Warnings sounded on robo-surgery risks

The march of robots into the operating room has hit a speed bump, with authors of an opinion piece in the journal JAMA warning that poor outcomes in robot surgery for cancer mean safeguards must be put in place to protect patients.

Surgical resident Kyle Sheetz and Professor of Surgery Justin Dimick, both from the Department of Surgery at the University of Michigan, Ann Arbor, US, seize on a recent safety communication from the US Food and Drug Administration (FDA) to call for urgent reforms to surgeon credentialing, patient informed consent and insurance coverage for robotic surgery.

In February, the FDA advised that, while robots have been cleared for use in some procedures, “The safety and effectiveness of robotically-assisted surgical devices for use in mastectomy procedures or prevention or treatment of cancer has not been established”.

The warning referred specifically to recent studies suggesting robot and other “minimally invasive” keyhole surgery may lead to worse outcomes for cancer of the cervix.

Robots have become a common sight in the operating room since being first approved by the FDA in 2000; the JAMA authors report in excess of 600,000 robotic surgeries in the US in 2017, a threefold increase in just a decade. It’s a technology used in surgery of the heart, bladder, prostate, bowel and kidney, among others.

The most well-known bot is called the da Vinci system, which looms over the operating table like a mechanical insect. The machine has four arms – one with a camera and three with pincers to hold scalpels, suture needles and other surgical instruments.

The surgeon sits a few metres away, head buried in a screen that delivers a high definition 3D view of the surgical field, hands manipulating exquisitely sensitive controls that generate movements in the robot’s arms.

The claimed benefits of robot surgery include less pain and blood loss, fewer complications such as infection, and smaller scars.

But, at least in some types of surgery, the claims come with a catch.

The FDA notes in its advisory that existing standards for robotic devices focus on complication rates at 30 days post-surgery. That time frame cannot, obviously, evaluate how well robotic surgery fares in longer term outcomes for cancer.

Which is where two recent studies published in the New England Journal of Medicine have delivered some unwelcome news.

The first, lead-authored by Alexander Melamed from the Harvard Medical School, US, looked at women who had their uterus removed because of cancer of the cervix, with either open surgery via an incision through the abdomen, or minimally invasive surgery, 80% performed with robots.

Four years after the operation, the death rate of women who had minimally invasive surgery was nearly double (9.3%) that of the women who had open surgery (5.3%).

The second study, led by Pedro Ramirez at the University of Texas MD Anderson Cancer Centre, randomised women with cervical cancer to have either open or minimally invasive surgery (16% of which were robot-assisted) to remove the womb. Minimally invasive surgery was linked to a nearly four-fold increase in cancer recurrence or death at three years.

According to a report in the New York Times, not only were those results bad enough for the researchers to stop the trial early, but Ramirez ceased doing the minimally invasive operation in October 2017, when the results first became known.

It’s not clear why outcomes are worse with keyhole-robot surgery, but Ramirez and colleagues write that manipulation of the uterus and the use of carbon dioxide to inflate the abdomen may cause cancer cells to spread from the cervix.

The JAMA authors note both these studies and call for wide-ranging action.

They say insurers should mandate that hard evidence of benefit be produced before robot surgery for cancer is funded. They claim that some surgeons start doing robotic surgery after being mentored in just two procedures and call for institutions to establish clear proficiency benchmarks before credentialing doctors.

Finally, they say surgeons must be transparent with patients about the evidence for the effectiveness of robot surgery, and their own experience with it, to ensure properly informed consent.

All this may leave readers wondering how doctors could possibly offer services for which the evidence of effectiveness is missing. The reality is that the FDA does not regulate the “off label” use of devices; that is, the use of authorised devices for unapproved purposes.

The New York Times reported that, “According to a spokeswoman, the FDA does not regulate the practice of medicine”, leaving it up to the discretion of practitioners to decide if a procedure offers benefit.

In the absence of FDA oversight, Sheetz and Dimick make very clear who must take up the regulatory slack.

“Payers, hospitals, and surgeons can take immediate steps to ensure that certain safeguards remain in place until the evidence for or against the use of robotic assisted surgery has time to mature,” they conclude.