Limitations of Assessing Value in Robotic Surgery for Prostate Cancer: What Data Should Patients and Physicians Use to Make the Best Decision?

摘要

The rise of robotic surgery for prostate cancer has been swift. The da Vinci surgical system (Intuitive Surgical, Sunnyvale, CA) was introduced just over a decade ago. Yet the majority of radical prostatectomies in the United States are already being performed robotically. Even at Johns Hopkins, a center known for devising and popularizing nerve-sparing open radical retropubic prostatectomy, the proportion of robot-assisted radical prostatectomies (RARPs) has increased dramatically, from 17% in 2003% to 57% in 2013. It has been suggested that the rapid rise of RARP is a result of direct-to-patient marketing and maneuvers by hospitals to increase market share and profits. Another equally, if not more, pertinent reason is that robotic assistance has rendered minimally invasive radical prostatectomy (MIRP) technically feasible for many surgeons, whereas pure laparoscopic radical prostatectomy without robotic assistance is technically daunting and has a steep learning curve. This suggests that the da Vinci surgical system has significant merit as an enabling technology. There are several ways to assess the value of a new medical technology, including patient and user satisfaction, complication risk, cost-effectiveness, and oncologic outcomes, especially in cancer treatment. RARP is more costly to perform than open radical prostatectomy (ORP) because of capital investments/maintenance as well as increased surgical supply costs. However, previous studies have failed to show meaningful differences in complications or cancer control between these two surgical techniques. In the article that accompanies this editorial, Gandaglia et al seek to clarify an important clinical and health-policy dilemma in prostate cancer surgery in an analysis of an updated SEER-Medicare cohort. They ask: in the contemporary era, when robotic technology has already become widespread, is robotic prostatectomy worth it? This is a difficult question to answer, and the authors should be commended for their careful study design and thoughtful analysis. In the study, a cohort of 5,915 patients who underwent RARP or ORP for localized prostate cancer were analyzed, and associations between surgical technique with complications and secondary cancer therapies were assessed. Adjustments were made for treatment year, socioeconomic factors, tumor grade and stage, and whether pelvic lymphadenectomy was performed. In their final adjusted analyses, overall complication rates between RARP and ORP were similar. Consistent with other previous reports, the blood transfusion rate was significantly less and median postoperative length of hospital stay was shorter with RARP. However, men who underwent RARP were more likely to incur genitourinary and miscellaneous medical complications with higher total health care expenditures within a year after surgery. Finally, the adjusted risk of receiving secondary cancer therapies was not significantly different after RARP versus ORP. Although the study results are intriguing, measuring complications and oncologic outcomes of radical prostatectomy poses certain challenges. Postoperative complications are closely related to baseline comorbidities, which are difficult to assess. In addition, the assessment of prostate cancer treatment requires long-term follow-up because of the protracted nature of the disease. Given these challenges, there are some notable limitations to the current study. It is obviously not a prospective, randomized trial of RARP versus ORP—although one is currently ongoing in Australia. Despite multivariable analysis and use of an instrument variable to adjust for confounders, there is still a significant potential for selection bias. For example, body mass index was not adjusted for in these analyses, which is noteworthy because body mass index is known to correlate with surgical difficulty/complications and cancer recurrence after radical prostatectomy. Perhaps a more concerning aspect of selection bias was patient age