Renal Disease

The urine protein/creatinine ratio as a reliable indicator of 24-h urine protein excretion across different levels of renal function and proteinuria: the TUNARI prospective study – BMC Nephrology

Time Stamp: November 20, 2024 7:00 PM
Source Node: 754795

The evaluation of proteinuria in patients with CKD is a great clinical challenge, since the definition of this term can be variable, and the progression of kidney disease can be slow and asymptomatic in the early stages [14]. Measurement of 24-hUP has been an important tool in the evaluation of renal disease, but the urine samples can be inconvenient and difficult to collect. Consequently, PCR has been viewed as a more practical alternative for the evaluation of nephropathies. The effective identification of urinary protein excretion is essential in the diagnosis of nephropathies, as well as in the prognosis and evaluation of the therapeutic response, especially in patients with glomerulopathies. The KDIGO guidelines strongly recommended PCR, both for patients with glomerulopathies and other nephropathies [2, 3, 21].

In this study, we found a high correlation between PCR and 24-hUP in patients with different levels of kidney function, even for lower levels of kidney function (< 30 mL/min). However, we observed that there was a lack of agreement between both methods at higher levels of proteinuria, which has already been described in other reports that compare the methods in different clinical conditions [4, 5, 23]. The analysis using the Bland–Altman plot revealed a loss of agreement at the most distant or extreme points on the graph. This could indicate that the measurement methods are not equally reliable or effective in all areas [23, 24]. It is important to consider these observations to better understand potential biases or inconsistencies in the comparison between methods and to make informed decisions about their use in specific situations.

The ROC curve analysis revealed a high effective of PCR to separate patients above and below a clinically significant level of 24-hUP excretion. A high sensitivity and specificity for PCR were observed both in general analysis and in subgroups of renal function. Consistent with the high sensitivity and specificity in patients across levels of renal function, a high Youden index was observed. Also, across levels of renal function, the capacity of PCR to rule in and rule out significant 24-hUP was demonstrated by the LR + above 6 and LR- below 0.15, respectively. Overall, the results indicate that PCR on isolated urine samples shows substantial discriminatory ability in patients with different levels of renal function, making it valuable for clinical decision-making [25, 26].

We used the ROC curve and AUC to compare the performance of PCR with 24-h UP, evaluating how different thresholds affect sensitivity and specificity. These metrics are key to compare the effective of both methods. With a cut-off value of 0.77 according to the Youden index, and an AUC of 0.95, along with a sensitivity of 91% and a specificity of 86.5%, these metrics are essential to correctly detect proteinuria in general. When assessed by renal function levels, the cut-off value ranged from 0.76 to 0.83 with an AUC greater than 0.94, the sensitivity was between 86.4% to 93.3% and the specificity was approximately 90%. By proteinuria levels, performance was significantly affected in the > 0.3–3.5 g/day group, the cut-off point was 0.50 with a sensitivity of 64.1% and specificity of 84.6, and the AUC was 0.76. However, it is important to consider that the Youden index has limitations, such as not identifying false positives and negatives. Therefore, LR improves this evaluation by measuring how PCR results can change the probability of disease, providing a direct clinical interpretation for diagnosis.

Despite the substantial correlation between PCR and 24-hUP, and the effective measures supporting the use of PCR as a proxy for 24-hUP, the reduction in agreement at higher 24-hUP levels should be considered in clinical decision-making. For example, in the case of an acute patient with suspected lupus nephritis/nephrotic syndrome, anasarca, rapid loss of kidney function, who needs immediate action, such as immunosuppressive pulse therapy, PCR is ideal to discriminate whether the patient is with proteinuria, or not, quickly and easily. This method has high specificity, above all, it has a high capacity to discriminate whether or not the patient is in the nephrotic range (> 3.5 g/day), a capacity for which it was originally designed [4, 5, 27].

In another context, where we need to follow up patients with stable conditions, treated with immunosuppression therapy and whit partial remission, we must be cautious when obtaining a result of a rapid decrease in proteinuria by the PCR method, especially when it is not accompanied by improvement of other clinical and laboratory parameters (e.g., improvement in serum albumin, improvement in renal function), since the agreement between both methods was compromised at high levels of proteinuria [4, 5, 16, 28]. In addition, we must avoid the interspersed use between consultations of both methods, since it can show a discrepancy between them, especially in patients receiving immunotherapy when the level of proteinuria is at levels of partial remission. One should rely on the 24-hUP for patients with higher levels of proteinuria.

In the case of a stable patient, in complete remission on immunosuppressants, the use of PCR to monitor the level of proteinuria on an outpatient basis is highly recommended. In an additional analysis where we stratified the level of proteinuria, we observed a moderate correlation between PCR and 24-hUP for patients with proteinuria levels ≤ 0.3 g/day (r_s = 0.70 and r² = 0.40), however, the correlation was low for patients with proteinuria between > 0.3–3.5 g/day (r_s = 0.53 and r² = 0.21). It is imperative to highlight that the correlation was compromised in patients with proteinuria > 3.5 g/day (r_s = 0.36), but the effective of proteinuria detection in this group was 100% (proteinuria > 3.5 g/day), which makes ROC curve analysis unfeasible and reinforces the purpose for which PCR was created. These findings underscore the discriminatory capacity of PCR to identify patients within the nephrotic range.

Potential methodological limitations of this study should be acknowledged. One limitation is that the study was conducted within the nephrology service of a single center, which limited the sample size to the patients available in that service. The sample was restricted to the number of patients in the service and some patients who had two urine samples taken in different periods of follow-up. We performed analyses comparing the first sample of all patients to all samples as well as the subgroup assessment (phase 1 vs phase 2) and no notable changes were observed in correlation, sensitivity or specificity. The inclusion of these additional samples helped us increase the power of our analyses and markedly reduced our likelihood of bias. This study presents a majority non-White population, which reduces external validity, but is also a strong point for internal validation in non-White populations. This study was conducted in Salvador, BA, which has the largest population of African descent outside of Africa [20, 29, 30]. The Black and Mixed-Race populations have been reported to have a much higher risk of end-stage renal disease compared to the White population [31, 32]. Furthermore, studies from the United States and Brazil show that younger populations of African descent experience higher prevalence rates, worse renal survival in certain glomerulopathies, particularly focal glomerulosclerosis [33, 34], and greater mortality in the case of lupus nephritis [34, 35]. This underscores the need for heightened attention and care for these younger demographic groups, who often face worse outcomes.

The results of this study are consistent with previous studies that have demonstrated the correlation between PCR and 24-hUP in patients with different levels of renal function [14]. Furthermore, these results highlight the usefulness of the PCR as an alternative tool for the measurement of 24-hUP for the detection of proteinuria in different stages of CKD. However, caution is necessary in the use of the PCR index to substitute for 24-hUP measurements in patients with CKD and nephrotic range proteinuria, due to the observed decrease in agreement between PCR and 24-hUP at higher levels of proteinuria. In some cases, it may be necessary to measure 24-hUP to perform certain more specific assessments in patients with CKD or in patients with glomerulopathies with high disease activity, always assessing the clinical context in general. We emphasize that PCR was initially validated for the diagnosis of patients with nephrotic proteinuria and was not designed to precisely measure proteinuria [4, 5, 27].