Randomized Clinical Trial Comparing On-clamp Versus Off-clamp Laparoscopic Partial Nephrectomy for Small Renal Masses (CLOCK II Laparoscopic Study): A Intention-to-treat Analysis of Perioperative Outcomes

Take Home Message In the setting of a randomized controlled trial, no differences were found in the perioperative and early functional outcomes between on- and off-clamp laparoscopic partial nephrectomy.


Introduction
The current guidelines recommend partial nephrectomy (PN) whenever feasible as a surgical treatment option for T1 renal masses suspicious for cancer [1]. Together with the complete removal of the mass (the ''leave no tumor behind'' concept [2]), PN has gained broader diffusion over radical nephrectomy with the ideal goal of preserving more renal function [3]. Several factors interplay in the residual renal function after PN [4]: patient-related variables, including age, comorbidities, and baseline kidney function, represent unmodifiable factors, while surgery-related variables, such as the technique for managing the renal pedicle, resecting the tumor, and suturing the renal defect, are modifiable factors [5].
Since the beginning of the ''PN saga,'' the ischemic injury deriving from long ischemia intervals during the clamping of renal artery has been named among the more detrimental factors responsible for the postoperative renal function decrease [6,7]. As such, it has been hypothesized that choosing an off-clamp approach might better preserve the ultimate renal function. If during open PN an ''off-clamp'' technique is often performed, given the direct handassisted control over the renal pedicle allowed to the surgeon [8], during laparoscopic PN (LPN), the clamping of renal artery has been considered crucial in order to obtain a bloodless field during the tumor excision. Challenges related to the pure laparoscopic approach increase the risk of longer ischemia times, particularly in less experienced hands. This is why the clamping of the main renal artery has been challenged by several ''alternative'' techniques aimed to shorten/avoid renal ischemia. The exasperation of the concept has led to the ''no ischemia'' approach first described by Guillonneau et al. [9] and duplicated by other authors [10,11]. Off-clamp LPN is a technically demanding approach, requiring consistent experience before embarking on it. Some authors questioned off-clamp LPN as potentially negatively impacting the perioperative outcomes of the surgery, increasing blood loss, worsening the vision of the operative field, and sponsoring the likelihood of complications and positive surgical margins.
Actually, a recent systematic literature review including pure LPN showed no impact on either surgical or oncological outcomes by the off-clamp technique [12], but the finding is debatable and based on the pooled analysis of mostly small sample size studies, affected by several confounders, including a selection bias (the off-clamp approach is more likely to be performed in low-complexity renal masses by more expert surgeons).
Our research group sought to contribute to the field with the aim of increasing the level of evidence about the topic. A randomized clinical trial (RCT), the ''CLOCK II'' study (CLamp vs Off Clamp of the Kidney during laparoscopic partial nephrectomy; ClinicalTrials.gov NCT 02287987) was conceived. The purpose of the study was to compare the perioperative outcomes of on-versus off-clamp LPN.

Study protocol
The CLOCK II laparoscopic study is a multicentric RCT, promoted by the

Randomization
The random sequence for the two comparison groups (on-vs off-clamp) was computer generated using the command ralloc in Stata 15. Randomized allocation with a 1:1 allocation ratio was assigned by a permuted block design, stratified by center [13,15]. Randomization was also stratified according to the complexity of the tumor based on the RENAL score (<7 vs !7). The allocation arm was notified by means of a dedicated e-form, managed by an independent software house. At any moment, the investigators were able to amend the indication given by the randomization and shift the patients to the alternative option for managing the renal artery.

Outcome measurements
Patients' demographics, baseline patient and tumor variables, and periand postoperative data were collected. Specifically, patients' comorbidities were classified according to the Charlson's index [16]. The resection technique was classified by the surface-intermediate-base score [17].
As stated, the surgical strategy for managing the renal pedicle was dictated by the study protocol. Perirenal fat and renal artery dissection were mandatory surgical steps in all patients.
Neither preoperative transarterial embolization nor intraoperative controlled hypotension was allowed. When randomized in the onclamp treatment group, tumor resection and inner renorrhaphy layer had to be done with the renal artery clamped. On the contrary, in the off-clamp arm, the renal artery had to remain unclamped for the whole procedure duration.
Since the start till the conclusion of the tumor resection/parenchymal renorrhaphy, the investigators were able to waive at any moment the indication given by the randomization and shift to the alternative renal artery management option. Details and reasons for deciding for a shift had to be reported mandatorily. As concerning the perioperative outcomes, the severity of the eventual intraoperative bleeding was categorized on a scale from 0 (=no bleeding) to 5 (=intense bleeding exceeding the suction cannula capacity). Postoperative complications were classified according to the Clavien-Dindo system [18].

Statistical analysis
The target sample size has been calculated using the Borm, Fransen, and Lemmens formula for analysis of covariance (ANCOVA). As reported in the published study protocol [15], we assumed from previous reports a standard deviation r of 20 ml/min/1.73 m 2 . For a = 5%, 1 À b = 80%, a q 2 of 0.6, and a clinically significant minimum difference d = 5 ml/ min/1.73 m 2 , the minimum required sample size was calculated to be 102 patients per arm. Actually, 113 patients per arm were considered the target enrolment after adjusting for a 10% chance of dropping out.
Reporting the intention-to-treat analysis of the patients' demographics, baseline characteristics, and perioperative outcomes is the purpose of the study.
Categorical variables were summarized as absolute and relative frequencies. Numerical variables were reported as means and standard deviations or medians and interquartile ranges (IQRs), as appropriate.

Results
As of September 2019, the CLOCK II laparoscopic study recruited 249 patients. Of them, 123 and 126 were randomized and allocated to the on-and off-clamp treatment groups, respectively. The number of enrolled patients exceeded the planned sample size [13]. At baseline, the on-and off-clamp treatment groups were well balanced as concerning the patients' demographics, coagulative function, comorbidities, renal function, and clinical tumor size and complexity. Complete baseline data are reported in Table 1.  Table 2.
Multivariable model analyses were performed to account for eventual unbalances. Each +1 increase in the RENAL score was found to increase protection from the occurrence of complications (OR 0.72, 95% confidence interval [CI] 0.54-0.97, p = 0.034). Each +1 cm increase in tumor size corresponded to an increase in the risk of blood transfusion (OR 1.39, 95% CI 1.14-1.70, p = 0.001). Retroperitoneal access protected from the occurrence of blood transfusions (OR 0.07, 95% CI 0.01-0.29, p < 0.001).
No specific impact of resection and suture techniques was found on the surgical perioperative outcomes. Finally, the higher the preoperative patient's eGFR, the higher the probability of acute kidney injury (considered as a decrease in eGFR of >25%) measured on the 1st and 5th postoperative days (OR 1.03, 95% CI 1.01-1.06, p = 0.002).

Discussion
To the best of our knowledge, this is the first randomized controlled trial comparing the off-versus on-clamp approach during pure LPN. Several reports aimed to investi-gate the differences between the two techniques for managing the renal pedicle during the tumor resection in nephronsparing surgery [12]. Those investigations were hindered by retrospective study design and imprecision of the outcome measurements. Systematic literature reviews pooled the limited evidence available and concluded that there is no impact of the technique used to manage the renal pedicle on either surgical or oncological outcomes [12,19]. Only two recently published studies sought to raise the bar in the available level of evidence on this topic. Anderson and coworkers [20] performed a prospective randomized trial comparing the on-versus off-clamp technique in the setting of robotic PN. Thirty-seven versus 34 patients were analyzed. Unfortunately, the significance was not achieved in the study since the observed difference in eGFR was much less than what the authors hypothesized a priori. Our group underlined one of the issues of the study-inability to detect any variation below 10% of eGFR due to statistical matters [21]. As such, no difference in the eGFR change was reported between the groups, even if off-clamp patients had larger variability in eGFR that remained undetected. In summary, beyond the comparable perioperative outcomes, the study once again reported the ''nonfinding'' of no differences in renal functional outcomes irrespective of the clamping approach. Antonelli et al. [22,23] followed the CLOCK I study. The on-and off-clamp approaches during robotic PN were confirmed to have a comparable safety profile [23]. Moreover, in the specific setting of bilateral kidneys with regular baseline function, the authors found no differences in functional outcomes [24].
We believe that our study contributes to further improving the quality of the evidence in this field. Notably, the randomization protocol worked properly, with no differences in all the preoperative baseline variables. Both clinical tumor size and tumor complexity as assessed by the RENAL score were comparable between the groups as well.
Confirming the lower level of evidence about the topic, the approaches were found comparable in the surgical perioperative outcomes, including blood loss, and complication and positive surgical margin rates.
As concerning the more advanced analysis performed to account for potential confounders, multivariable models found increasing RENAL score to be protective against the occurrence of complications (OR 0.72, 95% CI 0.54-0.97, p = 0.034). This could sound in contrast with literature evidence. As such, a recently published literature review confirmed the impact of higher nephrometry scores on the likelihood of having complications after PN [25]. A possible explanation is the average high experience of the surgeons who were involved in the present study. Moreover, the higher the nephrometry score, the higher the probability of being forced in performing a more enucleative technique to score with a nephron-sparing intent. A recent analysis by Minervini et al. [26] within the ''Surface-Intermediate-Base consortium'' found enucleation to be protective against complications if compared with the standard enucleoresection technique. Consistent with previous findings, larger tumor size increased the risk of blood transfusion (OR 1.39, 95% CI 1.14-1.70, p = 0.001) [5].
In our randomized study, retroperitoneal access was found to protect from blood transfusions (OR 0.07, 95% CI 0.01-0.29, p < 0.001), similar to the findings by Pavan et al. [27], who reported lower estimated blood loss for retroperitoneal PN in a recent systematic review and cumulative analysis of perioperative outcomes of transperitoneal versus retroperitoneal PN.
No specific impact of resection and suture techniques was found on the surgical perioperative outcomes. With  the available evidence, it is hard to test the impact of such variables per se, given their interplay with other diseaseand surgery-related factors [4,5,7,26,28].
Regarding the renal functional outcomes, better patient baseline renal function as assessed by eGFR predicted a higher probability of acute kidney injury on either the first or the fifth postoperative day (OR 1.03, 95% CI 1.01-1.06, p = 0.002). This is in line with many reports [29].
Despite the randomized controlled design, the present study is not devoid of limitations.
The most important is the fact that only expert surgeons and referral institutions participated in this study. Nevertheless, despite a rigorous randomization protocol, assigned clamping strategy was affected by subjective feelings during surgery, with the surgeon being able to shift from the offclamp to the on-clamp approach. The likelihood of shifting from pure off-clamp to on-clamp LPN relied on tumor size and complexity. Among the cases assigned to an off-clamp procedure, 41 (32.5%) were intraoperatively converted to on-clamp LPN, but the intraoperative need to convert the planned strategy seemed harmless on the postoperative course [30]. Moreover, as per the study protocol, enrolled patients had baseline eGFR >60 ml/min, a normal contralateral kidney, and a RENAL score of 10. This undoubtedly does not represent the daily practice and limits the generalizability of the findings. Another bias given by unmeasurable surgeon's skills, as well the surgeon's attitude and experience toward a specific clamping approach should be consid-  ered. Moreover, it could be hypothesized that the peculiar setting of a surgical RCT could have prompted some additional care to preserve patient's safety, influencing the results with respect to the daily clinical practice. Last, notwithstanding the adequate population size randomized, a small sample size of patients was analyzed relative to the number of variables collected. As such, our analysis excluded a significant relationship of clamping, resection, and renorrhaphy techniques with the outcomes of interest. However, the influence of these factors on intra-and postoperative events cannot be excluded completely, also considering the previous reports.

Conclusions
In the setting of an RCT comparing on-versus off-clamp LPN, no differences were found in the perioperative and early functional outcomes between the approaches.
Author contributions: Riccardo Bertolo had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Analysis and interpretation of data: Bertolo, Sandri.
Drafting of the manuscript: Bertolo.
Critical revision of the manuscript for important intellectual content: Bove, Antonelli, Minervini.
Administrative, technical, or material support: None.
Financial disclosures: Riccardo Bertolo certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None.
Funding/Support and role of the sponsor: None.