A Tox Talks Mystery: The Toxic Woman of Riverside

I’ve always enjoyed a good mystery. As a kid I tore voraciously through mystery novels and have watched all the way through House MD several times. But I’ve got nothing on my grandma, who not only has had a several decade head start on watching murder mystery shows, but shows no sign of slowing down. From Poirrot to Midsomer Murders, she’s seen nearly every mystery show that’s ever come out of the US or UK, and read nearly as many novels. I’ve joined her to watch a show or two during this wonderful period of fourth year where I’ve had some free time, which put me in the mood to share a medical mystery with you. So today, Tox Talks proudly presents The Toxic Woman of Riverside.

We open with a slow fade in of a cityscape. The words “Riverside, California” appear on screen. It’s raining, naturally, a brisk February evening. Cut to the inside of the Riverside General Hospital emergency department. Monitors beep, interspersed with the sounds of feet shuffling as nurses draw labs and doctors converse with their patients. “The Sign” by Ace of Base can be heard faintly in the background, as this scene takes place in 1994. Offscreen is heard the sound of doors opening as EMS wheels in a gurney. On it lies a woman gasping for breath, her name is Gloria Ramirez. She is just 31 years old, and has terminal metastatic cervical cancer. 

The emergency team springs into action. The patient is tachycardic with shallow breaths and a decreased level of consciousness. She is given benzodiazepines through the IV placed by the EMS crew. Her condition worsens and the team begins full-code CPR. A respiratory therapist oversees ventilation, and a nurse starts a second IV. After drawing blood, the nurse notes an ammonia-like odor coming from the patient and white crystals in the blood sample. She passes the blood sample to one of the medical residents at bedside, who also notes the crystals and unusual odor. At this point the nurse faints. Soon afterwards the resident also loses consciousness, and begins convulsing on the floor. The respiratory therapist and several others nearby begin to feel unwell, and the emergency medicine department chair orders the department to be evacuated.

The chair, Dr. Humberto Ochoa, and several others continued to care for Ms. Ramirez, administering several different medications and defibrillating the patient multiple times. Unfortunately the patient did not respond, and Gloria Ramirez was pronounced dead at 8:50 PM. A total of 23 people involved in her care became ill and six were hospitalized.The resident was admitted to the ICU and suffered hepatitis, pancreatitis, and knee osteonecrosis, but survived. There were no fatalities besides Ms. Ramirez.

Ultimately, the cause of death was determined to be cardiac dysrhythmia secondary to kidney failure. A HAZMAT team secured the scene and tested for common hazardous gasses; none were detected. An investigation of the hospital’s ventilation and plumbing systems, as well as the equipment and medications used in the emergency department revealed no obvious cause for the staff’s ailment. A toxicology panel was negative for many of the usual suspects, including opioids, lidocaine, or pesticides including carbamates and organophosphates; it was also negative for the number of aerosolized compounds. The only substance found at elevated levels (that was not administered to her by medical personnel) was acetone. What caused the mysterious illness that affected nearly two dozen healthcare workers? Was there more to the tragic death of this young woman? The forensic investigation failed to turn up any convincing answers, and samples of several organs procured from the autopsy were sent to outside labs for further analysis. But in the meantime, it was publicly announced by the California department of health services to be a case of sociogenic illness – mass hysteria.

Although the trail had run cold for the time being, there were a few details about the patient’s presentation that didn’t seem to add up. Firstly, there was the matter of the crystals in the patient’s blood sample, and the ammonia-like smell that the team noticed. When they removed the patient’s clothes for defibrillation, several people reported seeing an oily sheen on the patient’s skin, and reported a “fruity-garlicky” odor on her breath. She was not receiving chemotherapy or radiation treatment for her cancer, ruling out any possible metabolite from one of those treatments. But what if she had taken something else?

In this and many other cases, the toxicologist’s tool of choice is gas chromatography/mass spectroscopy. You may remember GCMS from organic chemistry, and if the CT scan is the Donut of Truth, the GCMS is the Panel of Veracity. For toxicology departments that are lucky enough to have the capabilities, GCMS is a very valuable diagnostic tool and can identify drastically more compounds than a typical tox panel. Ms. Ramirez’s GCMS below revealed a prominent spike correlating with dimethyl sulfone (DMSO₂), a metabolite of dimethyl sulfoxide, or DMSO.

DMSO is a clear, odorless liquid produced as a byproduct of the paper industry, industrially used as a solvent, reaction medium, and antifreeze. It has a range of potential pharmacological properties including analgesia and anti-inflammation; it currently has FDA approval to treat interstitial cystitis but has been used to treat scleroderma, osteoarthritis, rheumatoid arthritis, and amyloidosis. It also interestingly has the ability to easily penetrate organic membranes without damaging them, and it can enhance the absorption of other compounds. The possibility remains that Gloria Ramirez could have taken DMSO as a home remedy to treat the pain from her cancer, which may have been oxidized to dimethyl sulfone, and then again to the much more toxic dimethyl sulfate (DMSO₄). 

Dimethyl sulfate has been reported to cause dizziness, bradycardia, headache, eye irritation, blistering of the skin, severe pulmonary damage, cerebral edema, and a slew of other symptoms, several of which lined up with the symptoms experienced by the patient and healthcare workers. Grant et al from the Forensic Science Centre (FSC) at Livermore National Laboratory, after reanalysis of all available data seven months after the incident, theorized that when Gloria Ramirez received supplemental oxygen in the ambulance, the oxidation reactions were expedited and DMSO₂ was produced from the DMSO that she had been taking. The concentrated DMSO₂ precipitated in the blood draw syringe, and some reacted to form DMSO₄ which then poisoned the healthcare workers. And that’s it, case closed, bake ’em away, toys.

Except…some things still don’t hold water. A letter to the editor by de la Torre raises some excellent points which led me to “press X to doubt” the DMSO theory. To my eyes, the most salient points are as follows:

1. There was no mention of when and how much DMSO Ms. Ramirez had taken. She was noted to have been in renal failure, but no DMSO₂ was found in her kidneys, where it would have accumulated in high concentrations. Additionally, the toxic substance in question – dimethyl sulfate – was entirely absent from the GCMS analysis.
2. Dimethyl sulfate is known to cause pulmonary, renal, cardiac, and liver damage, yet there was no mention of autopsy findings in the report by Grant et al. It seems unlikely that the same toxin that could seriously harm multiple people in the same room would be completely missing from the patient that was ostensibly the source.
3. There is no explanation given why some, but not all, of the health care workers became ill. Grant et al hypothesize that the oxidation reactions which were kickstarted in the ambulance took too long to expose the paramedics to any dimethyl sulfate (despite the paramedics obtaining IV access and spilling a small amount of blood), but this too seems less than convincing.
4. The original report compares the ED staff’s symptoms to the known symptoms of dimethyl sulfate poisoning. Reviewing the comparison table reveals that some pairs range from questionable (eye irritation vs tear gas-like lacrimation) to dubious (flushing of skin vs severe blistering and necrosis of skin). All healthcare workers involved reported nausea and vomiting, symptoms which rarely if ever occur with dimethyl sulfate poisoning. And notably, the ED staff reported symptoms immediately after contact with Ms. Ramirez, despite the fact that most symptoms should have been delayed for approximately six hours or more. 
5. DMSO₄ rapidly hydrolyzes to methanol and sulfate at temperatures above 18°C/64°F. The explanation of the conversion from DMSO₂ to DMSO₄ in this case hinges on the fact that the blood drawn in the syringe would drop to a low enough temperature to prevent hydrolysis of DMSO₄, and the original authors admit that their mechanisms for the DMSO₂ to DMSO₄ conversion are hypothetical and require “extraordinary oxidizing conditions.” Given that DMSO has been used by a great number of patients (surely including some that were also quite sick and receiving supplemental oxygen), the fact that this has never been reported since casts further doubt on this theory of a novel, spontaneous chemical reaction.

So here we reach the end of the story. A few theories still remain plausible, such as the ammonia-like odor possibly emanating from breakdown products of acetaminophen, which was indeed found on Gloria Ramirez’s GCMS analysis. Contostavlos and Lichtenwalner hypothesize that the toxic agent was actually chloramine produced from the mixing of bleach and urea in a drain, which would have transiently released noxious fumes into the area while washing down the drain and being therefore untraceable. We may truly never know the answer to this mystery. Ultimately, this is a tragic tale of a young woman’s death, a number of healthcare workers suffering workplace-related illnesses, and a media circus that muddied the waters and prevented the patients’ families from being at peace during a very difficult time. It is also a lesson that you truly never know who or what might come into the emergency department, so maybe a bit of PPE won’t hurt. 

John Michael Sherman

+ posts

John Michael Sherman is a member of the UA College of Medicine - Phoenix Class of 2023. Born in Chicago, IL, he's been an Arizona resident for 15 years but is still heat-intolerant. He received a degree in Jazz Trumpet Performance from ASU and was a professional freelance musician for several years before starting medical school. He enjoys exploring new recipes in the kitchen, swing dancing with his wife, and (for some reason) riding his bike for long distances.

• The State of Emergency
• Tox Talks: Arizona Bark Scorpion
• Tox Talks: Acetaminophen Overdose