Cumulative Risk Factors May Be Prognostic in Advanced Ovarian Cancer

Partners | <b>US Oncology</b>

In an interview with Dana Chase, MD during the ESMO Congress 2021, Chase discussed unmet needs for frontline advanced ovarian cancer treatment and the retrospective analysis of cumulative risk factors and their potential impact on outcome in these patients.

Within 2 years of the initiation of frontline therapy for advanced ovarian cancer, greater than 70% of patients developed disease progression. This fact brings into question the risk factors that may be predictive of overall survival (OS), as well as the time to next treatment.

In a retrospective analysis presented in a poster during the European Society of Medical Oncology (ESMO) Congress 2021, 1251 patient files from the Flatiron Health electronic health record–derived de-identified US database were assessed. Per the protocol of the retrospective study, patients were aged 18 years or older with stage III or IV disease, had begun platinum-based chemotherapy, had an ECOG performance status of 0 or 1, and had ≥12 weeks of follow-up time after first-line treatment. Patients were classified by their risk level, BRCA mutational status, and prior surgery. Investigators evaluated these factors to determine their impact on TTNT and OS.

Overall, 4% of the population had moderate risk disease (0 risk factors), 96% had high-risk disease, and 24% had low-risk disease. Results showed that the more risk factors a patient had, the shorter the OS and TTNT. The population with the best outcomes were those who had only 2 risk factors. This group had a median OS of 11.7 months (95% CI, 9.9-13.2) and a median TTNT of 12.1 months (95% CI, 38.4-47.9).

During a presentation of the poster, Dana Chase, MD, said: “Confirmatory studies are needed to validate the clinical utility of cumulative risk factor assessment as a stratification factor for clinical trials of first line treatments for ovarian cancer.”

In an interview with Chase during the ESMO Congress 2021, a gynecologic oncologist at Arizona Oncology, a part of the US Oncology Network, she discussed unmet needs for frontline advanced ovarian cancer treatment and the retrospective analysis of cumulative risk factors and their potential impact on outcome in these patients.

TARGETED ONCOLOGY: In clinical practice, what are the common challenges you observe with frontline treatment of advanced ovarian cancer?

Chase: There's a couple ways to phrase this. The first situation is, you know, how does the woman present like her first visit your first encounter with her? How does she present with the disease, and what are her current challenges? So, there's definitely a type of woman that presents as a low-risk, advanced ovarian cancer versus a high-risk advanced ovarian cancer. We get pretty used to identifying these women when they come into clinic. There are factors like, do they have a fixed pelvic mass? Do they have multiple comorbidities? Do they have pleural effusions? What's their age? What's their family support?

So, a lot of this is taken into consideration, and that can present challenges to the different treatment options we have, like whether or not we decide to do surgery first, or decide to do chemotherapy first and do surgery later. That's one potential challenges at diagnosis, how do they present?

The second potential challenge is, how do they tolerate treatment? Are they having a lot of toxicities? For patients who’ve had surgery, are they having wound issues if they had while they're getting treatment? Potentially they need dose reductions, or they need dose delays with their chemotherapy. Getting them through chemotherapy can sometimes be very challenging. Then once they're in remission, the challenge is who to figure out is a high risk for recurrence and when are they going to recur?

The majority of patients, about 80%, do recur? To this date, we don't have a way of figuring out who's going to recur sooner versus later. We have some ideas but nothing's really concrete. This is potentially another challenge.

What do guidelines currently say about what to do for patients who progress on frontline therapy and what are standard options presenting to patients in your practice?

Currently, if a patient recurs within 6 months of her last platinum, they're given this title of platinum resistant. When a patient has recurred close to her last platinum, we usually either go for a clinical trial, there's a lot of clinical trials available for the platinum-resistance setting, or we do standard of care treatment, which is usually a nonplatinum-based chemotherapy with bevacizumab. The choice of that nonplatinum single-agent treatment is dependent upon the patient's toxicities that she accumulated with her last platinum therapy. Sometimes that will determine what nonplatinum agent you decide to give for the platinum resistance setting.

If the patient recurred over 6 months from her last platinum, she's considered platinum-sensitive, and that patient may be exposed to another platinum doublet. That platinum doublet could be platinum with a taxane or platinum with another agent, and bevacizumab may be incorporating with that regimen versus doing some other form of maintenance after the platinum is complete.

Right now, patients in the first line setting who go into remission when done with their platinum treatment, most are going to get either bevacizumab maintenance, or PARP inhibitor maintenance. Unfortunately, we just don't have a way to predict who's going to do well. So even though you give PARP inhibitor maintenance, or you give bevacizumab maintenance, or you just watch and wait, we don't really know how to say who's going to recur, or have progression within 6 months versus 6 to 12 versus.

You assessed survival outcomes in patients with advanced ovarian cancer based on cumulative risk factors. Can you discuss the study design and methods?

This is essentially an electronic database that consists of structured and unstructured data from identified patients, and it's called the Flat Iron Health database. Many investigators have used this database to do hypothesis generating retrospective reviews looking at a large number of patients within this database. You find a question, you asked that question of the database, and then the results can help you hypothesize about treatment patterns or patient experiences or outcome data.

Out study is a retrospective chart review. It's hypothesis generating and it's not going to change therapy, but hopefully, it will lead to questions that can be incorporated in prospective trials down the road. This was an analysis of patients diagnosed with ovarian cancer between January 1, 2011 to February 28, 2021. And they had to have had stage 3 or 4 disease and had to have initiated frontline platinum-based chemotherapy, and have at least 12 weeks of follow-up time after their first-line treatment.

In the analysis, we define kind of what a high risk versus a low-risk patient would be. And, you know, moderate risk patients would have stage 3 disease, no visible residual disease, primary debulking surgery, and BRCA mutation. And the high-risk patients would have at least 1 or more of these high-risk factors, stage IV disease, visible residual disease present, or no surgery at all, interval debulking, or no surgery, and then BRCA, wild type BRCA, unknown, or BRCA missing.There were these 2 cohorts that were defined as moderate risk patients and high-risk patients. Then we looked at how these risk factors, and the number of risk factors that affected time to the next treatment and overall survival.

What results were presented during ESMO this year?

What we found was that having more risk factors was associated with decreased time to next treatment, and worse overall survival.

The hypothesis is that these risk factors as they accumulate, they potentially lead to worse outcomes. And, you know, that is important for the design of future clinical trials in that you preset these high risk and lower risk categories to and use them as a stratification factor so that these groups are balanced within your treatment arms.

The other thing I want to point out is the majority of patients have about 2 high-risk factors. Only about 15% of the cohort had 4 high-risk factors, and only about 4% had no high-risk factors and 24% had 1. So, the majority sort fall into the 2 high-risk factors category. As shown in the Kaplan Meier curves, as you accumulate more of these high-risk features, you can clearly see that, that time to next treatment, and overall survival are impacted by the number of high-risk factors.

Do you have any key takeaways for community oncologists who are reading an article about this interview or watching a clip of this interview?

The take home message is definitely the number of high-risk factors that your patient has, can impact outcomes. For example, if you have a patient that has stage IV disease, she has visible residual disease present, or no surgery at all,, and she's had interval debulking surgery, and she's BRCA wild type, we could hypothesize that her treatment outcome is probably going to be inferior.

For me, when I know a patient like that, who I can maybe predict is not going to do as well, I really pay attention to quality of life, managing toxicity, and choice of therapy because potentially that patient is not going to be here with us for a long time. I think that's really important. We really need and what's missing is to study this prospectively. We need to identify these risk factors at diagnosis and kind of predict outcomes. But we do need better studies to tell us if we should treat these women differently or not. Physicians shouldn't necessarily hold therapy from patients, but maybe we would want to tailor that therapy to be more or better tolerated, and potentially study different treatment approaches in those higher-risk groups.

References:

Chase D, Perhanidis J, Gupta D, et al. Survival in patients with advanced ovarian cancer changes with cumulative number of risk factors, a US real-world analysis. Presented at: 2021 European Society for Medical Oncology Congress; September 16-21, 2021; virtual. Abstract 742P