This prospective observational study was performed between June 2017 to September 2019, after obtaining permission from the appropriate authority within the institute. A total of 100 consecutive patients of chronic kidney disease visiting Department of Urology, Superspeciality Hospital and Government Medical College, Nagpur, Maharashtra, India, and seeking operative treatment for urolithiasis were included. Written informed consent was obtained from all the patients. Sample size was calculated by nMaster 1.0 Sample size software developed by Christian Medical College, Vellore, Tamil Nadu, India. Patients were evaluated for their demographics like age, sex, weight, co-morbidity, significant medical and surgical history.

Inclusion and Exclusion criteria: All suspected patients underwent a plain CT scan. Patients less than 18 years of age or more than 70 years, immunodeficient patient, patient with bleeding disorders, patient with anatomic anomalies of kidney like horseshoe kidney, patients with prior renal surgery were excluded from the study. However, patients with co-morbidities like hypertension and diabetes were not excluded. All the CKD patients with the urolithiasis were included.

The included patients were classified into three groups; group 1 included stage III CKD patients, group 2 included stage IV CKD patients and group 3 stage included stage V CKD patient. All patients underwent an Technetium-99m ethylene dicysteine renal scan (EC scan) to look at the functional status of each kidney.

Patients with the features of obstruction and Acute Kidney Injury (AKI) underwent urine diversion with either PCN insertion or Double J stent (DJ stenting) till nadir serum creatinine was reached, before posting the patient for surgery. Stone analysis was done using the X-ray diffraction technique. The GFR was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) (2009) equation [7]. CKD was defined according to the eGFR as outlined by the National kidney foundation [8]. ERPF was calculated by measuring change in the gamma radiation of the radionucleotide 99mTm EC (Technetium-99m Ethylene Dicysteine) over fixed time using computer. After the surgery, the complete stone clearance was defined as the non visualisation of residual fragments in X-ray or ultrasound.

All PCNLs were performed in prone position. Initially a retrograde pyelogram was done and calyx was selected for renal puncture. Following renal puncture, a guidewire was parked in the renal pelvicalyceal system. Over the guidewire, the tract was dilated upto 24 to 26 F. Using 22F nephroscope the stone was visualised and broken with pneumatic lithoclast. A 6/26 DJ stent was kept in all cases.

In the perioperative period, all patients undergoing PCNL were monitored for short term parameters like blood transfusion, urinary tract infection with fever and sepsis, the incidence of AKI, hospital stay, residual calculus, stone composition. Long term impact was studied by the change in estimated Glomerular Filtration Rate (eGFR), Estimated Renal Plasma Flow (ERPF), stage migration, need for renal replacement therapy, and mortality.

In the long term impact of PCNL on renal function, eGFR levels were measured before surgical intervention and three months postsurgery. EC scan was repeated after three months in all the patients. All patients underwent an Ultrasound (USG) scan to look for the residual stone.

All statistical tests were done using Statistical Package for the Social Sciences (SPSS) version 27. 0. Statistical test applied were Kruskal Wallis Test for Age distribution, Chi-square test for gender distribution and frequency of AKI. Fisher’s-exact test was used for stone composition and distribution of diabetes, need for transfusion. Kruskal Wallis Test was used for the distribution of BMI, hospital stay, change in ERPF and eGFR. Stuart-Maxwell test was used to assess stage migration of CKD. Paired Wilcoxon test was used to explore the difference in ERPF and eGFR. Wilcoxon rank-sum test (Mann-Whitney U test) was used to compare the change in renal function with diabetes.

In patients enrolled in the study, 55 patients were Grade III CKD, 31 were Grade IV CKD and 14 patients were of stage V CKD and classified into groups 1, 2 and 3, respectively [Table/Fig-1]. The mean age of patients was 47.05, 46.0, 43.79 years in groups 1,2 and 3, respectively. Age distribution was not significantly different among these groups (p-value=0.968). The male to female ratio was 4:1. There is no significant difference in gender distribution among the three groups (p-value= 0.177). Thirteen patients were suffering from diabetes mellitus. There was no significant difference between these groups in terms of the distribution of diabetes (p-value=0.178). There was no significant difference between the groups in terms of BMI (Kg/m²) (p-value=0.626). The mean preoperative creatinine in non diabetic was 2.85±0.99 and in diabetic was 2.86±0.81.

The most common stone composition was calcium oxalate dihydrate (53%). Other stone compositions were calcium oxalate monohydrate (24%), phosphate (6%) and uric acid (12%). There was no difference in stone composition (p-value=0.476) in the three groups. Urine was sterile in 77% of the patient. The most common organism in infected urine was Pseudomonas aeruginosa (12%) followed by E. coli (7%) and Klebsiella (4%).

For the short term impact of PCNL, we evaluated hospital stay, blood transfusion, the incidence of AKI, urinary tract infection, residual stones [Table/Fig-2]. Around nine patients developed urinary tract infection with fever and sepsis. No significant difference was found with Fisher’s-exact test in the incidence of urinary tract infection in these groups (p-value=0.629). Around 26% of patients developed AKI in the postoperative period with patients in group 3 having significantly higher incidence (p-value=0.001). A 9% of patients had residual stones after surgery. Stone free rate was not significantly different among these groups (p-value=0.252). Blood transfusion rates were similar in these groups (p-value=0.066). Mean hospital stay was 2.38, 3.55, 4.00 days in groups 1, 2 and 3, respectively. Mean hospital stay was significantly higher with higher stages of CKD (p-value <0.001). Mean hospital stay was also found to be higher with larger stone load and multiple punctures. Staghorn calculus involving 2 or more calyx required 2 or more punctures, had longer operating time and hospital stay.

Change in renal function was evaluated by change ERPF and eGFR. The ERPF significantly increased by 1.96±4.87 (p-value=0.001) after surgery. Change in ERPF after surgery in groups 1, 2 and 3 were 1.43, 1.9, 4.15, respectively. All three groups showed a statistically significant increase in ERPF [Table/Fig-3]. Also, there was a significant difference in change in ERPF (mL/min) over time between the three groups. Mean eGFR significantly increased by 1.08±2.69 (p-value=0.001) after surgery. Change in eGFR after surgery in group 1, 2 and 3 were 1.38, 0.99, 0.07, respectively. three groups showed a statistically significant increase [Table/Fig-4]. Also, the overall p-value for comparison of the change in eGFR over time between the three groups (Generalised Estimating Equations Method) is 0.033.

Change in ERPF in DM was 5.75 and non diabetic was 1.39 [Table/Fig-5]. This change was statistically significant (p-value=0.009). Change in eGFR in DM was 2.14 and non diabetic was 0.92. There was a significant difference in the trend of eGFR (mL/min/m²) over time in both groups (p-value=0.037). Stage migration was seen as not statistically different among these stages of CKD (p-value=0.321) [Table/Fig-6]. A total of 21% of patients had a history of Haemodialysis (HD) during their presentation with AKI. A 5% patient posted for surgery required perioperative HD. During follow-up for one year 7% patient required Maintenance HD. There was no perioperative mortality. However, during the follow-up for one year three patients succumbed to death secondary to the complications of chronic kidney disease.

In one patient with 2 cm pelvic calculus was taken for PCNL but initial puncture in PCS could not be made even after multiple attempts and PCNL couldn’t be performed. Patient died on postoperative day 3 after developing complications of CKD. No patient was lost to follow-up.

A 73% of patients presented with AKI and required some form of diversion. Of these patients, more than 95.9% of patients were not known CKD patients but were diagnosed as CKD on serial follow-up. Around 14% of patients presented with severe CKD (stage V). Renal stones contribute to poor renal function but the exact contribution of stone on renal function is less defined [9]. Age and gender were equally distributed among different CKD groups. Mean creatinine value was slightly high in diabetic patients and the difference was not significant. Diabetes may be associated with early presentation of these patients and therefore, would compensate for the overall equivalent renal functions in these patients. Staghorn calculus is a large renal calculus that fills the whole pelvis with one renal calyx or pelvis into multiple calyces [10-12]. Staghorn calculus is associated with progressive renal deterioration [13]. CKD patients have overall higher infectious complications [9]. Similarly, in the present study, the overall incidence of postoperative fever and sepsis was higher with increasing stage of CKD. The difference was, however, not significant owing to a lower number of recruited patients. The most common urinary infection is E. coli [14] but owing to the diversion in hospital setting pseudomonas was found to be of higher incidence. Similar to the findings of Clinical Research Office of the Endourological Society (CROES) PCNL Global Study the mean hospital stay was significantly greater in higher stages of CKD [9]. Blood transfusion rates were higher in the higher stage CKD patients. The perioperative blood transfusion may be more influenced by the poor nutrition status and anaemia in these patients, stone load and operative time rather than the CKD stage itself. Gopalakrishnan G and Prasad GS stated greater haemolytic tendency, thinned out parenchyma, lower tamponade effect as a more important factor for perioperative blood transfusion [15]. Higher blood transfusion rates did not amount to increased residual calculus with higher stages of CKD contradicting the findings of Gambaro G et al., who stated increased risk of residual stone and repeat PCNL with higher CKD stage [16]. Numerous studies suggested that removal of stones in CKD patients may prevent renal deterioration [4,17,18]. Overall the detrimental effect of PCNL appears to be balanced with no perioperative mortality in the present study. Both mean eGFR and ERPF show a significant increase after surgery. The improvement eGFR was significantly different among the three stages of CKD. The ERPF improvement was, however, better in higher stages of CKD. But the clinical impact of this improvement appears to be insignificant. Kurien A et al., demonstrated, better the renal function preoperatively, the more will be the improvement after surgery [19]. Although these studies elaborate that the need for haemodialysis can be deferred with aggressive surgery [20] but comment could not be made in the present study due to shorter follow-up.

Limitations of the study included limited number of patients. A prospective randomised control trial would have made the results of the study more impactful. This study will pave the path for future research in the area and it will help in decision making for management of renal stones in CKD patients.

The PCNL in patients of chronic kidney disease has a minor but significant improvement in GFR with good clearance rates. Urinary infection and stone composition do not affect improvement in GFR. In diabetic patients also, PCNL appears to have a positive impact on GFR improvement in CKD patients.