Tracking the Evolution of Oncology Drug Prescribing

Do you know what your cancer drug is up to? Answering that question is not as simple as for medicines taken by well-defined patient groups, or principally for one disease. Yet among drug classes oncology treatments hold a unique position: A frightening disease of immense scientific interest, hundreds of pipeline drugs and nearly as many protocols for single and combination treatments, a complex distribution and reimbursement, and—as information service companies tell us—unprecedented opportunities for those with the right data.

Kalorama Information estimates the worldwide cancer drug market at $54 billion. Oncology’s share of global pharmaceutical demand, while only about 5% in dollar terms, is growing at more than 10% per year.

Much of oncology practice is guideline- or protocol-driven, but a good deal is based on cancer doctors’ experience, journal reports, and information from conferences. Off-label prescribing, which is relatively rare for some conditions, is quite common in cancer therapy.

Drug utilization data have long provided critical feedback for pharmaceutical marketing, product lifecycle management, and development of best clinical practices. Oncology drugs take the information game to a new level because cancer and its treatment are qualitatively and quantitatively different from other diseases. Knowing how many pills were sold in what area or prescribed by what medical specialty is not enough. Drug firms demand much more, and information providers have so far delivered detailed data to suit every need.

Oncology data come in a wide variety of types, including complied prescriber data, clinical practice data and processed claims. Providers range from add-on business services offered by wholesalers (McKesson’s Specialty Care division, and AmerisourceBergen Specialty Group), to the market-data services of IMS Health (Conshohocken, PA), SDI Health (Plymouth Meeting, PA), IntrinsiQ LLC (Waltham, MA), to oncology practice management firms like Synovate Healthcare (Lawrenceville, NJ), Accelerated Community Oncology Research Network (Memphis, TN) or IMPAC Medical Systems (Sunnyvale, CA), to business consulting firms like Mattson Jack Group (St. Louis). There are also patient registries and practice guidelines compiled at the American Society of Clinical Oncology and NIH.

A customer of IntrinsiQ’s notes, “I don’t think any one data source or vendor supplies an all-encompassing perfect set of data. There are strengths and limitations to each.” Moreover the robustness and freshness of the data are significant factors (some datasets can lag current practice by as much as six months), as well as the nature of the business questions for which answers are being sought. “Any vendors who continue to add depth and breadth to their data tend to be well regarded,” he says, noting that Intrinsiq started out providing software to physicians to help them avoid overdosing their patients with IV chemotherapeutics.” Now, “IntrinsiQ can capture the entire regimen the patient receives, including oral chemotherapies, how much and how frequently the patient receives each drug and for how long, diagnostic tests performed on that patient, how these tests influence treatment decisions, and what supportive care that patient received.”

Collecting both ways

Information-selling has become a natural side business for pharmaceutical distributors who are ideally positioned to parlay utilization data into marketing knowledge. Onmark (South San Francisco, CA), a group purchasing organization within McKesson Specialty Care Solutions, boasts 3,000 member clinics and 4,800 physicians who benefit from Onmark’s relationships with 18 manufacturers of about 100 oncology drugs. McKesson Specialty also distributes drugs for rheumatology, urology, ophthalmology, dermatology, and is investigating nephrology.

McKesson’s and Onmark’s information products include practice-level software, Lynx, and the IQ service for drug companies. Lynx, adopted by 1,500 oncology practices with an additional 250-300 installed in hospitals, streamlines charge capture, inventory management, dispensing, reporting, ordering and billing. More than 2,800 physicians use Lynx for approximately 50 cancer diagnoses.

Practices pay a nominal fee for Lynx if they agree to share data with McKesson, which massages the information and sells it to drug manufacturers through a service called Integrated Quality (IQ). IQ became possible as a result of the 1.3 million unique patient-treatment incidents captured by Lynx since 2000 (with additional records accumulating at a rate of 290,000/yr). Administered by Onmark, IQ provides insight on how oncologists use drugs individually and in combination. A physician feedback component helps to reinforce practice guidelines for “appropriate use.”

Onmark’s first IQ program, on treatment for white blood-cell depletion resulting from chemotherapy, was released in May. It found that just 65% of doctors prescribed a white cell booster in elderly patients treated for non-Hodgkins lymphoma while guidelines call for every patient to receive the drug. “Manufacturers of these agents view this information as an opportunity for 50% sales growth,” says Mike Cunningham, PharmD, VP and GM at McKesson Specialty Care Solutions.

Because IQ is an opt-in program, it is possible to re-survey oncologists to determine why they did not prescribe a treatment. This dialog informs physicians about treatment guidelines and provides them with comparisons between their practice and others.

Physicians benefit from compliance “because there are a lot of pay-for-quality initiatives afoot,” Cunningham observes. “More importantly, they can evaluate how well they’re doing from a quality perspective.” Drug companies, in turn, learn in near real time of any unfavorable clinical or reimbursement experiences with their products. “The worst thing that can happen is the provider has a negative experience and may not be willing to use a particular product in the future.”

IntrinsiQ (Waltham, MA) is another company that collects fees from information providers and pharmaceutical industry clients. Oncologists pay to use use Intellidose, IntrinsiQ’s oncology medical management system, while drug companies purchase the information Intellidose generates.

Oncologists use Intellidose track office-administered drugs from the time they enter the practice to when patients receive them. The software controls inventory, stores patient information, performs and double-checks dosing, archives notes related to care, provides charge capture for billing, and can even trigger ordering of medicines. Oncologists have long profited from purchasing and administering cancer drugs, but without inventory and reimbursement management the cost of high-priced medicines can overwhelm a practice.

IntrinsiQ’s customers include twenty-eight of the world’s thirty top pharmaceutical companies, according to Julie Maloney, director of client services, plus smaller firms. Each information package is tailored to the customer’s needs because “not every company looks at the market the same way,” Moloney says. “They’re not only looking at their own drug, but those of competitors.” That, says Moloney, distinguishes IntrinsiQ’s data sets from raw sales printouts. “We can tell them when, where, and how a drug is being used, on what kind of tumor, what stage of the disease, in which line of therapy, and how much drug was administered to individual patients.”

Intellidose tracks the drug prescribing and administration habits of more than 600 physicians across the United States, and logs two million drug administrations per year. The system also notes usage of ancillary treatments that mitigate the side effects from cancer drugs, such as nausea, lack of appetite, and blood cell depletion. Back-end software tracks drug use over time in patients sorted by demographics, cancer type, stage of disease, and course of treatment. Once the information enters the system it is anonymized, aggregated, projected to the national level, and delivered to clients.

Steve Gavel, Global Marketing Director for Oncology at IMS Health (Plymouth Meeting, PA), agrees that data alone provide a rather narrow view of the oncology marketplace. For this reason IMS clients who purchase the company’s data and require greater depth and interpretation also subscribe to IMS consulting services as an add-on feature. “Clients frequently have questions whose answers cannot be found in DDD alone,” Gavel says. He is referring to DDD Integrated Views, a database linking healthcare system purchasing dynamics with sales dollar volumes. “Many of our clients require supplementary consulting and primary market research to accompany their quantitative research.” Additional IMS product offerings, such as Oncology Analyzer, are often used with DDD and other quantitative tools to provide further insight.

Getting personal

Not every information services company obtains its data from cold, hard sales figures. With the right incentives it is possible to get oncologists to take time from their busy practices to tell research firms what they think.

NMCR Analytics’ (Frisco, TX) flagship product, Challenging Cases, captures live market research data from physicians recruited from parent company AmerisourceBergen’s network of self-registered oncologists. Doctors receive “fair market” non-cash compensation for participating anonymously in the nearly day-long exercises. Challenging Cases mimics the “tumor board” case-based format in which oncologists review patients and treatment options as they would in clinical practice.

Pharmaceutical company clients receive a demographic breakdown of participants and response statistics grouped by participant’s sex, type and size of practice, group purchasing organization affiliation, region of the country, and years post-training. Participants, who are similarly blinded to clients, are recruited from AmerisourceBergen Specialty Group’s network of self-registered oncologists.

As the Challenging Cases session continues the case morphs through a series of deliberate alterations. Questioning leads participants through a maze of variations on the now-hypothetical patient’s age, sex, ethnicity, treatment success or failure, relapses, economic and insurance status, other illnesses, daily living status, and other relevant factors.

Questions are designed by NMCR’s in-house clinical staff with or without drug company input. From the responses NMCR generates client-specific data or original research which it may publish on its own.

“Manufacturers are looking for nuances in market segmentation to gain a better understanding of how their products are utilized, in what patient types, and what trigger factors are important to treatment options,” says Kristine Lemke, group VP of NMCR Analytics.

Lemke acknowledges that traditional market surveillance and sales data have their place but are limited in value. “They tell you what the medical oncologist did. We get closer to the decision process by challenging the doctor to describe what he or she would do. Knowing what goes on in the hearts and minds of these doctors, when faced with certain patient types, is very valuable.” This approach also measures doctors’ uptake of drug company messages, for example if community doctors are aware of the latest data, whether they know how to utilize it, and if they have ever followed its recommendations.

While NMCR’s data are not specifically designed for formulating treatment guidelines and protocols, they have been used in dossiers to support reimbursement as well as changes in evidence-based oncology guidelines. “I can’t tell you the percentage of the time this happens but I know it’s been done,” Ms. Lemke told Pharmaceutical Commerce.

Lemke claims that Challenging Cases can predict prescribing patterns well before sales data confirm them. NMCR presented a poster at the 2008 ASCO meeting that showed a three-year analysis of the expected use of a particular cancer-drug combination in prostate cancer. This data captured “early intent to treat” market trends by medical oncologists before post-surveillance market information. The manufacturer’s actual sales data correlated nearly perfectly with NMCR’s prediction.

TargetRx also takes the prospective physician survey approach, but with a more traditional doctor-focused marketing angle. The company analyzes survey data to predict future prescribing and to inform drug firms on how they might change the anticipated outcome.

TargetRx describes the four forces that drive physician prescribing as the product, patient, payer, and promotion. “By capturing information from physicians on each of these forces, we tell pharmaceutical companies what’s happening, what the doctors think, where they learned it, and how that affects the likelihood they will prescribe a particular product,” explains Michael Luby, co-founder.

Unlike MNCR, TargetRx pays a cash honorarium to participating doctors. But that is apparently not what makes oncologists sign up because, as Luby notes, “a lot of people are throwing money at doctors.” Oncologists, he says, value that drug companies seek their opinions, that their ideas may affect how drugs are marketed to them, and that TargetRx surveys help them keep abreast of new therapies.

TargetRx does not release income figures but Luby says the firm’s oncology business has increased five-fold since last year. “Demand has been intense.” Several thousand oncologists have registered with the firm to participate in surveys.

Some of the most useful information from TargetRx surveys pertains to drugs within a year of launch. By then, oncologists will probably have heard of such products through journals, meetings, or education programs. Doctors’ opinions at this point, says Luby, are uncannily predictive for how the drug will be utilized within its first few years on the market. “It’s amazing how telling those signals are, even when your questions are open-ended and you don’t give the doctors a lot of data.” By six months pre-launch many oncologists will have already decided how to use the drug and in what combinations. Sponsors can use this information to design educational programs around the launch and beyond. This, says Luby, is when companies can decide whether to pursue the product’s natural positioning or to try a non-traditional approach.

The industry appetite for data—both business and clinical—is bottomless. When asked about the “ideal” information service, the Intrinsiq client noted above had this list—as a starting point:

• Patient demographics: Age, gender, race (some cancers/subtypes of cancer are more prevalent in specific races).
• Physician demographics: practice size, type, and location, physician specialty, number of years in practice.
• Patient characteristics and treatment: Agents/regimens the patient received, dosing regimen, duration of treatment, by stage and line, concomitant treatments (hormones, anti-emetics, supportive care, other treatments the patient is receiving (some cancer drugs are contra-indicated if the patient is already receiving other non-oncology products), diagnostic tests performed, surgical history for the cancer, radiation treatment history.
• Reimbursement details: Payor and copay levels, claims, denials, reasons for denial.
• Promotional information: Number of rep details, detail minutes; physician recall of rep message, physician impression of product, any impact on drug prescribing patterns as a result of rep visit.

But he notes that “of course, this resource would be prohibitively expensive!” and says that “the challenge for the pharma industry is to compile all this information, to build a landscape that most accurately reflects true physician attitudes, practices and patient treatments.” PC

BOX: QUANTIFYING THE PRESCRIBER-PATIENT DIALOGUE

There seems to be no limit to where data-miners will go to get their information. Verilogue (Horsham, PA) records and archives in-office conversations between oncologists and patients, capturing information “across the journey of care” according to CEO Jeff Kozloff. The goal is to provide drug companyies with better understanding of how oncologists and patients interact at different stages of disease and treatment.

Verilogue takes the audio files, from conversations lasting between two minutes and two hours, and transforms them through computer algorithms and linguistic analysis into data useful for crafting educational and marketing messages that improve caregiver-patient communication. Through one study, for example, Verilogue determined that physicians use language associated with war or athletic competition while patients prefer softer terms connoting faith, trust, and personal journey. Conclusion: Doctors need to show better interpersonal skills and adopt language that transmits feelings of confidence and trust.

Other goals include putting oncologists and patients on an equal footing in the examination room, uncovering communication gaps, and determining what language works and what does not. “There’s been a big focus by drug companies on relationship management and relationship marketing, how to support these to improve treatment outcomes,” Kozloff says.

Sharpening doctors’ psychosocial aptitude is not the only use for this information. Companies are eager to learn what drives the treatment continuum, for example the decision to move to a second-line therapy. Verilogue claims its product can help companies achieve that.