Adherence to PARP Inhibitor Therapy Among Women with Ovarian Cancer
Keywords: Oral anticancer agents, adherence, PARP inhibitors, ovarian cancer
Highlights
Approximately one-quarter of patients may be suboptimally adherent to PARP inhibitor (PARPi) therapy.
Adherence did not differ when adjusting for dose adjustments and delays in therapy.
Risk of adherence may differ based on type of PARPi and indication for treatment.
Abstract
Objective:
To evaluate medical adherence for patients with ovarian cancer who initiated treatment with a PARP inhibitor therapy, and to identify factors associated with nonadherence.
Methods:
We used the MarketScan Database to identify ovarian cancer patients who initiated PARP inhibitor therapy between January 1, 2008 and December 31, 2017. The primary outcome was adherence defined as ≥80% proportion of days covered (PDC). A secondary outcome included adherence assessed using the medication possession ratio (MPR). Multivariable logistic regression analysis was performed to assess the relation between PDC and explanatory variables. Sensitivity analysis evaluated the impact of dose adjustments and toxicity-related delays on adherence.
Results:
Among 170,976 patients diagnosed with ovarian cancer, 151 met inclusion criteria. The median time from diagnosis to initiating therapy was 33 months. Overall, 40 (26.5%) were non-adherent based on a PDC less than 80%. Non-adherent patients were more likely to receive niraparib and have a shorter duration of therapy (p < 0.05). No significant impact of age, comorbidities, insurance plan, or year of PARP inhibitor initiation on non-adherence was found. Sensitivity analysis showed non-adherence ranged from 11.3% to 41.1%, depending on the definition used. When accounting for possible dose adjustments, 21.2% of patients were non-adherent.
Conclusion:
This population-based study found that a quarter of ovarian cancer patients may be sub-optimally adherent to PARP inhibitor therapy. Future research should focus on identifying patients at risk for nonadherence and interventions to improve adherence.
1. Introduction
Traditionally, patient-administered oral anti-cancer agents have played a minor role in ovarian cancer treatment compared to cytotoxic parenteral therapies. The introduction of orally administered poly(ADP-ribose) polymerase (PARP) inhibitors has transformed ovarian cancer treatment. Since 2014, three PARPis have received FDA approval for the treatment of epithelial ovarian cancer, including fallopian tube and primary peritoneal cancer. Indications include maintenance therapy in recurrent platinum-sensitive cancer, treatment of recurrent disease, and, most recently, maintenance therapy in the frontline setting. While initial approvals were for BRCA-associated cancers, more recent indications include all patients regardless of homologous recombination deficiency (HRD) or BRCA status.
Oral anti-cancer agents offer advantages such as convenience and improved patient engagement, but may be associated with nonadherence, especially in the maintenance setting where patients may be less inclined to continue therapy in the absence of symptomatic disease. Adherence is defined as the proportion of prescribed doses taken over a specific period. It is influenced by patient factors (physical and cognitive limitations), social factors (financial resources, living environment), healthcare-related factors (communication of benefits), and therapy-related factors (regimen complexity, side effects). Cancer patients are particularly vulnerable to nonadherence due to age and polypharmacy.
As PARPi use becomes more prevalent, understanding adherence is crucial. This study examined factors associated with nonadherence among ovarian cancer patients receiving PARPi therapy.
2. Methods
A retrospective cohort study was conducted using the IBM MarketScan database, which includes de-identified inpatient, outpatient, and pharmaceutical claims for over 250 million unique patients. Patients with a diagnosis of ovarian cancer who underwent oophorectomy between 2008 and 2017 and initiated PARPi therapy (niraparib, olaparib, or rucaparib) were included. Continuous health plan enrollment and prescription drug coverage for at least six months before and after PARPi initiation were required. Patients with multiple PARPi claims on the same day, zero day supply, or less than 28 days of PARPi were excluded.
Primary Outcome:
Nonadherence to PARPi therapy, defined as a PDC lower than 80%. PDC was calculated as the number of days covered with a PARPi supply divided by the number of days from the first to last filled date plus the last fill's days of supply. MPR was also calculated as a secondary measure.
Sensitivity Analyses:
Evaluated different PDC thresholds (≤70% and ≤90%) and accounted for dose reductions and treatment delays.
Statistical Analysis:
Categorical variables were compared using Chi-square tests; continuous variables with Wilcoxon rank-sum tests. Multivariable logistic regression estimated adjusted odds ratios for non-adherence.
3. Results
3.1 Study Population
Of 170,976 patients with a diagnosis of ovarian cancer, 151 met all inclusion criteria. The median age was 57, and the median comorbidity score was 3. The median time from diagnosis to PARPi initiation was 33 months. At the time of PARPi initiation, 43.1% were prescribed 5–10 drugs, and 31.8% were prescribed more than 10 drugs. Most patients received olaparib (57.0%), followed by niraparib (29.8%) and rucaparib (13.3%). The median duration of treatment was 150 days.
3.2 Adherence
Based on a PDC less than 80%, 40 (26.5%) of 151 patients were non-adherent. Non-adherent patients were more likely to receive niraparib (42.5% vs. 25.2%, p = 0.02) and had a shorter duration of therapy (127 days vs. 165 days, p < 0.001). No significant differences were found in adherence by age, polypharmacy, comorbidity score, or insurance type. Multivariable models showed no significant predictors of nonadherence.
Using MPR, 34 (22.5%) were non-adherent. Only duration of therapy was associated with nonadherence (130 days vs. 168 days, p < 0.01).
When accounting for dose reductions and toxicity-related delays, 21.2% were non-adherent. Shorter duration of therapy remained associated with nonadherence (127 vs. 161 days, p = 0.003).
Sensitivity analyses showed nonadherence ranged from 11.3% (PDC <70%) to 41.1% (PDC <90%).
4. Discussion
The shift from intravenous to self-administered oral PARPi therapy raises concerns about medication adherence. Approximately one-quarter of women with ovarian cancer may be suboptimally adherent to PARPi therapy. Suboptimal adherence may compromise treatment outcomes or lead to treatment resistance. Clinicians may misinterpret disease progression as lack of therapeutic response, resulting in premature cessation of therapy.
Nonadherence was more common among patients prescribed niraparib. This may relate to its use in the maintenance setting or specific toxicity profiles, as patients on niraparib are more likely to experience grade 3 or 4 toxicities such as fatigue, anemia, and thrombocytopenia. The use of administrative data may misclassify intentional treatment delays as nonadherence.
Compared to studies of oral endocrine therapies in breast cancer, where adherence rates range widely and decrease with longer treatment duration, adherence to PARPi therapy in this real-world setting may be overestimated due to the relatively short treatment duration in this study.
High cost is another factor influencing adherence. PARPi therapy can cost around $14,000 per month, and out-of-pocket expenses vary depending on insurance. Financial assistance programs are commonly used but not captured in many databases.
Other factors associated with poor adherence to oral anticancer agents include age, marital status, comorbidities, regimen complexity, lack of awareness of benefit, and side effect profile. Qualitative studies highlight mental burden, perceived importance of the medication, difficulty managing side effects, and need for better provider communication as barriers to adherence.
Interventions to improve adherence include pill counts, self-report, pharmacy claims, electronic monitoring, and mobile health (mHealth) tools. While some interventions are effective, cost and feasibility limit their widespread use.
Strengths and Limitations:
This study used a large national database, enhancing generalizability. The sample size was small due to the rarity of ovarian cancer and specific drug use. Limitations include reliance on prescription refills as a proxy for adherence, inability to measure partial adherence, and lack of data on financial assistance programs.
5. Conclusion
The increasing use of PARP inhibitors in ovarian cancer has shifted chemotherapy administration to a self-managed, oral regimen, raising concerns about adherence. Approximately 25% of patients may be suboptimally adherent, with higher nonadherence among those prescribed niraparib and those with shorter treatment duration. No significant associations were found with age, comorbidities, insurance, or polypharmacy. Future research should focus on identifying patients at risk for nonadherence saruparib and developing interventions to improve adherence and outcomes.