Early compensatory increase in single-kidney estimated GFR after unilateral nephrectomy is associated with a lower long-term risk of estimated GFR decline

ABSTRACT

Background and hypothesis

A more pronounced short-term increase in single-kidney GFR (ΔskGFR) has been associated with better long-term kidney function in living kidney donors. Whether this also applies to non-donors is unknown. We evaluated whether ΔskGFR is associated with long-term risk of eGFR decline in individuals undergoing unilateral nephrectomy.

Methods

This study included 1777 participants from the SCREAM cohort who underwent radical unilateral nephrectomy in Stockholm between 2006 and 2021. The ΔskGFR was calculated as the early (1–6 months) post-nephrectomy eGFR minus 50% of the pre-nephrectomy eGFR. Multivariable Cox regression was used to study the association between Δsk-GFR and the subsequent risk of progressive eGFR decline, defined as composite of an eGFR decline >30% compared to the early (6 months) post-nephrectomy eGFR or kidney failure.

Results

Mean age at nephrectomy was 68 ± 11 years, 40% were female, 92% had kidney cancer, and median (IQR) pre-nephrectomy eGFR was 76 (61–89) ml/min/1.73 m². Median Δsk-GFR was 11 (7–20) ml/min/1.73 m². Pre-nephrectomy determinants of Δsk-GFR were age (inverse association) and pre-nephrectomy eGFR (positive association). During a median follow-up of 5 years (range 0.6–15 years), 178 participants developed progressive eGFR decline. Individuals with a Δsk-GFR above the median had a lower rate of progressive eGFR decline (adjusted HR: 0.58, 95% CI: 0.42–0.80), compared to those with a Δsk-GFR below the median, independent of baseline eGFR and age.

Conclusions

A stronger increase in single-kidney eGFR early after unilateral nephrectomy was associated with a lower long-term risk of progressive eGFR decline. Evaluation of Δsk-GFR could help identify patients at higher risk of progressive kidney function decline following unilateral nephrectomy.

KEY LEARNING POINTS.

What was known:

• After unilateral nephrectomy, the remaining kidney can increase GFR up to 65%–75% of the pre-nephrectomy GFR by increasing the single-nephron GFR in the remaining nephrons.

• Although initial concerns existed regarding potential kidney damage due to hyperfiltration following nephrectomy, it has been demonstrated that in kidney donors that a greater increase in GFR predicts better long-term GFR outcomes.

This study adds:

• In this study we show that also in patients that undergo unilateral nephrectomy for other reasons than donation—who are not selected based on health—a stronger increase in estimated GFR after nephrectomy (delta single-kidney GFR), is associated with less estimated GFR decline on the long-term.

Potential impact:

• Evaluating the Δsk-GFR after unilateral nephrectomy could aid in identifying patients at risk of progressive eGFR.

INTRODUCTION

Physical resilience is defined as “the ability to resist functional decline or recover functional health following a stressor” and is (partly) determined by reserve capacity, defined as “the potential capacity of a cell, tissue, or organ system to function beyond its basal level in response to alterations in physiologic demands” [1]. In the human kidney, resilience is demonstrated by maintenance of glomerular filtration rate (GFR) during early kidney damage or by recovery of GFR after an acute reduction of nephrons, for example unilateral nephrectomy [2, 3]. Since no new nephrons are formed during life, this is achieved by increasing the single-nephron GFR in the remaining nephrons [2].

An increase in single-nephron GFR may have deleterious effect as hyperfiltration, induced by an increased hydraulic pressure in the glomerulus, may lead to podocyte detachment, proteinuria, nephrosclerosis, further nephron loss, and subsequent adverse outcomes, which has been demonstrated in patients with diabetes nephropathy and obesity [4–8]. Conversely, we recently demonstrated that a higher post-donation increase in single-kidney GFR in living kidney donors is an independent predictor of better long-term kidney function [9], suggesting that glomerular hyperfiltration is not always problematic. Whether this principle only applies to the highly selected kidney donor population or may be extended to other populations is unknown.

Here, we studied the association of the short-term increase in single-kidney GFR with long-term kidney function in a population of patients undergoing unilateral nephrectomy for other reasons than kidney donation, mostly malignancy. We hypothesized that a higher post-nephrectomy increase in single-kidney GFR might reflect a higher reserve capacity in the remaining kidney, and that this would be associated with a lower risk of subsequent kidney function decline.

MATERIALS AND METHODS

Data source

We used data from the Stockholm Creatinine Measurements (SCREAM) project. SCREAM is a healthcare utilization cohort from the region of Stockholm, Sweden, covering the period from 2006 to 2021 [10]. A single healthcare provider in the Stockholm region provides universal and tax-funded healthcare to 20% to 25% of the population of Sweden. Using unique personal identification numbers, SCREAM linked regional and national administrative databases that hold complete information on demographics, healthcare utilization, laboratory tests undertaken, dispensed drugs, diagnoses, and vital status. The Regional Ethical Review Board in Stockholm approved the study (reference 2017/793–31); informed consent was not deemed necessary because all data were de-identified at the Swedish Board of Health and Welfare.

Study population and study design

A flowchart of the study population selection for this study is shown in Fig. S1, and a schematic overview of the study design is shown in Fig. 1. Inclusion criteria were adults (>18 years) undergoing radical unilateral nephrectomy (Table S1), with at least one creatinine measurement within 1 year before nephrectomy, at least one during the exposure period (4 weeks to 6 months post-nephrectomy), and at least two creatinine measurements during follow-up (after 6 months post-nephrectomy) (Fig. 1). Exclusion criteria were having a status of living kidney donor or history of kidney replacement therapy (KRT, ascertained by linkage with the Swedish Renal Registry). Additionally, we excluded patients with a diagnosis of urinary flow obstruction or urolithiasis, because patients that undergo unilateral nephrectomy for these reasons are likely to compensate kidney function of the contralateral healthy kidney prior to nephrectomy [11].

Figure 1:

Schematic overview of study design and data collection at each time point. Baseline characteristics were collected prior to nephrectomy, the exposure (Δsk-GFR) was calculated as short-term post-nephrectomy eGFR minus 50% of pre-nephrectomy eGFR. Follow-up period started after exposure period at 6 months. t₀ = time zero (start follow-up), t₁ = end of follow-up.

Exposure, covariates, and outcomes

For study participants, we extracted all serum and plasma creatinine measurements performed in connection to healthcare encounters, and used them to estimate glomerular filtration rate (eGFR) using the 2009 CKD-EPI equation without the race coefficient, which has been shown to be more accurate than the 2021 CKD-EPI equation in our data [12, 13]. Our exposure was the short-term post-nephrectomy increase in single-kidney GFR (Δsk-GFR), calculated as the eGFR between 1 and 6 months after nephrectomy minus 50% of the pre-nephrectomy eGFR [9]. The pre-nephrectomy eGFR value was defined as the median eGFR of all measurements in the year prior to the date of nephrectomy. The post-nephrectomy eGFR value was defined as the median eGFR of all measurements during the 1 to 6 months after nephrectomy (Fig. 1). The baseline for post-nephrectomy follow-up was set at month 6 after the nephrectomy date, and on this date all other study covariates were derived, and follow-up began. By means of sensitivity analysis, we also calculated the percentage increase in single-kidney GFR (Δsk-GFR%) from pre-nephrectomy eGFR and repeated the analyses using this variable as exposure.

Study covariates included age, sex, pre-nephrectomy eGFR, comorbidities, and ongoing medications (see Table S1 for detailed definitions and look back periods for ascertainment).

The primary study outcome was progressive eGFR decline, defined as a composite endpoint of a decline in eGFR >30% compared to baseline (6 months post-nephrectomy) eGFR, or kidney failure with replacement therapy (i.e. dialysis or kidney transplantation). A decline of 30% was chosen because a loss of 30% in eGFR would result in CKD according to the official CKD staging in most patients. To reduce outcome misclassification bias owing to intrinsic eGFR variability, and to confirm whether eGFR declines were sustained over time, we used a linear interpolation method [14]. In brief, for each study participant we fitted a linear regression line through all outpatient eGFR measurements after index date. To be considered a sustained eGFR decline of >30% relative to post-nephrectomy eGFR, the linear regression slope needed to be negative and the threshold of a 30% difference needed to be crossed before the last available measurement. The time-to-event outcome was then defined as the interpolated moment at which the linear regression line reached an eGFR 30% lower. Patients were followed until event, death, migration, or end of follow-up (31 December 2021), whichever occurred first. Date of death was retrieved from the Swedish cause of death register. The secondary study outcome was all-cause death.

Statistical analyses

Data were presented as mean with standard deviation or as median with interquartile range when appropriate for continuous variables and as number with percentage for categorical variables. In univariable linear regression analyses, we investigated whether age, sex, pre-nephrectomy eGFR, hypertension, diabetes, or cardiovascular disease were associated with the Δsk-GFR. Next, we graphically depicted the cumulative incidence of our outcome progressive eGFR decline over time for patients with low Δsk-GFR (above the median) vs. patients with high Δsk-GFR (below or equal to the median) using Kaplan–Meier plots. Using multivariable Cox proportional hazard's regression, we investigated the association between a high Δsk-GFR (defined by an increase above the median value) and the risk of developing progressive eGFR decline. We adjusted for age, sex, and pre-nephrectomy eGFR.

We explored potential effect modification by multiplicative interaction terms across subgroups of age; sex; and hypertension, diabetes, and cardiovascular disease status. Sensitivity analyses evaluated the robustness of our results by exploring alternative thresholds of Δsk-GFR (highest and lowest quartile instead of the median), a secondary outcome that was a composite of 50% eGFR decline and or kidney failure with replacement therapy (i.e. dialysis or kidney transplantation), and considering death as a competing risk using Fine and Gray models [15]. Furthermore, we repeated the main analyses using eGFR calculated according to the European Kidney Function Consortium (EKFC) equation [16]. Analyses were performed with R Software (R studio version 2022.07.2-576 ‘Spotted Wakerobin’, R version 4.2.1).

RESULTS

Population characteristics

A total of 1777 adults undergoing radical unilateral nephrectomy met the inclusion and exclusion criteria (see flow chart in Fig. S1). Mean age of the study population was 68 ± 11 years, pre-nephrectomy eGFR was 75±19 ml/min/1.73 m², and 703 (40%) were female. Most (92%) patients had a diagnosis of kidney cancer in 3 years prior to nephrectomy, and 20% had diabetes. As many as 62% had a clinical diagnosis of hypertension. For the 8% of patients without a clinical diagnosis of kidney cancer, the reason for nephrectomy was not clear. Mean exposure eGFR (i.e. eGFR between 1 and 6 months post-nephrectomy), was 51±17 ml/min/1.73 m². The median Δsk-GFR was 11 (interquartile range 7 to 20) ml/min/1.73 m². Patients were subsequently divided into categories according to low (≤11 ml/min/1.73 m²) or high (>11 ml/min/1.73 m²) Δsk-GFR value. Patient characteristics are shown in Table 1. Patients in the high Δsk-GFR group were younger and had a higher pre-nephrectomy eGFR than patients in the low Δsk-GFR group (Table 1). The number of available creatinine measurements per time point are shown in Table S2.

Table 1:

Baseline characteristics of the study population.

	Overall	Low Δsk-GFR ≤11 ml/min/1.73 m2	High Δsk-GFR >11 ml/min/1.73 m2
N (%)	1777 (100%)	865 (49%)	912 (51%)
Age, years	67.9 [11.3]	70.8 [9.9]	65.0 [12.0]
Age categories
19–40 years	32 (1.8%)	5 (0.6%)	27 (3.0%)
41–65 years	610 (35%)	212 (25%)	412 (45%)
65+ years	1106 (63%)	646 (75%)	473 (51%)
Female sex, N (%)	703 (40%)	343 (40%)	368 (40%)
Pre-nephrectomy eGFR, ml/min/1.73 m2	74.9 [19.3]	69.7 [19.3]	78.3 [20.7]
Pre-nephrectomy eGFR categories
≥60 ml/min/1.73 m2	1364 (78%)	623 (72%)	747 (82%)
30–59 ml/min/1.73 m2	361 (21%)	212 (25%)	149 (16%)
<30 ml/min/1.73 m2	23 (1.3%)	28 (3.2%)	16 (1.8%)
Exposure eGFR, ml/min/1.73 m2	50.6 [17.0]	39.8 [11.7]	60.8 [14.7]
Comorbidities, N (%)
Hypertension	1091 (62%)	606 (70%)	511 (56%)
Diabetes	344 (20%)	198 (23%)	151 (17%)
Cardiovascular disease	433 (25%)	260 (30%)	183 (20%)
Cancera	1654 (95%)	806 (93%)	862 (95%)
Kidney cancera	1615 (92%)	807 (93%)	819 (90%)
Kidney trauma	7 (0.4%)	2 (0.2%)	6 (0.7%)
Ongoing medications
Antihypertensives	1684 (96%)	843 (97%)	869 (95%)
Glucose-lowering drugs	256 (15%)	144 (17%)	116 (13%)

^aIn the 3 years prior to nephrectomy

Data presented as N (%) for binary variables and mean [standard deviation] for continuous variables.

For diagnosis and ATC codes used to extract comorbidities and medications, see Table S1.

Pre-nephrectomy associates of the Δsk-GFR

In univariable analyses, age (negatively), pre-nephrectomy eGFR (positively), a diagnosis of hypertension (negatively), diabetes (negatively), and cardiovascular disease (negatively) were significantly associated with the Δsk-GFR (Table 2). When including these variables in a multivariable model, only age (negative association) and pre-nephrectomy eGFR (positive association) remained as independent associates of the Δsk-GFR (model R² = 0.11).

Table 2:

Univariable and multivariable linear regression analysis of pre-nephrectomy predictors of Δsk-GFR.

	Univariable		Multivariable
	St.β	95% CI	St.β	95% CI
Age, per decade	−0.29	−0.33 to −0.24	−0.20	−0.26 to −0.15
Sex	−0.03	−0.08 to 0.01
Pre-nephrectomy eGFR, per 10 ml/min/1.73 m2	0.23	0.18 to 0.27	0.11	0.05 to 0.17
Hypertension	−0.16	−0.20 to −0.11	−0.05	−0.09 to 0.002
Diabetes	−0.08	−0.12 to −0.03	−0.03	−0.08 to 0.01
Cardiovascular disease	−0.13	−0.17 to −0.08	−0.04	−0.08 to 0.01

All variables with P < .2 were included in the multivariable model.

Δsk-GFR and outcomes

During a median follow-up 4.8 (range 0.6–15 years), 178 (10%) patients developed progressive eGFR decline (of which 153 reached 30% eGFR decline and 25 reached kidney failure with RRT) and 543 (31%) patients died before experiencing progressive eGFR decline. Causes of death are shown in Table S3. Figure 2 depicts the cumulative incidence of progressive eGFR decline events, which was higher for patients in the low Δsk-GFR compared vs. patients in the high Δsk-GFR group (P value log-rank test <.001). Compared to patients in the low Δsk-GFR group, those in the high Δsk-GFR group had a lower rate of progressive eGFR decline, independent of age, sex, and pre-nephrectomy eGFR [hazard ratio (HR): 0.58, 95% CI: 0.42–0.80, Table 3]. Results were similar after accounting for death as a competing risk. For our secondary outcome death, there was no significant association between the Δsk-GFR and the risk of death (Table S4). Survival probabilities for patients with high or low Δsk-GFR are shown in Fig. S2.

Figure 2:

Kaplan–Meier plot showing cumulative incidence of progressive eGFR decline for low Δsk-GFR vs. high Δsk-GFR. The Δsk-GFR was dichotomized based on the median value (11 ml/min/1.73 m²): low, Δsk-GFR ≤11 ml/min/1.73 m²; high, Δsk-GFR >11 ml/min/1.73 m². Progressive eGFR decline was a composite endpoint of a decline in eGFR >30% compared to baseline (6 months post-nephrectomy) eGFR, or initiation of KRT (i.e. dialysis or kidney transplantation).

Table 3:

Association between Δsk-GFR categories with the risk of progressive eGFR decline.

	Number of events/participants	IR (95% CI) per 1000 person years	HR (95% CI)a	Subdistribution HR (95% CI)b
Low Δsk-GFR	117/865 (14%)	24 (20 to 29)	1.00 (ref)	1.00 (ref)
High Δsk-GFR	61/912 (7%)	12 (9 to 15)	0.58 (0.42 to 0.80)	0.67 (0.48 to 0.94)

^aRisk of progressive eGFR decline with Cox regression, censoring for death and emigration. Model adjusted for pre-nephrectomy eGFR, age, and sex.

^bRisk of progressive eGFR decline with Fine and Gray models considering death as a competing event and censoring for emigration. Model adjusted for pre-nephrectomy eGFR, age, and sex.

The Δsk-GFR was dichotomized based on the median value (11 ml/min/1.73 m²): low, Δsk-GFR ≤11 ml/min/1.73 m²; high, Δsk-GFR >11 ml/min/1.73 m².

Abbreviation: IR = incidence rate.

Subgroup analyses

The association of the Δsk-GFR with progressive eGFR decline did not differ across strata of age, sex, hypertension, diabetes, or cardiovascular disease (Table 4). When the Δsk-GFR was dichotomized based on the highest quartile (≥19 ml/min/1.73 m²) vs. the rest of the cohort (<19 ml/min/1.73 m²), a higher Δsk-GFR was not significantly associated with a lower risk of progressive eGFR decline (Table S5). There were 38 events in the highest Δsk-GFR quartile vs. 140 events in the rest of the cohort (total 178 events). A Δsk-GFR above the lowest quartile (≥7 ml/min/1.73 m²) was significantly associated with a 54% lower risk of progressive eGFR decline (Table S5). There were 82 events in the lowest Δsk-GFR quartile vs. 96 events in the rest of the cohort (total 178 events).

Table 4:

Subgroup analyses by age, sex, and absolute post-nephrectomy eGFR.

	N Δsk-GFR ≤11 ml/min/1.73 m2	N Δsk-GFR >11 ml/min/1.73 m2	HR (95% CI), ref.: Δsk-GFR ≤11 ml/min/1.73 m2	P value for interaction
Sex
Female	343	368	0.71 (0.41 to 1.23)	.37
Male	522	544	0.52 (0.35 to 0.77)
Age
≥70 years	355	551	0.58 (0.36 to 0.94)	.95
<70 years	510	361	0.51 (0.33 to 0.79)
Hypertension
Yes	606	511	0.59 (0.41 to 0.84)	.87
No	259	401	0.62 (0.32 to 1.22)
Diabetes
Yes	198	151	0.60 (0.34 to 1.13)	.90
No	667	761	0.62 (0.41 to 0.87)
Cardiovascular disease
Yes	260	182	0.67 (0.39 to 1.16)	.57
No	605	730	0.55 (0.37 to 0.81)

Sensitivity analyses

We performed a sensitivity analysis in which we ran the main analysis (Cox regression model with outcome progressive eGFR decline) only in patients with a history of kidney cancer (N = 1626), which did not affect our results, as shown in Table S6. Similarly, we repeated the main analysis in a subgroup of 153 patients that reached 30% eGFR decline (excluding the 25 patients that reached the outcome kidney failure with replacement therapy) Table S7.

Second, we calculated the Δsk-GFR as percentage from pre-nephrectomy eGFR (Δsk-GFR%) and repeated the Cox regression analysis using this variable as exposure (Table S8). Here again, a higher Δsk-GFR% was significantly associated with a lower risk of progressive eGFR decline after unilateral nephrectomy.

Then, we repeated the analyses using a secondary outcome that was a composite of 50% eGFR decline and or kidney failure with replacement therapy (i.e. dialysis or kidney transplantation), reached by 40 patients in total (Table S9), which yielded similar results.

Last, we calculated eGFR according to the EKFC equation and repeated the analyses (Table S10). Mean pre-nephrectomy eGFR according to the EKFC equation was 64 ml/min/1.73 m² and the median Δsk-GFR according the EKFC equation was 12 ml/min/1.73 m². Here, a higher Δsk-GFR was also significantly associated with a lower risk of progressive eGFR decline (composite of 30% eGFR decline and or kidney failure with replacement therapy), independent of age and sex and pre-nephrectomy eGFR.

DISCUSSION

The purpose of this study was to investigate the association between the short-term increase in single-kidney GFR (Δsk-GFR) after radical unilateral nephrectomy and the long-term risk of subsequent kidney function loss (>30% reduction in eGFR or kidney failure). The main result was that patients with a higher Δsk-GFR had a lower rate of progressive eGFR decline, independent of age, sex, and pre-nephrectomy eGFR. Individuals with a very limited increase in single-kidney eGFR, possibly reflecting very little renal reserve, seemed to be particularly at risk of subsequent kidney function loss. This suggests that more compensatory GFR increase after loss of kidney mass (nephrectomy) might reflect a better reserve more than malignant hyperfiltration. These results might aid in identifying patients at risk of developing progressive eGFR decline after unilateral nephrectomy.

Whether an increase in single-kidney eGFR is linked with beneficial or adverse outcomes has been subject of discussion, and may be context dependent. Animal studies have shown that hyperfiltration in remnant nephrons after subtotal nephrectomy can lead to glomerular damage, proteinuria, nephrosclerosis, and subsequent progressive nephron loss in 5/6-nephrectomy rat models [8, 17, 18]. These findings were the foundation of the “Brenner hypothesis,” stating that after substantial loss of nephrons, hyperfiltration in remnant nephrons, mediated by increased intraglomerular pressure, leads to a vicious circle of further nephron loss and progressive kidney function decline [7]. Hyperfiltration-mediated kidney damage is for example seen in clinical diseases such as early stages of diabetic nephropathy [19]. In keeping with this hypothesis, concerns may be raised about compensatory hyperfiltration in the remaining kidney after unilateral nephrectomy. However, other (potentially less harmful) mechanisms may be involved in compensatory hyperfiltration as well, such as suppression of growth inhibitory genes [20, 21]. In pregnant women for example, another situation in which hyperfiltration occurs, it has been demonstrated that hyperfiltration to 120–150 ml/min/1.73 m² results in better pregnancy outcomes and favorable outcomes after unilateral nephrectomy in healthy kidney donors support this hypothesis [22]. In line, results of the current study show that a stronger increase in single-kidney GFR after unilateral nephrectomy, indicating more hyperfiltration, is associated with progressive eGFR decline. Possibly, there is a range of nephron loss that the kidney can tolerate without inducing pathophysiological pathways leading to further nephron loss. The extent of damage present in the remaining nephrons may also be of importance, and therefore a comparison of the current study population with the healthy donor population in our previous work is of interest.

The results of this study align with previous findings about the post-donation increase in single-kidney GFR in living kidney donors [9], despite differences in patient populations. Compared to previous results, patients in the current study had lower pre-nephrectomy eGFR, were older, had more comorbidities, and most were diagnosed with kidney cancer and may have received chemotherapy prior to nephrectomy. In addition, the outcome measures of the present study were different from the previous study. Given these differences, a comparison between both studies should be interpreted with caution. It has been shown that in >60% of patients undergoing unilateral nephrectomy for renal cell carcinoma, the renal parenchyma and vasculature show evident pathologic abnormalities [23]. Consequently, it could be hypothesized that these patients are more prone to hyperfiltration, accompanied by increased glomerular pressure, and resulting in glomerular damage and subsequent adverse outcomes. However, even in this population, we found that a stronger increase in single-kidney GFR post-nephrectomy is independently associated with a reduced risk of progressive kidney function decline. While it is possible that some patients experienced malignant post-nephrectomy hyperfiltration leading to adverse outcomes [23], overall, our data suggest that more pronounced hyperfiltration was not an indicator of unfavorable outcomes. In line, a previous study showed that renal blood flow after unilateral nephrectomy in patients with renal cell carcinoma increased at 1 week and 1 month after nephrectomy and returned to pre-nephrectomy values at 3 months [24]. A more pronounced Δsk-GFR may reflect a better reserve capacity of the kidney, although further studies are needed to identify potential differences between benign and malignant compensatory pathways. Moreover, future studies should include (new-onset) albuminuria as an outcome.

Most patients in this study that underwent radical unilateral nephrectomy had a history of kidney cancer. This might raise concerns about pre-nephrectomy compensation of the contralateral healthy kidney for reduced single-kidney function of the affected kidney, which may affect applicability of our Δsk-GFR equation. Previous studies show that increased tumor size (>7 cm) negatively affects post-nephrectomy compensation of the contralateral remaining kidney, indicating that compensation may have (partly) occurred prior to nephrectomy [25, 26]. Yet, Song et al. found that patients undergoing unilateral nephrectomy for kidney cancer had a volume ratio of contralateral healthy kidney compared to diseased kidney of 1.03 in a population with mean tumor size of 6 cm, suggesting equal kidney size [11], but still the affected kidney could be compensating. In the group undergoing unilateral nephrectomy due to urolithiasis, strictures, pyelonephritis, or tuberculosis in the same study, this ratio was 2.81, which supports our method of excluding patients with urolithiasis or urinary flow obstruction [11]. A limitation of our study is that we did not have information on tumor size or kidney volume a, which could influence compensation of the contralateral healthy kidney. However, the distribution of the Δsk-GFR in the current study was highly comparable to the Δsk-GFR study in living kidney donors [9], which suggests that our equation of the Δsk-GFR is not (strongly) affected by pre-nephrectomy compensation of the contralateral healthy kidney. Additionally, remnant kidney function is positively associated with the Δsk-GFR in most studies in patients with kidney cancer, which contradicts pre-nephrectomy compensation of the remaining kidney [27–29].

Other independent associates of Δsk-GFR were age (inverse association) and pre-nephrectomy eGFR (positive association), in line with previous findings in both radical nephrectomy patients and kidney donors [28, 30–35]. All previous studies including kidney volume of the remnant kidney found an independent and positive association with post-nephrectomy compensation of GFR in both populations [9, 29–31, 34]. Some studies also identified BMI, hypertension, sex, and presence of cysts as determinants [29–32, 35]. However, compensation was defined differently in all studies and the overall explained variance by the previously mentioned determinants was low. Our study underlines the relevance of identifying more predictors of the Δsk-GFR to identify patients at risk of progressive kidney function decline.

Strengths of this study include the complete health care coverage from a region with universal tax-funded health care and the availability of pre-nephrectomy, exposure, and follow-up measurements in 1777 patients. Moreover, linear interpolation of eGFR during follow-up minimizes the risk of falsely detecting 30% eGFR decline by a transient drop in eGFR. However, it should be acknowledged that due to the retrospective and observational design of the study, no pre-specified follow-up time points were available and, possibly, patients with stronger kidney function decline were likely followed up more extensively. Yet, the incidence rates of creatinine testing in patients with high Δsk-GFR vs. low Δsk-GFR and rates were comparable. Second, a significant number of patients were excluded because they did not have two creatinine measurements available during follow-up, which could introduce immortal time bias. Also, measured GFR data were not available in the SCREAM registry, which meant we had to rely on estimated GFR data that could have been affected by non-GFR determinants such as muscle mass. Likewise, data on differences in single-kidney size/function within individuals were not available, which could have affected the calculation of the Δsk-GFR. Yet, the influence of differences in kidney size/function on the Δsk-GFR might be minimal, as shown previously [9]. Last, our data could have been supported by kidney biopsy data in which parameters such as glomerular hypertrophy and estimated nephron number could have been measured [2].

In conclusion, we found that a stronger Δsk-GFR after radical unilateral nephrectomy is independently associated with a lower risk of long-term progressive kidney function decline. A higher Δsk-GFR may therefore be an expression of resilience, possibly driven by the reserve capacity of the kidney. Future studies are needed to investigate post-nephrectomy adaptive mechanisms in both healthy individuals and patients, thereby improving understanding of the reserve capacity of the kidney. Such studies may provide important insight in kidney physiology, and shape protocols to provide better follow-up and care after unilateral nephrectomy.

FUNDING

E.L.F. is funded by the (NWO-VENI project 09150162310058). M.H.d.B. is funded by the European Union (ELIMINATE-CKD, ERC-COG 101125516). J.J.C. and the SCREAM cohort are supported by the Swedish Research Council (2023-01807), the US National Institutes of Health (NIH) (R01DK115534), the Swedish Heart and Lung Foundation (20230371), and Region Stockholm (ALF Medicine, FoUI-986028).

AUTHORS’ CONTRIBUTIONS

Conceptualization, J.v.d.W., F.M., E.L.F., M.v.L., M.E., S.P.B., M.H.d.B., and J.J.C.; methodology, J.v.d.W., F.M., M.H.d.B., and J.J.C.; software, F.M., J.J.C.; formal analysis, J.v.d.W., F.M., E.L.F., M.H.d.B., and J.J.C.; investigation, J.v.d.W., E.L.F., M.v.L., M.E., S.P.B., M.H.d.B., and J.J.C.; resources, S.P.B., M.H.d.B., and J.J.C.; data curation, F.M., E.L.F., and J.J.C.; writing, original draft preparation, J.v.d.W., M.H.d.B., and J.J.C.; writing, review, and editing, J.v.d.W., F.M., E.L.F., M.v.L., M.E., S.P.B., M.H.d.B., and J.J.C.; and visualization, J.v.d.W., F.M., M.H.d.B., and J.J.C. All authors have read and agreed to the published version of the manuscript.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available upon reasonable request from the corresponding author (M.H.d.B.). The data are not publicly available due to the privacy of the research participants.

CONFLICT OF INTEREST STATEMENT

None declared.