Today is a rainy cool day in August. I am on 24/7 home call on the pediatric neurology service for 7 days and somehow I have been scheduled to work 17 days straight. My allergies are really getting the better of me. I am constantly sneezing and sniffling, and I am a walking zombie as I recently finished night float and I am struggling to get reacquainted with the days. Yet somehow the volume has been less this time on the pediatric neurology service and I am less frustrated and depressed than last year. My brother in law who is going in to interventional radiology came and went lack week and questioned my decision to pick small town living over big city excitement, and then for a brief moment I did too.
Today I write about the case for empiric treatment. I will go back 1 month as my first day is senior on the inpatient adult neurology consult service on what was supposed to be a lazy in and out Sunday. The medicine ICU had admitted an alcoholic woman with a questionable seizure history with a chief complaint of altered mental status and staring spells/seizures. Her baseline functionality was "normal." The senior neurology resident who saw her last night said she likely had wernicke's-korsakoff encephalopathy and that she needed IV thiamine and folate. She was hospital day 2 and still not waking up with this treatment. Her spinal fluid was negative for any infectious processes and her MRI brain was normal. She recently had an EEG which showed no evidence of seizure.
I walked in the patient's room and she was staring out the window and appeared to be responding to some sort of external stimulus. She was delirious. She was virtually nonverbal, she was moving all 4 extremities but she certainly was not following commands. She was spiking fevers and her heart rate was elevated. Her nurse asked me very innocently: "What's going on, why isn't she getting better?" I said the simplest explaination is usually the right one, and this could all potentially be explained by alcohol withdrawal. I told the primary team to give her scheduled valium as she was clearly having delirium tremens. Her last drink was 48 hours prior when she had 40 ounces of hard liquor...
2PM rolled around and I was hoping to roll out of the hospital and have a little alcohol myself. Then I got a page from the MICU resident that the nurse witnessed a seizure from the patient. I said she needs ativan and scheduled valium. Ativan was given, but not scheduled valium. 20 minutes later I receive a page that the nurse was witnessing the patient constantly seizing. My attending physician was skeptical, as over half of the calls we receive for seizures are not seizures. I called the EEG tech and told her that we had an emergent need for continuous EEG monitoring. She was not thrilled as it was supposed to be a lazy Sunday for her as well, and now I was holding her up to start continuous EEG monitoring on a patient who had an unremarkable EEG just 1 day prior.
By the time I get to the patient's room, the patient was being intubated. She had developed a dangerous metabolic lactic acidosis. She was constantly having what appeared to be very real focal onset seizures with secondary generalzation with at least 10 witnessed seizures. This had been ongoing for 90 minutes. This meets criteria for status epilepticus. The definitive treatment for status epilepticus is to induce a medical coma with heavy sedation such as propofol, pentobarbitol, or versed to end the seizures. The patient was intubated and then connected to continuous EEG monitoring. The nurse was starting to get nervous. After the patient was intubated, I requested that we capture just 1 seizure on EEG, and then start the propofol drip. The patient had one last seizure, and then the propofol was started and she fell into a deep sleep.
The next day the patient's EEG was read out as negative for seizures. I questioned this read and the attending physician reviewed the tracing and agreed with me to change the impression as she clearly had a seizure right before the propofol was started. After 24 hours, her propofol was weaned off. She appeared to be purposefully following commands with her arms and blinking appropriately to command. The next day she was extubated, but her mental status worsened. She was nonresponsive, not moving her extremities, and in scientific terms, she was a vegetable. I requested an MRI brain to be ordered to see if she had suffered anoxic brain injury during her multiple seizures. The MRI was normal.
Days passed, then weeks. The patient was still not waking up. Her medicine intern asked me, why isn't she waking up? So did my attending. Simple, she was in convulsive status epilepticus for 90 minutes from alcohol withdrawal was my response. I wanted to say that if they had been more aggressive with scheduled benzodiazepines, perhaps this would have been preventable. I was bold enough to say that perhaps if the patient stayed home and continued to drink 40 ounces of liquor a day, she wouldn't be in this state.
A new younger neurology attending came on service. She was less satisfied with my simple explaination. She demanded that the patient receive a full metabolic and autoimmune workup. She was tested for hashimotos encephalitis and various autoimmune conditions such as lupus and sjogren's syndrome. This was negative. She had a whole body CT scan to assess for cancer; this was negative. She was tested for autoimmune paraneoplastic conditions such as NMDA receptor encephalitis. 3 weeks later, all her labs came back negative. An EEG was repeated a 3rd time; this was negative.
I cringed for a moment inside. Why couldn't my simple explaination that the patient has sufferred permanent bilateral cortical damage from prolonged alcohol withdrawal seizures be enough? Why did we have to spend thousands and thousands of dollars to search for a needle in a haystack diagnosis. What if this patient were my family member? I suppose I would also want every potentially treatable cause to be explored even if the propability is infinitessimally low that these potential diagnoses were going on.
Around hospital day 7 when the patient was still not waking up; I raised the possibility of "empiric treatment." Empiric treatment is a term used when we treat for a condition even if there is no established diagnosis. It's sometimes a shot in the dark. It is done all the time in patients who come in with sepsis who are covered with multiple broad spectrum antibiotics and antiviral medications until their blood/urine/spinal fluid cultures reveal a source. In the case of my patient, we would be empirically treating her with IV steroids for something such as hashimotos encephalitis. Another option would be empirically treating her with IV IG immunoglobulins to treat for an NMDA receptor encephalitis as the lab test for this takes 1 month to come back.
Our team decided not to start empiric treatment with anything. The probability for hashimotos encephalitis was low, and there was no underlying malignancy to suggest an autoimmune process such as NMDA receptor encephalitis. But nonetheless the labs were sent even though the suspicion for these conditions were low. So we waited for 3 weeks patiently for these labs to come back; and they came back negative.
During that 3 week period, I sat down with the family. I told them that the simplest explaination is that the patient suffered permanent brain damage from prolonged alcohol withdrawal seizures. She had shown minimal if any signs of improvement, and in my humble opinion there was no reason to expect her to recover further. I wanted to use the term persistent vegetative state as the patient had no purposeful movement of her arms, legs, or eyes, and she had preserved sleep wake cycles but I chose to avoid this controversial term. I explained that we were searching for very rare autoimmune conditions which can cause treatable/reversible comas, but that I expected the results to be negative. I told them that I did not expect the patient to show significant improvement, and that there are no further treatments indicated. The patient's large African American family was upset. They cried, they gave me angry looks as if I had taken something inconceivably valuable away from them, and then her son who was in his late twenties like myself gave me an angry look and questioned if we were doing everything in our ability. Some of the family members thanked me for my forthright explaination. I certainly don't know everything and I expressed that while this is the most likely explaination, there is always the possibility I am incorrect. I said it wouldn't surprise me if she were able to follow simple commands down the road, but there is no reason to expect she will make a significant recovery or return to the state she was in before. I've learned to give myself this out as people with end stage illnesses occasionally have unpredictable hospital courses and occasionally have inexplicable recoveries. It surprised me that this discussion had not happened sooner. I told them that I felt it is important to honestly convey our thought process even if the news is not good news so that they could know what to expect moving forward. I patted the son on the shoulder and left the room so I could finish my notes, go home and write more notes, and respond to other mundane messages from my clinic patients who wanted medication refills and more pain medication.
The patient never did recover, but later on I would re-examine her and she clearly did have purposeful eye movements which was a slight improvement. She would follow with her eyes when I called out her name. She was still unable to do anything else. She was being fed through a feeding tube in her stomach. She was able to breath on her own. My young and new attending had requested a 3rd brain MRI even though the previous 2 were normal. I was unhappy about this, but then the results showed something I didn't expect. There was damage to the brain through the caudate nuclei which can sometimes be responsible for arousal, and there was damage of the motor tracts through the ventral pons. The MRI was suggestive of a condition called central pontine myelinosis (as well as extrapontine myelinosis). The patient was "locked in." She probably could understand the people around her, but she was unable to move her arms and legs because of the damage to the motor tracts in her ventral pons. There is no real treatment for this condition. I rotated off service and someone else was left to discuss the ramifications of this with the family. I am still confused about how the patient developed central pontine myelinosis. Typically this condition is caused by rapid fluctuations of sodium but her sodium and electrolytes were mostly normal through her hospitalization. Why didn't her previous 2 MRIs reveal this diagnosis earlier?
This remains one of the many medical mysteries that I cannot completely explain.
But let me go back to the principle of empiric treatment. Wouldn't it have been easier to just treat the patient with IV steroids at the beginning? Why waste thousands of dollars in expensive labs and hospital days waiting for labs to come back. Why not just give 5 days of IV steroids which are generally pretty safe? If that didn't work, couldn't we empirically treat with IV IG for an autoimmune receptor encephalitis? Again 5 days of IV IG isn't necessarily cheap and there are some side effects but it is usually tolerated pretty well. I guess there was never really a reason to suspect the patient had hashimotos encephalitis or NMDA receptor encephalitis. So why even test for it? These questions are difficult, and again in retrospect I think if this patient just stayed home and continued to drink, she would be in better shape today and she would not have racked up what might be a half million dollars in hospital bills. The questions I am posing in this blog entry are difficult, and would be approached differently by different people, and at the end of the day there may not be a clear right answer.
