乌鸦传媒

ORLANDO -- Three experienced surgeons needed more than 1,600 cases to achieve acceptable outcomes with robotic-assisted laparoscopic prostatectomy (RALP), data from a retrospective chart review showed.

The surgeons did not consistently meet the current standard of less than 10% positive surgical margins until they had completed almost half of all RALP procedures performed during the six-year study period, according to a study that will be presented here at the Genitourinary Cancers Symposium.

Operative time reached a plateau after about 750 cases but started to increase again near the end of the six years.

The results suggest that proficiency in RALP involves a much longer learning curve than previously recognized.

"Robotic prostatectomy is gaining in popularity and being done by a large number of surgeons who thus do a relatively few number of cases," Prasanna Sooriakumaran, MD, of Weill Cornell Medical College in New York City, said at a news briefing.

"While this may be sufficient to achieve safe results, a long learning curve must be overcome in order to optimize results for the patient, in terms of positive margins. We thus recommend that this operation should not be done by all urologists in small community hospitals," Sooriakumaran said.

"It should be concentrated into those high-volume centers of excellence, where the operation can be done by surgeons who do a large number of cases very frequently, in order that they can achieve the best possible cancer-control results for their patients," he said.

Of the estimated 90,000 radical prostatectomies performed in the U.S. each year, robotics is used in more than 70,000 cases. More than 70% of RALP procedures are performed by surgeons who do fewer than 100 cases a year, according to Sooriakumaran.

Studies suggest urologic surgeons require only 25 to 40 cases to master the safety of RALP, he continued. However, the learning curve to achieve expert-level results has not been established.

In an effort to help establish that, he and his colleagues examined medical records of 3,794 men who underwent the robotic-assisted procedures from 2003 to 2009 at three institutions -- the University of Pennsylvania in Philadelphia, the Karolinska Institute in Stockholm, and Weill Cornell. One surgeon at each institution performed all of the procedures.

The principal outcomes of interest were the rate of positive surgical margins and operative time. Surgical performance was evaluated after every 50 RALP procedures.

For all cases, positive-margin rate started at about 20% and then gradually declined to less than 10% after 1,600 total cases.

At the end of the six-year review period, rate remained in a downward trajectory, suggesting continuation of the learning curve, said Sooriakumaran.

For the subgroup of patients with extracapsular disease (pT3), the positive-margin rate began at about 40% and remained stable throughout the study period.

"That was to be expected," said Sooriakumaran. "If the cancer has spread beyond the prostate and only the prostate is removed, then it's quite likely that some cancer cells will be left behind."

Operative time exceeded three hours at the outset, was approaching two hours after 750 cases, where it stabilized for another 700 or so cases before starting to rise again at the end of the study period.

Expertise in chemotherapy and radiation therapy for prostate cancer also requires a learning curve, said Nicholas Vogelzang, MD, who moderated the news conference. However, he acknowledged the rapid growth rate of robotic procedures.

"I think this data will make everyone pause and say, 'maybe it's time to go a little bit farther before we rush into this surgery,'" said Vogelzang, of Comprehensive Cancers Centers of Nevada in Las Vegas.

"I would remind you that surgeons who perform open prostatectomy will sometimes have done 8,000 to 10,000 cases and have also been shown to have a similar learning curve," Vogelzang added. "Experience really does matter."

The take-home message from the study, according to Sooriakumaran is that RALP is "indeed a simple procedure, in terms of the safety aspect. In terms of achieving the best possible results, it is not a simple procedure, and it takes a lot of skill and experience."

The lack of "tactile feedback" with robotics explains much of the learning curve, Sooriakumaran continued. With conventional laparoscopic or open prostatectomy, the surgeon can feel the tissue and better assess the adequacy of resection, he explained.

Comment