U.S. & World
GREG BERK, MD: Hi, my name is Dr. Greg Berk from the New York Hospital, Cornell Medical Center. I'm pleased to have Dr. Anne Mohrbacher, Assistant Professor of Medicine in the Division of Hematology at University of Southern California here today to address the issue of physician barriers to patient recruitment in clinical trials. Welcome, Dr. Mohrbacher.
ANN MOHRBACHER, MD: Good morning.
GREG BERK, MD: Dr. Mohrbacher, I'd like to maybe open this discussion up with a general sense from you as to how you feel physician awareness and oncology awareness is directly toward the involvement of patients in clinical trials.
ANN MOHRBACHER, MD: I think there is quite a disparity between university-based centers and the community that I think is now coming to a point where we now have to cross barriers and bring those two systems closer together. I think the community oncologist tends to think of trials as something they did when they were back in fellowship and sort of remain in the university arena. I think we're approaching an era now where clinical trials really have to be out there and in the community. Perhaps the university will be the coordinating center, but not the only place where a patient can go for their treatment trials. I think our goal for this millennium is to involve the community physician in the clinical trials process.
GREG BERK, MD: That's very helpful. Dr. Mohrbacher, there is perception amongst both university-based as well as community-based oncologists that putting patients on clinical trials creates a whole new tier or work level, much more work, not just for the physician directly, but his staff. Can you address that perception?
ANN MOHRBACHER, MD: I do think that's true and again that's a correct perception. I think for community physicians, really the only practical way at the present time to enroll a patient on trial is to send them off to a cancer center and sort of give up control over that patient's case. Because the HMO and the private practice group and so doesn't have the infrastructure to run a trial. No physician can sit and do the detail work that's necessary on a trial. Although it's quite simple work to do, it's time consuming. That includes needing to check eligibility lists, read through the protocol, make sure an up-to-date version of the protocol is approved, schedule quite a variety of tests, and follow-up on the results of the test and make sure they are in charts. That really requires having a data manager and usually in addition, a research nurse, which is beyond the scope of most private practice groups. Even HMOs have been sort of slow about committing to that kind of personnel.
In reality when I enroll patients on trials, my main work is thinking in the first place of whether the patient should consider a trial. I then discuss the trial with them, offer them the time to read the informed consent and then go over with them whether they feel it would be appropriate to participate in that trial. After that, I turn the workload over to the data manager and research nurse, so there is actually no additional work for me. That wouldn't be the case out in the community center. That's where we need to think a little more creatively about how we're going to bridge that gap.
I feel personally the only way to do this is to have data managers and RNs who are linked to a central study site, such as a university or academic cancer center. They will be doing outreach to the community groups or working for an HMO in close affiliation with the university to basically do that footwork for the trial. It is unrealistic to expect a physician to pursue those details for a study outside of the immediate university and even in the university setting. We don't do it ourselves. We have our support personnel do it.
GREG BERK, MD: There are many community-based oncologists both in small groups and in large groups who are really, truly interested in getting involved with clinical trials. As you know, there are many good studies carried out primarily by the pharmaceutical companies, that are taking place in the community setting.
ANN MOHRBACHER, MD: I'm always impressed, for example, by how cardiology studies get done out there in the community, compared to oncology trials, which traditionally have been almost exclusively based at academic centers. It's probably just a matter of rethinking how we do things. They tend to have centralized data collection sites, computer access, and phone access so the physician really has to do only the most minimal paperwork or detail work on the study before it's turned over to a central sight that follows up on it. That is a very helpful model that I think we need to start thinking about in an oncology setting.
GREG BERK, MD: I think the physician in the community settings that are interested in doing, many of them are already aware of this. But some are not. Many of these pharmaceutical-sponsored studies, much more so than the cooperative-group studies, may very well, in the budget of these protocols, have the budget built in for additional data managers, research nurses, etc.
I've seen in the New York area, more and more sort of community practices pop up doing studies, which is impressive to me, actually.
ANN MOHRBACHER, MD: Yes. There is still a little bit of a logistical problem in that. An experienced data manager and research nurse is—there is no course in college that produces someone ready to go for you that you can just hire with an ad. It usually is someone who is self-trained in a system. They got into the area and learned by doing trials under the apprenticeship system. That is not that easy if you have trials that have intermittent enrollment. I think that's where my argument for a bit of centralization of those personnel comes in. Then the money would help fund those individuals.
It's hard unless you have a critical mass of patients who are going on a trial. This is particularly true of hematology trials because they tend to be a scattering of low numbers of patients because the diseases are less common and there are many different things to treat them with. It hard to have the critical mass of patients to know that you can sustain a full-time research nurse or a full-time data manager.
To find people who are really good at that is a whole other project. You'll find that one good very experienced research nurse and data manager is worth two novices easily. Spreading that expertise around is sometime the problem and again, I think [it is] an argument for a little centralization having more experienced people supervise and train more junior people who are perhaps out there in the community outreach.
I think one of the other big issues about clinical trials is, even where available, how do the physicians come to know about them? The current system is very much a patchwork that relies extensively on the physician's initiative for seeking out those trials. They have a patient who has gone through two lines of treatment and they're looking for something else that they can offer that patient. I think that's where the physician's desire to get involved in these trials comes in.
It's difficult to identify trials that might be appropriate for patients. They are scattered throughout individual Web sites affiliated with the centers that offer the trial. You might have to check the National Cancer Institute site, the Central Watch site, which is a more global listing, but by no means complete. [Check] individual universities, perhaps in the area where the patient lives, or pharmaceutical company Web sites to try to seek out what trials might be out there.
I know other groups such as Hutchinson tried to approach this by direct mailing with a brief synopsis of eligibility criteria to physicians in the area to make them aware of new trials that were opening up and might be available for their patients. Then there are special interest groups like the Lymphoma Research Foundation, Cure for Lymphoma, or other disease-oriented special interest groups that will try to maintain their own listings or make announcements of new clinical trials.
The process is cumbersome. I still find that sometimes patients are approaching me with information about—they heard of a trial in such and such place. It becomes incumbent on me to sort of track that trial down. It's really not a smooth process.
We don't have a global access system to all clinical trials available everywhere. That would sort of be a very desirable thing to make it more convenient for physician to consider getting their patients enrolled in trials and also seeking out trials that they might want to join—if they're not already part of a large intergroup such as SWOG or ECOG in the case of oncology trials.
GREG BERK, MD: Anne, I'm an active member of CALGB. Our institution has been a member of CALGB since its founding. We have at Cornell, over the years, put much more emphasis—and in recent years—in a lot of these pharmaceutical-sponsored studies—
ANN MOHRBACHER, MD: So have we.
GREG BERK, MD: —as CALGB funding has sort of dwindled down. But at the same time, I have heard from many of the companies involved in clinical trials that the cost of doing clinical trials in the university setting and in the big cancer centers continues to escalate. Do you have any handle on...
ANN MOHRBACHER, MD: I was just reading a little synopsis on this from the Cancer Journal, I believe, about how the intergroup trials are grossly underfunded relative to the amount of work it takes in terms of personnel—data managers and research nurses. That ends up being a motivation to steer away from those trials. The pharmaceutical companies vary tremendously, but generally offer just an adequate amount of money. I think it's got to be somewhere in range of being adequate to cover the personnel, plus a little extra for sort of the time and trouble factor. But they can't be so highly paid that it becomes a purely monetary motivation to enroll patients on trials. I think that would sort of cloud the ethical issues of why we encourage patients to consider trials.
But it really is sort of a question of what is an adequate amount of money. As in our case, we are expecting that most of the tests, CT scans, and so on done for the trial are part of routine care for the patient. We're not expecting the trial to cover those costs. Insurance companies have certainly been becoming more flexible and open to this idea.
First of all, I think some of them realize that as the drug, the main agent, is usually provided free in the majority of trials, and that offsets the fact that they might do 20 percent more tests than in routine medical care. Most of them are not particularly objecting to patients going on clinical trials at this point, I'm finding.
But it does sort of raise the question of what the money is for. My mind is that the sponsor should provide for the mailing of samples that they need to do any esoteric tests under their own center in a central lab, rather than, of course, trying to pay that out to individual centers. Most of the expense of the trial should be focused on the cost of the data manager's and research nurse's efforts. That is usually paid on a per patient basis. It really depends on the complexity of the trial once you start introducing in terms of how many tests have to be interpreted, how much data has to be entered, how many meetings there have to be to go over the charts.
But once you start entering issues about what's going to be covered or not in terms of clinical care, it's impossible to estimate what a trial should cost. That's very difficult.
On the other hand, a national intergroup paying say $500 or $1000 per patient and requiring—quite frankly I find usually the intergroup trials require the most data collection. They add in the most tests, have the most mailings of special samples, and then pay an inadequate amount to cover the time that it takes that data manager or research nurse to participate in the trial. It's somewhere between the two.
I think we are able to make fairly accurate estimates of what time commitment. A nurse can look at a trial now and say, "This one is going to take me about as much time as this other one, and I was able to handle 20 patients a month. My salary is X for a year." You can pretty much piece it out of the slice of the pie, how much that trial should be per patient. I think most pharmaceutical trials are paying a reasonable amount in correlation to the work load of the trial. It's not excessive. If it were excessive, then that sort of again starts to raise some ethical concerns.
GREG BERK, MD: Very good. Anne, I'm wondering if you could comment on what's happening with managed care's approach to clinical trials in the oncology setting—at least from your standpoint out in California.
ANN MOHRBACHER, MD: Traditionally they have been very opposed and everyone has had a global veto line saying, "no experimental trials." "No experimental therapy will be covered by this insurance."
Many of them, even when they have that clause, have in fact, been perfectly happy to have patients on trials. I believe that they've been aware of the fact that we've put their patients on trials. I think they just want to have a line-item veto to be able to refuse it when they want to, if an individual trial proposed something extremely expensive, like a month-long inpatient hospitalization.
In fact, most mainstream insurers including PPO and HMO plans have been fairly open to this particularly if it's an outpatient treatment. As long as the tests don't strike them as being wildly out of line with the routine tests of standard care.
When they start to require inpatient stays, then they need to compare it against what the patient would otherwise be receiving at this point in their care. Very often we can justify it on that basis.
We will often end up writing a supportive letter for why we feel this is reasonable. The drug doesn't cost them money. The inpatient hospitalization is no more or less than what they would have taken as an alternative standard treatment at this time. Given the cost of chemotherapeutic drugs, that ends up being a significant argument. When you pay $5,000 for Taxol, for example, for a single dose and you're now talking about a one-day hospitalization with the drug that's free. It doesn't look that expensive anymore.
What I'd like to see, though, are large institutions like Kaiser and some of the larger HMOs who clearly have some interest now in participating in clinical trials. There are some special advantages in having them participate. They've got a large patient population and relatively healthy patient population with a central computerized database. That is a special advantage that you don't have in most other medical centers.
I think they are actually an ideal setting for doing perhaps some of the late phase II and phase III trials. I don't see them particularly as a phase I type setting to offer clinical trials in where they have to deal with unknown toxicities. I think they want to deal with something that is already a bit of a known entity where the treatment issues have been simplified, where there aren't dose escalation problems or additional lab testing.
I think some of the larger HMOs would be the perfect place to do some of the trials that we look to do in the future. I even have particular ones in mind, and occasionally can get cooperation, even from a very traditional HMO, if I have a drug I can offer that's either oral or has very brief outpatient administration. We have worked out sort of informal agreements. They will agree to do the blood tests, the CT scans, the MUGA scan, and the patient will only pay for the drug administration for the day they come to our hospital—the actual agent. If it's IV, it usually has to be administered at the site.
That has worked out in a couple of patients cases. But I'd like to see it done much more systematically. I think the HMOs could be a very powerful tool for advancing research in cancer medicine. It's more of a commitment to the idea of perhaps hiring some data managers and research nurses. It allows some more creative arrangements with research personnel from another center, interacting with their staff to request tests and enrollment in trials
GREG BERK, MD: On that note, I'd like to actually close. I really appreciate your comments on this important subject.
ANN MOHRBACHER, MD: Well, it's an important area. As I've pointed out, with breast cancer, for example, one of our most prevalent cancers in this country, only three percent of patients participate in clinical trials. We're going to be stuck at our present cure rate for the next 10 to 15 years if we don't increase the participation in clinical trials. It really has to become a routine part of cancer care—whether that be in an HMO, PPO, or an academic center. I think that should be one of top priorities in cancer research.
We've got lots of creative ideas that are sitting there percolating very slowly through the system. We need to get out there and get enrollment up in these trials. I think there are a lot of safe and very innovative drugs out there that could be perfectly well-handled in a variety of settings from the community right on up to the university.
That's one of my big hopes for this decade—that we can really improve the speed with which we do clinical trials research.
GREG BERK, MD: I think your comments are really very timely and really helpful in that regard. And thanks again.
ANN MOHRBACHER, MD: Thank you.