CAP TODAY Pathology/Laboratory Medicine/Laboratory Management
‘It is incumbent on oncologists to obtain autopsies’
Here is part of what George Lundberg, MD, chief medical officer and editor in chief, CollabRx, and editor in chief of the “Curious Dr. George” blog on the CollabRx website, posted April 19 about (and from) Dr. Richard Horowitz:
American pathology lost one of its greatest leaders (and I lost one of my best friends) when Richard Horowitz died on March 15, 2017 in Los Angeles, Calif. Still of sharp mind and keen humor, he died with dignity and grace, in a manner of his own choice during home hospice care, of metastatic non-small-cell adenocarcinoma of the lung.
Richard and I met across an autopsy table at the old LA County General Hospital in summer 1967. We bonded and remained colleagues and friends who shared many professional beliefs based on personal experiences for 50 years.
Richard was born in Vienna, Austria, on May 17, 1931. He left Austria with his parents to flee Hitler’s scourge in 1939.
Richard’s total course of illness after initial diagnosis (malignant pleural effusion discovered at a routine annual medical checkup with established widespread metastases) was nine months. He tried “precision oncology”; his cancer was found to harbor an EGFR mutation, so he was begun on erlotinib. He experienced adverse effects of such severity that he decided to decline further “curative” therapy of any sort and quickly moved into palliative home-hospice care.
His final (of many) contributions to the CollabRx discussion group posted this year on Feb. 25 [in response to a Q&A on developments in precision medicine for treating cancer]. It reads:
You are making pronouncements and decisions based on insufficient knowledge. Until the use of autopsies becomes the standard of whether the new therapy worked or how the new therapy’s side effects caused the death, we do not have adequate data. I previously sent the following:
1. A letter sent to the Wall Street Journal (published on Sept. 22, 2016): “ . . . The autopsy is a credible outcome measure; nothing else can attest as convincingly to the accuracy of a diagnosis or the efficacy of a therapy. Few, if any, clinical trials utilize the autopsy to test their hypotheses. . . .”
2. A short composite of the many autopsies I have personally done: The patient has stage IV lung cancer; all standard therapy has failed. The patient is coerced into treatment, first with targeted therapy and later with immunotherapy. Soon he experiences diarrhea—the oncologist “handles” that with loperamide, which results in annoying constipation. Then after a few days, there is marked increase in dyspnea—is it a progression of the disease, perhaps carcinomatous pneumonia or therapy-related (autoimmune) interstitial pneumonitis? Well, that can certainly be treated with steroids. Oh, the oncologist forgot to tell the patient that he needs CNS radiation because of brain metastases. So the patient is given a course of radiation therapy—unfortunately, there is significant cerebral edema. Again the oncologist ameliorates that with steroids; however, the cognitive impairment and confusion persist. About the same time there are cardiac arrhythmias—are they due to metastases to the heart or due to “autoimmune” myocarditis? No worry, add more steroids. Regrettably, a mixed bacterial and fungal pneumonia develops and that, of course, is treated with powerful antibiotics. Within a brief period of time another bout of diarrhea, this time due to C. difficile, develops and progresses into a dire megacolon that appears about to perforate. The patient is taken to surgery; the colon has, in fact, perforated and a segment is resected. In the surgical ICU the early signs of sepsis appear; soon septic shock ensues and the patient dies after prolonged intensive, but futile, care. The surgeon requests an autopsy, the oncologist does not attend the autopsy and does not answer the call when informed of the cause of death.
In September 2015 the National Academy of Sciences/IOM released its report “Improving Diagnosis in Health Care.” The report listed eight goals and multiple recommendations. Goal 4 was to develop and deploy approaches to identify, learn from, and reduce diagnostic errors and near misses in clinical practice, and recommendation 4C was that HHS should provide funding . . . to conduct routine postmortem examinations on a representative sample of patient deaths. It is incumbent on oncologists to obtain autopsies—then they will know if their “magic bullet” worked or killed. —R. E. Horowitz, MD