Impact of preoperative MRI on surgical outcomes of robot-assisted laparoscopic prostatectomy: a single-center retrospective cohort of 844 patients

Abstract

Background:

Multiparametric magnetic resonance imaging (mpMRI) refines local staging in prostate cancer, but its effect on hard surgical endpoints after robot-assisted laparoscopic prostatectomy (RALP) remains uncertain.

Objectives:

To evaluate whether systematic preoperative mpMRI implementation was associated with improved surgical margin status and early biochemical control in a mature robotic RALP program.

Design:

Single-center retrospective cohort study comparing two calendar periods (pre‑mpMRI era vs mpMRI era).

Methods:

We included 844 consecutive men who underwent RALP for cT1–cT2 N0 M0 prostate adenocarcinoma between January 2007 and December 2019. Patients were classified into a pre-mpMRI era (2007–2012) and a post-mpMRI era (2014–2019), excluding 2013 as a transition year. From 2014, standardized 3T mpMRI interpreted by a single uroradiologist was systematically integrated into side-specific surgical planning, including nerve-sparing decisions. The primary endpoint was any positive surgical margin (PSM); the secondary endpoint was biochemical recurrence (BCR, prostate-specific antigen (PSA) ⩾ 0.2 ng/mL) within 24 months. Logistic regression (overall and stratified by pT2/pT3) assessed associations between era and PSM, with multivariable adjustment for age, PSA, tumor volume, biopsy Gleason score, D’Amico risk group, and pathological stage. BCR-free survival was analyzed with Kaplan–Meier curves, log-rank tests, and Cox models censored at 24 months.

Results:

Of 844 men, 393 were operated on in 2007–2012 and 451 in 2014–2019. Compared with the pre-mpMRI era, patients in the mpMRI era were older (median 67 vs 64 years) and had larger tumor volume (3.6 vs 2.2 cc) and less favorable Gleason and D’Amico profiles, while pT distribution was similar. Margin status was available for 802 patients; PSM decreased from 25% (97/381) to 15% (65/421) in the mpMRI era (absolute reduction ~10 percentage points). In the overall cohort, surgery in the mpMRI era was associated with lower odds of PSM (adjusted odds ratio 0.51; 95% CI 0.34–0.74; p < 0.001). In pT2 disease, the association was directionally favorable but non-significant (adjusted odds ratio (OR) 0.71; 95% CI 0.42–1.17; p = 0.20), whereas in pT3 tumors the mpMRI era was associated with a marked reduction in PSM (adjusted OR 0.35; 95% CI 0.19–0.63; p < 0.001). Among 807 patients with PSA follow-up, 24-month BCR-free survival was 0.896 (95% CI 0.861–0.923) in 2007–2012 and 0.846 (95% CI 0.801–0.881) in 2014–2019 (log-rank p = 0.0377); the corresponding unadjusted hazard ratio for BCR was 1.55 (95% CI 1.02–2.35), which attenuated and lost significance after adjustment for age and D’Amico risk (adjusted hazard ratio 1.10; p = 0.656).

Conclusion:

In a high-volume center with a stable, experienced robotic surgeon, the preoperative mpMRI era has been linked to better surgical outcomes, particularly for patients with more pT3 disease. While this temporal association supports a potential contribution of systematic preoperative mpMRI to surgical precision, residual temporal confounding cannot be excluded. Early BCR-free survival within 24 months did not differ significantly between eras after adjustment for baseline risk. In the future, more precise 3D MRI mapping could further enhance the preservation of neurovascular bundles without compromising the oncological prognosis.

Trial registration:

Not applicable (retrospective analysis of routinely collected clinical data).

Keywords

mpMRI prostate cancer robot-assisted laparoscopic prostatectomy surgical margin

Introduction

Prostate cancer is among the most frequently diagnosed malignancies in men; globally, it has the second-highest age-standardized incidence rate among male cancers, after lung cancer. In Switzerland, it is the most frequently diagnosed cancer in men.¹ For patients with localized disease and sufficient life expectancy, radical prostatectomy remains a cornerstone of curative treatment. Over the past two decades, robot-assisted laparoscopic prostatectomy (RALP) has progressively become the predominant surgical approach in many high-volume centers.^2,3 By providing high-definition, three-dimensional visualization and enhanced instrument dexterity, RALP facilitates precise dissection around the prostate and adjacent neurovascular structures. These technical advantages aim to achieve oncologic control while preserving urinary continence and sexual function. In daily practice, however, these oncological and functional goals often pull in opposite directions during preoperative planning and intraoperative decision-making. In men with a high risk of extracapsular extension (ECE), a wider resection that compromises structures crucial for continence and erectile function, such as the neurovascular bundles or urethral length, may be justified to secure negative margins. In contrast, in patients whose disease is confined to the gland, a more tailored dissection can be pursued to spare these structures. A refined understanding of tumor location and aggressiveness is, therefore, essential to guide the extent of resection, allowing surgeons to balance oncologic safety with functional preservation when performing RALP in contemporary practice.

Several clinical parameters, including serum prostate-specific antigen (PSA) level, Gleason score, digital rectal examination findings, and the proportion of positive biopsy cores, are routinely used to predict ECE and lymph node involvement, and thereby support preoperative risk stratification and surgical planning.⁴ When this initial staging is suboptimal, surgeons may either perform overly conservative resections that result in positive surgical margins (PSMs) or unnecessarily wide excisions that compromise functional outcomes. There is therefore a clear need for more precise preoperative evaluation to guide surgical planning, refine the extent of nerve-sparing, and ultimately improve the balance between oncological control and preservation of quality of life.

Multiparametric magnetic resonance imaging (mpMRI) has emerged as a major tool in the diagnostic and staging work-up for prostate cancer.⁵ Multiple studies have shown that mpMRI improves lesion detection and localization, better estimates tumor size, and enhances the assessment of ECE and seminal vesicle invasion compared with conventional imaging.^6
–8 Incorporation of mpMRI findings into preoperative planning has been associated, in various series, with modifications in the surgical strategy, such as adjustment of the nerve-sparing technique or wider resection in areas suspected of capsular breach.^9
–12

However, the impact of preoperative mpMRI on hard surgical outcomes, particularly positive surgical margin rates after radical prostatectomy, remains incompletely defined. In this context, a mature robotic program with standardized imaging and pathology offers a unique setting to reassess the impact of systematic preoperative mpMRI on margin status. We aimed to evaluate whether the introduction of systematic preoperative mpMRI was associated with a reduction in PSM after RALP, overall and stratified by pathological stage (pT2 vs pT3), and to explore its association with early biochemical outcomes.

Material and methods

Study design and setting

This single-center retrospective cohort study included consecutive patients undergoing RALP for biopsy-proven prostate adenocarcinoma at Clinique Générale Beaulieu (Geneva, Switzerland) between January 2007 and December 2019.

Study population

Eligible patients had clinically localized disease at preoperative staging (stages cT1–cT2 N0 M0 according to the TNM classification). Exclusion criteria were metastatic disease at baseline, salvage prostatectomy after prior definitive local therapy, neoadjuvant systemic therapy prior to surgery, and missing information on the exposure variable or surgery date.

Exposure: Implementation of systematic preoperative mpMRI

The exposure of interest was the implementation of routine preoperative mpMRI integrated into surgical planning. Patients were categorized according to calendar period using the year of surgery as the index date: The pre-mpMRI group consisted of patients operated between 2007 and 2012, while the post-mpMRI group included those operated between 2014 and 2019. The year 2013 was excluded as a transition year to minimize exposure misclassification, as mpMRI use was being implemented during that period.

Preoperative assessment and risk stratification

Preoperative variables collected included age at surgery, serum PSA levels, biopsy Gleason scores, and clinical tumor stage (cT), as assessed through rectal examination and imaging studies. For descriptive purposes, biopsy Gleason scores were grouped as 3 + 3 or less, 3 + 4, 4 + 3, and ⩾8. D’Amico risk classification was applied, categorizing patients into low risk (PSA < 10 ng/mL and Gleason score ⩽ 6), intermediate risk (PSA between 10 and 20 ng/mL or Gleason score of 7), and high risk (PSA > 20 ng/mL or Gleason score ⩾ 8).

MpMRI acquisition and reporting

From 2014 onward, mpMRI findings were systematically integrated into both the clinical assessment and the surgical plan for all patients in the mpMRI era. For each case, the operating surgeon reviewed the mpMRI report and images with the dedicated uroradiologist, focusing on lesion location, size, PI-RADS score, and radiological signs of extracapsular extension or seminal vesicle invasion. Tumor foci were mapped to specific prostate regions, and the probability and laterality of capsular breach were discussed before surgery. On this basis, side-specific nerve-sparing was tailored: in regions with low suspicion of extracapsular extension, interfascial or intrafascial dissection was favored to maximize functional preservation, whereas in areas with suspected capsular breach or long capsular contact, wider extrafascial resection or deliberate non–nerve-sparing was planned. mpMRI findings also contributed, together with clinical risk stratification, to decisions regarding pelvic lymph node dissection in higher-risk cases. Thus, mpMRI acted as a structured roadmap for surgery, rather than a purely diagnostic test, and was used consistently to individualize the surgical template on a segmental, side-specific basis.

Surgical technique and operator characteristics

RALP was performed using the da Vinci Surgical System (Intuitive Surgical, Sunnyvale, CA, USA). A three-arm system was used early in the program and was updated to a four-arm higher-definition system in January 2007, before the inclusion period considered in the present analysis, thereby limiting the potential impact of subsequent technological changes on surgical margins. All procedures were performed by a single high-volume surgeon within an established robotic program. The operating surgeon had extensive prior experience with laparoscopic and robot-assisted radical prostatectomy before the study period, thereby limiting the influence of an early learning curve on outcomes. Although minor refinements in technique inevitably occurred over more than a decade, no major change in the basic surgical template (apical dissection, bladder neck handling, and nerve-sparing planes) was implemented during the study period

Pathological assessment

Pathological evaluation of the prostate specimens was conducted following the Stanford protocol,¹³ with formalin fixation and paraffin embedding, and all slides were reviewed by the same dedicated genitourinary pathologist. The pathologic data collected included tumor stage (pT), pathologic Gleason score, presence of PSMs, and lymph node invasion (pN). PSMs were defined as cancer glands observed at the inked surface of prostate specimens and were further classified based on location (apical, posterolateral, basal, or bladder neck) and extent (focal ⩽3 mm or extensive > 3 mm). Prostatectomy tumor volume (cc) was recorded when available.

Follow-up and biochemical recurrence

Postoperative follow-up included regular PSA monitoring. For patients with undetectable PSA (<0.1 ng/mL) at 6 weeks postoperatively, PSA levels were measured every 6 months during the first 2 years, and annually thereafter. Biochemical recurrence (BCR) was defined by a PSA level > 0.2 ng/mL. Patients were contacted between 2015 and 2020 to update their records and PSA levels. Follow-up time for time-to-event analyses was calculated from the date of surgery to the date of BCR, last PSA measurement, death, or end of study period (31/12/2020), whichever occurred first. Patients without follow-up during the record update period and without confirmed death were considered lost to follow-up. Patients without any postoperative PSA follow-up were excluded from time-to-event analyses.

Missing data

Missing values were retained in the cohort. For descriptive analyses, the number of missing observations is reported for each variable; percentages are calculated using available (non-missing) data as the denominator unless otherwise stated. For regression analyses, a complete-case approach was used: patients with missing outcome, exposure, or covariates included in a given model were excluded from that specific model. The analysis sample size (N) is reported for each model.

Pathological margin status was routinely assessed; however, patients with missing margin status were retained in the cohort and excluded only from analyses involving the PSM endpoint.

Outcomes and statistical analysis

The primary outcome was any PSM (positive vs negative), and the secondary outcome was time to BCR. Baseline characteristics were summarized using mean (SD) or median (IQR) for continuous variables and counts (%) for categorical variables. Between-group comparisons used chi-square tests for categorical variables and t-tests for continuous variables.

The association between mpMRI implementation period and PSM was assessed using logistic regression, reporting odds ratios (ORs) with 95% confidence intervals. Multivariate models were subsequently adjusted for potential confounders, including age at diagnosis, PSA level, tumor volume, biopsy Gleason score, D’Amico risk group, and pathologic stage (pT). The same analysis was repeated for data stratified by pathologic stage (pT2, pT3, and pT2 + pT3). Statistical significance was set at p < 0.05.

BCR-free survival was estimated using Kaplan–Meier methods and compared between periods using the log-rank test. Cox proportional hazards regression was used to estimate hazard ratios (HRs) for BCR associated with period, first unadjusted and then adjusted for age at surgery and D’Amico risk group. Because follow-up duration differed substantially between periods, the primary time-to-event comparisons (Kaplan–Meier/log-rank and Cox models) were performed with administrative censoring at 24 months.

Sample size considerations

No formal a priori sample size calculation was performed, as this retrospective study included all consecutive eligible patients who underwent RALP at our institution between 2007 and 2019. The sample size was therefore determined by the available cohort rather than by a predefined target. With 802 patients contributing to surgical margin analyses (162 PSM events, 20%), the study met commonly used empirical criteria for multivariable modeling, including more than 10 events per predictor parameter in the adjusted logistic regression models. In addition, a post-hoc calculation indicated that, for the primary endpoint, the achieved power to detect the observed 10-percentage-point absolute difference in PSM rates between eras (25% vs 15%) was approximately 95% with a two-sided α of 0.05.

The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement (Supplemental Material).¹⁴

Results

Study cohort and baseline characteristics

A total of 844 men underwent RALP during the study period (2007–2012: n = 393; 2014–2019: n = 451). Baseline characteristics by calendar period are shown in Table 1. Compared with the pre-mpMRI period, men operated in the mpMRI period were older (median 67 vs 64 years, p < 0.001) and presented with larger tumor volumes among those with available data (median 3.6 vs 2.2 cc, p < 0.001; tumor volume missing for 105 patients overall). The distribution of biopsy Gleason categories shifted toward higher-risk patterns in the mpMRI period (p < 0.001), and D’Amico risk groups were less favorable with fewer low-risk and more intermediate/high-risk cancers (p < 0.001). Pathological stage distribution was broadly similar between periods (p = 0.3).

Table 1.

Cohort characteristics by period of diagnosis: 2007–2012 versus 2014–2019.

Characteristic	Overall (N = 844)	2007–2012 (N = 393)	2014–2019 (N = 451)	p-Value*
Age at diagnosis	66 (60, 70)	64 (59, 68)	67 (62, 71)	<0.001
PSA level (ng/mL)	7.0 (5.0, 9.8)	6.9 (5.0, 9.4)	7.0 (5.0, 10.4)	0.3
Missing	13	8	5	—
Gleason score				<0.001
3 + 3	369 (45%)	237 (61%)	132 (30%)	—
3 + 4	229 (28%)	74 (19%)	155 (35%)	—
4 + 3	139 (17%)	41 (11%)	98 (22%)	—
More	90 (11%)	37 (9.5%)	53 (12%)	—
Missing	17	4	13	—
D’Amico risk group				<0.001
Low risk	294 (35%)	192 (50%)	102 (23%)	—
Intermediate risk	415 (50%)	145 (38%)	270 (61%)	—
High risk	120 (14%)	49 (13%)	71 (16%)	—
Missing	15	7	8	—
Pathologic stage (pT)				0.3
pT2a	73 (8.8%)	37 (9.5%)	36 (8.2%)	—
pT2b	21 (2.5%)	12 (3.1%)	9 (2.1%)	—
pT2c	483 (58%)	233 (60%)	250 (57%)	—
pT3a	138 (17%)	63 (16%)	75 (17%)	—
pT3b	110 (13%)	42 (11%)	68 (16%)	—
pT4	1 (0.1%)	1 (0.3%)	0 (0%)	—
Missing	18	5	13	—
Tumor volume (cc)	3.0 (1.3, 5.7)	2.2 (1.0, 4.2)	3.6 (1.6, 6.1)	<0.001
Unknown	105	64	41	—
Surgical margins				<0.001
Negative	640 (80%)	284 (75%)	356 (85%)	—
Positive	162 (20%)	97 (25%)	65 (15%)	—
Missing	42	12	30	—

Missing data are excluded from the proportion and statistical tests.

PSA, prostate-specific antigen.

Positive surgical margins (primary endpoint)

Margin status was available for 802/844 patients (95.0%) and was used for analyses involving this endpoint. The proportion of PSMs was lower in the mpMRI period than in the pre-mpMRI period (65/421 (15%) vs 97/381 (25%), p < 0.001), corresponding to an absolute reduction of ~10 percentage points (Table 1).

In univariate logistic regression, surgery in the mpMRI period was associated with lower odds of PSM (OR 0.53; 95% CI 0.38–0.76; p < 0.001). This association remained statistically significant in the adjusted model (adjusted OR 0.51; 95% CI 0.34–0.74; p < 0.001) (Table 2), indicating improved margin status in the mpMRI period despite an older and higher-risk case-mix. As this was a period-based comparison, residual confounding related to temporal changes in practice cannot be fully excluded.

Table 2.

Multivariate logistic regression analysis for surgical margin status: entire cohort, pT2, and pT3 subgroups.

Covariables	All cohort OR (95% CI) N = 787	p-Value	pT2 OR (95% CI) N = 550	p-Value	pT3 OR (95% CI) N = 234	p-Value
Year of diagnosis
2007–2012 (reference)	1.00	—	1.00	—	1.00	—
2014–2019	0.51 (0.34, 0.74)	<0.001	0.71 (0.42, 1.17)	0.2	0.35 (0.19, 0.63)	<0.001
Age at diagnosis	0.99 (0.96, 1.01)	0.3	0.98 (0.94, 1.01)	0.2	0.98 (0.93, 1.02)	0.3
D’Amico risk group
Low risk (reference)	1.00	—	1.00	—	1.00	—
Intermediate risk	1.29 (0.85, 1.96)	0.2	1.01 (0.59, 1.71)	>0.9	0.93 (0.42, 2.10)	0.9
High risk	2.59 (1.52, 4.41)	<0.001	1.39 (0.58, 3.07)	0.4	1.87 (0.77, 4.64)	0.2

OR, odds ratio.

Stratified analysis by pathological stage

Among patients with pT2 disease, there was a nonsignificant trend toward fewer PSMs in the mpMRI period (unadjusted OR 0.66; 95% CI 0.41–1.05; p = 0.083), which remained nonsignificant after adjustment (adjusted OR 0.71; 95% CI 0.42–1.17; p = 0.20).

In contrast, among patients with pT3 disease, the mpMRI period was associated with a marked reduction in PSM: unadjusted OR 0.32 (95% CI 0.18–0.56; p < 0.001) and adjusted OR 0.35 (95% CI 0.19–0.63; p < 0.001), suggesting that the benefit in margin status was primarily observed in men with extraprostatic disease (Table 2).

Biochemical recurrence

Postoperative PSA follow-up data were available for 807/844 patients, and these patients were included in time-to-event analyses (Figure 1). Follow-up duration differed substantially between periods (median time under observation: 76.5 months in 2007–2012 vs 19.7 months in 2014–2019). To limit bias due to differential follow-up, analyses were restricted to the first 24 months after surgery (administrative censoring at 24 months).

Figure 1.

Biochemical recurrence-free survival after radical prostatectomy by surgical period (2007–2012 vs 2014–2019), Kaplan–Meier curves administratively censored at 24 months.

Within 24 months, 91 biochemical recurrences were observed (39 in 2007–2012; 52 in 2014–2019). Kaplan–Meier estimates of BCR-free survival at 24 months were 0.896 (95% CI 0.861–0.923) in 2007–2012 and 0.846 (95% CI 0.801–0.881) in 2014–2019 (Figure 1). The survival curves differed significantly (log-rank χ² = 4.32, p = 0.0377).

In univariable Cox regression, the mpMRI period was associated with a higher hazard of BCR within 24 months (HR 1.55; 95% CI 1.02–2.35; p = 0.039). After adjustment for baseline risk factors, age at surgery, and D’Amico risk groups, the association was attenuated and no longer statistically significant (adjusted HR 1.10; p = 0.656), suggesting that the higher early recurrence in the mpMRI period was largely explained by differences in baseline risk profile rather than the period itself.

Discussion

In this single-center retrospective cohort of 844 men undergoing RALP, the systematic introduction of preoperative mpMRI from 2014 onward was associated with a clinically and statistically significant reduction in positive surgical margins from 25% to 15%, despite an older and less favorable case-mix in the mpMRI era. This effect was most pronounced in pT3 tumors, whereas the impact in pT2 disease was modest and not statistically significant. Because exposure was defined by calendar period, the estimated association reflects an mpMRI-era care pathway rather than mpMRI use at the individual level, and may partly capture concurrent temporal changes in practice.

Most contemporary series evaluating preoperative mpMRI before radical prostatectomy, including those by Rud et al., Joyce et al., and Noor et al., indicate that MRI improves local staging and guides nerve-sparing decisions. However, its impact on “hard” oncologic endpoints such as PSMs or early biochemical recurrence remains heterogeneous, with several reports showing only modest or no benefit.^12,15,16 In our cohort, the systematic use of standardized preoperative 3T mpMRI, read by a single experienced uroradiologist and systematically integrated into the planning of a stable robotic team, was associated with a marked reduction in overall PSM rates, with the effect being most pronounced in pT3 disease despite an older, higher-risk case-mix in the post-MRI era. These observations are broadly in line with studies that have reported lower PSM rates in high-risk or locally advanced tumors when MRI findings are explicitly used to tailor resection margins, as reported by Peng et al.¹¹ and Hansomwong et al.¹⁷ In contrast, they differ from series such as Joyce et al.,¹² in which MRI improved staging but did not translate into clear differences in PSM or early oncologic outcomes. Several factors may explain these discrepancies, including differences in case-mix, the systematic versus selective use of mpMRI, variability in MRI protocols and radiological expertise, the consistency with which imaging is translated into side-specific surgical planning, and heterogeneity in surgical experience, team workflows, and perioperative pathways across centers. First, our population included a higher proportion of intermediate- and high-risk cancers, a clinical scenario in which more accurate assessment of extracapsular extension may have a greater impact on margin control than in predominantly low-risk cohorts. A second factor may be the imaging strategy: in our center, MRI was systematic, protocolized, and interpreted by the same radiologist, whereas in other studies, MRI has often been selective, performed across multiple centers, or read by radiologists with variable expertise, potentially diluting its surgical value. Third, all procedures in our series were performed within a mature robotic program led by a high-volume surgeon, so that MRI-driven adjustments (targeted wider excision, asymmetric nerve sparing) could be consistently implemented, while in mono- or multicenter series with teams with heterogeneous experience and implementation of mpMRI-guided planning, the incremental benefit of MRI may be less apparent.^18,19 Finally, differences in endpoint definitions (e.g., handling of focal vs extensive PSM or varying PSA thresholds and follow-up durations for defining biochemical recurrence) may further contribute to the variability among published results. The case-mix in the MRI era suggests that the observed improvements in surgical precision and margin control in our study are unlikely to be driven by the selection of “easier” cases, but rather by the combination of standardized imaging and a mature robotic program. The observed improvement in PSM rates over time is unlikely to be explained by mpMRI alone. In parallel, gradual advances in robotic systems, perioperative management, and pathological processing could also have influenced outcomes. While these elements enhance the real-world relevance of our cohort, they inevitably act as unmeasured confounders in this before-and-after comparison. Our multivariable models accounted for key clinical and pathological factors, but residual confounding by temporal improvements in surgical care cannot be fully excluded. Rather than proving, our initial hypothesis is that standardized preoperative mpMRI, interpreted by an experienced radiologist and integrated into the planning of a stable robotic team, could improve oncological precision at the time of prostatectomy.

Baseline PSM rates in pT2 disease are already relatively low in experienced centers in the range of 10%–20%.²⁰ In our cohort, the effect of preoperative mpMRI in this subgroup was more modest and not statistically significant (OR 0.71; 95% CI 0.42–1.17). When the tumor is truly pT2, there is usually no macroscopic extracapsular component for mpMRI to visualize, so preoperative imaging mainly refines intraprostatic localization rather than redefining the external dissection plane. This pattern has been described in other cohorts, where preoperative mpMRI produced either small or no differences in PSM among men with organ-confined disease or low–intermediate-risk profiles. Joyce et al. and Druskin et al. found that mpMRI improved staging and lesion localization, but did not meaningfully lower PSM rates overall or within lower-stage subgroups, suggesting that mpMRI alone may not substantially change the margin.^12,21 Similarly, our findings in pT2 disease are in line with those of a large retrospective series including predominantly organ-confined disease, in which Gietelink et al. reported no reduction in PSM after the introduction of mpMRI in a high-volume RARP program.²² Also, the dissection in pT2 cases often follows the same anatomic envelope regardless of MRI, and margins depend more on surgical technique, apical anatomy, and individual variations in capsular thickness than on preoperative imaging. Our finding of a nonsignificant, directionally favorable odds ratio in pT2 disease, therefore, suggests that mpMRI may still be useful to optimize side-specific nerve-sparing and to avoid unnecessarily wide resections, but that its capacity to further lower already low PSM rates in organ-confined tumors appears limited.

The apparent benefit of preoperative MRI was most evident in pT3 tumors. In this subgroup, the odds of positive margins in the MRI era were reduced by approximately two-thirds (OR 0.35; 95% CI 0.19–0.63). In the literature, the impact of preoperative mpMRI on PSM appears most pronounced in high-risk or locally advanced disease, where improved staging of extraprostatic extension can guide wider resection or selective nerve-sparing. Series focusing on high-risk cohorts have suggested that mpMRI-informed planning may reduce margin positivity: Kukreja et al. reported that mpMRI in high-risk prostate cancer led to substantial modifications of surgical strategy with a potential decrease in PSM compared with historical controls,⁹ while Humke et al. reported that, in high-risk patients, mpMRI achieved reasonable accuracy for detecting ⩾pT3 disease and, when combined with intraoperative frozen section, allowed nerve-sparing in most men with a very low PSM rate at the level of secondary resections.²³ Hansomwong et al.¹⁷ observed a modest reduction in global and apical margins with preoperative MRI, but without a clear focus on pT3-specific benefit. These findings suggest that the added value of MRI is greatest when extracapsular extension or seminal vesicle involvement is likely, providing the surgeon with more precise information to guide the balance between adequate resection and preservation of the neurovascular structures. The oncologic benefit of preoperative mpMRI seems concentrated in higher-stage disease, refining previous largely neutral data from unselected cohorts and reinforcing the rationale for systematic, expert-interpreted mpMRI in patients at risk of pT3 tumors, even in a cohort with progressively less favorable disease characteristics. Nonetheless, our findings should be interpreted as a hypothesis-generating, indicating that mpMRI-guided planning may particularly benefit pT3 cancers, rather than as definitive proof of a causal effect.

Regarding biochemical outcomes, interpretation requires particular caution. We used a pragmatic definition of biochemical recurrence as the first postoperative PSA ⩾0.2 ng/mL, which may be more sensitive but potentially less specific than definitions requiring confirmation. Differences in PSA surveillance intensity between eras could also influence the detection of early events. Follow-up duration differed substantially between periods; therefore, time-to-event comparisons were primarily restricted to the first 24 months after surgery via administrative censoring. Within this window, BCR-free survival differed between periods in unadjusted analyses (log-rank p ≈ 0.04; univariable Cox HR > 1), but this association was attenuated after adjustment for baseline risk, suggesting that differences in tumor risk profile rather than the period itself largely explained early recurrence patterns. While margin status is an important surrogate of surgical precision, early biochemical recurrence also reflects tumor biology and baseline risk, which differed substantially between eras.

Beyond traditional clinicopathological predictors, several groups have recently proposed mpMRI-based tools to estimate the risk of PSMs before surgery. An mpMRI-based nomogram combining clinical and imaging parameters has shown moderate but clinically relevant ability to predict PSMs and may help refine preoperative planning and patient counseling. In parallel, an mpMRI-based grading system for preoperative PSM risk stratification has also been developed, further supporting the role of structured mpMRI assessment in guiding the extent of resection.²⁴

More broadly, systematic reviews and meta-analyses underline that, while mpMRI can substantially influence the surgical template and nerve-sparing decisions, its capacity to consistently reduce PSM rates across all risk groups remains heterogeneous and not fully established. In this context, the present study adds real-world evidence from a mature robotic program in which standardized mpMRI interpretation and its integration into side-specific surgical planning were associated with a meaningful reduction in overall and pT3 PSMs.²⁵

This study is not without its limitations, but it also presents several distinctive features that strengthen the interpretation of its findings. We included a relatively large cohort treated in a high-volume center. All procedures were performed by a single experienced surgeon using a stable technique, and a homogeneous, prospectively implemented mpMRI protocol interpreted by a dedicated uroradiologist was applied throughout the study. These elements reduce inter-operator variability and make the link between imaging strategy and surgical outcome more internally consistent. However, our analysis remains retrospective and monocenter, with a non-randomized design that exposes it to residual confounding and limits the generalizability of the results beyond similar practice settings. Because the pathologic stage is determined postoperatively and may lie on the causal pathway or reflect stage migration, models adjusting for pT should be interpreted as exploratory. A key limitation of our study is the inherent temporal bias related to comparing two calendar periods. Improvements in robotic technology, perioperative care, and the surgeon’s cumulative experience over time could have contributed to the observed reduction in PSM independently of mpMRI. In our center, the robotic program was already mature at the beginning of the inclusion period, and all procedures were performed by a single experienced surgeon with a stable technique, which likely mitigates, but does not eliminate, the influence of a learning curve. Nevertheless, the non-randomized, before-and-after design precludes firm causal inference, and the observed association between the mpMRI era and lower PSM rates should be interpreted as hypothesis-generating rather than definitive proof of a direct effect of mpMRI. Follow-up is necessarily shorter, and statistical power is more limited for the most recent mpMRI-era patients, especially for late biochemical recurrence. Differential follow-up between eras is a key methodological constraint for biochemical endpoints and motivated the 24-month restriction, which limits inference on long-term recurrence. Pathological assessment, although standardized locally, was not centrally reviewed, which may contribute to inter-study variability in margin rates and staging. Missing data were handled using complete-case analyses for regression models, which may introduce bias if missingness is not random; however, missingness was limited for key endpoints/exposure. These strengths and limitations must, therefore, be carefully considered when interpreting the apparent impact of systematic preoperative mpMRI on surgical margins and oncological outcomes in our cohort. These results justify future prospective multicenter studies evaluating standardized mpMRI-guided planning algorithms on margins and recurrence across pathological stages, the integration of advanced tools such as 3D mpMRI mapping or radiomics-based risk scores to better define high-risk segments, and dedicated work on functional outcomes (continence and erectile function) to determine whether mpMRI-driven adjustments truly optimize the trade-off between cancer control and quality of life.

Conclusion

In this single-center retrospective cohort of men undergoing RALP, the introduction of systematic preoperative mpMRI was associated with a reduction in overall positive surgical margins from 25% to 15%. The apparent benefit was most pronounced in pT3 disease, whereas early biochemical recurrence-free survival, once adjusted for tumor risk, did not show a clear difference between eras. However, the non-randomized, longitudinal design and potential unmeasured temporal confounders, including subtle improvements in technique and technology, mean that our results should be viewed as hypothesis-generating. Further prospective, ideally multicenter studies are needed to identify which mpMRI features most reliably guide nerve-sparing and wider excision strategies, and to determine their long-term impact on both oncological control and functional outcomes.

Supplemental Material

sj-doc-1-tau-10.1177_17562872261452717 – Supplemental material for Impact of preoperative MRI on surgical outcomes of robot-assisted laparoscopic prostatectomy: a single-center retrospective cohort of 844 patients

Supplemental material, sj-doc-1-tau-10.1177_17562872261452717 for Impact of preoperative MRI on surgical outcomes of robot-assisted laparoscopic prostatectomy: a single-center retrospective cohort of 844 patients by Charles Henry Rochat, Robin Schaffar, Jacques Martin Randriantsalama, Ildiko Quinodoz Szalay, Georges Antoine de Boccard and Martina Martins Favre in Therapeutic Advances in Urology

Footnotes

Acknowledgements

The authors thank the nursing and operating room staff at Clinique Générale Beaulieu for their dedicated care of the patients included in this study.

Declarations

ORCID iD

Jacques Martin Randriantsalama

Supplemental material

Supplemental material for this article is available online.

References

Bray

Laversanne

Sung

, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024; 74(3): 229–263.

Lowrance

Eastham

Savage

, et al. Contemporary open and robotic radical prostatectomy practice patterns among urologists in the United States. J Urol 2012; 187(6): 2087–2093.

Rochat

Sauvain

Dubernard

, et al. Mid-term biochemical recurrence-free outcomes following robotic versus laparoscopic radical prostatectomy. J Robot Surg 2011; 5(4): 251–257.

Campbell

Raymond

O’Callaghan

, et al. Optimum tools for predicting clinical outcomes in prostate cancer patients undergoing radical prostatectomy: a systematic review of prognostic accuracy and validity. Clin Genitourin Cancer 2017; 15(5): e827–e834.

Triquell

Foulon

Salleron

, et al. Multiparametric MRI for staging of prostate cancer. Cancers (Basel) 2022; 14(16): 3966.

Zhang

Yang

Fan

, et al. Accuracy of multiparametric magnetic resonance imaging in predicting extracapsular extension of prostate cancer: a systematic review and meta-analysis. Br J Radiol 2019; 92(1095): 20181537.

Lin

Johnson

Fennessy

, et al. Prostate cancer local staging with magnetic resonance imaging. Radiol Clin North Am 2024; 62(1): 93–108.

Sitharthan

Kang

Treacy

, et al. The sensitivity and specificity of multiparametric magnetic resonance imaging and prostate-specific membrane antigen positron emission tomography/computed tomography for predicting seminal vesicle invasion in clinically significant prostate cancer: a multicenter retrospective study. J Clin Med 2024; 13(15): 4424.

Baack Kukreja

Bathala

Reichard

, et al. Impact of preoperative prostate magnetic resonance imaging on the surgical management of high-risk prostate cancer. Prostate Cancer Prostatic Dis 2020; 23(1): 172–178.

10.

McEvoy

Raeside

Chaim

, et al. Preoperative prostate MRI: a road map for surgery. AJR Am J Roentgenol 2018; 211(2): 383–391.

11.

Peng

, et al. Effect of preoperative PI-RADS assessment on pathological outcomes in patients who underwent radical prostatectomy. Cancer Imaging 2023; 23(1): 60.

12.

Joyce

Brown

Ghiraldi

, et al. Effect of preoperative multiparametric magnetic resonance imaging on oncologic and functional outcomes following radical prostatectomy. Eur Urol Open Sci 2023; 47: 87–93.

13.

Lecornet

Ahmed

, et al. The accuracy of different biopsy strategies for the detection of clinically important prostate cancer: a computer simulation. J Urol 2012; 188(3): 974–980.

14.

von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol 2008; 61(4): 344–349.

15.

Rud

Baco

Klotz

, et al. Does preoperative magnetic resonance imaging reduce the rate of positive surgical margins at radical prostatectomy in a randomised clinical trial? Eur Urol 2015; 68(3): 487–494.

16.

Noor

Baco

Lauritzen

, et al. The effect of preoperative magnetic resonance imaging on long-term oncological outcomes following radical prostatectomy: a 12-year follow-up of a randomized controlled trial. Eur Urol Open Sci 2025; 79: 79–86.

17.

Hansomwong

Saksirisampant

Isharwal

, et al. Role of preoperative magnetic resonance imaging on the surgical outcomes of radical prostatectomy: does preoperative tumour recognition reduce the positive surgical margin in a specific location? Experience from a Thailand prostate cancer specialized center. Asian J Urol 2023; 10(4): 494–501.

18.

Atug

Castle

Srivastav

, et al. Positive surgical margins in robotic-assisted radical prostatectomy: impact of learning curve on oncologic outcomes. Eur Urol 2006; 49(5): 866–872.

19.

Bravi

Tin

Vertosick

, et al. The impact of surgeon experience on the risk of surgical margins and biochemical recurrence after robot-assisted radical prostatectomy: a learning curve study. J Urol 2019; 202(1): 108–113.

20.

Kim

JW.

Recent advances in radical prostatectomy: a narrative review of surgical innovations and outcomes. Cancers (Basel) 2025; 17(5): 902.

21.

Druskin

Liu

Young

, et al. Prostate MRI prior to radical prostatectomy: effects on nerve sparing and pathological margin status. Res Rep Urol 2017; 9: 55–63.

22.

Gietelink

Jansen

BHE

Oprea-Lager

, et al. Preoperative multiparametric MRI does not lower positive surgical margin rate in a large series of patients undergoing robot-assisted radical prostatectomy. J Robot Surg 2022; 16(2): 273–278.

23.

Humke

Hoeh

Preisser

, et al. Concordance between preoperative mpMRI and pathological stage and its influence on nerve-sparing surgery in patients with high-risk prostate cancer. Curr Oncol 2022; 29(4): 2385–2394.

24.

Meng

Chen

Zhang

, et al. Multiparametric MRI-based nomograms in predicting positive surgical margins of prostate cancer after laparoscopic radical prostatectomy. Front Oncol 2022; 12: 973285.

25.

Wang

Zhu

, et al. The diagnostic performance of tumor stage on MRI for predicting prostate cancer-positive surgical margins: a systematic review and meta-analysis. Diagnostics (Basel) 2023; 13(15): 2497.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.08 MB

0.00 MB