Interview with Dr. Ross Levatter
[Dr. Levatter is in solo practice in radiology. He performs second opinion teleradiology interpretations. This interview was performed during late November, 2011.]
Interviewer: How long have you been in practice, Dr. Levatter
Dr. Ross Levatter: Well, I’ve practiced radiology, or medical imaging, for over two decades. As you know, we in medicine say we’re “in practice” because we’re always practicing, always learning new things.
I: Is that a necessity in your field?
RL: It’s true in all of medicine, I’d guess, but it’s certainly true in medical imaging. Just look at the name. When I was in training in the 1980s, it was called “RADiology” because it involved interpreting images that came about by RADiating people with x-rays. But now many of our images come from other sources, not requiring radiation, like MRI or Ultrasound. Each new modality requires learning new tricks, if you will. In addition, many of the standard work-ups of just two decades ago have fallen by the wayside. When I was in training, colon cancers were picked up by barium enemas, small bowel obstructions were picked up by fluoroscopic exams of the small bowel, and kidney problems were evaluated with something called an IVP, or intravenous pyelogram, which involved taking a series of xrays after injecting the patient with a contrast agent, or “dye”, in their arm vein. All of these problems are now worked up with abdominal CTs, which have become a workhorse of our specialty. It is one of my personal areas of subspecialty interest and expertise.
I: What are your other areas of expertise?
RL: Well, it goes to show you can’t predict a career in advance. My fellowship training was in Nuclear Medicine, but as time went on I made a name for myself in GI tract imaging and more generally in abdominal imaging. I’ve also become very interested in breast imaging (mammography, US, MRI) and more generally oncologic, or cancer, imaging, after a close relation succumbed to the disease. And over the last decade, ironically, PET imaging--a part of my nuclear medicine fellowship training which at the time was viewed as a research tool only for large university settings--has become a mainstay of oncologic imaging. A fusion technique, PET/CT imaging, allows me to combine the best of both worlds, my interest and enjoyment in both nuclear medicine and CT along with my feeling that, in helping people with cancer, I’m making a significant contribution.
I: Can you share some interesting cases with us?
RL: Well, of course medical imaging is a visual science, and in this interview context I can only describe what I saw rather than point things out on the images themselves. But if you’d like I can verbally share some of my “best cases” with you.
I: Please do.
RL: Well, keep in mind, as I said, my field has been advancing dramatically from a technical perspective in the last two decades. Things that were tough and challenging to diagnose in the 1980s can be much easier today. But one of my favorite cases occurred in 1990. I had just started work in Phoenix and was asked to present some cases at “tumor board.” This is a monthly conference where surgeons, oncologists, pathologists, and radiologists get together to discuss challenging current cancer cases. The CT on the patient of interest showed a large solid mass in the right upper abdomen, just below the liver. It took up over 50% of the abdomen and was completely nondescript--solid and uniform in appearance. The official report merely described it as a mass and recommended biopsy. As I discussed the case, I made a number of points not in the original report. First, the large mass was not the only finding of interest. In addition, the kidneys and ureters--the tubes connecting the kidneys to the bladder--were abnormal, grossly dilated. In addition, there was a very subtle but significant finding. It’s taught in my field that “the hardest finding to make is of the normal structure that’s not present.” So we’re all taught to develop what’s called a “systematic search pattern.” In doing my search, I noticed there was no muscle in the front of the abdominal wall. So, to sum up: we had a man in his 40s presenting with grossly dilated kidneys and ureters, no abdominal wall muscle, and a large abdominal mass. What can it all mean? Part of the fun of being a radiologist is trying to put together clues like that. It turns out there’s a disease called “Prune-belly Syndrome.” The term prune-belly comes from the lack of abdominal muscles, causing the abdomen to wrinkle like a prune. The disease is congenital, that is, you’re born with it. It has 3 components: dilated urine-containing structures, absent abdominal wall muscle, and undescended testes. When the baby boy is in the womb, his testicles are in his abdomen. Late in pregnancy, they drop to their normal location, MOST OF THE TIME. But in this disease, they don’t. Now, it’s a medical fact undescended testes have an increased risk of becoming cancerous, and the most common testicular cancer is a “seminoma.” So I concluded, based on the images alone, that this 43 year old man had an intra-abdominal testicular seminoma as a complication of his life-long prune-belly syndrome. This is exceedingly rare, but if I were right, the patient didn’t need a biopsy; he needed chemotherapy. I had just started working at this hospital and no other radiologist, in my group of 20 was as confident as I, so they biopsied it. Intra-abdominal seminoma. In more than 20 years, I’ve never seen another case.
I: Wow. That’s fascinating. Can you give us an example of a case that’s easier to diagnose now, with the improved technology you mentioned?
RL: Sure. In the early 1990s, a middle aged man had been in our hospital for several weeks. He was getting sicker, as evidenced by the fact that, when I first ran into his imaging on call one night, I was reading his third abdominal CT in the prior 2.5 weeks. Step one is to review the earlier exams. I noticed the first study showed an infarct (area of dead tissue) in the right kidney, which was called, and a subtle inflammation of the mesentery (or intestinal root, connecting the bowel to the back of the abdomen, where the blood flow comes from) which was not called. In the second study, I noticed the subtle inflammation of the mesentery was now becoming more well-defined, and a new lesion was present in the spleen. The mesenteric problem was called abnormal bowel by the reading radiologist, but I thought that was unlikely because I could not connect it, image to image, to the remaining bowel. Now we take 1 mm thick slices on routine CT, but back then we took 8-10 mm thick slices. Our “building blocks” of the body, if you will, were 8 to 10 times thicker. So of course our resolution was less. It’s much easier today to answer questions like “is this bowel or something else” on CT. Back then it could be a real challenge, but as I said it happened to be an area of subspecialty interest to me. The second interpreter also said the splenic lesion was a cyst, but I thought its sector, or pie-shaped, nature of the lesion was rather classic of an infarct. Cysts tend to be round. Now on my third study the mesenteric lesion was becoming more mass-like and partially filled with contrast. Putting together the evolution of this inflammatory mass over a short time span with the fact the patient was infarcting both his kidney and his spleen, I realized the patient was throwing emboli (blood clots) from some source. By carefully tracking the lesion from image to image, I confirmed the mesenteric mass arose from a mesenteric vessel, and diagnosed the lesion as a mycotic aneurysm of the SMA (superior mesenteric artery). And here’s my point about changing technology: in the intervening two decades, technological advances have made CT the modality of choice in assessing vessels for aneurysms, and 3D reconstructions would now make this diagnosis trivial. But at the time, few radiologists would hazard such a diagnosis with the equipment then available. Back then everyone “knew” that CT was good for evaluating solid organs--the liver, the spleen, the kidneys--but that angiography, an invasive study of the blood vessels done in an OR-like setting, is required to diagnose aneurysms. Mycotic aneurysms arise from bacterial infection of a blood vessel wall, weakening the wall so that it balloons out. The SMA comes off the aorta in the abdomen and supplies the gut. Mycotic aneurysms are uncommon and mycotic aneurysms of the SMA are reportable. Because it was such a rare diagnosis, and because urgent surgery was called for if I were right, I brought the patient back down (it was after midnight) and confirmed the diagnosis with vascular US of the abdomen.
I: Is that something most radiologists would do?
RL: I’d have to say “No.” Most radiologists see themselves as consultants who interpret whatever image another physician orders. And of course that’s the normal expectation and how I practice as well 99% of the time. But I had a number of health problems myself when I was growing up--it’s one of the reasons I developed an interest in medicine--and so I sometimes act more as a patient activist than is typically done by doctors in my specialty. It’s a personal choice--I’m not suggesting everyone should do it this way--and in fact today as government regulations on hospitals have become more onerous it has become impossible to do what I did 20 years ago. Today I wouldn’t be allowed to perform a free US on someone without an order after midnight simply to confirm my CT diagnosis. But I’m glad I was able to do it back then, because it confirmed my diagnosis and led to urgent surgery, saving the patient from losing any bowel. The cause of the emboli, by the way, was an infected heart valve, which was also replaced. Like I say, this would be a much easier diagnosis today, with computer post-processing, 3D imaging, and much thinner CT slices, but fortunately as a result of my subspecialty interest in abdominal imaging I was able to make the diagnosis back in the “primitive” days of the early 1990s.
I: Are there benefits to subspecialty radiology?
RL: Oh, I think so. Definitely. Radiology is a burgeoning field. It’s hard to imagine a patient in the hospital these days who get NO imaging. It’s ubiquitous. There is breakdown by body part (neuroimaging, head-and-neck imaging, chest imaging, abdominal imaging, pelvic imaging, breast imaging, musculoskeletal imaging); breakdown by age and gender (women’s imaging, pediatric imaging); breakdown by technique (interventional imaging, nuclear medicine). The best general radiologists are very good in most all of these areas (in the top 30-40%). But a subspecialist who devotes himself near full-time to one or a few things can hopefully be in the top 10%. Does it matter to the patient if the reader of their study is in the top 40% or the top 10%? Often not. But occasionally yes. So I think it can be important. Let me tell you about another case, this more recent, that I think makes that point. About 2006, I was asked for a second opinion on a patient who was clinically doing very poorly. Hours earlier, another radiologist had read out her abdomen CT as essentially normal. I noted the pancreatic body, which sits just behind the stomach, was subtly edematous and inflamed, and there was a minimal though abnormal amount of fluid in the lesser sac, a potential space between the stomach and pancreas. Further, I noticed the back wall of the stomach was dramatically thickened, and contained gas. Now gas in the lumen of the stomach, where the food goes, is perfectly normal, but there should never be gas in the WALL of the stomach! In addition, the bright enhancing line representing the lining of the stomach was discontinuous. That should never happen either. I diagnosed perforated stomach ulcer with secondary pancreatitis, which was proven at emergency endoscopy. I should note here that many general radiologists would be uncomfortable making this diagnosis on CT. They were taught that CT is not the way to diagnose such things. In the old day one did barium GI exams of the stomach; more recently one goes straight to endoscopy (the GI doctor sedates you and puts a telescopic tube down your throat into your stomach and looks around). But by 2006, CT quality had improved to the point that if you looked for GI problems and knew what to look for, you’d occasionally find them. Perhaps that’s a difference between a good general radiologist and a subspecialist imager.
I: Does it really matter, the difference between a good general and a good subspecialist radiologist?
RL: Sometimes--and I don’t mean to sound overwrought here--it can be a matter of life and death. I recall a case a few years back involving a young woman in her late 3rd trimester of pregnancy. She was flown down by emergency helicopter from northern Arizona after a motor vehicle accident that killed her boyfriend and young son. An emergency CT for trauma was done and the initial radiologist, a general rad, described a small non-surgical laceration of the liver but thought everything else was OK. He specifically declared the fetus to be fine. A few hours later I was asked to look at the case, because the fetus was in distress. The placenta was grossly torn, what the obstetricians call abruptio placentae, which meant the fetus was not getting a proper blood supply. This led to an emergency C-section, but by that time it sadly delivered a dead fetus.
I: Sad case.
RL: Well, I should leave you on a happy note. The development over the last decade of teleradiology has made it easier to send your images to the best imager for your problem. In the past, if you’re in a smaller city or town, with a radiology group that has maybe 3 or 4 radiologists, it’s really not feasible to have all subspecialists. First, there are more than 3 or 4 subspecialists. Second, people take vacations, so you really need more than one of each subspecialist per group. Finally, there really aren’t enough cases to “feed” the subspecialist. If you’re practice is only doing 20 complex musculoskeletal MRIs each week, you can’t keep an MSK imaging specialist busy enough. But now the technology is available to send your images anywhere. If you’re in Kansas but the best person to read your case is in Phoenix or Los Angeles, you can upload your images and he can download them in mere minutes. Geographic proximity no longer has to determine who reads your studies. That’s potentially good for everyone.
I: Why don’t we see more of it?
RL: Mostly the restricting effects of government regulations. As things now stand, if you had your CT scan done in Topeka and you wanted it interpreted by an expert in Phoenix, you have two options. You can hand-deliver them to the Phoenix radiologist. In that case, he would only need to be licensed in Arizona. Or you can send the images over the Internet. But in THAT case, that radiologist would have to be licensed BOTH in Kansas AND in Arizona. Since state licenses are costly and create an enormous time-burden to maintain, that makes it harder for teleradiology to expand than would otherwise be the case.
I: Can we speak to you again. You’ve been most helpful in explaining what radiologists do.
RL: Thank you. I always enjoy discussing interesting cases. If your readers enjoyed this discussion, feel free to come back.
