European Urology

Renal function after a partial nephrectomy (PN) depends on the residual kidney quality and quantity [1]. Unfortunately, surgeons can change neither tumour size nor the baseline glomerular filtration rate (GFR), but they can minimise the length of ischemia and maximise the number of nephrons left in place. The efforts of the urologist are in this direction, and new operative techniques were proposed to limit the length of ischemia (early unclamping, zero ischemia, clampless) and the sacrifice of innocent nephrons (pure enucleation technique) [2] and [3]. Such goals are reasonable and intuitive, and in reaching them, we offer the best possible solution to our patients. When we try to speculate on the mechanisms of kidney damage or to measure and follow kidney damage during the time, however, things become less clear.

There are two important topics when discussing the different nuances of PN: first, the length of ischemia and the critical threshold for observing increased renal ischemic damage; second, the evolution of such damage and the tools used to assess it during the postoperative follow-up.

Many authors have suggested that the critical threshold for warm ischemia time (WIT) is approximately 20–25 min [4] and [5], although every minute is crucial in determining postoperative renal ischemic damage [5]. However, it is unclear how this damage and loss of renal function (LRF) can be correctly measured postoperatively. Many markers of renal injury have been proposed (alanine aminopeptidase, γ-glutamyl transpeptidase, lysozyme, neutrophil gelatinase-associated lipocalin). Unfortunately, most are able to demonstrate renal damage but not to measure or correlate renal damage with the amount of excised tissue or the length of ischemia, and they are not currently used routinely in clinical practice [4].

Traditionally, serum creatinine (SCr) and its derived formulas, such as the estimated GFR (eGFR) and creatinine clearance, have been widely used tools to quantify residual renal function (and, consequently, LRF). Recently, many authors (including our group) suggested that SCr is not reliable in the post-PN setting [4], [5], and [6], and the vast majority of authors consider eGFR to be the best method for measuring ischemia-induced damage in this setting [4]. However, eGFR is limited because the LRF of the operated kidney could be masked by the functioning contralateral kidney. Conversely, a solitary kidney treated with a PN is not always a good model for evaluating renal damage [7].

In our present and previous studies, eGFR was not found to be useful for the identification of LRF after laparoscopic PN [5], [6], and [8]. However, our study population was composed of relatively young and healthy patients with normal renal function, and in this condition, LRF after LPN can be too minimal to be identifiable by eGFR. It is reasonable to suggest that eGFR could be useful for quantifying a postoperative LRF in a large, more heterogeneous series of patients (including patients presenting with poor baseline renal function) or in patients with a solitary kidney.

In this scenario, it is clear that something else is required. The gold standard method for measuring the GFR in humans is determining urinary inulin clearance during constant intravenous inulin infusion and timed urine collections, but this method is not suitable for clinical practice. Renal scintigraphy (RS) is known as the best way to evaluate renal function independently of SCr levels [9]. RS performed with ⁹⁹mTc-mercaptoacetyltriglycine (⁹⁹mTc-MAG-3) is a reliable and largely available method for evaluating kidney function, and we have previously confirmed its usefulness [4], [5], and [6]. However, RS does have some drawbacks: It measures the estimated renal plasmatic flow, not the GFR, and it strictly depends on the region of interest that is chosen during the examination [9]. As an example of the limits of RS, Karam and Wood [10] questioned why the LRF in our data series (6%) [8] is lower than that reported by Gill et al. (10%) in a subgroup of patients treated with zero-ischemia LPN [2]. This comparison is intriguing, but Gill et al.’s study included only 11 patients with a very wide rate of kidney excision (≤50%) [2]. As such, it is not comparable with our results. Finally, in our opinion, RS is not the ideal tool for measuring kidney function, but it is certainly the best available tool, especially for studies such as ours in which the goal is to follow renal function over time.

Karam and Wood underline some important questions [10]: Should we now abandon using eGFR and start routinely obtaining MAG-3 renal scans on all patients? It is clear that eGFR should not be abandoned in all cases, but if a surgeon wants to evaluate the detrimental effects of LPN in a patient with both kidneys (and this is not basic in all cases) or for research purposes, RS is the method to use.

Is our relative decline of 6% in SRF clinically significant? The decline may not be clinically significant because RS is able to detect minimal changes in kidney function that may not have an impact on patient lives. However, it is questionable also whether a decrease of a few millilitres per minute in the eGFR is significant in a patient with both kidneys and if the variation will have an impact on the patient's life.

Finally, how can we generalise these results to an unselected population? Our patients were relatively young and healthy with normal renal function, and they presented with a small renal lesion. Is not this the picture of a typical patient treated surgically? As an example, in our data series, the vast majority of the patients had these characteristics, whereas only a few patients had a solitary kidney or poor baseline renal function. Thus the question could be how can we generalise the results in patients with a solitary kidney to a patient with both kidneys?

In conclusion, we are convinced that an LRF after LPN mostly occurs within 3 mo, depending (even if not exclusively) on the WIT, and remains stable over time.

Will all of the criticisms surrounding LRF and RS in the setting of PN fade away due to our paper? Obviously not, but we believe that the final words of Karam and Wood contain the answer to all of these criticisms: Multicentric prospective data from an unselected population will illuminate the best methods for patient treatment and monitoring.

Conflicts of interest

The authors have nothing to disclose.