In this journal section, we speak with cancer care practitioners about their clinical pathways and the pathways programs being used in their practice, how they are being applied in a particular disease state, and what challenges persist regarding treatment decision-making and personalized medicine integration.
In this “Pathways in Practice” installment, we speak with Robert M Wenham, MD, MS, FACOG, FACS, chair, department of gynecologic oncology, Moffitt Cancer Center (Tampa, FL) about the special challenges related to the creation of a clinical pathway for ovarian cancer. Driven and consistently guided by their own physicians and with input from across multiple services teams, Moffitt creates their own clinical pathways from start to finish in-house and updates them regularly. Dr Wenham has been involved in Moffitt’s clinical pathways program from its earliest stages. Dr Wenham has a strong interest in research and is a principal investigator for numerous clinical trials to improve cancer care for women. His research activities primarily include the use of novel drugs for the treatment of gynecologic cancers.
Please describe how clinical pathways are developed and used at Moffitt?
The development process I’m going to separate from the implementation process. Interestingly, I was appointed the original pathway lead here at Moffitt by our visionary physician-in-chief (and now current CEO), Alan F List, MD, about 9 years ago to try and figure out how to develop these pathways. The first pathway to be put down on paper actually was the ovarian cancer pathway, and it was on the back of a napkin at about 30,000 ft in an airplane. At that time, there were many different ideas that were circulating in the field about what a pathway should be and how our pathway might differ from other attempts to develop pathways. The big thing I found out when making the pathways is that it is one thing to write down what you do and make a little algorithm sheet, but it is a whole other thing to come to a consensus about what a pathway’s purpose is, the terminology that should be used, how it should be implemented, and the maintenance of pathways.
So we formed a group to ty to figure these issues out, and it became apparent that people thought everybody was on the same page but, in reality, there were some conflicting thoughts about the vision of the pathways program. One of the assumptions from some people on the business side was that pathways were going to be used as a way to homogenize practice, to decrease variability, to decrease costs, to inform payers of their potential risk, and to use as a negotiation thing. But from the clinical members, they primarily thought of using them to be a platform of discussion, of consensus, but also for understanding disagreement, and to being a research tool. For example, every time there is a branch in a pathway, it represents a divergence and means that there is a set of questions around why you go on one branch vs the other and that brings up opportunities for collecting information and doing specific research.
There were definitely some who thought that a pathway designed for improvement in quality may not decrease variability; it may actually increase variability and increase costs, so we had to figure out what our priorities were. So when our pathways were first designed, I and others went into it with more of an altruistic standpoint. We decided that we would establish how we deliver care, the reasons why we do that, and ignore certain things such as cost, etc, initially, so we can outline what we would consider ideal care for particular patients and try to make it as algorithmized as possible in medicine, understanding that there are going to be differences of opinion.
The other thing that I thought was important was annotating the pathways. Having a bunch of stick drawings is not sufficient. For someone to pick up a pathway and either understand the rationale or to modify the pathway at a later time, there had to be companion documents that explained the logic and had references to the literature that was appropriate and also would discuss why patients may or may not meet what a pathway shows.
Another item we thought was important was defining the terminology that would be used around the pathways. We tried to avoid words like “deviate” from pathways because “deviate” means, often, that there is deviance—and there is not really deviance in that sense, there is just disagreement. So rather than “adherence,” it was “alignment” or “agreement.” Those term choices are powerful because they send a signal or undertone that these are not punitive pathways; they are meant to help develop clinical practice.
Would you say that the pathways you have created and the idea of cost savings behind clinical pathways are less to do with consideration of the actual cost of therapy on the pathways and more to do with simply reducing variability of care, and as a result, reducing unwarranted costs?
Yeah. I think that’s right. We need to reduce unwarranted variability because when you have unwarranted variability in a system, that can also lead to things other than just cost; it can lead to error. For all those reasons, many people said that pathways may be useful. On the other hand, we aren’t oblivious to the fact that there may be times when cost may be a deciding factor. After we finish making the pathways and justifying it on an efficacy and tolerability standpoint, we then append—where there was division or controversy—some of the cost information. And certainly, I think they have been trying to track how one decision or another affects cost. But that’s a moving target, right? Cost is not a fixed thing. And a drug today may be expensive and tomorrow may be cheap, and it’s hard to plan an entire pathway with that being your primary motivator.
Shifting gears to focus our discussion more on ovarian pathways specifically, does Moffitt have a clinical pathway or a section in the ovarian clinical pathway for platinum-sensitive recurrent ovarian cancer, and if so, what is the approach to stratifying platinum-sensitive patients to guide therapeutic decision-making?
Yes, we have an ovarian cancer clinical pathway that has subsections/divergences based on platinum status. At Moffitt, we always select a physician to be the primary pathway author, and I am the ovarian primary author. This means I work with our clinical pathways development/revision team (non-physicians) on any new updates needed for the ovarian pathways. Karen K Fields, MD, medical director of the Moffitt Oncology Network, is the clinical overseer of this team. We typically meet quarterly and review the pathway to see if there are any changes to be made. Then the drafted changes are sent out to all members across the facility to get their multidisciplinary input, and then ultimately approved for inclusion.
So for initial presentation, we do have a platinum-sensitive path and a platinum-resistant path. I think if you had to divide the pathway into 4 major parts for ovary, it would be (1) initial diagnosis or suspicion of ovarian cancer, (2) primary therapy, (3) those who have the platinum-sensitive recurrence, and (4) platinum-resistant recurrence. And then built in there are these auxiliary things which I consider surveillance: genetic referral, clinical trials, etc. So, 4 different major areas.
The platinum-sensitive and platinum-resistant disease treatments were a fairly slow-moving and changing entity until the last 4 years or so. The approval now of both bevacizumab and of 3 different poly ADP ribose polymerase (PARP) inhibitors in those settings has quickly altered how we think about triaging these patients into various treatments. Even now we have to amend initial frontline therapy to include PARP inhibitors for maintenance of BRCA mutation cancers.
Probably the more difficult question for oncologists treating patients with ovarian cancer is what to do when a patient has their recurrence in a platinum-sensitive setting. Previously, it was trying to decide between a few different chemotherapy regimens, all of them having platinum because it was platinum-sensitive, so it was either platinum plus a second agent—most commonly paclitaxel, docetaxel, pegylated liposomal doxorubicin, or gemcitabine as pairing agents. To be honest, if you look across the literature, there really are not hugely significant differences between those chemotherapies. It was more schedule-based and toxicity-based and looking at what kind of chemotherapy the patient had before and what their tolerance was.
The approval of bevacizumab occurred at about the same time when we saw the approval of PARP inhibitors. The problem is, right now, that we don’t have approval of a combined bevacizumab and PARP inhibitor strategy for platinum-sensitive disease; if you’re going to make the decision to put somebody on bevacizumab, which was studied in basically a gene-agnostic population, you do that up front with the chemotherapy and then you continue the bevacizumab for a long duration after chemotherapy as maintenance. If you’re going to choose to do the PARP inhibitor route, then the PARP inhibitor is delivered after chemotherapy, and it’s for patients who have a response to chemotherapy. When bevacizumab is given before, you don’t know whether the patient is going to have a response or not. But for the PARP inhibitor, you give the chemotherapy, and then you look for a complete or partial response; and only then you can put them on maintenance.
To make it even more confusing, we know that there is some difference in the degree of response between patients who have certain mutations in their homologous re-combination repair genes, those who have assays that look like they’re homologous recombination repair-deficient, and then those patients who seem to have neither based on differences in responses. So the question is, do you start trying to guess ahead of time whether or not your patient is going to have a response and put them on a PARP inhibitor or do you put them on the bevacizumab agnostically? But then, if you did that and they respond and they happen to have, let’s say, a genetic mutation, do you then switch them over to a PARP inhibitor? Do you continue the bevacizumab and treat with a PARP inhibitor for recurrent disease?
This is what made creating the pathway complicated, and it partly can depend on what you were going to do in platinum-resistant disease later on. The two tie together somewhat in that many people were feeling like the real role of bevacizumab might be with platinum resistance because it had already been shown to significantly improve progression-free survival time in combination with chemotherapy in the platinum-resistant setting. But then, other data emerged about preserving the effect of bevacizumab throughout different settings like front-line to platinum- sensitive, then platinum-resistant.
Overall, it has become more confusing for the average oncologist deciding what to do with a patient with platinum-sensitive recurrence.
