Medicine is a very humbling field and sometimes when we see the presentation of a condition, particularly in neurology, we don’t know exactly what is going on. Which is to say we have a general idea of a category of conditions: infectious, inflammatory, neuromuscular, oncologic, genetic … but we don’t have a more specific diagnosis. Yet.
Sometimes it just takes a bit more time for things to declare themselves. It’s easier to be "right" as a second or third opinion when the initial workup has been completed, more time has passed, new symptoms appear, or more diagnostic resources or specialists are at your disposal. The response to certain therapies helps narrow the differential diagnosis as well. Did they get better on steroids? Did they fail to improve with antibiotics? The time course, associated symptoms, detailed exam, and then ancillary testing often confirm the diagnosis.
One can think about an atypical presentation of a more common condition (like confusional migraines) or a more classic presentation of rare condition (infant botulism). Hmm, or an atypical presentation of a rare disease or a yet-to-be-described condition ...
Early in my career, when I began caring for children with demyelinating diseases, there was a subset of kids who all behaved similarly: recurrent episodes of central nervous system demyelination, very steroid responsive, who did not have multiple sclerosis or neuromyelitis optica. Despite not having a more specific diagnosis, we treated with steroid-sparing immunosuppressive agents. Around 2017, Mayo began offering research testing for a new antibody. Five out of the five initial samples I sent were positive. The lab actually ran them twice because they thought it was a mistake. Nope. All these kids just had the same yet-to-be-diagnosed condition: anti-MOG associated disease (MOGAD).
So what do we do when faced with an unknown diagnosis?
First, say to the patient/family: I don’t know, but we will keep trying to find out. Then, start over from the beginning. Literally say: To make sure I understand what’s going on, and I haven’t missed anything, let’s go back to the beginning of your condition/symptoms. Ask, again about family history, travel, associated symptoms, supplements, etc. Listen differently for new clues and nuggets of information; don’t make assumptions.
Rereview all the labs, clicking on all the outside hospital records, and scanned reports. Go back to the original source and personally review documents. When there is an unknown diagnosis, lots and lots of random labs are sent. Sometimes the ER or the ICU or the inpatient team sends them, and someone finally gets routed the results when they are no longer involved in the patient’s care. I’ve seen many cases where the diagnosis has already been confirmed, but no one saw the test results, particularly for genetic testing.
If there is the ability to get a biopsy of the affected organ, this is the time to consider it. Often the main diagnostic conditions are infectious, oncologic, or inflammatory/autoimmune, each of which are treated very differently. A biopsy may not be conclusive of a specific diagnosis but can be immensely helpful to exclude these groups of diseases.
Next, ask the family and/or patient what they think is going on, or what they are worried about. It might not necessarily lead down the correct path, but it may help generate new ideas. And allow you to truly reassure them if they name serious conditions already eliminated or very unlikely. For example, patients may worry they have a brain tumor when a brain MRI has already been normal.
With patient permission, review case with colleagues — locally, regionally, nationally, internationally. Many subspecialties have listservs or opportunities for case reviews. While these reviews are very helpful, I often think the quality of the opinion is directly related to how personally they are involved in consultation and care. Referring for a second opinion with prior test results and images available for review can be invaluable.
Lastly, there are centers for rare disease consultation, like the NIH undiagnosed disease network, The Undiagnosed Diseases Network "is a research program that combines basic and clinical research services to understand health conditions for individuals and their families who have sought a clinical diagnosis without success." Physicians and/or patients can submit cases for review.
In the meantime, without a specific diagnosis it’s tempting to avoid starting disease-specific treatments, but that might not be possible. Doing "nothing" is a decision. When empirically treating a condition, close and ongoing re-assessment is critical. If patient’s condition is progressing or not responding to treatment as expected, there are several possibilities: 1) the patient is not taking treatment as prescribed, 2) the condition is not responding or is more severe than initially expected, or 3) we are on the wrong track. We see this a lot. "Trying to fit a square peg into a round hole," I can hear my mentor saying. When patients aren’t responding to current therapies, the same concepts apply: Start over again from the beginning, consult colleagues, modify treatment, and assess for effects.
What can patients do when affected by an undiagnosed condition? Advocate for themselves, keep records, review their own test results, ask questions, get scans on a disc. It may be tempting to seek many different opinions, but it’s often helpful to have someone "steering the ship" or following the disease course over time. Physicians need to understand these patients bring additional frustrations, hopelessness, or distrust of the medical community. But it is extra rewarding to solve these complex cases, and not just in a dorky, academic medical center kind of way.
Dr Jennifer Rubin is a child neurologist with specialized training and interest in acquired demyelinating diseases (multiple sclerosis, transverse myelitis, neuromyelitis optica) and genetic white matter diseases (leukodystrophies). She also serves as chairperson of the hospital’s Institutional Review Board (IRB).
Illustration by April Brust
