Development of Clinical Pathways and Patient Characteristics

Caroline Block, MD provides an insight into the development and management of clinical pathways.

Transcript:

Caroline Block, MD: With regard to the development of clinical pathways, we started doing them at Dana-Farber [Cancer Institute] 4 years ago. We have used them in each disease center to make sure we’re approaching treatment in a more homogenous fashion, that we prioritize what treatment to use in a given situation, and that all the major disease sites at our institution use clinical pathways.

When clinical pathways were developed, there were several goals of having them. One was so that, as new treatments came out, we could discuss as a group how we incorporate these new treatments. It’s rapid in the field of oncology now as I’m sure many people know. We had, in the last 6 weeks, 3 new drugs approved in breast cancer, so it is important to be able to meet and decide where and how we are them, what line of therapy they will be used in, and what toxicities are important to point out. Efficacy is still high on the list. If the drug is not very effective, we’re probably not going to add it to our pathways. But if it’s effective in a given system, we will certainly try to add it. Toxicity is always included as well. It’s hard to say that 1 is important but not the other. They’re both important. If the drug is efficacious, we’ll look at toxicity. If there are different drugs in that same category, then that’s where toxicity becomes even more important.

Every time we discuss a new drug or review our clinical pathways and decide whether to make changes, we also look at cost. We don’t typically know the cost until the drug is out there in the market, but we always talk about that as well as 1 of the considerations.

From an IT [information technology] standpoint, early on, the clinical pathways were in a separate app, so providers had to go in and click a separate app to open that up. You could do it from your schedule, so you didn’t have to put in the patient date of birth or medical record number, but you still had to plug in the patient’s stage, the biomarkers, and the oncotype. It was more tedious, if you will. For the more recent platform we’ve been using for the last year, you don’t have to put any of that in; it’s embedded in the electronic medical record, which has made it much easier to use. They’re just on the menu. There’s a Dana-Farber clinical pathways category that you can click on, which opens clinical pathways right up within the medical record. The patient’s information is in there, and you don’t have to enter all that data. You can just go to what category is appropriate. If this is an adjuvant, stage II, HER2 [human epidermal growth factor receptor 2]–positive breast cancer, then you go to that section, but you don’t have to click there and enter the T [tumor] and N [node] and M [metastasis] stage and that sort of thing. That’s made it a lot easier to use. People used to complain, appropriately, about the clicks and all the data entry involved, so we’ve taken that away. We’ve tried to make it more up-front, user-friendly, right there, and easy-to-use.

We also added a carrot because we do want to make sure there’s a reward for people going in to use clinical pathways besides the information there: The consent form prints from it, so once you’ve finalized the pathway that you’ve chosen, the consent form is populated. You can edit it if you’d like, and you print it. It saves you from having to do that in a whole other system, so that is probably what providers have liked the best about it. If they go through this fairly quick process, they come out at the end with a nice consent form ready for them.

How we link in subtypes and markers is that the pathways are, in general, based on subtypes, not just stage. Is this early stage breast cancer, local recurrent, or metastatic? Those are 3 different categories, but, within each category, it’ll be organized by hormone receptor positive/HER2 negative, or triple-negative, or HER2 positive. Each category has those 3 biomarker subsets.

Within a biomarker subset, for the nonmetastatic patients, there will be different stages. We’ll treat stage I, HER2 positive different from stage II or III HER2 positive. The section you choose is dictated by the stage and the biomarker. In a given section, if there are other appropriate markers that are important for a decision, there will be notations about that: either a reminder note that this is a good time to check this particular marker or that this treatment is only effective in the presence of a marker. For instance, if it’s metastatic, triple-negative breast cancer, we will state that the tumor needs to be PD-L1 positive for the atezolizumab and nano albumin–bound paclitaxel to be an option. If it’s PD-L1 negative, there’s a different branch for that. For metastatic ER [estrogen receptor]–positive breast cancer, if patients are considering use of a PI3K inhibitor, such as alpelisib, then the tumor needs to have a PI3K mutation.

It’s less common. We mention these, but we don’t have a separate branch because it doesn’t come up very much. For breast cancers that have microsatellite instability that are MSA [mammary serum antigen] high, then pembrolizumab is a treatment option. We don’t see it very often in breast cancer, so we have it as a note rather than a separate branch of treatment.

The process for getting approval to use an agent not on the clinical pathway is not any different from what it was before clinical pathways. So far, we don’t link that to approval. Providers, as always, are able to go off clinical pathways to use a treatment. Last week, for instance, I started a patient with metastatic HER2-positive breast cancer on eribulin, Herceptin [trastuzamab], and pertuzumab. The 1 we have that’s already approved as a regimen is just eribulin and trastuzumab. There’s nothing that prevents providers choosing something off clinical pathways and writing what they want to use.

We have a separate mechanism where we have a group that you email to say, “I’d like to use this. This is why. Here’s the reference.” That process happens in parallel. In clinical pathways itself, the way we have it set up, you can go off the clinical pathway.

Our goal is 80% or above of the time, people are on clinical pathway, meaning that we’ve covered the treatment options well. If someone is off pathway 30% to 40% of the time in a given month, there’s no penalty for that. You don’t get called out for it, but we do look at it. If providers are going off pathway a lot, then maybe our clinical pathways aren’t what they need to be. We need to relook at that and decide if we need to modify things and make it so our pathways are reflecting what we’re actually doing more.

Over the last year, since we’ve changed platforms, we work with a software company called Royal Philips. They provide the actual software IT platform. All the content is developed at Dana-Farber. We look at things like NCCN [National Comprehensive Cancer Network] Guidelines, but honestly, that’s done after the fact. It’s never part of our actual process of deciding if we are going to use this drug, where we are going to use it, and what information we’re going to put in when we add it to our pathways. The content is driven by each disease center.

In the breast oncology group, we have a minimum of 3 meetings a year when we’ll look at our clinical pathway usage and off-pathway rate as well as what we are doing and what we need to change. We’ll also look at new drugs as they come and figure out where they should be added in the clinical pathway.

We do have ways we can look at some patient characteristics and preferences in clinical pathways, although there’s more work to be done, and we would like to expand that going forward. For instance, in the adjuvant clinical pathway regimen, we always give options for nonanthracycline regimens and options for nontaxane regimens for patients who have contraindications to cardiotoxic agents or have preexisting neuropathy.

We will also specifically state if a regimen has a toxicity that providers and patients need to be aware of. We just approved and put on our pathway trastuzumab-deruxtecan, and we boldly mention the interstitial pneumonitis, since that is not uncommon and can be fatal. That certainly needs to be considered strongly when you’re considering that pathway.

What we haven’t been able to do yet but we want to do is to have it a bit more clickable. Especially in the advanced breast cancer setting, where toxicity is so important–efficacy is important, but so is quality of life and minimizing toxicity as much as you can–you could click on a link that says, “This patient doesn’t want to lose their hair.” What are my best options if the patient doesn’t want to lose their hair as long as possible? If this person already has bad neuropathy from a prior treatment, then I want to get rid of anything on the list that might worsen their neuropathy.

We’ve talked about it. We just don’t have the IT ability to prioritize those sorts of things and then sort a list based on toxicities. But that’s certainly high in everyone’s minds to be able to do so.