A Radiologist Discusses His Experience with "Don't Touch Lesions" (DTLs)
Posted May 09 2011 12:00am
It is my understanding that the term "don't touch lesion" was first coined by Dr. Clyde Helms (see: "Don't Touch" Lesions ). He is a well-respected and prolific musculoskeletal radiologist who currently serves as chief of his division at Duke University. “Don’t touch lesions" (DTLs) are imaging abnormalities that a radiologist should readily recognize as benign and therefore should not be biopsied or surgically removed. Many of these lesions are pathognomonic based on their imaging appearance. Therefore, a definitive diagnosis can be made by the radiologist based solely on imaging studies. DTLs include the following: fibrous cortical defect (sometimes referred to as non-ossifying fibroma), periosteal desmoid, fibrous dysplasia, enchondroma, bone infarct, subchondral cyst, and heterotopic ossification.
Radiologists such a myself are taught to provide clinicians with a differential diagnosis in our imaging reports whenever we see a radiographic abnormality. We see a lesion, generate the differential, and leave it to the clinician to “correlate clinically” and decide what action to take next or to decide what to label the lesion. Unfortunately, such advice is often unhelpful and, in the case of DTLs, it can be misleading and actually harmful. When a definitive diagnosis is not provided in the report for a DTL, many problems can occur, not the least of which is patient anxiety. Also, additional and more costly imaging studies are often recommended by the radiologist to pursue a diagnosis. Some of these studies such as nuclear medicine bone scan and computed tomography will not only increase the cost of care but also expose the patient to an additional dose of radiation.
Dr. Helms divides DTLs into three categories: (1) post-traumatic lesions; (2) normal variants; and (3) lesions that are not normal but definitely benign. Post-traumatic lesions include remote fracture deformities, avulsion fractures, and heterotopic ossification. Normal variants include pseudocyst of the calcaneus, pseudolesion at the undersurface of the medial aspect of the clavicle at the insertion site of the costoclavicular ligament, and cortical desmoid at the posteromedial aspect of the distal femur. Lesions that are normal and should be categorized as definitely benign include bone islands (i.e, enostosis), bone infarcts, and hemispheric spondylosclerosis.
One of the most satisfying cases that I personally have been involved in was that of a 13 year old girl from Saudi Arabia. Her uncle, an American surgeon and colleague, showed me her images and told me that her leg was going to be amputated for a tumor in her distal femur. I reviewed her films and noticed a cortically-based sclerotic focus in the posteromedial aspect of her distal femur. This lesion was "hot" on the nuclear medicine bone scan. I was told that a biopsy of the lesion has already been performed and was diagnosed as osteosarcoma.
The first thing I noted was that the lesion was non-aggressive in appearance. Secondly, it was in the characteristic location of a cortical (i.e, periosteal) desmoid. Thirdly, and on close inspection, the bone scan showed increased activity in the posteromedial aspect of BOTH distal femurs. I told my colleague that I knew, without a doubt, that this lesion was benign and advised him to cancel the planned amputation. I further asked him to recommend a radiograph of the contralateral, asymptomatic knee of the patient. A few weeks later he informed me that the young girl's doctors had heeded my advice and that imaging was performed on her asymptomatic knee. The results of this test confirmed that she had the identical normal variant cortical desmoid in her other femur. I am very glad to say the planned amputation had been avoided.