The prospective randomized double-blinded pilot study was undertaken after getting approval from the Institutional Ethical Committee-Human Research (IEC-HR), in its meeting held on 13-10-2018 and informed written consent from all participating patients. It is prospectively registered with CTRI (Ref. No. CTRI/2018/12/016693; registered on 17/12/2018). 

Adult patients of age between 16 and 65 years scheduled for PCNL surgery under general anesthesia were included in the study. Patients receiving renal replacement therapy or with transplanted kidney; or clinical evidence of significant dysrhythmia, cardiac failure, stroke, coronary heart disease (with left ventricular ejection fraction [LVEF] < 50% where echocardiography was indicated) were excluded.

Out of the total 33 patients assessed for eligibility, 1 patient who was receiving renal replacement therapy was excluded (Figure 1). Total of 32 patients were randomized into 2 groups, Group C and Group S, with 15 and 17 patients in each group respectively and received intervention. None of the patients were lost to follow up.

The baseline characteristics for both groups were statistically similar (P > 0.05, 95% Confidence Interval) (Table 1). The incidence of various systemic co-morbidities in group C and group S was as follows: hypertension (2 versus 4), diabetes mellitus (0 versus 2), hypothyroidism (1 each), obesity (i.e., BMI > 30 kg/m²) (0 versus 1) and human immunodeficiency virus infection (0 versus 1) respectively. More than 1 co-morbidity was present in some patients.

Incidence of early postoperative AKI (within 48 hours postoperatively) was clinically higher, but statistically similar, for group S as compared to group C (8/17 = 47.1% versus 4/15 = 26.7%) (P = 0.234, 95% Confidence Interval). The diagnosis of postoperative AKI is based on changes in serum creatinine as per the KDIGO guidelines.

The distribution of patients according to severity of AKI was also statistically similar between both groups (P = 0.312, 95% Confidence Interval). From among those who developed AKI, the percentage of patients in group C and group S with grade 1 AKI was 3/4 (75%) and 7/8 (87.5%); and for grade 2: 1/4 (25%) and 1/8 (12.5%). None of the patients developed grade 3 AKI.

The postoperative change in serum creatinine and estimated glomerular filtration rate (eGFR), on 1^st as well as 2^nd postoperative day, were also statistically similar between group C and group S (P > 0.05) (Table 2). However, the increase in eGFR and 24-hour urine output were clinically greater on first postoperative day for group S (Table 2). Although the hourly estimates of postoperative urine output were not available, the 24-hour volumes were noted and found to be statistically similar between both groups (Table 2).

Volume of intraoperative fluids was significantly higher for group C versus group S (P = 0.000, 95% Confidence Interval) (Table 3). Despite this, the need of vasopressor was significantly greater and hypotension (defined as fall in mean arterial presure by >20% from baseline) was clinically higher, though statistically similar for group C versus group S (P = 0.659) (Table 3). The serum lactate remained statistically similar between both groups prior to induction and at the end of surgery (P> 0.05) (Table 3).

The hemoglobin level was similar between group C and group S on each of pre-defined times i.e., preoperatively, postoperative day 1, as well as on postoperative day 2: 12.0 ± 1.9 versus 12.5 ± 2.3 gm/dl; 10.5 ± 1.6 versus 11.2 ± 2.0 gm/dl; and 10.6 ± 1.7 versus 11.3 ± 2.0 gm/dl respectively. (P = 0.338).

An intergroup comparison showed similar SVV in both groups at each of the predefined time points (P = 0.141), although it was clinically lower for group C as compared to group S (Figure 2). However, the SVV remained above threshold value of 10% in supine position and 14% in prone position on all times. The mean heart rate, invasive blood pressure as well as cardiac index were statistically similar between both groups at all observed time points (P = 0.150, 0.524 and 0.816 respectively) (Figure 3, Figure 4).

Figure 2

Figure 3

Figure 4

The surgical characteristics are represented in Table 4. Duration of surgery and volume of irrigation fluid were similar between both groups (P > 0.05). An intraoperative change of surgical plan to forgo the PCNL after URS with DJ stenting was made in 2 patients of each group (P = 1.000). Number of patients requiring multiple tracts during PCNL, and those with single functioning kidney preoperatively was statistically similar between both groups (P> 0.05) (Table 4). None of the patients in either group had intraoperative hemorrhage, required blood transfusion, or developed a pneumothorax (Table 4).

The electrolyte (serum sodium and potassium) as well as metabolic (pH, serum HCO3, and PaCO2) parameters were similar between group C and group S prior to induction as well as at the end of surgery (P > 0.05, 95% Confidence Interval) (Table 5).

Table 5

Characteristic	Group C (n = 15)	Group S (n = 17)	P value	SMD (95% CI)
Serum sodium (pre-induction) (mmol/l)	139.1 ± 6.7	140.4 ± 7.6	0. 719*	0.156 (-0.567 to 0.88)
Serum sodium (end of surgery) (mmol/l)	144.7 ± 9.5	145.0 ± 10.6		-0.025 (-0.748 to 0.699)
Serum potassium (pre-induction) (mmol/l)	3.6 ± 0.5	3.5 ± 0.5	0. 223*	0.003 (-0.721 to 0.726)
Serum potassium (end of surgery) (mmol/l)	3.7 ± 0.5	3.4 ± 0.5		-0.677 (-1.4 to 0.047)
pH (pre-induction)	7.39 ± 0.06	7.39 ± 0.04	0.359*	-0.077 (-0.8 to 0.647)
pH (end of surgery)	7.37 ± 0.10	7.33 ± 0.08		-0.497 (-1.221 to 0.226)
HCO3 (pre-induction) (mmol/l)	19.9 ± 3.8	18.9 ± 2.8	0.759*	-0.304 (-1.027 to 0.42)
HCO3 (end of surgery) (mmol/l)	17.9 ± 3.0	18.4 ± 3.2		0.143 (-0.58 to 0.867)
PaCO2 (pre-induction) (mmHg)	33.1 ± 7.7	32.1 ± 4.6	0.867*	-0.25 (-1 to 0.499)
PaCO2 (end of surgery) (mmHg)	31.6 ± 12.3	33.3 ± 8.8		0.241 (-0.508 to 0.991)

Values are mean ± SD. * P-value for inter-group comparison using the repeated measure ANOVA test, SMD-Standardised mean difference.

This randomized controlled pilot study aimed to evaluate and compare early postoperative AKI in patients undergoing PCNL surgery, following intraoperative use of conventional versus SVV directed fluid therapy. It was a pilot study, and there appears to be a clinical trend of greater incidence of AKI with use of SVV guided therapy despite better intraoperative hemodynamic stability and greater improvement in postoperative urine output and eGFR in the first postoperative day (P > 0.05).

The incidence of postoperative AKI in the control group was 26.7%, wherein conventional intraoperative fluid therapy was used. Previous studies noting postoperative AKI following PCNL surgery are few in number. 2, 4, 5 Herein the incidence was noted to be 11%, 5 16.2%, 2 and 25% 4 respectively, albeit without any mention of intraoperative fluid therapy. Thus, our observed incidence of postoperative AKI in the control group (26.7%) is similar to the previously reported figure of 25%. 4 Postoperative AKI was seen in 47.1% patients of our treatment group, wherein SVV guided intraoperative fluid therapy was used. We could not locate any previous evidence evaluating intraoperative SVV guided fluid therapy for effect on postoperative AKI. There is however some earlier data during non-PCNL surgery, showing decreased renal complications and increased urine output following intraoperative SVV directed fluid therapy. 10, 13 Indeed, we also noted greater improvement in postoperative urine output and eGFR on the first postoperative day, although it was not statistically significant. The previous studies did not evaluate occurrence of AKI per se and hence the same cannot be compared.

Given the much greater incidence of AKI with SVV guided therapy, it is tempting to conclude its detrimental effect on renal function. However, there are reasons to desist against a conclusive result at present. Firstly, this is only a pilot study and thus not adequately powered. Secondly, the higher incidence of AKI is contradicted by the clinical trends for better renal function on first postoperative day (urine output and estimated glomerular filtration rate) as well as greater intraoperative hemodynamic stability (lesser hypotension and vasopressor usage) with the use of SVV guided fluid therapy. Intraoperative hypotension is well-known to be perhaps the commonest risk factor for causing perioperative AKI. 2, 12 It is possible then, that the higher incidence of AKI based on serum creatinine could merely be a reflection of relatively lower volumes of fluid therapy used with SVV guided therapy. The role of serum creatinine for diagnosing AKI is itself riddled with limitations, though no substitute has been discovered despite years of research. Lastly, despite randomization, patients with risk factors for postoperative AKI were greater in those who received SVV directed therapy. These associated risk factors included presence of a single functioning kidney, preoperative hypertension, diabetes mellitus and obesity. 2, 4, 5

We noted that majority of the patients with postoperative AKI, in either group, had a mild disease (stage 1), with none developing the most severe form (stage 3). Previously, only 1 out of 3 studies evaluating postoperative AKI following PCNL surgery noted its severity distribution. 4 Herein, the severity distribution showed more severe stages of AKI, probably because the study was conducted solely in patients with solitary kidney undergoing PCNL surgery, an independently associated risk factor of AKI. In our patients, those with solitary functioning kidney were 0% and 12% among those receiving conventional/SVV guided fluid therapy respectively.

Thus, our findings suggest a clinical utility of SVV guided fluid therapy for maintaining hemodynamic stability during PCNL surgery. Effect of SVV guided fluid therapy on postoperative renal function however will need further research, being guided by the present observations.

We thus recommend that using SVV guided intraoperative fluid therapy in patients undergoing PCNL surgery under general anesthesia is feasible and safe in terms of maintaining fluid balance as well as hemodynamic stability. The failure of better intraoperative hemodynamic stability to translate into greater preservation of renal functions with SVV directed therapy will need further research.