30% CANCER DRUGS ARE USED OFF-LABEL 

About one third of intravenous chemotherapy is prescribed off-label to cancer patients in the United States, according to a study published online February 19 in the Journal of Clinical Oncology.

The study authors analyzed the use of 10 common chemotherapies in 35 states and found that 70% of use conforms to the approved drug labeling. This rate of on-label use is more frequent than "previous oncology-specific estimates suggest," write the authors, led by Rena Conti, MD, from the University of Chicago in Illinois.

Factors other than endorsement by the US Food and Drug Administration (FDA) guide overall chemotherapy use, Dr. Conti said in an email to Medscape Medical News.

"Our main interest...was understanding how much use and spending is driven by FDA indication and how much is driven by expert opinion," she said. "It is clear from our results that expert opinion and physician judgment accounts for a sizable percentage of these drugs' use."

The finding that only 30% of the prescribing was off-label is "reassuring," writes Monika Krzyzanowska, MD, from University of Toronto in Ontario, Canada, in an accompanying editorial.

But the study also turned up some disconcerting findings about the high rate of off-label prescribing of some agents, especially gemcitabine (Gemzar, Lilly), rituximab (Rituxan, Genentech), and bevacizumab (Avastin, Genentech), Dr. Krzyzanowska notes.

This study provides a "benchmark" for measuring the use and related cost of the off-label prescribing of "expensive cancer drugs," she explains. It is "the most comprehensive...contemporaneous snapshot of off-label prescribing in the United States," she writes.

Dr. Conti and her colleagues limited their study to 10 patent-protected chemotherapies because of their "high price" and "uncertain clinical risks." Only intravenous chemotherapies, not oral drugs, were analyzed. All of the study data came from 2010.

In a novel twist, the authors used a national — not local or regional — private dataset (Intellidose from IntrinisiQ) from 35 states that included 122 medical oncology private practices with 570 oncologists, 19,500 patients, and 135,000 chemotherapy administrations.

The usage rates and cost of on-label use varied significantly by drug.

Estimates of on-label use ranged from lows of 32% for gemcitabine, 33% for rituximab, and 48% for bevacizumab to highs of 92% for bortezomib (Velcade, Millennium/Takeda) and pemetrexed (Alimta, Lilly) and 99% for trastuzumab (Herceptin, Genentech).

The national sales of the 10 chemotherapies totaled nearly $12 billion in 2010. Overall, on-label use totaled $7.3 billion and off-label use totaled $4.5 billion.

Bevacizumab was the single largest contributor to these off-label sales, accounting for nearly $2 billion. In addition, rituximab accounted for $1.4 billion in off-label costs. Combined, the 2 Genentech drugs accounted for about three quarters of the off-label cost total.

Not all off-label prescribing is "bad," Dr. Krzyzanowska points out, but some is "more appropriate" than others. The authors acknowledge as much, and made an effort to "dig deeper" into the prescribing by categorizing each off-label use as either conforming or not conforming to National Comprehensive Cancer Network (NCCN) guidelines. This effort "gives us a more complete picture about the extent of inappropriate prescribing," writes Dr. Krzyzanowska.

The authors found that 14% of the chemotherapy use conformed to an NCCN-supported off-label indication, and 10% of off-label use was associated with an FDA-approved cancer site but an NCCN-unsupported cancer stage and/or line of therapy.

This last finding gave one expert pause.

"It is concerning that 10% of the time it would appear that patients are getting chemotherapy that is not consistent with the evidence," said Jennifer Malin, MD, PhD, in an email to Medscape Medical News. Dr. Malin, who is from the University of California, Los Angeles and WellPoint, a managed care company headquartered in Indianapolis, Indiana, was not involved in the study.

However, "it is reassuring that much of the chemotherapy that is prescribed is consistent with the FDA label or NCCN guidelines," she said.

Of the $4.5 billion off-label total for 2010, $2 billion was NCCN supported and $2.5 billion was not.

The authors believe that the magnitude of this NCCN-supported off-label prescribing is one of the "important implications" of the study.

"Compendia are of significant value to clinicians because they provide evidence in support of treatments' use," the study authors note.

The problem is that compendia such as the NCCN's are not strictly evidence based, and partly rely on the opinion of experts, some of whom have conflicts of interest. In short, the NCCN is not the FDA in terms of the rigor of review, they explain.

Dr. Krzyzanowska highlights one recommendation from the NCCN guidelines that has no evidence base. Bevacizumab was recommended in the 2010 NCCN guideline for the adjuvant treatment of colorectal cancer. The drug had proven beneficial in the metastatic setting, which fuelled its use in earlier-stage disease. However, there was no evidence for this adjuvant use in 2010; in fact, a major study of the drug that had a negative finding in the adjuvant setting was publicly presented in 2009, she points out.

Study Probably Overestimates On-Label Use

This study might be a bit optimistic about on-label use, said Dr. Conti. "Our study likely overestimates on-label use and, therefore, likely underestimates off-label use," she said.

Dr. Conti explained that the data used to determine usage was only somewhat granular; it included cancer indication, stage, and line of therapy in classifying the use as either off or on label.

The data have "limitations," she said. "We could not use them to distinguish on-label use that required patient and tumor characteristics, beyond cancer stage and treatment line, and could not examine use that is indicated in combination therapy."

For example, the FDA labeling for trastuzumab indicates use in combination with paclitaxel for the first-line treatment of HER2-overexpressing metastatic breast cancer, Dr. Conti noted.

"While we were able to determine whether the patient was undergoing first-line therapy for metastatic breast cancer, we did not have the data to determine whether the woman's tumor had this genetic signature, nor did we have the data to examine whether trastuzumab was used in combination with paclitaxel. Therefore, we classified the use of trastuzumab for the treatment of first-line breast cancer in all women with metastatic breast cancer as on-label," she said.

This study was supported in part by the National Cancer Institute. Some of the authors report financial ties to pharmaceutical companies, including some makers of the 10 chemotherapies in the study.

J Clin Oncol. Published online February 19, 2013. Abstract, Editorial