Is Nonpapillary Puncture a Feasible Alternative to Papillary Access in Percutaneous Nephrolithotomy? A Systematic Review and Meta-Analysis

Introduction

Percutaneous nephrolithotomy (PCNL) is widely recognized as the gold standard for the treatment of renal calculi larger than 2 cm, as well as complex stones such as staghorn calculi. ¹ However, since PCNL requires percutaneous penetration of the renal parenchyma, it carries a considerable risk of bleeding—making the selection of puncture path a core critical step in the effective implementation of the procedure. ² Based on anatomical studies of renal vasculature and the cadaveric research by Professor Sampiao,^3,4 the optimal and safest puncture path with the lowest bleeding risk is considered to be a vertical puncture through the papilla of the selected calix, targeting Brödel’s avascular plane. This remains the mainstream consensus in current clinical practice.

Despite its theoretical safety advantage, papillary puncture has inherent drawbacks. It is technically demanding, resulting in a prolonged learning curve for novice urologists.^5,6 Moreover, puncture through papillae of different calices presents specific limitations: upper calix puncture carries a risk of pleural injury; middle calix puncture fails to address lower calix stones with steep angulation; and lower calix puncture makes antegrade ureteral stent placement challenging. Consequently, in clinical practice, management of large or multicentric caliceal stones often requires multiple access tracts to achieve a high stone-free rate (SFR), increasing procedural complexity and potential morbidity. ⁷

In response to these limitations, researchers have proposed the concept of nonpapillary puncture. Theoretically, this approach offers several advantages: a larger target area, lower technical difficulty, and shorter operative time; the ability to manipulate stones from the caliceal neck to the calix, preventing stone migration; enhanced nephroscope maneuverability, which reduces the need for multiple access tracts when treating large, multicentric, or staghorn calculi; and easier identification of the ureter for antegrade stent placement.

Kyriazis et al. first validated the clinical safety and feasibility of nonpapillary puncture in a prospective cohort of 137 patients, reporting shorter operative time, higher SFR, minimal bleeding, and low complication rates. ⁷ Subsequent studies have confirmed the feasibility of nonpapillary puncture in patients with staghorn calculi and renal malformations, and consistent conclusions have been drawn for nonpapillary puncture combined with miniaturized access tracts.^8–10 However, conflicting evidence emerged from a porcine study by Hao et al., which found a higher bleeding risk with nonpapillary puncture compared with papillary puncture. ¹¹

In recent years, a growing number of comparative studies have been conducted to further explore the feasibility of nonpapillary puncture. Studies by Kallidonis et al., Budak et al., Tahra et al., Kashi et al., Farshid et al., and Amer et al. have consistently shown no significant differences between the two approaches in terms of bleeding, complication rate, operative time, or SFR, supporting the safety and feasibility of nonpapillary puncture.^12–18 In contrast, a recent study by Li et al. and a retrospective analysis of data from our own institution (the First Affiliated Hospital of Nanjing Medical University) suggested that nonpapillary puncture was associated with a higher bleeding risk and complication rate compared with papillary puncture. ¹⁹

Given the persisting absence of consensus regarding the comparative safety and efficacy of nonpapillary vs papillary access in PCNL, we conducted this systematic review and meta-analysis to synthesize the available evidence. Our primary objective was to evaluate the impact of access site on postoperative hemoglobin drop, with secondary outcomes including complication rates, SFRs, and operative times, aiming to provide evidence-based guidance for clinical decision-making.

Methods

Results

Literature search and study characteristics

A total of 847 studies were retrieved through the systematic literature search. After sequential screening based on the inclusion and exclusion criteria, eight published studies and a retrospective analysis of data from our own institution (the First Affiliated Hospital of Nanjing Medical University) were finally included in this meta-analysis, and the detailed literature search process is illustrated in Figure 1.

OPEN IN VIEWER

FIG. 1.

Flow diagram of study selection process.

Among the included studies, two were RCTs and seven were observational studies (one prospective and six retrospective), with publication years ranging from 2017 to 2025. A total of 1229 patients were enrolled, with 787 cases in the papillary access group and 442 cases in the nonpapillary access group. The basic characteristics of the included studies are summarized in Table 1. Specifically, postoperative hemoglobin drop was reported in eight published studies, but only five provided complete data on mean and standard deviation (SD); overall complication rates were documented in seven studies; SFRs were reported in seven studies; and operative time was described in all the nine included studies, with complete mean and SD data available in seven studies (Table 2).

OPEN IN VIEWER

Table 1.

Main Characteristics of Studies Included in This Meta-Analysis

Study	Year	Country	Ethnicity	Design	Number of patients		Mean age (SD)		Sex, M/F		BMI (kg/m2)		Stone size (mm)		Assessments
					Papillary	Nonpapillary	Papillary	Nonpapillary	Papillary	Nonpapillary	Papillary	Nonpapillary	Papillary	Nonpapillary
Kallidonis	2017	Greece	European	RCT	27	28	58.11 ± 12.72	53.68 ± 14.29	19/18	11/17	25.83 ± 2.89	26.07 ± 2.63	14.25 ± 5.82	14.86 ± 6.6	a,b,d
Budak	2018	Turkey	European	Retrospective	104	91	47.4 ± 14.4	50.4 ± 12.6	66/38	51/40	NM	NM	NM	NM	b,c
Tahra	2020	Turkey	European	Retrospective	207	69	51.6 ± 13.54	52.2 ± 12.43	66/141	24/45	28.3 ± 5.2	27.9 ± 3.6	24.6 ± 46	23.8 ± 51	b–d
Kashi	2022	Iran	Asian	Prospective	67	67	50.75 ± 13.30	48.43 ± 13.10	52/15	49/18	NM	NM	27.23 ± 13.71	30.65 ± 15.20	a–d
Farshid	2025	Iran	Asian	Retrospective	55	48	42.7 ± 8.7	41.6 ± 8.6	30/25	33/15	28.99 ± 5.34	26.73 ± 4.03	NM	NM	a–d
Amer	2025	Egypt	African	RCT	35	35	43.11 ± 13.33	42.66 ± 12.63	18/17	19/16	25.74 ± 3.33	24.67 ± 3.79	26.3 ± 4.3	27.2 ± 5.2	b–d
Chen	2025	China	Asian	Retrospective	82	28	56 ± 10	56 ± 11	40/42	20/8	24 ± 3	25 ± 4	2.40 (1.90, 3.15)	2.80 (2.23, 3.50)	b,c
Li	2025	China	Asian	Retrospective	182	36	47.6 ± 14.2	48.3 ± 11.2	87/95	16/20	24.9 ± 7.9	26.9/8.7	43 ± 16	45 ± 17	a,c,d
Meng	2025	China	Asian	Retrospective	28	40	57.04 ± 8.19	49.23 ± 11.79	26/2	33/7	24.09 ± 3.08	23.73 ± 3.29	33.21 ± 20.79	44.76 ± 22.44	a,d

a

Hb (hemoglobin) drop (g/dL).

b

Overall complications.

c

Stone-free rate (SFR).

d

Operative time (minutes).

BMI = body mass index; RCT = randomized controlled trial; SD = standard deviation.

OPEN IN VIEWER

Table 2.

Summary of Outcome Measures

Study	Number of patients		Hb drop {g/dL, MD ± SD/median [M(Q)]}		Overall complications		Stone-free rate		Operative time {min, MD ± SD/MD [M(Q)]}
	Papillary	Nonpapillary	Papillary	Nonpapillary	Papillary	Nonpapillary	Papillary	Nonpapillary	Papillary	Nonpapillary
Kallidonis	27	28	1.54 ± 1.29	1.35 ± 0.79	2	2	NM	NM	51.97 ± 16.1	43.21 ± 12.38
Budak	104	91	1.8 (−0.90 to 5.80)	1.6 (−0.80 to 8.70)	27	17	74	67	102.5 (40–245)	100 (45–200)
Tahra	207	69	NM	NM	15	5	179	59	58.3 ± 14.3	56.8 ± 15.3
Kashi	67	67	1.978 ± 1.292	1.873 ± 1.126	12	12	55	51	90.22 ± 22.52	87.58 ± 25.80
Farshid	55	48	1.14 ± 1.38	1.1 ± 1.49	8	11	49	43	81.5 ± 12.9	76.08 ± 14.8
Amer	35	35	14.3	14.4	8	6	29	31	92.11 ± 25.03	82.91 ± 21.33
Chen	82	28	0.8 (−0.08 to 1.575)	0.45 (−0.25 to 0.93)	15	6	69	21	127.5 (105.0–162.3)	144.0 (100.0–182.0)
Li	182	36	1.04 ± 0.47	1.84 ± 0.54	NM	NM	148	25	51.6 ± 18.7	59.2 ± 17.6
Meng	28	40	3.99 ± 1.72	4.21 ± 1.62	NM	NM	NM	NM	76.07 ± 44.31	93.35 ± 35.22

Stone-free rate was defined as no residual stones or residual stone ≤2–4 mm and assessed by CT/KUB at 1–3 months postoperatively in studies by Budak et al., Tahra et al., Amer et al., Chen et al., and Li et al. and defined as no residual stones or residual stone <3 mm and assessed by CT/KUB/ultrasonography within 2 days to 2 weeks postoperatively in studies by Kashi et al. and Farshid et al.

CT = computed tomography; KUB = kidney-ureter-bladder radiograph; MD = mean difference; NM = not mentioned.

The results of quality assessment are presented in Figure 2. For the two included RCTs, the risk of bias was low in the domains of random sequence generation, completeness of outcome data, outcome measurement, and selective outcome reporting (Fig. 2A). All seven observational studies achieved an NOS score of ≥7, thus being classified as high-quality studies (Fig. 2B).

OPEN IN VIEWER

FIG. 2.

Literature quality assessment results. (A) Risk-of-bias assessment of the two randomized controlled trials (RCTs) using the Cochrane Risk of Bias Tool 2.0. Green circles (+) indicate low risk of bias, yellow circles (−) indicate some concerns. (B) Newcastle-Ottawa Scale (NOS) quality scores for the seven observational studies. The asterisks (*) represent individual awarded points within the three NOS domains (Selection, Comparability, Exposure/outcome assessment). All observational studies reached a total NOS score ≥7 and were categorized as high-quality studies.

Comparative analysis of nonpapillary vs papillary access in PCNL: Hemoglobin drop, complications, SFR, and operative time

Postoperative hemoglobin drop was evaluated across five studies using a random-effects model (Fig. 3A). No statistically significant difference was observed between the nonpapillary and papillary access groups (MD = 0.159, 95% CI = −0.365 to 0.683, p = 0.5519), but high heterogeneity was detected (I² = 86%, p < 0.0001). This high heterogeneity was primarily attributed to the study by Li et al. (2025) (Fig. 3B). After excluding the primary source of heterogeneity, a fixed-effects model still showed no significant between-group difference (MD = −0.075, 95% CI = −0.345 to 0.194, p = 0.5842), with heterogeneity completely eliminated (I² = 0%, p = 0.8711; Fig. 3C). No publication bias was identified (Egger’s test p = 0.2884; Begg’s test p = 0.4969; Fig. 3D). In addition, subgroup analyses by region (Fig. 3E) and study design (Fig. 3F) also revealed no significant differences between subgroups.

OPEN IN VIEWER

FIG. 3.

Meta-analysis of postoperative hemoglobin drop between nonpapillary and papillary access groups. (A) Forest plot of five studies with complete mean and standard deviation (SD) data. (B) Sensitivity analysis (leave-one-out method). (C) Forest plot after excluding Li et al. (2025). (D) Funnel plot for publication bias assessment. (E) Subgroup analysis by region. (F) Subgroup analysis by study design.

For the overall complication rates, seven studies were pooled using a fixed-effects model. The RR of complications in the nonpapillary group relative to the papillary group was 0.949 (95% CI = 0.701–1.286, p = 0.7371; Fig. 4A), indicating no significant difference. No heterogeneity (I² = 0%, p = 0.8117) or publication bias (Egger’s test p = 0.4321; Begg’s test p = 0.8806; Fig. 4B) was observed. Table 3 presents the occurrence of complications recorded in the included literature. For blood transfusion rates, seven studies were analyzed using a fixed-effects model (Fig. 4C). The RR of transfusion in the nonpapillary group was 0.835 (95% CI = 0.483–1.443, p = 0.5174), with no significant difference between the groups. No heterogeneity (I² = 0%, p = 0.7748) or publication bias (Egger’s test p = 0.9167; Begg’s test p = 0.6523; Fig. 4D) was detected.

OPEN IN VIEWER

FIG. 4.

Meta-analysis of overall complication and blood transfusion rates between the nonpapillary and papillary access groups. (A) Forest plot of overall complication rate (seven studies). (B) Funnel plot for publication bias of overall complications. (C) Forest plot of blood transfusion rate (seven studies). (D) Funnel plot for publication bias of blood transfusion.

OPEN IN VIEWER

Table 3.

The Incidence of Postoperative Complications

Study	Number of patients		Blood transfusion		Hematuria		Fever		Pseudoaneurysm		Urinary leakage		Pulmonary embolism		Urosepsis		Pleural injury		Colon damage
	Papillary	Nonpapillary	Papillary	Nonpapillary	Papillary	Nonpapillary	Papillary	Nonpapillary	Papillary	Nonpapillary	Papillary	Nonpapillary	Papillary	Nonpapillary	Papillary	Nonpapillary	Papillary	Nonpapillary	Papillary	Nonpapillary
Kallidonis	27	28	1	0	0	1	0	1	1	0	NM	NM	NM	NM	NM	NM	1	0	NM	NM
Budak	104	91	8	5	NM	NM	7	5	NM	NM	2	1	1	0	1	1	1	1	NM	NM
Tahra	207	69	8	1	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM
Kashi	67	67	8	6	NM	NM	2	3	NM	NM	2	3	NM	NM	NM	NM	NM	NM	NM	NM
Farshid	55	48	NM	NM	NM	NM	3	6	NM	NM	2	3	NM	NM	NM	NM	3	2	0	0
Amer	35	35	3	4	NM	NM	8	6	NM	NM	1	0	NM	NM	0	1	NM	NM	NM	NM
Chen	82	28	2	2	NM	NM	10	2	NM	NM	NM	NM	NM	NM	2	0	NM	NM	NM	NM
Li	182	36	4	1	NM	NM	5	4	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM
Meng	28	40	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM	NM

Regarding the SFR, a fixed-effects model was applied to seven studies (Fig. 5A). Among the seven included studies that reported an SFR, five studies (Budak et al., Tahra et al., Amer et al., Chen et al., Li et al.) assessed the SFR at 1–3 months postoperatively, primarily using CT, with some cases evaluated by kidney-ureter-bladder radiograph (KUB). SFR was defined as no residual stones or residual stone ≤2–4 mm. The remaining two studies (Kashi et al., Farshid et al.) assessed the SFR within a shorter term (2 days to 2 weeks postoperatively), using CT, KUB, or ultrasonography, with SFR defined as no residual stones or residual stone <3 mm.

OPEN IN VIEWER

FIG. 5.

Meta-analysis of stone-free rate between the nonpapillary and papillary access groups. (A) Forest plot of seven studies. (B) Funnel plot for publication bias assessment.

The RR of achieving a stone-free status in the nonpapillary group was 0.970 (95% CI = 0.909–1.035, p = 0.3527), demonstrating no significant difference between the groups. No heterogeneity (I² = 0%, p = 0.6927) or publication bias (Egger’s test p = 0.2995; Begg’s test p = 0.4527; Fig. 5B) was observed.

Operative time was analyzed across seven studies using a random-effects model. No significant difference was found between the nonpapillary and papillary access groups (MD = −1.852, 95% CI = −6.745 to 3.040, p = 0.4580; Fig. 6A), with moderate heterogeneity (I² = 68%, p = 0.0046). Sensitivity analysis confirmed the robustness of this finding (Fig. 6B). Subgroup analyses by region (Fig. 6C) and stone size (Fig. 6E) did not identify significant sources of heterogeneity. Notably, in the prospective study subgroup (three studies), the nonpapillary access group was associated with a significantly shorter operative time compared with the papillary access group (MD = −6.608, 95% CI = −11.573 to −1.643, p = 0.0091; Fig. 6D).

OPEN IN VIEWER

FIG. 6.

Meta-analysis of operative time between the nonpapillary and papillary access groups. (A) Forest plot of seven studies with complete mean and standard deviation (SD) data. (B) Sensitivity analysis (leave-one-out method). (C) Subgroup analysis by region. (D) Subgroup analysis by study design. (E) Subgroup analysis by stone size. (F) Funnel plot for publication bias assessment.

Discussion

This meta-analysis, encompassing nine studies and 1229 patients, demonstrates that nonpapillary and papillary access for PCNL is comparable in terms of postoperative hemoglobin drop, complication rates, and SFRs. Notably, although no significant overall difference in operative time was observed between the two approaches, subgroup analysis of prospective studies revealed that nonpapillary access was associated with a significantly shorter operative time. This discrepancy between the subgroup and overall results likely reflects the methodological rigor of prospective designs: prospective studies typically standardize procedural workflows (e.g., puncture localization, tract dilation) more consistently than retrospective studies, which may mask the technical efficiency of nonpapillary access. The shorter operative time in prospective nonpapillary cohorts is plausible given the approach’s inherent advantages: a larger target area (reducing puncture attempts), direct trajectory to calculi (minimizing nephroscope manipulation), and easier antegrade stent placement—all of which streamline the surgical process, particularly in the hands of experienced operators. This finding supports the clinical utility of nonpapillary access for optimizing procedural efficiency, especially in standardized, high-volume settings.

The primary outcome of postoperative hemoglobin drop showed no significant difference between groups, even after excluding the outlier study (Li et al., 2025) that accounted for high heterogeneity. Several factors may collectively contribute to this equivalent bleeding risk. First, variations in puncture localization methods across studies should be emphasized: Most European studies adopted X-ray-guided puncture, whereas Asian studies (those conducted in China) predominantly relied on ultrasound guidance. Doppler ultrasound offers a unique advantage of real-time blood flow signal detection, enabling surgeons to proactively avoid puncturing blood-rich regions intraoperatively.^21–23 This technical difference may have mitigated potential bleeding disparities between the two access paths, contributing to the nonsignificant pooled result. Second, the anatomical basis of PCNL puncture further explains the comparable bleeding risk: The majority of percutaneous access is performed in the prone position, with the target region mostly located in Brödel’s avascular plane—a relatively blood-sparse area of the renal cortex. ²⁴ Regardless of whether papillary or nonpapillary access is adopted, puncturing within this anatomical plane minimizes vascular injury, which directly aligns with our finding of no significant difference in postoperative hemoglobin drop between the two groups.

Third, it is crucial to recognize that bleeding risk in PCNL is multifactorial; the puncture path is merely one component. Other influential factors include tract dilation technique, access tract diameter, intraoperative manipulation (e.g., stone fragmentation intensity, renal pelvis irritation), and inherent individual differences in renal vascular anatomy.^25–28 These confounding variables may have interacted and collectively masked potential subtle differences in bleeding risk between the two puncture approaches, further influencing the judgment of the outcome. Fourth, the ambiguity of puncture accuracy in clinical practice should be considered. In real-world clinical settings, achieving absolute papillary puncture is often challenging. Many cases presumed to be papillary puncture may have been inadvertently converted to nonpapillary access because of the absence of intraoperative CT verification to confirm the puncture site. This clinical reality has been implicitly validated in practice and may partially explain why no significant difference was observed between the two groups in our meta-analysis.

Although papillary puncture is theoretically the safest approach because of its avascular characteristics, it has inherent limitations: It is technically demanding and unfriendly to novice urologists, and exhibits disadvantages in managing multiple caliceal stones and placing ureteral stents via lower caliceal puncture. With the rapid advancement of artificial intelligence (AI) technology, the concept of papillary puncture may be gradually weakened in the near future.^29,30 It is plausible that AI, combined with preoperative CT angiography for vascular and anatomical assessment, could integrate with ultrasound to achieve a truly precise fused-image-guided puncture—optimizing access safety and efficiency without being constrained by the traditional distinction between papillary and nonpapillary paths.

The absence of heterogeneity in complication and SFRs further supports the equivalence of the two techniques, confirming that neither approach confers a safety or efficacy advantage in these domains. The moderate heterogeneity observed in operative time, which was not explained by region or study design, likely reflects variations in surgeon experience, stone complexity, intraoperative decision-making, localization method-related efficiency differences, and technical proficiency in puncture approaches across studies.

For urologists, these data confirm that both access techniques are safe and effective, allowing for greater flexibility in surgical planning. Nonpapillary access can be appropriately used for peripheral or complex stones, whereas papillary access remains a reliable option for central caliceal stones. The choice of localization method (ultrasound vs X-ray) should be guided by institutional expertise, stone characteristics, and the need for real-time blood flow assessment—Doppler ultrasound may be preferred in settings where avoiding vascular-rich regions is prioritized. In addition, clinicians should acknowledge the ambiguity of puncture site confirmation in routine practice and the multifactorial nature of bleeding risk. Looking ahead, embracing AI-integrated imaging technology may revolutionize PCNL access planning, transcending traditional puncture classifications to achieve personalized, precision-guided procedures. Future research should focus on large, multicenter RCTs with standardized operative protocols (including uniform localization, dilation techniques, and intraoperative verification of puncture sites) to clarify the interactive effects of puncture path, localization method, and other procedural factors on bleeding risk. In addition, studies exploring AI-assisted fused-image-guided puncture are warranted to validate its clinical utility.

This study has limitations inherent to meta-analyses of observational data: The majority of included studies were retrospective, introducing potential selection bias. In addition, significant heterogeneity in some outcomes (e.g., postoperative hemoglobin drop) may be partially attributed to unmeasured variations in localization methods, puncture accuracy, and other procedural factors, which limits generalizability. The small number of RCTs included reduces the statistical power to detect subtle differences between groups. Individual patient data were not available, precluding adjustment for confounding variables such as stone burden, tract diameter, dilation technique, and individual renal vascular anatomy—factors that may influence bleeding risk independently of puncture path. Finally, another limitation is the inclusion of unpublished data from our team’s ongoing study. Although these data were extracted from a well-designed retrospective cohort with standardized CRFs, and sensitivity analysis excluding this dataset did not alter the primary conclusions, the absence of external peer review for this ongoing study may introduce potential bias in data interpretation. We have provided detailed data of this ongoing study in Table 1 to ensure transparency, and the full study will be submitted for publication separately to avoid data duplication.

Conclusion

Nonpapillary access should be recognized as a viable and valuable option in PCNL practice, offering comparable safety and efficacy to papillary access with potential procedural efficiency benefits. The findings of this meta-analysis provide evidence-based guidance to individualize surgical planning, ultimately improving patient care in the management of large and complex renal calculi.

Authors’ Contributions

X.M. and Y.X.: Protocol/project development. Z.S. and X.W.: Data collection or management. C.J. and X.M.: Data analysis and article writing/editing.