How an Antipsychotic Became Widely Used as an Antiemetic

But Was Never Licensed

Helen Leask

March 06, 2020

It's a fall day on Manhattan's East Side in 1998. Inside the Memorial Sloan Kettering Cancer Center, a large woman lies in a darkened room that's cluttered with stuffed toys and oversize photographs of her children.

A week ago, this 41-year-old freelance writer started chemotherapy for acute lymphocytic leukemia, and she is prostrate with nausea and vomiting. Her physicians have tried dexamethasone, prochlorperazine, metoclopramide, lorazepam, cisapride, and ondansetron, but she continues to vomit.

Her daily drug cocktail is now fluoxetine, ondansetron, lorazepam, granulocyte colony-stimulating factor, antibiotics, ranitidine, and docusate. She is becoming increasingly agitated and enraged, crying and constantly calling out for her nurses.

Psychiatry is called.

That day, attending psychiatrist Andy Roth, MD, was accompanied by clinical fellow Bill Pirl, MD. The clinical situation they faced was complicated by the patient's history of depression, anxiety, and borderline personality disorder.

The two clinicians began by increasing the patient's fluoxetine dosage from 40 to 60 mg qd. For 3 days, the patient remained agitated and labile and continued to vomit.

They then decided to try olanzapine (multiple brands).

Olanzapine had been approved 2 years earlier (1996) by the US Food and Drug Administration (FDA) for the treatment of schizophrenia and related psychoses.

Psychiatrists were very interested in the new drug. It was one of the first so-called second-generation, or atypical, antipsychotic. The new drug class promised relief of psychosis with less risk of the terrifying extrapyramidal symptoms (EPS) seen with first-generation drugs such as haloperidol.

This interest was turning into commercial success. By 1998, olanzepine would earn the company $1.44 billion in worldwide sales. It was shaping up to be the new Eli Lilly blockbuster, stepping into the shoes of its previous top seller, Prozac (fluoxetine), which was soon to lose patent protection.

Olanzapine had certainly caught the attention of the psychiatry team at Memorial Sloan Kettering. It led the world in the psychiatric support of cancer patients, and the medical discipline of psycho-oncology was born there, created by the legendary psychiatrist Jimmie Holland, MD, a decade earlier. With that background, Pirl and Roth were well equipped to help an oncology patient with psychiatric symptoms.

They prescribed olanzapine 5 mg at bedtime.

Then something happened that they didn't expect. By the next day, all symptoms of nausea and vomiting had vanished.

Pirl recalled the exciting moment in an email interview: "I had prescribed olanzapine for a hospitalized patient with cancer for mood stabilization but noticed that it also had a dramatic effect on her nausea and vomiting."

At the time, Holland was editor of the journal Psycho-Oncology, and she encouraged Pirl and Roth to report their discovery. By December 1998, the case report had been submitted to the journal.

"Remission of Chemotherapy-Induced Emesis With Concurrent Olanzapine Treatment: A Case Report," published in the February 10, 2000, issue, presented the first such finding.

At the time, the Sloan Kettering psychiatrists attributed olanzapine's antiemetic effects to its dual blockade of the dopamine D2 and serotonin 5-HT3 receptors.

First-generation antipsychotics were known to be effective antiemetics through dopamine D2 blockade, but their extrapyramidal symptoms made them a choice of last resort. One of the most powerful known antiemetics, ondansetron, targeted serotonin 5-HT3 receptors in the gut.

Olanzapine, the authors speculated in the case report, "may not only have some theoretical efficacy advantage over typical antipsychotics for nausea by blocking both dopamine and 5-HT3 receptors, but it is also less likely to cause extrapyramidal symptoms."

Word of the case report reached Steve Passik, PhD, a clinical psychologist who had interned at Sloan-Kettering under Roth and had then spent 10 years working there, so he had strong connections to Holland's team in New York City.

Now a psychologist in the Indiana cancer system, Passik was not surprised to hear of olanzapine's success for the vomiting New York patient. Working on a parallel track to the Sloan Kettering team, he had also become convinced that olanzapine had huge potential for relieving the misery of cancer patients undergoing chemotherapy.

He recalled his experiences working in partnership with psychiatry colleagues in the late '90s. "One of the beauties of [olanzapine] sometimes we use it for confusional states, for anxiety states, but also it treats nausea and stimulates appetite a bit. When you think how sick some of these folks are and how many medicines they are taking...the fact that you can sometimes give them some benefit for three, four, five symptoms in a cluster without giving them five medicines — the psychiatrists I worked with really found it useful."

Around the time the case study was published, Passik had started work on a pilot study of olanzapine with colleagues at Community Cancer Care, Indianapolis, Indiana, and the Markey Cancer Center, Lexington, Kentucky.

Passik said that he had talked to Eli Lilly "around 1998" about the potential of the use of olanzapine in cancer. The company agreed to support the pilot trial, which was published in June 2002 in the Journal of Pain and Symptom Management.

The open-label pilot study, the first formal test of olanzapine as an antiemetic, concerned itself with nausea in advanced cancer, not treatment-related nausea. The researchers decided not to challenge the existing gold-standard treatment for acute chemotherpy-induced nausea and vomiting (CINV) until they had stronger evidence that olanzapine was, indeed, effective against cancer-related nausea without causing significant EPS.

In that pilot study, 15 patients with pain and nausea from advanced cancer (most commonly, breast, lung, or ovarian cancers) each received escalating treatment for 8 days: 2 days each with placebo and olanzapine 2.5 mg, 5.0 mg, and 10 mg. Rescue medication of additional doses of olanzapine 2.5 mg was allowed. All the patients were white, and 11 of 15 were women.

The researchers assessed olanzapine's impact on the patients' nausea using the nausea item of the Functional Assessment of Cancer Treatment ― General Section (FACT-G). Patients could score their nausea from 0 ("not at all"), through 1 ("a little bit"), 2 ("somewhat"), 3 ("quite a bit") to 4 ("very much so").

Despite the small number of patients involved, the results of Passik's pilot study were convincing. The patients' nausea significantly improved with all the doses of olanzapine. At baseline, half the patients rated their nausea as "quite a bit" or "very much." With the 5-mg dose, only one patient had "quite a bit" of nausea, and none had "very much" nausea (2.5 mg, P < .04; 5 mg and 10 mg, P < .0001).

The next step was a clinical trial.

Passik believed that olanzapine had most potential for delayed emesis ― nausea and vomiting that occurs within 24 to 48 hours of chemotherapy. Published studies indicated rates of 50% to 70%, and delayed emesis was inadequately addressed by the drugs then in use. It represented a huge, unmet need in cancer. Many patients were unwilling or unable to endure the next round of nausea and vomiting, and they stopped receiving effective treatment too soon.

Passik and the pilot-study team laid the groundwork for a delayed-emesis trial by conducting a retrospective chart review. The team utilized the substantial experience they had by then gained with their patients in the Indiana cancer system.

The chart review, supported in part by Eli Lilly, was published in May 2003 in the Journal of Pain and Symptom Management. During 95 cycles of chemotherapy with olanzapine, there were only 21 instances of nausea (22.1%) and 10 instances of vomiting (10.5%). Side effects were "rarely noted" ― 7.1% of patients experienced sedation, and 3.6% experienced edema, and there was no restlessness.

Believing that had taken the clinical science as far as his expertise would go, Passik contacted Rudi Navari, MD, an expert in CINV who had been writing on the topic since 1989.

Fortuitously, Navari had recently joined the faculty of the University of Notre Dame as director of the Walther Cancer Research Center and was soon appointed director of the Indiana University School of Medicine–South Bend.

Navari was able to marshal unprecedented academic and clinical resources across Indiana to explore the potential of olanzapine for delayed emesis.

"He was just blown away," recalled Passik. Unlike with typical antipsychotics, which often rendered patients "flattened" and nervous, Passik said, "Patients really wanted to take [olanzapine], it made them feel better, and we saw almost no akathisia at these low doses."

Navari recalled being approached by Passik at the start of the formal clinical trials program of olanzapine for CINV.

"Passik and colleagues in 2003 found that olanzapine had antinausea properties and approached me about using it for prevention of chemotherapy-induced nausea and vomiting. We did a phase 1 trial and found that 10 mg/day for 4 days was safe and effective," Navari said.

That phase 1 trial brought together the clinical team that would take the research forward, anchored by the Hoosier Oncology Group of the Walther Cancer Institute, a clinical trials collaborative.

The phase 1 trial, which was similar in structure and had similar findings to Passik's pilot study, was published in August 2004 in Cancer Investigation with Passik as first author. The researchers concluded that the maximum tolerated dose was 10 mg.

There was no stopping olanzapine's momentum after that. As the phase 1 trial closed, the phase 2 trial began. Results of the phase 2 study were submitted to Supportive Care in Cancer in May 2004. Navari was first author. The study was was published in February 2005.

The phase 2 trial involved 30 patients who were scheduled to receive either moderately emetogenic chemotherapy (carboplatin, irinotecan, cyclophosphamide, doxorubicin) or highly emetogenic chemotherapy (cisplatin). The patients received olanzapine 5 mg for 2 days prior to each cycle of chemotherapy. On the day of chemotherapy, they received 10 mg, and this dose was continued for 2 to 4 days. All patients also received a standard antiemetic regimen of dexamethasone and granisetron. Patients underwent a maximum of six cycles of chemo, with the number of cycles determined by the investigator.

The authors concluded that "olanzapine was very effective in controlling acute and delayed [CINV]." The complete response rates for emesis with the olanzapine regimen were 100% in the acute phase, 80% in the delayed phase, and 85% in the moderately emetogenic phase. For the highly emetogenic regimen, nausea control was 100% at all stages. For the moderately emetogenic regimen, nausea control was 85% in the acute phase and 65% in the delayed phase.

The authors commented that response rates and nausea control for olanzapine were "equal to or greater than" published data for aprepitant and palonosetron. At these doses of olanzapine, the investigators did not observe "significant" sedation, weight gain, or induction of hyperglycemia.

Olanzapine 5 or 10 mg was shaping up to be a serious contender in comparison with the currently available antiemetics. It was apparently effective for both acute and delayed vomiting and nausea, with acceptable side effects. It was also, according to Navari, only 10% to 20% the price of the first-generation 5-HT3 receptor antagonists ondansetron, granisetron, and dolasetron.

Company Was Not Involved in New Indication

All this work was taking place right in Eli Lilly's back yard in Indianapolis. Was the company interested in applying for a new indication for their blockbuster antipsychotic?

Navari recalls talking to Eli Lilly about the potential for olanzapine in CINV, but he said the company showed no interest in either supporting further clinical development in CINV or applying for a new indication.

"We approached them 2005, 2006, 2007, when it was pretty apparent from our phase 1 and phase 2 clinical trials that olanzapine was a very effective drug for the prevention of chemotherapy-induced nausea and vomiting," said Navari.

"I think it was a huge mistake for them not to go for another indication, but that was their decision," Navari said.

Medscape Medical News endeavored to speak to someone who had been at the company at the time about this issue, but without success.

Eli Lilly's annual reports show that from 2003 to 2007, the company's scientific executives were Steven Paul, MD, head of Lilly Research Labs and executive vice-president of science and technology; Alan Breier, MD, chief medical officer; and Thomas Verhoeven, PhD, vice-president of product research development. Bill Chin, MD, joined the company as vice-president of discovery research and clinical investigation in 2005.

Paul remains an expert in central nervous system (CNS) drug development and is currently CEO of Karuna Therapeutics in Boston, Massachusetts. His biographical note on the Karuna website states, "At Lilly he was responsible for the company's overall research and development efforts and helped oversee the development of CNS drugs such as Zyprexa."

Medscape Medical News requested an interview with Paul through his Karuna office. He responded through public relations representative Jenn Gordon: "Despite a deep respect for Medscape and desire to help out, Dr Paul needs to decline this interview opportunity. He was involved in the development of Zyprexa for psychosis / schizophrenia, not CINV, so does not feel he is the right person for you to speak to for this particular story."

Medcape Medical News also approached Breier, who holds several senior positions at Indiana University School of Medicine, including vice-chair for clinical research, and Verhoeven, currently President's Fellow for pharmaceuticals development and partnerships, Purdue University, West Lafayette, Indiana. Neither responded to repeated requests for an interview.

Chin, now professor emeritus of Harvard Medical School and executive vice-president of frequency therapeutics, responded by email: "While I was head of Research at Lilly for many years, the ones involved in the development of olanzapine were not among them. So I'm afraid I can't help."

Pharmaceutical companies are, of course, under no obligation to develop an existing product for a new indication, however worthy it may seem.

According to a 2013 Forbes analysis of 95 companies, drug development costs average $350 million per drug and rise to several billion dollars. Research and development may or may not lead to a successful product, and costs must be recouped before the patent expires.

In 2005, when Navari first approached Eli Lilly, the US patent on olanzapine was due to expire in 2011, so there were 6 years of patent life left. What was the likely return on investment in developing an application for cancer, a new therapeutic area for Eli Lilly?

Without information from Eli Lilly decision makers, one can only speculate.

According to Navari, one of the advantages of olanzapine for antiemesis was that it cost only 10% to 20% the cost of the new 5-HT3 antagonists. At the time, each of these drugs was each generating annual sales of $1 billion or more.

A back-of-an-envelope calculation suggests that olanzapine for CINV would have yielded at least $100 million per year for the company after clinical development and FDA approval ― assuming that any patent life remained.

By contrast, in 2005 alone, Zyprexa generated $4.2 billion in worldwide sales as an antipsychotic and had 6 years of patent life left.

Passik, speculating on the reason for the lack of support by Eli Lilly, said, "[The company was] trying to do something revolutionary in mental health.... Maybe the feeling was, we've got bigger fish to fry and the nausea thing was small potatoes...."

Navari suggested that in addition, the timing might have been poor. When he first approached the company about the new indication in 2005, Eli Lilly was in the process of setting aside $700 million to settle lawsuits against the drug.

"We talked to [the company] about getting a new indication, but at the time, they were getting a lot of patient lawsuits due to olanzapine causing hyperglycemia and the onset of type 2 diabetes when you use it in very high doses, like 30, 40 mg a day for 6 months," Navari commented. "We, of course, used much lower doses for short periods of time, so we didn't get that. But I think because at the time we went to them they were getting these lawsuits, they were less interested in using olanzapine for another indication."

Other Sources of Funding

Whatever the reason for Eli Lilly's decision, once it was clear that the company had no interest in pursuing the CINV indication, Navari and colleagues turned to other sources of funding to continue their work.

The first phase 3 trial of the use of olanzapine in CINV was financed by the Walther Cancer Foundation and the Reich Endowment for the Care of the Whole Patient. In the trial, olanzapine went head to head with the current guideline gold standard: dexamethasone, palonosetron, and aprepitant during and after chemo.

This phase 3 study, published in Supportive Oncology in 2011, showed that a regimen consisting of single doses of dexamethasone and palonosetron prior to chemo and olanzapine alone during and after chemo was as effective as the gold standard for acute and delayed vomiting. Importantly, it was significantly more effective for delayed nausea.

This key advantage for olanzapine continues to this day.

Alex Molasiotis, RN, PhD, chair of the antiemetics working group of Multinational Association of Supportive Care in Cancer (MASCC), commented to Medscape Medical News in an interview: "With the new antiemetics over the last decade...we have managed quite well with the vomiting. What we haven't really done well with is the nausea. This is a much more difficult issue for patients because with vomiting, it comes out and the patient feels instant relief, but to feel constantly nauseous.... It really has major impact on them, and with that, we haven't done well. Olanzapine helps them with the nausea much better than the standard emetics that we use."

In 2013, a decade after he started work on olanzapine for CINV, Navari and colleagues were rewarded for their determination: a $2.1 million grant from the National Cancer Institute for a national phase 3 pivotal trial.

The results of this pivotal trial were published in the New England Journal of Medicine in July 2016 and secured a place for olanzapine in the antiemetic guidelines from several large organizations, including the American Society of Clinical Oncology (ASCO), the National Comprehensive Cancer Network, and MASCC.

This not only vindicated the team's long support for olanzapine for CINV but also finally provided a legal safety net for physicians using the drug off label to help their cancer patients. "Since it's been incorporated into the guidelines as an off-label use, physicians don't have any issue in terms of using it, and neither does the FDA," Navari explained.

In the ASCO antiemetics guidelines, olanzapine is recommended for high-emetic-risk chemotherapy in combination with an NK1-receptor antagonist, a 5-HT3 receptor antagonist, and dexamethasone, with continued use of olanzapine on days 2 through 4.

Repurposing Drugs in Oncology (ReDO)

Pan Pantziarka, PhD, is familiar with the challenges faced by researchers who are enthusiastic about repurposing a drug in the face of lackluster enthusiasm from the manufacturer.

Pantziarka heads up the Repurposing Drugs in Oncology (ReDO) project, an international collaboration between US-based GlobalCures and Belgium's Anticancer Fund that seeks new indications for existing medications. ReDO backs up researchers' hunches by providing grants of €50,000 to €I million to conduct research into drugs that have lost commercial backing.

ReDO most commonly steps in to support drugs that have lost their patent. "Olanzapine is a bit of a strange case," Pantziarka said.

The story gets stranger. At the heart of ReDO is a curated open-access database of 310 noncancer drugs that could work against cancer ( According to the ReDO database, olanzapine has anticancer properties.

"It crosses the blood-brain barrier, which makes it interesting for brain tumors," Pantziarka said. "So it's a drug that could help patients with their nausea and vomiting, but it could potentially also have some positive effects on their cancer.... Twenty years on, it's just tragic that it's not approved."

Molasiotis, Navari, Pantziarka, Passik, Pirl, and Roth have disclosed no relevant financial relationships.

For more from Medscape Oncology, follow us on Twitter: @MedscapeOnc.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.