TAPE STARTS – [00:00]


Marlena Jbara: Diabetic pedal osteomyelitis concept and controversies part 2. My name is Marlena Jbara. I am an MSK radiologist from Staten Island University Hospital, Northwell Health. Disclosures: I or related party have no financial relationships to disclose. Learning objectives: This lectures will again review the current epidemiologic trends and implications related to care and imaging of the diabetic foot. The current Infectious Disease Society of America and American College of Radiology guidelines will be reviewed with imaging findings of x-ray, MRI and PET-MRI in patients afflicted with a diabetic foot. We will review the epidemiology and pathophysiology of the diabetic foot. Review international and national guidelines with respect to imaging of the diabetic foot. We will review pertinent imaging findings and pitfalls of diabetic foot pathology with regards to x-ray, MRI and PET-MRI. Outline: In Part 2, we are going to follow-up on the pathophysiology with an update on the role of imaging. We are going to look at MRI, nuclear medicine and future imaging pathways. We are going to sort the statistics and then finally, we are going to finish with American College of Radiology appropriateness criteria with regards to imaging of the diabetic foot.

[02:00]
So picking up from where we left off on Part 1, the downside of x-rays in the diagnosis of osteomyelitis is that requires 30% to 50% of bone loss to appreciate demineralization and that takes approximately two to three weeks of active disease. Simply, doesn't work when you are looking to make the diagnosis early and accurately. And that follows with sensitivities of ranges from 43% to 75% and specificities of 75% to 83%, bringing us to the upside of MRI. With MRI, we are going to use this technique to differentiate osteomyelitis from reactive edema. We can differentiate neuropathy from osteomyelitis and further apply imaging concepts to differentiating neuropathy from infected neuropathy and we can follow with this in an article by Toledano, MR evaluation of bone marrow changes in a diabetic foot, a practical approach. Why MRI? It defines the extent of bone and soft tissue infection and helps to define and limit the extent of resection. With a tensed sensitivity ranging from 82% to 100% and a test specificity averaging 81% with a range of 75% to 96%, this is an excellent test for the diagnosis of osteomyelitis. MRI osteomyelitis versus reactive edema. We will use some primary signs such as the ulcer being contiguous with the bone and T1 weighted confluent geographic low signal.

[04:00]

Now what I mean by this is that an MRI takes advantage of different tissue properties of fat and water and their precession frequencies. One of the ways that we listen to this frequency is T1 weighting and anything that has a high fat content will be very bright. Alternatively, any infiltration of that fat such as osteomyelitis with infiltration of the bone marrow will produce a confluent geographic low signal. Secondary signs of osteomyelitis versus reactive edema include periosteal reaction, skin ulceration, sinus tract leading from the skin ulceration, of course cellulitis and abscess. Criteria for diagnosis of osteomyelitis by MRI. An excellent article regarding the diagnostic utility of T1 weighted MRI characteristics in evaluation of osteomyelitis of the foot was undertaken by Johnson. A retrospective evaluation from 1996 to 2002 included 80 patients with MRI diagnosis of osteomyelitis. All of these patients went to surgery within six weeks of that MRI and postsurgical histology found MRI to be truly positive for osteomyelitis in 59 patients and falsely positive in 21 patients. T1 low confluent signal in a geographic pattern was present in 100% of surgically proven cases of osteomyelitis, the true positives. The majority of false positive had a hazy reticulated pattern independent of T2 or post contrast enhancement and none had surgically proven osteomyelitis.

[06:09]

Powerful paper. So let's turn to our images of the fifth toe where we are asked to exclude osteomyelitis in a person who has a diabetic foot ulcer. Again, we are here with our ghost sign including eraser of the fifth distal phalangeal cortex with some periosteal reaction along its distal medial border. Moving on to the MRI, I bring back our eraser to remind us of the fact that on the bottom right picture on this T1 weighted image, you can barely see the outline of the confluent infiltration of the fifth distal phalanx on the sagittal image, which you can see clearly see outlined by this light bulb T2 weighted bright signal at the top left image denoting osteomyelitis and the ghost sign. Here is another example in this lateral x-ray in the back field noticing these hammertoe deformities and perhaps the soft tissue swelling at the dorsal aspect of the fourth proximal phalangeal head at its distal aspect. This close-up image on the bottom left shows more clearly the sub-periosteal bone resorption and eraser of the cortex. Moving onto four representative images in this patients forefoot MRI, we can see on the top left the proximal phalangeal head T1 low signal marrow infiltration and its extension into the middle phalanx further defining the extent of this disease.

[08:00]

Notice the T2 weighted image on the right denoting the high signal crossing these boundaries and of course the short axis cartilage on the bottom left image denoting T1 confluent low signal marrow infiltration equivalent to osteomyelitis and the corroborating short axis T2 image on the right. Of course, there are secondary signs in MRI. Here on three beautiful representative images that are taken out of Donovan and Schweitzer's excellent article on the use of MR imaging in diagnosing diabetes related pedal osteomyelitis and radiographics in 2010. You can see in the midfoot, there is a low signal replacement of the metatarsal fat pad denoting a pressure point of course with preservation of the fifth metatarsal head fatty marrow. In the central image, we can see the plantar declination of the talus in someone who is undergoing a collapsing Charcot neuropathic osteoarthropathy. Noticing on this post contrast sagittal T1 weighted image, we have curvilinear rim enhancement surrounding linear low signal focus consistent with an adventitial bursa. And on the right, we see long axis imaging through the great toe with a deep soft tissue ulcer bordering its lateral aspect. In addition on this post contrast T1 fat suppressed images, we see linear enhancement along the flexor hallucis tendon denoting pyo-tenosynovitis. In terms of the use of contrast to define extent of infection and MRI, here is a nice example denoting ulcer, abscess and fistula.

[10:06]

So on the coronal T1 weighted image through the ankle, we noticed the fatty marrow in the distal tibia, talus and calcaneus. We can even see parts of the deltoid and the spring ligament. We can see confluent T1 low signal soft tissue replacing the plantar soft tissues to the calcaneus with the addition of intravenous gadolinium contrast with fat suppression. We now can highlight the confluent areas into an area of soft tissue ulcer extending to a larger abscess and of course all of the remaining high signal changes denoting phlegmonous infection without the presence of drainable collection. And on the far right, we see this long axis image through the hind foot where we can see the soft tissue ulcer opening to a large rim enhancing abscess, which as a deeper dissection to the plantar calcaneus with osteitis and less than two to three marrow millimeter infiltration on this long axis image. And of course, we can see the presence of necrotizing infections studying the flexor tendon in this example where the gas produced by the bacteria create susceptibility artifacts denoted by the magnet with small areas of bloom. And we can see that here along the long axis of this plantar area flexor tendon with gangrene. Moving on to MRI, neuropathy versus osteomyelitis, the ulceration is the key and MRI will be performed to evaluate extent of disease.

[12:01]

So in the neuro-osteoarthropathic patient, a subchondral periarticular or polyarticular process without a soft tissue ulcer is statistically likely to be uninfected neuroarthropathy given the fact that the ulceration is required for the ingress of bacteria. This is going to be most common at Chopart at the hind foot and Lisfranc at the midfoot. Whereas osteomyelitis, we are going to be looking at a T1 weighted low signal marrow process adjacent to a skin ulcer with cellulitis, abscess or sinus tract. First presentations are generally located distal to the tarsometatarsal joints at the calcaneus and malleoli or other pressure areas. Looking at this example, a patient comes in with a talonavicular fragmented dislocation and clinical question remains, is there osteomyelitis? And here we can see that the talar head had medially subluxed from the navicula, which is fragmented, has dorsal enthesopathy and soft tissue swelling and has subtalar disarticulation. In this three representative transaxial T1 weighted images looking at Charcot versus osteomyelitis, one of the first things we can notice is that the soft tissue plate is maintained. The skin being of linear low signal followed by a small lower level of subcutaneous fat, we can see nicely outlines this process though all of the bones have T1 low signal change, subchondral cyst and loss of periarticular synovitis, a finding that can be seen in neuropathic osteoarthropathy.

[14:00]

In addition, the pattern of periarticular enhancement for Charcot patients surrounds the entire process rather than infiltrating in the individual joint spaces. And lastly, we can look at this chart looking at Charcot versus infected Charcot noticing a few interesting things. In terms of subchondral degenerative cyst and intraarticular bodies that may be present in a patient with non-infected neuroarthropathy that may disappear on followup when you are looking at a patient who now has infected neuroarthropathy. Of course, we have mentioned the ghost sign may be absent in a non-infected neuroarthropathy. Usually, things are more sclerotic, debris. In terms of infected neuroarthropathy, there may be erosive changes in subcortical lucency. Soft tissue enhancement again in Charcot versus infected Charcot is extensive and infected Charcot rather than defiant to the juxta-articular soft tissue in Charcot patients. Of course, the ghost sign bones that disappear on T1 and reappear on T2 are strong indication of osteomyelitis. Moving to this next example of Charcot and osteomyelitis, we can see on these two long axis frontal and oblique projections through the foot, a soft tissue ulceration along the lateral border of the fifth metatarsal with retained small pluck of hardware there in his cuboid mass to the base of the fourth metatarsal. Moving on to transaxial imaging through this area, we can note on T1 weighted image that there is a soft tissue ulcer as you can see from the skin defect and there is some soft tissue infiltrative changes but there is not T1 marrow confluent low signal within the adjacent cuboid.

[16:10]

And even in the presence of reactive changes within the cuboid, at best this may represent osteitis or inflammatory infectious changes of the cortex without deeper marrow infiltration at this time. In this case here of Charcot and osteomyelitis, we can see again the neuropathic midfoot changes with a collapsing pes planovalgus deformity. A large cuboid plantar soft tissue ulceration and again marrow T1 infiltration of the cortex with an adjacent small periosteal fluid plane denoting osteomyelitis in the Charcot patient. So moving on to bone scan, here I put out a summary statistical analysis slide that denotes meta analysis where they have taken 110 full text articles and extrapolated 17 to be included in this analysis removing things that had lack of relevance or didn't contain primary data or didn't recruit on the basis of suspicion, patients who didn't have the disease were all removed from the analysis and they found that there is low specificity for bone scan given patients may have fractures, postsurgical changes, Charcot and crossover to seeing gout and inflammatory arthritis in patients you suspect osteomyelitis. The upside is if the bone scan is completely negative, it rules out disease.

[18:01]

So it's a strong negative predictor. And looking at this, you can see in the chart here that the technetium labeled 99 bone scan has specificities that range from 18% to 100%. And MRI has found to outperform in the summary statistics where [indecipherable] [18:26] denotes that MRI is more sensitive and specific where you can see comparatively the MRI having more than 90% pulled sensitivity versus bone scan at 81% and a pulled specificity of 79% versus bone scan of 28%. If MRI is contraindicated then what else is there. Well, we have mentioned we can label technetium-99 bone scan but this can be an expensive process, time consuming and of course we just mentioned it's less specific than MRI. We can label the white cells with Indium-111 or technetium-99 labeled WBCs and again we can increase our bone scan sensitivity to 90% but the sensitivity will remain low at about 70%. And moving on to new and exciting techniques, we can use 18 fluorodeoxyglucose labeled red cells and perform a PET scan doing physiologic imaging. In addition, we can further go on to fuse that physiologic image with an anatomic image such as an MRI to further increase the sensitivity to detect osteomyelitis. And how this works is that your blood is drawn at the time you come for you imaging. It is then mixed with a fluorodeoxyglucose substrate, which will label your red cells.

[20:03]

This will be re-injected into you and the cells that are taking up great metabolic needs will of course absorb the fluorodeoxyglucose and things that we are looking for such as cancer and infection will have certain standard uptake value. We will then go on to image these areas and fuse this to an anatomic image to localize whether or not this is present within the bone. An inflammatory cells will preferentially metabolize glucose. And the power of FDG-PET is that the standard uptake value may even differentiate Charcot from an infected Charcot. This example on the right you can see the low resolution, physiologic image with uptake all around the hallux metatarsal. And you can see that we can put color on that and make that prettier and of course we can fuse that to a CT image where we can see that this is all periarticular, periosseous infection around the bone but not within the bone. And of course in this example down on the bottom right, we can see that there was a separate area and the talus had a standard uptake value that was infected and of course there is this area down below that's showing the periosseous infection. And the positive predictive value for MRI in these cases versus FDG-PET. FDG-PET is outperforming at 78% positive predictive value to be able to establish whether or not the disease is present where MRI may be 56% in this nice article out of molecular imaging and biology in 2010 by Nawaz. The power of FDG-PET further in diagnosis pedal osteomyelitis in diabetic patients can be in seen in a nice article in Journal of Nuclear Medicine in May 2015 where we can see representative images through this person's tibia and ankle on the PET scan showing this low resolution uptake fused to a CAT scan.

[22:16]

And you can see that this is within the bone and also along the adjacent soft tissue and this was 100% osteomyelitis surgically proven. We can also use hybrid imaging in the form of MRI and this transaxial leg imaging showing that the detection of osteomyelitis can be made with comparing MRI, white blood cell and scintigraphy showing that all had similar sensitivity but FDG-PET and 99 labeled white cells that offer the highest specificity. The American College of Radiology provides appropriateness criteria for the best accurate clinical imaging pathway to take on in patients with suspected osteomyelitis and diabetes. Those patients in this clinical scenario that have soft tissue swelling without neuropathic arthropathy or ulcer. They suggest that you perform an x-ray of the foot as an initial study where radiographs and MRI are complementary and both are indicated. The results of the initial x-ray exam do not preclude the necessity for additional studies. And that is so because as we noted before 30% to 50% of bone mineralization needs to be loss for us to denote the presence of bone erosion. Therefore, following this with an MRI of the foot without and with contrast where MRI and x-rays are complementary, both are indicated and MRI is going to be used preoperatively to identify the extent of involvement and to map devitalized areas.

[24:01]

An MRI of the foot without contrast can be obtained if there is a person who is renally insufficient or a dialysis patient and that radiographs and MRIs again noted to be complementary and both are indicated. And note that technetium labelled or Indium labelled 111 WBC nuclear medicine scans are usually not appropriate and they may be appropriate in certain circumstances such as if the MRI is contraindicated or unavailable in a postsurgical patient or someone with a complicated wound. The American College of Radiology gives a variety of clinical scenarios ranging from soft tissue swelling with neuropathic arthropathy without ulcer again noting that x-ray and MRI are your main stem treatment in diagnosis for treatment where the initial study is x-ray but it does not preclude the need for radiographs and MRIs together as they are complementary. And in this clinical scenario, a CT of the foot without contrast can be obtained if MRI is contraindicated and of course still nuclear labelled bone scan still may be appropriate for circumstance where MRI is contraindicated or unavailable. And moving onto a patient who has a soft tissue ulcer, again x-ray and MRI of the foot is the preferred imaging pathway to make the diagnosis of osteomyelitis in these important patients. Moving on to those patients who have neuropathic arthropathy and also notice it does not change the evidence based medicine supports, the addition of getting x-rays and then following that with an MRI to evaluate and map devitalized areas and extent of disease. And even with all that we know the prognosis of diabetic osteomyelitis, the prognosis remains poor.

[26:03]

And in the world statistics with more than a million diabetics, which will require limb amputation yearly, one major amputation is said to be performed worldwide every 30 seconds. And with limb amputation, the mortality rises. So patients who undergo a below-knee amputation may find their mortality rises from 13% to 40% at one year and from 39% to 80% at five years. In summary, I have provided you with an update on the role of imaging. We have sorted the statistics, we have looked at x-ray, MRI today, taken a glimpse of FDG-PET and hybrid imaging for tomorrow and went over the American College of Radiology appropriateness criteria as a reference tool for you when you are imaging your patients with the diabetic foot to provide a reliable clinical pathway to provide the best care possible for your patient. It has been my pleasure to be with you for this last 30 minutes and I thank you.



TAPE ENDS - [27:11]