Why may real world data not always reflect what was learned in a clinical trial?

As a myeloma doctor. Doctor Banerjee uses examples of real world data from multiple myeloma. However, this information on real world data can apply to all blood cancers.

Why may real world data not always reflect what was learned in a clinical trial?

patients who enroll in our clinical trials, I am tremendously appreciative of their time and effort.

You know, they I joke to patients, they're putting their blood, soil and tears into, you know, this clinical trial for the betterment of myeloma care for all future patients.

But they're not always representative of some of the other patients that I see. So I think that's the biggest difference. So two things can happen.

One, you know, patients who have the wherewithal to be at an academic center at a major trial center were they’re able to consider clinical trials are often healthier, have more resources, have more health literacy.

and, you know, in clinical trials, even though we reimburse patients, they often cost money for the logistics of getting to the academic center and are always reimbursed properly. So it's it's tricky. Being on a clinical trial is not easy. As I alluded to earlier, clinical trials involve a lot more time of the patient. Unfortunately, many more tests like ultrasounds and echos and bone marrow biopsies. Often more time in clinic.

As I alluded to, clinical trials often use older styles, for example, requiring a 24 hour urine assessment every three months. That's extra time stuck urinating into a jug and bring it back to the clinic. That's not for everybody. So I think that's one part of it. Of the patients who, whom we approach for clinical trials or even approach our academic center for clinical Trials are different.

And the other part of it, you know, again, that's not I can't easily fix that. I wish that I could the other part of it that we are trying to fix as a field is that clinical trial protocols are often restrictive clinical trial sponsors, meaning whoever is paying for the trial to be run sometimes, often drug companies not always.

Sometimes we have IITs which our investigator initiated trial, meaning that it’s my idea I'm using my research funding to pay for this. In all of those cases, however, we want to choose patients. We want to be able to see whether a drug is safe, whether a drug is effective. And unfortunately, patients who don't have the lab parameters to easily be able to prove either or have issues with either lab parameters to their blood work to improve safety or prove efficacy often are not allowed onto the trials.

What do I mean by that?

For example, we want to make sure that this new drug doesn't make people's blood counts get too low. And so we have good ways to assess for that. But that requires the patient going on to the trial have good blood work to begin with. So patients who have cytopenias, meaning low blood counts, often are not eligible for clinical trials, which is unfortunate. Or efficacy, we want to make sure that drug is actually working, making the Mspike go down, making the plasmacytomas shrink, making the Kappa and lambda ratio more normal or something along those lines. Patients who have oligosecretory or nonsecretory myeloma meaning they don't have those blood parameters that we can easily track are all of a sudden excluded from these clinical trials, and it's unfortunate. So I think that, between those two, that explains a lot of why trials and real world data often reach different conclusions.

Just because clinical trials, the patients are healthier in a lot of ways.

And so and a good example of this would be with teclistamab or elrenatamab. So teclistamab is a good example.

It's a BCMA bispecific antibody. And in the majestic one study we saw very impressive durations of responses a partial response of 18 months. So patients where that drug worked, it worked for a good solid a year and a half on average. and it worked in over 60, 70% of patients. Several real world data sets have emerged since then.

I've looked at this in real world patients receiving teclistamab also known as tecvayli commercially. And we actually it's encouraging actually to see that the duration, the depth of response is the same. The response rate is the same. Still, about two thirds of patients, maybe 50 to 60% of patients, can expect to see a clinical benefit from teclistamab.

But for whatever reason, compared to the trials, we have not really seen the same duration of response in the real world setting, where the drug doesn't work for as long. Why is that? Difficult to say?

My guess is that real world patients, you know, in the real world on these clinical trials, their T cells aren't as healthy. T cells need to work for teclistamab to work. In real life, teclistamab doses are often held for infections or complications, or the patient might miss a dose because of this or another reason.

So maybe in the real world, the actual dose intensity compared to the very regimented clinical trial is not as rigid. And so if doses get missed or unable to be given, and therefore the drug doesn't work as well or for as long, which set of data are right that impossible to say. And the clinical trials are very important to show that the drug works to get approved. Real world data are what I look to right.

Because real world data reflects how I treat my patients typically and how my patients are able to be treated. And so I do looked to real world data very, importantly in that regard, even though the data aren't as rich as a clinical trial, those patients are more likely to be the patient walking into my clinic.

And so I think that's, you know, if if there's a discrepancy between trial data and real world data it’s the real world data that I think is more applicable.