Q&A: Justifying the Cost of PET/MR

Steven Mendelsohn, MD, predicts that the PET/MR will become the benchmark gold standard in the next five to ten years. He calls it the “most revolutionary machine since the advent of the MRI.”

The statement might be strong, but studies continue to prove that PET/MR could be a game changer. But changing the game, so to speak, has been a slow process. The first simultaneous PET/MR system, the Siemens Biograph mMR, received FDA approval in 2011. Since then, Siemens has installed just 12 of these systems in the U.S. There are only two other FDA-approved PET/MRs, the Philips Ingenuity TF PET/MR (approved November 2011) and GE’s SIGNA PET/MR (approved November 2014). GE, to date, has installed nine of their SIGNA PET/MR systems in the U.S.

The problem, a common one with disruptive technology in radiology, is the cost.  The machines are expensive and reimbursement is unreliable.

Mendelsohn’s New York-based Zwanger-Pesiri is one of the few practices with PET/MR. They started scanning their first clinical patients in November 2014 with the Siemens Biograph mMR. Here, Mendelsohn tells Diagnostic Imaging about their decision to adopt the technology and what the process has been like.

What initially sparked your interest in PET/MR?

PET/MR is a combination of two of the best modalities in detecting and staging cancer and I thought the combination of both had a promising future. Secondly, it completely replaces the radiation from the CT part of PET/CT, which saves about 100 chest X-rays worth of radiation. Thirdly, it turns out that the PET component of the PET/MR is much higher quality than the PET in any existing PET/CT on the market, whereby we can significantly reduce the amount of isotope we administer, which decreases our typical FDG dose, for example from 15 millicurie to about 7 or 8 millicurie. This enables us to save another 50 chest X-rays worth of radiation. So, for every PET/MR we do instead of PET/CT, we are saving the patient 150 chest X-rays worth of radiation.[[{"type":"media","view_mode":"media_crop","fid":"34420","attributes":{"alt":"Steven Mendelsohn, MD","class":"media-image media-image-right","id":"media_crop_8134601807580","media_crop_h":"0","media_crop_image_style":"-1","media_crop_instance":"3662","media_crop_rotate":"0","media_crop_scale_h":"0","media_crop_scale_w":"0","media_crop_w":"0","media_crop_x":"0","media_crop_y":"0","style":"height: 203px; width: 160px; float: right;","title":"Steven Mendelsohn, MD","typeof":"foaf:Image"}}]]

How did you originally justify the cost?

You can’t justify the cost, at least in my mind’s eye. If you are looking at an ROI, you’re not going to find it. The way I can justify the cost of spending money is that it will save lives, and the patients have already seen it makes a dramatic difference. People who might think they only have a single, solitary lesion for local infection, we do the PET/MRI and we can find metastatic lesions that change treatment and hopefully helps prolong their life. The real reason to do it is noneconomic, the real reason is to save lives, alter the treatment, and do good for patients. Eventually, you may be able to make an economic base when there is a CPT code and third party insurers and Medicare cover it, and it should get a significantly higher rate than PET/CT because the machine is much more expensive. But, for now, the reimbursement is pretty much the same as a PET scan, even though PET/MR costs probably four or five times as much as a PET/CT.

It takes much longer to interpret it because a PET/CT is typically read by only one radiologist who reads both the PET part and the CT part. The PET/MR is interpreted by at least three radiologists because we have the PET component, which is a nuclear medicine doctor; we have extensive body MRIs that require the body MRI specialists; and we also have MRI of the brain and spine that require neuroradiologists. So the whole process is much more expensive. To save lives for our patients is fantastic. In this day and age, I cannot make a pitch that it will generate revenue at this point.

So there hasn’t been an ROI on the technology? What about revenue?

Oh we’ve seen revenue, but ROI is negative. We are not making money on it. We have spent much more money on it than any of the reimbursements we receive for performing the studies. If you look at the fact that it demonstrates to our physicians and the public that we are still at the leading edge in quality, then there is collateral benefit because all of the doctors in the New York area can see that Zwanger-Pesiri is on the leading-edge technologically, as we have been for the last 10 years. We are very much leaders in achieving highest quality and, quite frankly, I prioritize achieving highest quality over maximizing potential revenue. Eventually, after the insurance companies are finished declaring it research and investigational, they will recognize the value and reimburse for it, but they’ll want to wait out as long as they possibly can. What amazes me is the combination of the highest quality PET scanner on the market with a 3 Tesla MRI, which is the highest quality MRI on the market, and insurance companies are hiding behind paying saying ‘it’s research and investigational.’ This will continue for a number of years, basically until public outcry and patients demand coverage.

What has the reimbursement process been like?

The reimbursement is very difficult. Some of the insurance companies will pay us for the PET scan component, others will not. For a number of studies we do, we get zero reimbursement.

What kind of training did implementing this technology require for staff?

The main thing is that it’s MRI with PET and many people initially perceive this as a PET scanner that happens to use MRI for attenuation correction, where it’s being run and managed by the nuclear medicine people who are most comfortable with PET. The reality is, it’s a very complex MRI machine that needs to be run by very highly trained, highly skilled MRI technologists, because that’s where the true benefit is: you’re able to do a high quality MRI in a very time-efficient manner. For every study, we do a complete body MRI, which is not what anyone else in the country does, everyone else does just an MRI limited or what is called an attenuation correction, and then they’ll do one relevant body part MRI of one portion. We do the entire body: the brain, the head and neck, the chest, the abdomen, the pelvis, the cervical spine, thoracic spine, and lumbar spine, both prior to and following gadolinium. It’s very extensive MRI, but we are able to do it very efficiently because we have acquired a lot of ability, skills, and background over the years developing and implementing 3 Tesla MRI protocols, so we are able to very efficiently cover the whole body within that 45-minute scan time.

What measures did you take to get referring physicians on board and how have they responded?

Most of the referring physicians are the same doctors that order PET/CT scans. The oncologists, the pulmonologists, and some of the gastroenterologists, and it’s a continuing education process. We have meetings and dinners. We send them educational information. We have full portfolios of articles in the literature we provide them with. We provide them with interesting cases that demonstrate certain pathologies, so it’s the beginning of a long, arduous educational process. It’s very hard to move the referring physicians out of their comfort zones. Their comfort zone is now with PET/CT and for them to transition is difficult. Most of the oncologic protocols are based on PET/CT and they wonder how they can change their protocol with PET/MR. Well, they can do both PET/CT and PET/MR on the same patient. They will know the additional lesions that we see on the PET/MR. With PET/MR, you are saving all of the radiation dose, so the initial study can do both modalities and then all of the follow-ups can be with PET/MR.

But it’s a learning curve and we have experienced this before when MRI first came out in the late 1980s and early 1990s, we went through the same thing with doctors for them to basically accept that this is a modality, it’s valuable, it works, and when they see the value in it, and they only need one or two patients to see its tremendous value, and then that one doctor will become a convert. So it’s one at a time and right now there is not much literature, all of the literature is actually in the radiology journals, especially the journals of nuclear medicine. Very little literature has made it out into the oncology journals or the thoracic surgery journals, so we need the literature to be in the journals that these clinical doctors read, not just that the radiologists read.

What was your biggest struggle with the installation process?

Believe it or not, there was not much of a struggle. Siemens was very easy to work with. Their construction specifications were very straightforward: it requires an MRI room with appropriate lead shielding because that would typically be in a PET scan room. So, the amount of lead, and weight requirements on it was pretty easy. To be quite honest, it was easy construction; the equipment room is slightly larger than the typical MRI equipment room. It was no more difficult than any of our other typical MRI installations.

What would you say is your biggest struggle now?

Our biggest struggle now will be getting the insurance carriers to stop pushing the patients to PET/CT and over-radiating them, just outright accept that PET/MR is equivalent to PET/CT, and pay for it based on what they pay for PET/CT. They need to stop victimizing the patients with excess radiation. So we can eventually say ‘look its better than PET/CT’ and they’ll pay. But for right now, they should stop their arguments and not paying for it altogether and therefore forcing patients to get a PET/CT. They have no basis or justification to say that it’s anything less than equivalent. At a minimum, they can outright say its equivalent to a PET/CT, pay for it as a PET/CT regardless.

Is there anything else you would like to add?

We’ve done quite a number of patients so far and we are amazed at how much more information we get. For example, we have had a number of patients already with subcentimeter liver metastases that do not show up on CT and are too small to even light up on PET, but they get picked up on the MRI component. So we are seeing a good number of subcentimeter liver metastases that otherwise would have been completely missed. We’ve also picked up brain metastases very early on. CT or PET is not good for detecting small brain metastases and clinical doctors don’t request an MRI when there are no neurologic symptoms. Brain metastases, depending on the location of the brain, only need to be a centimeter or more in size to develop neurologic symptoms, so these patients never get an MRI. So here we are doing an MRI anyhow of the brain and it’s a post-gadolinium MRI so it is really highly sensitive and we are picking up a number of unsuspected brain metastases in patients. Perhaps that may or may not change their medical treatment or oncologic treatment, but we are picking them up, we are detecting them earlier.

Want to see PET/MR in action? Check out our slideshow.