanced hepatocellular carcinoma. Liver Cancer. 2021;11(1):9– 21.

6. Wu S, Kim C, Baer L, Zhu X. Bevacizumab increases risk for severe proteinuria in cancer patients.JAm Soc Nephrol. 2010;21(8): 1381–9.

7. Estrada CC, Maldonado A, Mallipattu SK. Therapeutic inhibition of VEGF signaling and associated nephrotoxicities.JAm Soc Nephrol. 2019;30(2): 187–200.

8. Herrmann SM, Perazella MA. Immune checkpoint inhibitors and immune-related adverse renal events. Kidney Int Rep. 2020;5(8): 1139–48.

9. Ikeda K, Kudo M, Kawazoe S, Osaki Y, Ikeda M, Okusaka T,. Phase2study of lenvatinib in patients with advanced hepatocellular carcinoma.JGastroenterol. 2017;52(4): 512–9.

10. National Cancer Institute. Protocol development. Cancer therapy evaluation program: Common Terminology Criteria for adverse events(CTCAE) version 5.0[cited 2024 January 26]. Available from: https://ctep.cancer.gov/protocoldevelopment/electronic_applications/docs/CTCAE_v5_Quick_Reference_5x7.pdf

11. Manohar S, Jhaveri KD, Perazella MA. Immunotherapy-related acute kidney injury. Adv Chronic Kidney Dis. 2021;28(5): 429–37.e1.

12. Seethapathy H, Herrmann SM, Sise ME. Immune checkpoint inhibitors and kidney toxicity: advances in diagnosis and management. Kidney Med. 2021;3(6): 1074–81.

13. Kandula P, Agarwal R. Proteinuria and hypertension with tyrosine kinase inhibitors. Kidney Int. 2011;80(12): 1271–7.

14. Hatanaka T, Hiraoka A, Tada T, Hirooka M, Kariyama K, Tani J,. Association of early bevacizumab interruption with efficacy of atezolizumab plus bevacizumab for advanced hepatocellular carcinoma:alandmark analysis. Hepatol Res. 2022;52(5): 462–70.

15. Nakagawa M, Inoue M, Ogasawara S, Maruta S, Okubo T, Itokawa N,. Clinical effects and emerging issues of atezolizumab plus bevacizumab in patients with advanced hepatocellular carcinoma from Japanese real-world practice. Cancer. 2023;129(4): 590–9.

16. Zhang ZF, Wang T, Liu LH, Guo HQ. Risks of proteinuria associated with vascular endothelial growth factor receptor tyrosine kinase inhibitors in cancer patients:asystematic review and meta-analysis. PLoS One. 2014;9(3): e90135.

17. Jagieła J, Bartnicki P, Rysz J. Nephrotoxicity asacomplication of chemotherapy and immunotherapy in the treatment of colorectal cancer, melanoma and non-small cell lung cancer. IntJMol Sci. 2021;22(9): 4618.

18. Ando Y, Kawaoka T, Kosaka M, Shirane Y, Johira Y, Miura R,. Risk factors for early onset of proteinuria in patients receiving atezolizumab plus bevacizumab for unresectable hepatocellular carcinoma. Liver Cancer. 2023;12(3): 251–61.

19. Ikesue H, Yamamoto H, Hirabatake M, Hashida T, Chung H, Inokuma T,. Risk factors of proteinuria in patients with hepatocellular carcinoma receiving lenvatinib. Biol Pharm Bull. 2022;45(3): 333–8.

20. Korean Liver Cancer Association KLCA and National Cancer Center NCC Korea. 2022 KLCA-NCC Korea practice guidelines for the management of hepatocellular carcinoma. Clin Mol Hepatol. 2022;28(4): 583–705.

21. Galle PR, Forner A, Llovet JM, Mazzaferro V, Piscaglia F, Raoul J-L,. EASL clinical practice guidelines: management of hepatocellular carcinoma.JHepatol. 2018;69(1): 182–236.

22. Singal AG, Llovet JM, Yarchoan M, Mehta N, Heimbach JK, Dawson LA,. AASLD Practice Guidance on prevention, diagnosis, and treatment of hepatocellular carcinoma. Hepatology. 2023;78(6).

23. Ginsberg JM, Chang BS, Matarese RA, Garella S. Use of single voided urine samples to estimate quantitative proteinuria.NEnglJMed. 1983;309(25): 1543–6.

24. Kristal B, Shasha SM, Labin L, Cohen A. Estimation of quantitative proteinuria by using the protein-creatinine ratio in random urine samples. AmJNephrol. 1988;8(3): 198–203.

25. Evans TRJ, Kudo M, Finn RS, Han KH, Cheng AL, Ikeda M,. Urine protein:creatinine ratio vs 24-hour urine protein for proteinuria management: analysis from the phase3REFLECT study of lenvatinib vs sorafenib in hepatocellular carcinoma. BrJCancer. 2019;121(3): 218–21.

26. Kai M, Hikita H, Kazuki M, Tahata Y, Shinkai K, Doi A,. Clinical factors associated with the therapeutic efficacy of atezolizumab plus bevacizumab in patients with unresectable hepatocellular carcinoma:amulticenter prospective observational study. PLoS One. 2024;19(1): e0294590.

27. Kudo M, Tsuchiya K, Shao YY, Finn RS, Galle PR, Ducreux M,. Impact of bevacizumab being skipped due to adverse events of special interest for bevacizumab in patients with unresectable hepatocellular carcinoma treated with atezolizumab plus bevacizumab: an exploratory analysis of the phase III IMbrave150 study. Liver Cancer. 2024;13(4): 401–12.

28. Nagaoki Y, Hyogo H, Ando Y, Kosaka Y, Uchikawa S, Nishida Y,. Increasing incidence of non-HBV- and non-HCV-related hepatocellular carcinoma: single-institution 20-year study. BMC Gastroenterol. 2021;21(1): 306.

29. Shah PA, Patil R, Harrison SA. NAFLD‐related hepatocellular carcinoma: the growing challenge. Hepatology. 2023;77(1): 323–38.

30. Nervo A, Retta F, Ragni A, Piovesan A, Mella A, Biancone L,. Nephrotoxicity in advanced thyroid cancer treated with tyrosine kinase inhibitors: an update. Crit Rev Oncol Hematol. 2021;168:103533.

31. Ben Khaled N, Möller M, Jochheim LS, Leyh C, Ehmer U, Böttcher K,. Atezolizumab/bevacizumab or lenvatinib in hepatocellular carcinoma: multicenter real-world study with focus on bleeding and thromboembolic events. JHEP Rep. 2024;6(6): 101065.

32. Rao Ullur A, Côté G, Pelletier K, Kitchlu A. Immunotherapy in oncology and the kidneys:aclinical review of the evaluation and management of kidney immune-related adverse events. Clin Kidney J. 2023;16(6): 939–51.

33. Meraz-Muñoz A, Amir E, Ng P, Avila-Casado C, Ragobar C, Chan C,. Acute kidney injury associated with immune checkpoint inhibitor therapy: incidence, risk factors and outcomes.JImmunother Cancer. 2020;8(1): e000467.

34. Kitchlu A, Jhaveri KD, Wadhwani S, Deshpande P, Harel Z, Kishibe T,.Asystematic review of immune checkpoint inhibitor-associated glomerular disease. Kidney Int Rep. 2021;6(1): 66–77.

35. Seethapathy H, Zhao S, Chute DF, Zubiri L, Oppong Y, Strohbehn I,. The incidence, causes, and risk factors of acute kidney injury in patients receiving immune checkpoint inhibitors. ClinJAm Soc Nephrol. 2019;14(12): 1692–700.

36. Hidaka H, Uojima H, Nakazawa T, Shao X, Hara Y, Iwasaki S,. Portal hemodynamic effects of lenvatinib in patients with advanced hepatocellular carcinoma:aprospective cohort study. Hepatol Res. 2020;50(9): 1083–90.

37. Lin T-Y, Su T-H. Progression of portal hypertension after atezolizumab plus bevacizumab for hepatocellular carcinoma-reportacase and literature review.JFormos Med Assoc. 2024;123(8): 916–9.