General demographic characteristics of the population.
We included 154 patients who underwent RB between April 2014 and September 2020. The main demographic characteristics are detailed in Table 1. The mean age was 58.1 years old, and 83 of the patients were female (53.9%). Twenty patients were diabetic (13.0%), 73 had hypertension (47.4%) and 30 had CKD (19.5%), with a mean GFR of 55 ml/min/1.73 m². Six patients were stage 2, 16 were stage 3A, 2 were stage 3B, and 1 was stage 4. The mean BMI was 25.2 kg/m². The main locations of cancer were gynecologic (n = 34, 22%), lungs (n = 25, 16%), kidneys (n = 19, 12%), and skin (n = 19, 12%).
Table 1: Demographic characteristics of the population
Characteristics
|
|
Age, mean (year)
Female, n (%)
Baseline SCr, median (IQR) (µmol/l)
|
58.1
83 (53.9%)
132 (88.5-168.2)
|
Comorbidités
Diabetes, n (%)
|
20 (13.0%)
|
Hypertension, n (%)
|
73 (47.4%)
|
Chronic kidney failure, n (%)
Cancer
Pelvic gynecologic cancer, n (%)
Pulmonary cancer, n (%)
|
30 (19.5%)
34 (22%)
25 (16%)
|
Urinary cancer, n (%)
Skin cancer, n (%)
Prior chemotherapies
1
2
3 or more
Indication for consultation and RB
HTA, n (%) Pu, n (%)
AKI, n (%)
Mixed, n (%)
|
19 (12.3%)
18 (11.7%)
58 (37.7%)
44 (28.6%)
52 (33.7%)
49 (31.8%)
112 (72.7%)
95 (61.8%)
90 (58.4%)
|
Clinicobiological presentation at the time of PBR.
The reason for the referral of the patients for an onconephrology consultation included acute kidney injury for 95 patients (61.8%), with a mean creatinine level for these patients at 179 µmol/L (vs. 77 µmol/L in patients without acute renal failure at the time of RB); proteinuria in 112 patients (72.7%); and uncontrolled or de novo hypertension for 49 patients (31.8%). Thirty-six patients had isolated AKI, 25 had isolated proteinuria, and 2 had isolated hypertension. Forty-three patients were referred for the combination of AKI and proteinuria; 30 for the combination of hypertension and proteinuria; 5 for the combination of AKI and hypertension; and 12 for the combination of AKI, hypertension, and proteinuria. Seven patients had biological hemolysis but had many missing data (50.6%). Among the patients tested, hematuria was found in 38/110 patients (35%), and leukocyturia was found in 40/106 patients (38%).
Treatment
All patients were receiving anticancer therapy at the time of RB. The number of previous treatment of anticancer therapy ranged from 1 to 7, with a median of 1.98, and a majority of patients had received a single line (37.6%) (Supplementary Figure S1). Fifty-six patients (36%) received cytotoxic chemotherapies at the time of RB, including alkylating agents in 22 and antimetabolites in 20 patients. Targeted therapies were used in 98 patients (64%), including tyrosine kinase inhibitors in 46 patients and monoclonal antibodies in 52 patients. Immunotherapy was ongoing in 42 patients (27%) and included anti-PD1 in 27 patients and anti-PDL1 in 10 patients (Supplementary Table S4). Twenty-two patients had previously received nephrotoxic treatments before RB, including iodinated contrast media for 8 patients, biphosphonates for 4 patients, NSAIDs for 4 patients, and unspecified for 4 patients.
Histological lesions
The basic histological lesions were essentially chronic and predominated in the interstitial, tubular and vascular areas (Figure 1). The main described lesions were interstitial fibrosis (n=114, 74%), followed by arteriolar hyalinosis (n=90, 58%), fibrous endarteritis (n=87, 56%), and tubular atrophy (n=85, 55%). Acute lesions were mainly found in the tubular area with acute tubular necrosis (ATN) (n=79, 51%), interstitial inflammation (n= 46, 30%), glomeruli with glomerular thrombi (n=46, double contours on 45 RB, mesangiolysis on 39, and glomerular thrombi on 22 RB) and immunofluorescence deposits (n=47).
Cytotoxic chemotherapies were mainly associated with ATN (17 patients: 9 patients receiving alkylating agents and 8 patients receiving antimetabolites) (Figure 2). Patients undergoing targeted therapies mostly presented a thrombotic microangiopathy pattern (42 patients: 10 patients receiving tyrosine kinase inhibitors, 32 patients receiving monoclonal antibodies). Patients undergoing immunotherapy mainly presented ATN (17/42 patients, 40%), and only 6 had AIN (6/42, 14%).
Onconephrology diagnosis
After RB, an onconephrological diagnosis for the underlying etiology of kidney injury was provided as the diagnostic category, as summarized in Table 2. Cancer treatment toxicity was the diagnostic category chosen for 121 patients (78%). The majority were thrombotic microangiopathies (46 patients), acute tubular necrosis (34 patients) and acute interstitial nephritis (9 patients). The other most common diagnostic categories were the toxicity of noncancer drug treatments, including iodinated contrast products, in 12 patients (7.8%); complications of nondrug cancer management in 10 patients (2.6%); direct complications of the cancer in 12 patients (7.8%), namely, paraneoplastic glomerulopathies, including MN in 2 patients; functional causes in 13 patients (8.4%); obstructive causes in 6 patients (3.9%); and renal disease unrelated to cancer or treatment in 32 patients (20.7%).
Table 2: Diagnostic category retained after RB
Diagnostic category
|
n (%)
|
AKI (%)
|
Pu (%)
|
HTA (%)
|
Cancer treatment toxicity
Functional cause
Kidney disease unrelated to cancer
Other treatment toxicity (non anticancer)
Direct cancer complication
Consequence of surgery or radiotherapy in the renal area
Obstructive cause
No conclusion
|
124 (80.5%)
13 (8.4%)
32 (20.8%)
12 (7.8%)
12 (7.8%)
3 (1.9%)
6 (3.9%)
3 (1.9%)
|
69 (55.6%)
9 (69.2%)
25 (78.1%)
12 (100%)
7(58.3%)
3 (100%)
6 (100%)
|
92 (74.2%)
10 (76.9%)
22 (68.8%)
9 (75%)
11 (91.7%)
2 (66.7%)
5 (83.3%)
|
44 (12.9%)
2 (15.4%)
8 (25%)
1 (8.3%)
3 (25%)
1 (33.3%)
1 (16.7%)
|
Classification of histological lesions according to diagnostic category:
The most common etiologies according to diagnostic category are summarized in Figure 3. Supplementary Table 3 reports the basic histological lesions according to the diagnostic category. As shown in Figure 3, 84 patients had only one underlying cause of kidney injury, while 47 had multiple causes. For 3 patients, no diagnosis could be made because of the RB findings were inconclusive.
Cluster analysis
Cluster analysis allowed the identification of 4 groups (dendogram, Supplementary Figure S4). Clusters 2 and 4 corresponded histologically to TMA, with or without arteriolar thrombi for Cluster 4 and Cluster 2, respectively. The treatments most frequently used in Cluster 2 were anti-VEGF, PDL1 inhibitors and alkylating agents. The treatments used in Cluster 4 were anti-VEGF, multikinase inhibitors and antimetabolites. These patients were not necessarily more often hypertensive. Cluster 3 identified patients with the most interstitial fibrosis and vascular lesions along with more severe interstitial inflammation. The most commonly used treatment was immunotherapy. It was also frequently used in Cluster 1 in combination with alkylates or antimetabolites, but interstitial inflammation was minimal. No clinical or biological data distinguished Clusters 1 and 3 (Figure 4). No difference between clusters was found (Supplementary Figure S5, 6, 7).