A previously healthy 41-year-old professional arrives by ambulance at a hospital, in inexplicably dire health. Within a week, she has died, her organs have been harvested and the remains have been cremated. There's no public indication of an investigation -- or of her apparent cause of death, cyanide -- until a week later.
Inexplicable? Not really, according to longtime forensic pathologists, including two who are involved in the case of Autumn M. Klein, the UPMC and University of Pittsburgh School of Medicine researcher whose April 20 death has spurred an investigation. An unusual -- albeit notorious -- poison like cyanide can escape detection until autopsy, and even sometimes thereafter.
"If you have reason to believe ... that drugs may have been involved [in a mysterious failure of health], then you would move toward a drug screening," said Cyril Wecht, the former Allegheny County coroner. "But I do not believe that you would be doing a cyanide test. I don't believe that is going to be ordered, that anybody is going to be thinking about cyanide."
Dr. Wecht said he has been hired by Jerry Johnson, the attorney representing Robert J. Ferrante, Dr. Klein's husband. Neither Mr. Johnson nor Mr. Ferrante could be reached Saturday.
Barring the telltale, but not universal, signs of the smell of almonds and reddened skin, would a doctor likely order a cyanide test on a struggling patient?
"Not off the bat, probably not," said Sanford H. Edberg, also a former county coroner and a former medical examiner for New York City. "If everything else is negative, then you usually start thinking of more esoteric [possibilities], like cyanide."
Dr. Klein was the chief of her division at UPMC, a leading researcher into the treatment of neurological problems during pregnancy, the wife of a professor of neurological surgery and the mother of a 6-year-old daughter. The manner of her death hasn't been determined, and county Medical Examiner Karl Williams said Saturday that he is awaiting toxicology reports.
Pittsburgh police confirmed Wednesday that toxic levels of cyanide were found in her system. Officers, with the FBI's assistance, searched the family's Oakland home Friday evening, removing a computer and vacuum cleaners, plus towing cars. The bureau had no update Saturday.
The case has become a national news story and has spurred speculation about why investigators seemed to be scrambling to assemble evidence nearly two weeks after her death.
Dr. Wecht said that when a patient comes into a hospital with an unexplained, apparently life-threatening condition, doctors draw blood, conduct a set of standard tests, then move on to cardiac enzymes, electrolytes, alcohol and drug tests, then finally diagnostics focused on the kidneys and liver as they try to rule out causes.
Dr. Edberg said when there's reason for suspicion, there "are various panels you do for poisons, especially poisons that cause unconsciousness," like alcohol, cocaine and heroin. He said he has seen just "one or two" cyanide poisonings in his career.
As a patient clung to life, as Dr. Klein did throughout April 18 and 19, the likelihood of cyanide being the culprit would wane, added Dr. Edberg, now a forensic pathology consultant. "I'm really puzzled why she lived that long with cyanide in her blood," he said.
Would doctors call the police if they found toxins or were just flummoxed?
"Leaving out trauma, and even if you found drugs, you don't notify law enforcement in my opinion," Dr. Wecht said. "If you find a toxic substance, you might report that to the police."
Dr. Edberg countered that police can be brought in prior to death.
"If you start thinking about poisons, then you would call in law enforcement," he said. "You'd want to know what's contained on the scene" where the ailment was first noted.
A UPMC spokeswoman declined to say when or whether doctors at UPMC Presbyterian conducted a cyanide test, citing the ongoing investigation.
Dr. Williams declined to talk about the specifics of the case, but said it would be very unusual for his office -- or for police -- to become involved prior to death, absent obvious indications of criminality. "Cases that are clearly homicide, clearly accidents, clearly suicide, have to come to my office, and that's the bulk of my cases," he said.
The medical examiner's office does have to sign off prior to cremation. Dr. Klein was cremated some time before April 24, and around a week before the investigation became public.
Dr. Williams said that the office doesn't stop a cremation unless it has a reason to. He would not discuss when or how that sign-off occurred in Dr. Klein's case.
By the time Dr. Klein was cremated, the medical examiner would have had the samples it needed to determine the cause of death.
"Routinely, you draw blood, bile and urine," said Dr. Wecht, and send those to a toxicology lab. The examiner will also take tissue samples from organs including the liver, and sometimes vitreous humor from the eyes, spinal fluid and other fluids, he said.
There are cases where a blood sample can either show a false reading for cyanide poisoning or fail to show an actual case of poisoning, experts said.
John Trestrail, a criminal poisoning expert, said when our bodies encounter natural traces of cyanide, as in cigarette smoke, an enzyme converts it into a safe compound called thiocyanate. If a postmortem blood sample sits around for too long, he said, some of that thiocyanate can degrade into cyanide, making it appear there was cyanide poisoning when there wasn't.
On the other hand, said Texas Tech University toxicology expert Stephen Borron, if poisoning blood samples aren't refrigerated and have too much air in the top of the tube, cyanide can evaporate into the airspace and leak out of the tube when it is opened, making a poisoned blood sample look clean.
The fact that the deceased was cremated might affect investigators' ability to probe further, said Dr. Edberg. "You might want to go back and do further tests on the body," he said. "You might want to test the hair samples for instance. You might want to test the bone."
Rich Lord: firstname.lastname@example.org, 412-263-1542 or Twitter @richelord. Staff writer Mark Roth contributed.