Adherence to Targeted Oral Anticancer Medications

Abstract

The use of targeted oral anticancer medications (OAMs) is becoming increasingly prevalent in cancer care. Approximately 25–30% of the oncology drug pipeline involves oral agents and there are now over 50 OAMs approved by the Food and Drug Administration. This change represents a major shift in management of patients with cancer from directly observed, intermittent intravenous therapy to self-administered, oral chronic therapy. The increased prevalence of OAMs raises the issue of adherence in oncology, including understanding the challenges of adherence to OAMs. This review focuses on studies of adherence for patients taking molecularly targeted OAMs for breast cancer, chronic myelogenous leukemia (CML), gastrointestinal stromal tumors (GIST), non-small cell lung cancer (NSCLC), and renal cell carcinoma (RCC). We then discuss barriers to adherence and studies performed to date testing interventions for improving adherence. Finally, we discuss future areas of investigation needed to define and improve adherence to OAMs in targeted therapy for cancer.

Introduction

Over the last decade, the rise in the use of targeted oral anticancer medications (OAMs) represents a major shift in management of patients with cancer from directly observed, intermittent intravenous therapy to self-administered, oral chronic therapy (Gebbia et al., 2012; Partridge et al., 2002; Ruddy et al., 2009). Through recent understanding of genetic, genomic, and molecular changes involved in tumor progression, many oral anticancer therapies have been developed to target abnormal proteins and signaling pathways specific to cancer cells. In 2008, it was estimated that 25–30% of the oncology drug pipeline involved oral agents, most of them targeted, with approximately 40% of all OAMs having been approved within the last seven years (Weingart et al., 2008). Table 1 lists recently approved targeted OAMs with many more in clinical development.

Table 1.

Selected FDA Approved Targeted Oral Anticancer Medications.

Generic Name	Trade Name	Year Approved*	Target Gene or Receptor	Indication
Dasatinib	Sprycell	2006	BCR-ABL	Philadelphia chromosome-positive (Ph+) chronic myeloid leukemia (CML)
Imatinib	Gleevec	2001	BCR-ABL	Ph+ CML Gastrointestinal stromal tumors (GIST)
Nilotinib	Tasigna	2007	BCR-ABL	Ph+ CML
Bosutinib	Bosulif	2012	BCR-ABL, Src	Ph+ CML
Ponatinib	Iclusig	2012	BCR-ABL	ALL and CML
Vemurafenib	Zelboraf	2012	BRAF V600E	Melanoma
Vismodegib	Erivedge	2012	SMO	Basal cell carcinoma
Ruxolitinib	Jakafi	2011	JAK1/2	Myelofibrosis
Gefinitib	Iressa	2003	EGFR	Non-small cell lung cancer (NSCLC)
Erlotinib	Tarceva	2004	EGFR	NSCLC and pancreatic cancer
Crizotinib	Xalkori	2011	EML4-ALK	NSCLC
Abiraterone	Zytiga	2011	CYP17A1	Prostate cancer
Enzalutamide	Xtandi	2012	AR	Prostate cancer
Regorafenib	Stivarga	2012	VEGFR2-TIE2	Colon cancer
Lenalidomide	Revlimid	2006	Anti-tumor, immunomodulatory	Multiple myeloma
Lapatinib	Tykerb	2007	EGFR, HER2/neu	Breast cancer
Sunitinib	Sutent	2006	PDGFR, VEGF, cKIT, RET, CSF-1R, flt3	GIST Renal cell carcinoma (RCC)
Sorafenib	Nexavar	2006	VEGFR, PDGFR, C-Raf, B-Raf, MAP Kinase, cKIT	RCC Hepatocellular carcinoma
Pazopanib	Votrient	2009	VEGF, c-kit, PDGFR	RCC Soft tissue sarcoma
Axitinib	Inlyta	2012	VEGFR1–3, PDGFR, cKIT	RCC
Everolimus	Afinitor	2009	mTOR	Breast, RCC, soft tissue sarcoma and renal angiomyolipoma

Abbreviations: BCR-ABL, fusion of Abelson (Abl) tyrosine kinase gene at chromosome 9 and break point cluster (Bcr) gene at chromosome 22; EGFR, epidermal growth factor receptor; EML4-ALK, rearrangement of echinoderm microtubule-associated protein-like 4anaplastic lymphoma kinase; HER2/neu, one of four membrane proteins in EGFR family; PDGFR, platelet-derived growth factor receptor; VEGFR, vascular endothelial growth factor receptor; RET, proto-oncogene, encodes receptor tyrosine kinase for the neurotrophic factor family; CSF-1R, colony stimulating factor 1; flt3, encodes receptor tyrosine kinase that regulates hematopoiesis; MAP Kinase, family of serine/threonine proteins responsible for regulating cellular activities, such as apoptosis; c-kit, tyrosine kinase stem cell factor receptor; SMO, Smoothened, a transmembrane protein involved in Hedgehog signal transduction; mTOR, mammalian target of rapamycin inhibitor; BRAF, gene encoding for B-Raf, member of Raf kinase family

^*year approved for initial indication.

Patients prefer oral medications over intravenous therapy. A study addressing self-reported patient preference for oral drugs versus intravenous palliative chemotherapy found that 92 of 102 assessable patients preferred oral chemotherapy, most importantly due to convenience and with the understanding that efficacy would not be sacrificed (Liu et al., 1997). Oncologists are also amenable to targeted OAMs, particularly in the palliative and adjuvant disease setting where quality of life is paramount (Benjamin et al., 2012). The rise in use of targeted OAMs has brought medication adherence to the forefront in oncology; for example, suboptimal adherence to targeted OAMs has already been shown to lead to worsened event free survival (EFS) in patients with chronic myelogenous leukemia (CML) (Ganesan et al., 2011). This review will summarize studies of adherence to targeted OAMs, barriers to adherence to targeted OAMs, and interventions to improve adherence in cancer patients.

Definition and Assessment of Adherence

Adherence, defined as the degree to which one conforms to recommendations about day-to-day treatment by the provider with respect to the timing, dosage, and frequency (Cramer et al., 2008), has long been an important issue for patients with other chronic diseases including diabetes (Skelly et al., 2009), hypertension (Bosworth et al., 2009), heart disease (Murphy et al., 2009), asthma (Horne, 2006), and HIV/AIDS (Gardner et al., 2010; Lima et al., 2009). Costs to the U.S. health-care system of nonadherence for patients with chronic disorders are estimated to be $300 billion/year (DiMatteo, 2004). Furthermore, multiple studies have shown that good adherence is associated with a reduction in healthcare costs (Sokol et al., 2005).

Techniques to quantify adherence are imperfect (Osterberg and Blaschke, 2005), and there is no consensus for a “gold standard” measurement of adherence (Table 2) (Font et al., 2012). Furthermore, no single standard definition of an acceptable adherence threshold exists in the literature, although a widely used one has been 80% (Banning, 2012; Spoelstra and Given, 2011). However, in an era of molecularly targeted drugs, 80% adherence may not be an appropriate benchmark because even small deviations from full (100%) adherence may result in resistance and treatment failure, as evidenced by studies of adherence in individuals with CML (Gater et al., 2012).

Table 2.

Measures of Adherence.

Measure	Pros	Cons
Direct
Direct observation	Most accurate	Not feasible in real-world practice
Serum drug levels	Objective measure of recent exposure to drug	Can be manipulated; acceptable ranges often unknown; assays not widely available
Indirect
Pill counts	Inexpensive	Difficult in real-world practice; easy to manipulate; may overestimate adherence; demeaning
MEMS (microelectronic event monitoring system)	Accurate data on when one opens the bottle; may be combined with reminder systems	Not easily feasible in real-world practice; expensive
Refill records	Objective higher level data; good for research purposes	Report fill rate and not actual intake; impractical for daily use
Biomarkers	May be important intermediaries to outcomes (e.g., hypertension with TKI use)	Few developed and validated
Outcomes	Most important variable	Difficult to discern nuances of adherence outside of clear extremes
Indirect and Subjective
Self-report	Quick; can use past validated instruments; does not require clinician time	Subject to significant bias such as the Hawthorne effect and overestimates adherence
Assessment by others	Inexpensive; allows for a dialogue	Hawthorne effect; time consuming
Diaries	Inexpensive; actively involves the patient	Subject to manipulation; demeaning; time consuming

Adherence to Targeted OAMs

In oncology, adherence to OAMs has been evaluated in a broad range of cancers (Mazzeo et al., 2011; Waterhouse et al., 1993) with adherence rates ranging from less than 20% to 100% (Partridge et al., 2003). The most robust data come from studies of women with breast cancer taking oral hormonal therapy (Banning, 2012) and patients with CML (Jabbour et al., 2012) taking imatinib. Seminal work in breast cancer (Fink et al., 2004; Hershman et al., 2010; Lash et al., 2006; Partridge et al., 2008; 2003; Waterhouse et al., 1993) has provided the basis for research examining adherence to targeted OAMs, particularly for “chronically critically ill” individuals with cancer (e.g., lung cancer, renal cell carcinoma). Research concerning adherence for women with breast cancer taking oral hormonal therapies has been well summarized; therefore, in this review, we focus on studies of adherence to molecularly targeted OAMs for persons with cancer (Table 3).

Table 3.

Salient Studies of Adherence to Targeted Oral Anticancer Medications.

Cancer	# of Patients	Oral Therapy	Adherence Measure	Adherence Definition	Outcome	Study
CML	267	Imatinib	MPR	Continuous	77.7% over 1 year; lower MPR associated with higher cost	Darkow et al., 2007
CML	169	Imatinib	BAAS; PC	Number of positive answers out of 4	14% perfectly adherent; 32.7% with at least one positive answer over 90 days	Noens et al., 2009
CML	592	Imatinib	MPR	Continuous	79% mean MPR over 12 months with 41% MPR <85%	Wu et al., 2010
CML	516	Imatinib	Patients’ clinic records	Interrupt >1 week	29.6% nonadherence rate; better survival in those adherent (5 year EFS 76.7% vs. 59.8%)	Ganesan et al., 2011
CML	87	Imatinib	MEMS x3 months	Continuous	Median adherence=97.6% (range 24–104%); 26% with <90%; 21% ≤85% →18 fold higher rate of losing CCyR	Marin et al., 2010 Ibrahim et al.,2011
CML	430	Imatinib	MPR	Continuous	60% adherent based on 85% threshold.	St. Charles et al., 2009
CML	328 550	Nilotinib Dasatinib	MPR from two claims databases	Continuous	Dasatinib users (73.9%) were less adherent than nilotinib (80.0%) users.	Guerin et al., 2012
CML	53 197	Nilotinib Dasatinib	MPR from research database	Continuous	Nilotinib users were almost 2× more likely to have poor adherence than dasatinib users.	Yood et al., 2012
NSCLC	65	Erlotinib	Self-report, MEMS, plasma erlotinib levels	Continuous	Ongoing	Timmers et al., 2011
NSCLC	30	Erlotinib	Self-report	Continuous	Individuals had mostly medium (n=11) to high (n=14) adherence rates.	Lucca et al., 2012
NSCLC	200	Erlotinib	BAAS, VAS, PC, missed appointments	Continuous	Disease control was higher in the intervention cohort (63%) compared with control (44%). The number of adverse events and patient-reported adherence were highly correlated (r=0.105; p=0.0001).	Gebbia et al., 2013
RCC	1080	Various	MPR from	Continuous	81% had adherence rates ≥80%	Hess et al., 2011
RCC	49	Various	MEMS	Continuous	Adherence=98.9%	Wolter et al., 2012
GIST	28	Imatinib	BAAS	No positive answers out of 4	24–29% at 90 days	Mazzeo et al., 2011
CML & GIST	~4,000	Imatinib	MPR	Continuous	MPR in GIST was 73%.	Tsang et al., 2006

Abbreviations: MPR, medication possession ratio; NA, not available; D/C rate, rate of complete medication discontinuation; PC, pill count; BAAS, Basel Assessment of Adherence Scale with Immunosuppressive Medications; MEMS, Medication Event Monitoring System; CCyR, complete cytogenetic response; VAS, Visual Analog Scale.

Breast cancer

Two targeted OAMs approved for women with breast cancer are lapatinib and everolimus. Lapatinib is a dual inhibitor of the epidermal growth factor receptor (EGFR) family and is approved for combination use in women with metastatic breast cancer (Opdam et al., 2012). The most common side effects of lapatinib include diarrhea, nausea, vomiting, palmar plantar erythrodesthesia, and rash. Everolimus is a mammalian target of the rapamycin (mTOR) inhibitor and its most common side effects are stomatitis, hypertension, edema, fatigue, and cytopenias (Vinayak and Carlson, 2013). Although no published studies of adherence to everolimus in breast cancer were found, two studies examined adherence to lapatinib. In a study of 69 women, adherence was assessed using self-report, medication diaries, and pharmacy controlled drug boxes (Addeo et al., 2011). Adherence was 82%, with a 65% rate of dosing violations (not further specified). In a second retrospective study using MarketScan® (a large database of pharmacy records for individuals with private insurance) data, refill records were used as a surrogate marker for adherence to lapatinib (Kartashov et al., 2012). Adherence measured by medication possession ratio (MPR), the proportion of days in the measured period covered by prescription claims, was 87% (n=666), with 22% of patients having an MPR <80%. Nonadherence was associated with concurrent IV chemotherapy use and more physician visits, with a trend toward higher cost.

Chronic myelogenous leukemia (CML)

Perhaps the best example of the profound shift to use of targeted OAMs is CML and imatinib mesylate (imatinib), a tyrosine kinase inhibitor (TKI) that blocks the adenosine triphosphate-binding site of the BCR-ABL tyrosine kinase. Imatinib has been shown to be effective in treating chronic and accelerated phases of CML as well as blast crisis (Jabbour and Kantarjian, 2012). Common side effects of imatinib include edema/fluid retention, fatigue, rash, nausea, anemia and leukopenia, muscle cramps, and transaminitis (Breccia et al., 2012).

In the last five years, many reports examining adherence to imatinib have been published with adherence rates ranging from 60% (St Charles et al., 2009) to 97% (Noens et al., 2009). Generally, adherence was determined by MPR, but thresholds for defining “adequate adherence” varied (Darkow et al., 2007). The most commonly used thresholds were 85% (0–100% scale) (St Charles et al., 2009; Wu et al., 2010) and 90% (Ibrahim et al., 2011; Marin et al., 2010). Predictors of adherence to imatinib in CML such as gender, age, number of concomitant medications, dose of imatinib, time since diagnosis, living alone, time on therapy, physician characteristics and interaction with the patient, patient knowledge and self-beliefs about their disease, prescription co-payment amount, and selfreported functional status and quality of life have been mixed between studies and well summarized previously (Gater et al., 2012).

Importantly, adherence to imatinib has been found to correlate to response to treatment. In a retrospective study in India of patients received imatinib free of charge, 29.6% were nonadherent (defined as not returning to the clinic for one week or more to receive drug) and had a decreased 5-year EFS (combination of cytogenetic and hematologic milestones, including death) compared to their adherent counterparts (p=0.011) (Ganesan et al., 2011). Univariate analysis demonstrated that prolonged symptom duration before diagnosis, treatment with hydroxyurea for more than one month prior to imatinib therapy, and nonadherence were associated with worse EFS, but only nonadherence was significant in the multivariate analysis (HR 1.6; p=0.048). Another study found a strong correlation between adherence measured by a medication event monitoring system (MEMS) (≤90% or >90%) and 6-year probability of a major molecular response (p=0.001) and complete molecular response (p=0.002) (Marin et al., 2010). Others have found similar results with Ibrahim et al. (2011) showing adherence rate and failure to achieve molecular response being the only predictors for loss of complete cytogenetic response and discontinuation of imatinib therapy in CML.

Dasatinib and nilotinib are second generation tyrosine kinase inhibitors that primarily target BCR-ABL. Nilotinib is approved for treatment of drug resistant CML and dasatinib is approved for treatment of CML after imatinib therapy and for individuals with Philadelphia chromosome-positive acute lymphocytic leukemia. Both dasatinib and nilotinib have side effect profiles similar to imatinib, with a few important rare exceptions (Wei et al., 2010). When examining real-world adherence to second-line therapies with dasatinib and nilotinib, one study found that patients taking nilotinib were two times more likely to be nonadherent than patients taking dasatinib 100 mg daily (Yood et al., 2012). However, a second study found a higher mean MPR for individuals taking nilotinib (adjusted difference=0.061; p=0.002) compared to dasatinib, regardless of dose (Guerin et al., 2012). Both studies measured adherence using an MPR calculated from large claims databases and it is possible that the data may not reflect actual consumption of medication, skipped doses, or division of doses (i.e., 100 mg in two 50 mg doses).

Gastrointestinal stromal tumors (GIST)

GISTs are mesenchymal neoplasms originating in the gastrointestinal tract that share molecular activation of either the KIT or PDGFRA proto-oncogenes. As a result, TKIs have been used in the metastatic and adjuvant settings, leading to improved progression free survival and overall survival. Imatinib is approved in the adjuvant and metastatic settings and sunitinib in the metastatic setting only. Duration of therapy is three years for individuals completing therapy after surgical resection and indefinite for individuals with metastatic disease (Dasanu, 2012).

Previous work has shown that for individuals with advanced GIST, imatinib trough plasma levels correlate with time to progression and objective response (Demetri et al., 2009; von Mehren and Widmer, 2011), supporting the importance of adherence. Tsang et al. (2006) found the overall MPR for patients with GIST was 73% and by 14 months of treatment, only 23% of CML and GIST patients were 100% persistent with their therapy. In the companion ADAGIO study (GIST), adherence was prospectively evaluated over 90 days in patients with GIST (N=28) taking imatinib (Mazzeo et al., 2011). There was a 29% nonadherent rate four weeks prior to baseline and 24% at follow-up. Further supporting the importance of persistence of therapy, Le Cesne et al. (2010) randomized 50 patients with advanced GIST after three years of no progression on imatinib therapy to either continue or interrupt treatment. The primary endpoint was progression free survival (PFS) and after a median follow-up of 35 months post-randomization, two-year PFS was 80% in the continual arm and 16% in the interruption arm (p<0.0001).

Several descriptive studies of patients with GIST have examined the role of toxicity on medication use and adherence. In one survey (N=173), decreasing toxicity from severe to moderate levels was most important to persons taking imatinib rather than completely eliminating all toxicity. Reducing heart failure from moderate to mild and diarrhea from severe to moderate had the largest effects on subjects’ evaluation of adherence (Hauber et al., 2011). An ethnographic study (N=50) revealed the most common strategies for remaining adherent included obtaining family support, setting reminders, taking medicine at routine times, and storing medicine in prominent places (Macdonald et al., 2012).

Non-small cell lung cancer (NSCLC)

Erlotinib and gefitinib (no longer approved for use in the U.S.) were developed to inhibit EGFR over-expression in individuals with NSCLC. Erlotinib is approved for second- and third-line treatment of advanced NSCLC and maintenance therapy for advanced stage NSCLC after initial treatment of chemotherapy. Erlotinib is also recommended for first-line therapy for individuals with EGFR mutations (Majem and Pallares, 2013). Common side effects include rash, diarrhea, fatigue, and anorexia (Hotta and Kiura, 2011).

Three studies examined adherence to erlotinib for patients with NSCLC and two of those studies examined interventions to improve adherence to erlotinib. Gebbia et al. (2013) evaluated the impact of a treatment-monitoring intervention on adherence for patients with advanced NSCLC receiving erlotinib as second-line therapy in two cohorts: 1) a retrospective non-interventional phase monitoring 50 participants without a treatment management strategy; and, 2) a prospective interventional phase following 150 participants who received a treatment-management program, including identification of a caregiver, patient/caregiver education and training about treatment and side effects of therapy, a calendar for follow-up visits, and a dedicated facsimile phone line to receive instructions or use of a fast-track visit system. Adherence was measured using multiple methods and generally patient self-reported adherence was higher than adherence measured by pill counts. Disease control rate (complete response plus partial response plus stable disease) was 44% in the first cohort and 63% in the second cohort (p=0.0368). Also, a significant correlation was found between the number of adverse events and adherence (r=0.176, p=0.035).

Lucca et al. (2012) tested an educational intervention to enhance knowledge and adherence to erlotinib while monitoring for side effects in 30 patients. Adherence behaviors were measured with the 8-item Morisky Medication Adherence Scale (MMAS-8); however, it is unclear if the tool was adapted for patients with cancer. MMAS-8 adherence scores were medium to high, and the mean number of erlotinib adverse events was 2.48 per patient; 22% reported four or more side effects. An ongoing prospective observational cohort study (Timmers et al., 2011) is examining adherence over 16 weeks for patients with NSCLC. Adherence is measured using MEMS, several questionnaires, and plasma levels of erlotinib. Findings from this study will provide information about adherence and short-term persistence to erlotinib.

Renal cell carcinoma (RCC)

For persons with metastatic RCC, four oral VEGF inhibitors (sorafenib, sunitinib, pazopanib, axitinib) and one oral mTOR inhibitor (everolimus) are available for use with more in clinical trials. Common side effects include hypertension and hand-foot syndrome (Mendez-Vidal et al., 2012). Complete responses, although rare, have been noted and some patients are now living for over three years with sequential treatments (Albiges et al., 2012; Posadas and Figlin, 2012; Sonpavde et al., 2012). At the same time, cost of these medications for patients and the medical community can be high, totaling more than tens of thousands of dollars per year per patient (Shih et al., 2011). Furthermore, ongoing trials are examining the role of targeted OAMs as adjuvant treatment for non-metastatic RCC, possibly leading to significant expansion of the use of these agents in the future (ClinicalTrials.gov, NCT01235962).

Few studies have investigated adherence for patients with metastatic RCC. Hess et al. (2011) examined adherence to sorafenib, sunitinib, and everolimus and found 81% had adherence rates of 80% or higher. Limitations of this study involve the inclusion of infused agents such as temsirolimus, bevacizumab, and interferon; use of medication fill rates via claims records rather than actual patient reported intake; and, the majority of patients had commercial healthcare insurance. At the 2012 ASCO Annual Meeting, Wolter et al. (2012) presented preliminary data of a prospective observational study measuring adherence using MEMS technology. Adherence was 98.9% and while the described adherence rate may be high, several significant limitations of the above study are worth noting, including a small sample size, free medications supplied to patients throughout the study, a European only cohort, and limited qualitative information. With a presumed transition to perceiving RCC as a chronic disease (Escudier, 2012), it becomes vitally important to identify at-risk populations and factors leading to nonadherence so targeted interventions can be developed.

Barriers to Adherence

Potential barriers to adherence to OAMs in general have been identified and are summarized in Figure 1 as well as in previous reviews (Gater et al., 2012; Ruddy et al., 2009). Barriers to adherence specifically for targeted OAMs are still being elucidated, but in particular may include cost, side effects, and timing with food.

Figure 1.

Schematic of factors and barriers involved in adherence to targeted OAMs and possible ways to improve adherence.

Patients with advanced cancer are living longer and thus the expense of long-term targeted OAMs may lead to a substantial burden, in particular for older adults. Targeted OAMs routinely cost thousands of dollars per month with variable levels of co-payments and premiums incurred by patients (Shih et al., 2011; Weingart et al., 2008; Winkeljohn, 2007). Findings from a grounded theory study (n=13) examining the process of medication-taking for individuals with NSCLC taking erlotinib suggest that increased cost can lead to premature therapy discontinuation as nearly all participants referenced the high cost of erlotinib (Wickersham et al., 2012). Examining the association of OAMs’ cost and adherence rates in Canada, Ebrahim et al. (2012) surveyed 453 patients and found self-reported adherence to be 80% with 51% of patients having costs of ≥$100/month.

The interaction of food with targeted OAMs is important. Most OAMs are labeled to be taken without food because they often have large positive food effects with much greater bioavailability when taken with food (Szmulewitz and Ratain, 2013). This can lead to complex dosing schedules (pill needs to be taken some hours before or after a meal) which are difficult to sustain over a long period of time. A prospective study of 77 patients taking OAMs for various malignancies showed that 43% of those taking an OAM with a significant food-drug effect did not know last time they ate before taking their OAM, 23% did not know that their OAM had a food-drug effect, 21% intentionally skipped or cut back on their OAM with 38% of those not informing their physicians and 20% had some difficulty understanding the directions on the bottle (Muluneh et al., 2012).

Finally, side effects are also common with targeted OAMs and vary depending on the drug. In the same grounded theory study above, all participants referenced side effects with the most common being rash and diarrhea; in one case, these side effects were referred to as “social inhibitors” (Wickersham et al., 2012). Another qualitative study examined reasons for nonadherence for individuals with CML and found both intentional (e.g., side effects) and unintentional (e.g., forgetfulness) reasons for nonadherence (Eliasson et al., 2011). In general, a conceptual model of adherence developed in the setting of tyrosine kinase inhibitor use in persons with CML involves an interplay of predisposing factors (patient, disease, treatment, and physician characteristics), patient interactions with their physician, and patients’ knowledge and beliefs, which can be expanded further to other targeted OAMs (Gater et al., 2012).

Interventions to Improve Medication Adherence in Persons with Cancer

Several studies examined interventions to improve adherence to OAMs with others underway. Spoelstra et al. (2013) developed a nursing intervention to improve adherence using a Symptom Management Toolkit®, based on a modified health belief model approach, and an automated voice response (AVR) reminder system. Participants were randomized to one of three groups: (AVR) system alone (n=40), AVR with strategies to manage symptoms and adherence (n=40), or AVR with strategies to manage adherence (n=39). Adherence was measured using medical record audit, patient self-report, and pharmacy report. Participants had primarily breast, colon, or lung cancers, received non-hormonal agents for cancer treatment, and adherence was defined as 80–100% over the past seven days. Findings showed 42% were nonadherent, with missed doses increasing with regimen complexity. Symptom severity declined over time in all groups and no difference was found in adherence rates; higher adherence was related with lower symptom severity across groups.

Another study reported preliminary results of an OAM management clinic for patients with breast, colon, rectal, and lung cancer (Wong et al., 2012). Thirty patients enrolled into the clinic by their oncologist were retrospectively analyzed for variables such as depression, adherence, and persistence to cancer treatment. The most common drug prescribed was capecitabine and patients on average had 12.7 concurrent medications. Interventions to improve adverse drug events, nonadherence, drug interactions, and medications errors decreased over time with complete resolution or improved response seen in 67% of patients.

An Irish study of 101 patients examined perceptions of education and safety of OAMs (Graham et al., 2012). Fifteen percent of patients took targeted OAMs and the rest received conventional chemotherapy. When starting OAMs, 17% did not understand the medication; this was improved by physician (p=0.03) or hospital-based nurse (p=0.04) and provision of information booklets (p=0.04). Patients were unaware of drug-drug interactions in 30% of cases and 20% were not aware of any safety issues. Patients who had been given information leaflets were significantly more aware of safety including careful handling (p<0.001), storage conditions (p=0.02), and safe disposal (p<0.001). Patients attending nurse-led oral chemotherapy clinics were significantly more aware of safety issues (p=0.04) and had improved adherence.

Finally, Sommers et al. (2012) conducted a feasibility study of a telephone education guide to improve adherence in 30 patients with gastrointestinal cancer. Participants received oral chemotherapy regimens, including therapy with targeted OAMs. The intervention included physician and nurse-delivered education in the clinic followed by telephone support from clinic nurses. Adherence was measured using the MMAS-8 adapted for oncology and a medication diary. Findings showed that the adapted MMAS-8 was a feasible measure of adherence with high self-reported rates (mean 7.89 on 0–8 scale) consistent with medication diaries. Figure 1 summarizes some other possible interventions for patients on targeted OAMs.

Conclusion and Future Directions

Patient adherence to medication is essential to optimize clinical outcomes, minimize toxicity, decrease bias in clinical trials, and reduce healthcare costs. The field of oncology is rapidly evolving, with oral anticancer medications becoming a common treatment modality. Given the often narrow therapeutic margins of some targeted OAMs, their enormous costs, and significant side effects, it is critical to understand barriers to adherence for individuals taking these drugs and ways to maximize adherence. To date, studies examining adherence to OAMs have provided valuable information, but most of the work has been conducted in patients with CML. Future research will need to expand to include individuals with other cancers such as NSCLC, RCC, prostate cancer, GIST, breast cancer, melanoma, and others for which oral drugs are being developed. Future directions may include development of oncology-specific adherence tools, defining an optimal threshold for adherence for persons with cancer taking targeted OAMs, and development of interventions for improving adherence and the implementation of systematic programs for all patients, including the rapidly growing older adult population.