Adherence to oral agents and supportive medication in oncology

Oral anticancer therapy has shifted much of cancer treatment from the infusion chair to the patient’s home as 25-30% of all compounds in the oncology development pipeline are administered orally.¹ But that convenience comes with a new clinical challenge: patients are no longer under direct observation while taking each dose. The assumption that patients reliably adhere to a potentially life-saving oral regimen — given the gravity of their diagnosis — has proven consistently false. A systematic review of oral cancer therapies reported wide variability in adherence (46–100%), with rates declining over time; In some studies, adherence was as low as 16%.²

As a result, adherence has become a core determinant of outcomes for targeted therapies, endocrine therapy, and several supportive treatments in oncology.^3,4,5 In chronic myeloid leukaemia (CML), adherence below 90% abolishes the chance of achieving complete molecular response.⁶ In breast cancer, 49% patients remain adherent to the full course of adjuvant endocrine therapy (AET).⁷ Even modest deviations from prescribed regimens can significantly impact disease control.⁸ Additionally, inappropriate adherence behaviors—including overadherence and excessive persistence beyond recommended therapy—may increase the risk of toxicity and adverse effects, ultimately compromising treatment safety and outcomes.⁹ In cancer-associated thrombosis, non-compliance is one of the leading identifiable causes of recurrent venous thromboembolism (VTE).¹⁰

Measuring adherence to oral oncology agents: A major challenge

For HCPs, the issue is not whether adherence matters, but how to measure it, when to suspect non-adherence, and which interventions are worth the effort. The clinical imperative is clear: adherence monitoring must be built into the architecture of oncology care, not reserved for patients who fail to respond.

Adherence measurement in oncology requires a nuanced approach due to dose modifications, treatment interruptions, and toxicity-related adjustments. In current practice, adherence measurement relies on a layered combination of objective and subjective tools, each with specific trade-offs: Commonly used metrics include:

• Claims-based metrics
  • Proportion of Days Covered (PDC): It is the most common adherence measure in routine research, calculated as days supplied divided by days in the observation window. PDC is recommended for administrative data studies and is widely used across oral cancer therapy research, including studies of tyrosine kinase inhibitors(TKIs), antiandrogens, and endocrine therapy.^5,11
  • Medication Possession Ratio (MPR): Calculated from pharmacy claims as total days’ supply dispensed divided by the observation period. A threshold of >80% commonly used to define adherence.^12,13 MPR can overestimate adherence if refills are early or if stockpiling occurs.¹⁴

For oncology HCPs, the strengths of these methods are their scalability and objectivity. Their weaknesses are equally important: they measure medication acquisition, not ingestion, and they cannot distinguish intentional dose omission from administrative refill delays.^5,15

• Self-report metrics

Self-report tools are useful in clinic because they are quick and can identify barriers that claims data miss.¹ The Morisky Medication Adherence Scale (MMAS-8) is the validated self-report instrument used in oncology. In CML, BAASIS self-report significantly outperforms clinician-estimated visual analogue scales (VAS) as a measure of adherence, study, found that clinician-reported VAS should not be relied upon for adherence assessment.^1,16

The Simplified Medication Adherence Questionnaire (SMAQ) can detect common patterns such as forgetfulness, stopping therapy when feeling worse, and missed doses over short recall windows.¹⁷ In one regional breast cancer cohort study, the SMAQ classified 65% of patients as adherent, and more side effects were associated with poorer adherence.¹⁷

• Electronic monitoring

Microelectronic monitoring system (MEMS) caps record bottle opening events and are considered the gold standard in research settings for implementation adherence. In CML, MEMS data from the landmark Marin et al. study established the pivotal 90% threshold below which complete molecular response was not observed.⁶

A practical compromise is to combine refill data, patient interview, and side-effect assessment. This triangulation is often better than relying on one imperfect measure.^4,16

Adherence Metrics Comparison

Metric	Definition	Strengths	Limitations
PDC14	Proportion of days covered by medication	Reflects continuity; preferred metric	Does not capture dosing accuracy
MPR13,14	Medication supplied over time	Easy to calculate	Overestimates adherence
Pill Count1	Remaining pills vs expected use	Simple, low cost	Manipulable by patients
Self-report1	Patient-reported adherence	Quick and practical	Recall and desirability bias
Electronic Monitoring1	Digital tracking of usage	High accuracy	Costly, limited availability

Adjuvant endocrine therapy: Long-term adherence and clinical consequences

Adjuvant endocrine therapy (AET), including tamoxifen and aromatase inhibitors, is a significant treatment for hormone receptor–positive breast cancer. The treatment is preventive and often lasts 5 to 10 years. Despite its proven efficacy, adherence remains suboptimal, with discontinuation rates approaching 50% at 5 years.^18,21 A systematic review of 26 studies found a mean five-year adherence rate of just 66.2%, with adherence declining by an average of 25.5% from year one to year five.¹⁹

Non-adherence is not simply a quality-of-life issue: early discontinuation is associated with increase in mortality in patients with hormone receptor–positive breast cancer and non-adherence more broadly with a 49% increase.²⁰ Barriers to adherence include treatment-related adverse effects (e.g., arthralgia, vasomotor symptoms), the long duration of therapy, and patient perceptions regarding the balance between benefits and risks.²¹ In a recent cohort study, adherence for AET for breast cancer measured by SMAQ was 65%, and side-effect burden was strongly associated with non-adherence.¹⁷

AIs demonstrate equal or superior adherence to tamoxifen in real-world settings, despite their well-known arthralgic side effects. Five-year non-persistence was 31.0% for AIs versus 47.2% for tamoxifen in meta-regression analyses.²³ Where tamoxifen is prescribed for breast cancer, awareness of CYP2D6 interactions with common co-medications is warranted, as oncologists are frequently unaware of these interactions — one study found 58% of such prescriptions were undocumented in oncology records.²⁴

Key metrics for AET adherence in practice

For AET in breast cancer, both PDC and MPR are widely used to measure adherence and can be extracted from pharmacy dispensing or electronic health record data, reviewed at each annual follow-up. For AET, persistence may be more important than exact adherence percentage.^11,12

Detecting non-Adherence to tyrosine kinase inhibitors in CML

In Chronic Myeloid Leukemia (CML), tyrosine kinase inhibitors (TKIs) dramatically improve survival; however, their effectiveness is highly dependent on adherence.⁴ They are taken daily for years, and treatment interruption can affect molecular response. Poor adherence has been linked to loss of cytogenetic response and treatment failure, making it both a therapeutic and monitoring issue.^4,16

A population-based study of 2,870 Korean patients with CML demonstrated a strong association between higher TKI adherence and improved outcomes, with high adherence (MPR ≥0.95) yielding markedly superior 5-year overall survival (97.2%) compared with moderate (85.2%) and low adherence (70.2%).¹² These findings underscore adherence as a key modifiable determinant of survival in CML.

The clinical challenge is detecting non-adherence before it translates into loss of molecular response. Non-adherence can be detected through both indirect and direct measures such as PDC/MPR, patient self-reports or biological markers, particularly BCR-ABL (Breakpoint Cluster Region – Abelsontranscript) levels.^2,6

Several evidence-based signals should prompt an adherence conversation:

• Failure to achieve milestones: Absence of a major molecular response or loss of a previously achieved complete cytogenetic response without a clear biological explanation could be indicative of non-adherence.⁶ Studies using MEMS data has established that no molecular response was observed with adherence ≤80%, and no complete molecular response with adherence ≤90%.⁶
• Unexplained BCR-ABL transcript rise: A rising BCR-ABL:ABL ratio, particularly in the absence of new mutations, is a highly actionable clinical indicator of non-adherence that should precede any decision to escalate TKI therapy.⁶
• Patient self-overestimation: Approximately 60% of CML patients who miss at least one dose in a given month believe they ‘always’ follow their treatment recommendations.²⁵ This systematic overestimation means direct questioning must be structured, non-judgmental, and specific, using validated tools rather than clinician impression or open-ended questioning.
• High-risk demographic signals: Younger age, female sex, third-line TKI therapy, and lack of membership in a patient support group have been independently associated with electronic monitoring non-adherence in CML.¹⁶ Clinicians should proactively monitor these groups more intensively rather than waiting for molecular failure.

Approximately 40% of CML patients on imatinib were categorised as non-adherent in large claims database analyses.²⁶ Before attributing disease progression to pharmacological resistance, HCPs should systematically rule out non-adherence — including considering a supervised dosing period or structured monitoring interval.

Supportive medications: The overlooked adherence problem in oncology

Supportive medicines such as antiemetics, steroids, acid suppression and antidiarrheals can determine whether a patient can stay on their anticancer regimen. If these drugs are not taken properly, toxicity control fails, and the patient may stop cancer treatment altogether. This is especially relevant with oral regimens that require self-management across multiple drugs.^27,28

A 90-patient clinical study comparing chemotherapy adherence to supportive drug adherence found that adherence to chemotherapy was 21% higher than adherence to supportive medications.²⁹ This finding carries an important implication: patients prioritise their ‘cancer drug’ over the supportive agents that make it tolerable.

Non-adherence to antiemetic regimens is a common cause of uncontrolled chemotherapy-induced nausea and vomiting (CINV), which can lead to dehydration, hospitalization, and treatment delays.³⁰

Practical monitoring strategies for supportive medications: When monitoring supportive medicines, the goal is not to police every pill but to identify whether poor supportive adherence is undermining cancer therapy. Use patient-reported outcome tools at every cycle visit rather than relying on clinician-estimated toxicity.³⁰ Proactively simplify regimens, address dexamethasone use explicitly in patients on checkpoint inhibitors, where steroid-sparing strategies are relevant.^31,32

Multicomponent interventions integrating nurses, pharmacists, and digital solutions have demonstrated the strongest evidence for improving adherence to oral anticancer medicines broadly, and this model applies with equal force to the supportive care regimen.³³

Oral anticoagulants in cancer-associated thrombosis: Adherence as a safety variable

Venous thromboembolism (VTE) is the second leading cause of death in patients with cancer.³⁴ Despite effective anticoagulation, patients carry a 10–20% risk of recurrent VTE, and non-compliance is consistently identified as one of the principal reasons for recurrence alongside inadequate dosing, thrombocytopenia, and disease progression.¹⁰

Current guidelines recommend either direct oral anticoagulants (DOAC) or subcutaneous heparin for the treatment of cancer-associated thrombosis.¹⁰ In one study, adherence measured by PDC at an 80% threshold was similarly high for heparin and DOACs, around 95%, although persistence was longer with DOACs.¹¹

Assessing adherence to DOACs:

• Direct adherence questioning: Integrating a standardized question about missed doses into routine oncology reviews is the most practical first-line strategy. Refill gap analysis using pharmacy dispensing data can supplement clinical questioning.¹¹
• Clinical consequence monitoring: New or recurrent thrombotic symptoms should prompt adherence assessment before considering dose escalation or switching agents.^10,11
• Anti-Xa assay in selected situations: Anti-FXa assays calibrated to specific agents may be used for suspected non-adherence, providing an objective confirmation of drug exposure.³⁵

VTE recurrence due to non-adherence is a marker of avoidable therapeutic failure.¹⁰ Embedding adherence assessment into each oncology encounter reflects the shared-care nature of this clinical problem.

Adherence monitoring strategies across oncology settings

Clinical setting	Preferred monitoring approach	Key indicator of non-adherence
Breast cancer (AET)17	MPR ≥80%	Early discontinuation
CML (TKIs)6	BCR-ABL monitoring	Suboptimal molecular response
Supportive care36	Patient-reported outcomes (PROs)	Uncontrolled symptoms
Cancer-associated thrombosis11	Refill data, clinical follow-up	Recurrent thrombosis/bleeding

Clinical takeaway for HCPs in oncology

Adherence to oral oncology therapy is a silent determinant of outcomes across every therapeutic category, from curative-intent targeted agents to the supportive medications that make chemotherapy tolerable. The field lacks a universal objective marker, but a multimodal assessment strategy — combining pharmacy refill analysis, validated self-report, molecular response monitoring, and structured clinical questioning — provides an actionable framework for clinical practice.

Key imperatives for HCPs are:

• Measure adherence proactively, not reactively, before treatment failure
• Recognize that patients consistently overestimate their own adherence; structured non-judgmental questioning is more reliable than open-ended enquiry
• Use PDC ≥80% or MPR ≥85% as the standard reporting threshold in pharmacy-based analyses for oral anticancer agents^2,11
• Adherence to supportive medications is systematically lower than anticancer drug adherence — embed PRO-based symptom monitoring into every chemotherapy cycle
• In CAT, the oral route confers an adherence advantage; integrate adherence assessment into every oncology visit and use anti-FXa assays selectively when adherence is uncertain
• Multicomponent interventions — pharmacist-led counselling, digital tools, and structured nursing support — have the strongest evidence base for improving adherence to oral oncology regimens

To learn and explore behavioral interventions used to enhance adherence to cancer therapies, please read our article Behavioral interventions to enhance adherence in oncology – Acare HCP Global

This article was written with the assistance of generative AI technology and reviewed for accuracy.