Polypharmacy in Older Adults with Cancer

This paper reviews the adverse outcomes associated with polypharmacy and presents polypharmacy definitions offered by the geriatrics literature, examining the strengths and weaknesses of the various definitions, as well as exploring the relationships among these definitions and what is known about the prevalence and impact of polypharmacy in the geriatric-oncology population.

Learning Objectives

After completing this course, the reader will be able to:

1. Differentiate the multiple definitions of polypharmacy in order to be able to recognize it in your patient population.

2. Discuss the current data available in evaluating polypharmacy specifically in older adults with cancer and incorporate the data in your evaluation of older patients.

3. Summarize the agents or drug classes that may be deemed inappropriate in older adults to avoid prescribing medications for older patients that may lead to adverse drug events.

This article is available for continuing medical education credit at CME.TheOncologist.com

Abstract

The definition of “polypharmacy” ranges from the use of a large number of medications; the use of potentially inappropriate medications, which can increase the risk for adverse drug events; medication underuse despite instructions to the contrary; and medication duplication. Older adults are particularly at risk because they often present with several medical conditions requiring pharmacotherapy. Cancer-related therapy adds to this risk in older adults, but few studies have been conducted in this patient population. In this review, we outline the adverse outcomes associated with polypharmacy and present polypharmacy definitions offered by the geriatrics literature. We also examine the strengths and weaknesses of these definitions and explore the relationships among these definitions and what is known about the prevalence and impact of polypharmacy.

Introduction

The term “polypharmacy” can be defined in several ways, including an increased number of medications; the use of potentially inappropriate medications, which can increase the risk for adverse drug events; medication underuse; and medication duplication [1, 2]. Older adults are more likely to experience polypharmacy because they tend to have more medical conditions requiring pharmacotherapy [3–7]. The prevalence of polypharmacy in older adults ranges from 13% to 92% [8–39], depending on the definition of polypharmacy used and the characteristics of the study population evaluated (Table 1). Several adverse outcomes have been linked to polypharmacy, including increases in health care costs and adverse drug events, often leading to increased morbidity [7, 17, 18, 20, 24, 36, 39–62]. However, the evidence for a strong association between polypharmacy and an increased risk of mortality independent of other concomitant risk factors such as comorbidity remains unclear [30, 37, 39, 63–65].

Table 1.

Representative studies evaluating the prevalence of polypharmacy

Representative U.S.-based studies from January 1994 to September 2009.

Abbreviations: ER, emergency room; HEDIS DAE, Healthcare Effectiveness Data and Information Set Drugs to Avoid in the Elderly list; MAI, Medication Appropriateness Index; PC, prospective cohort; RC, retrospective cohort; VA, Veterans Affairs; XS, cross-sectional.

^aMedications deemed inappropriate for those with underlying cognitive dysfunction.

Cancer-related therapy adds to the risk of polypharmacy in older adults, as many new medications may be prescribed, including cancer therapy and supportive medications [66–70], but studies reporting on polypharmacy specifically in older adults with cancer remain sparse [71–74]. This review offers definitions of polypharmacy proposed in the geriatrics literature, examines the strengths and weaknesses of these definitions, and explores the relationship among these definitions and what is known about the prevalence and impact of polypharmacy. In addition, we describe tools for evaluating polypharmacy in daily practice and propose research into how this information can be applied in the geriatric oncology population.

Defining Polypharmacy

Several definitions of polypharmacy have been proposed (Table 2). It was initially defined as the number of medications being used concomitantly [75, 76]. Over time, the definition of polypharmacy shifted to include specific medications or scenarios thought to be more clinically relevant, such as the use of potentially inappropriate medications associated with a high risk of adverse effects in older adults [26, 77]. For example, two patients in their 70s both could be taking five prescription medications, yet their risk for an adverse drug event would be markedly different. The first hypothetical patient with breast cancer, hypertension, and coronary artery disease could be taking aspirin, atorvastatin, metoprolol, lisinopril, and anastrozole. The other could have breast cancer along with depression, atrial fibrillation, and peripheral arterial disease, and be taking amitriptyline, diazepam, warfarin, aspirin, and capecitabine. The second patient could potentially be at increased risk compared with the first patient because of (a) potentially sedating medications (amitriptyline and diazepam); (b) an anticholinergic medication (amitriptyline); and (c) medications that concomitantly augment bleeding risk because of a specific drug-chemotherapy interaction (capecitabine increasing the anticoagulant effects of warfarin) [78].

Table 2.

Definitions of polypharmacy

Abbreviations: OTC, over-the-counter; PIM, potentially inappropriate medication.

The clinical significance of distinguishing the “number of medications” from the actual medications taken has not gone unnoticed. A recent review pointed out that many studies use the terms “polypharmacy” and “inappropriate drug use” interchangeably [79]. This confusion is further highlighted in a review that has shown that the definition of polypharmacy in studies can be related either to the number or to the type of medications taken (i.e., medications with a high risk of adverse drug events or unnecessary medications), both of which can lead to an adverse drug event [80]. Table 2 illustrates the multifaceted components of how to define polypharmacy. The inherent difficulties of multiple definitions of polypharmacy become more evident when they are compared [81]. These definitions are described below.

Evaluating Polypharmacy: Current Methods

Number of Medications

Many community-dwelling older adults take multiple prescription medications [1, 2]. The likelihood of older adults receiving prescriptions from multiple providers compounds the risk of polypharmacy [4, 7]. In addition, an increasing number of medications has been associated with a higher frequency of potentially inappropriate medication use [15, 27, 39, 82]. A large number of medications may also place older adults at risk for drug-related complications, seen in a variety of clinical settings [42, 44, 55, 61, 62, 83].

The number of medications is also associated with a higher risk of a more subtle adverse drug event: medication nonadherence [34, 84–86]. This association may be related to the finding that many medication discrepancies (i.e., a discrepancy between what is prescribed and what is actually being taken) are identified in those receiving higher numbers of outpatient prescriptions [87]. As a result, nonadherence is a potential issue for older adults, especially because it has been associated with adverse health-related outcomes, including increased emergency room visits, hospitalization rates, and the potential for increased morbidity and mortality [88].

Nonprescription medication use, excluding herbal or complementary agents, should also be accounted for when considering the number of medications. Approximately 48% to 63% of older adults take at least one vitamin/mineral; 26% to 36% take an herbal, complementary, or alternative medication; and up to 50% take two to four over-the-counter medications on a regular basis [1, 89]. The likelihood of taking such agents increases with age [90]. The use of nonprescription medications increases not only the total number of medications taken but also the risk for drug interactions [90–92]. However, older adults with multiple medical conditions may require this level of pharmacotherapy. As a result, additional definitions of polypharmacy have been developed, including the use of “unnecessary” or potentially inappropriate medication use described below.

Potentially Inappropriate Medications

Two of the approaches most frequently used to evaluate potentially inappropriate medication use in older adults have been the Beers criteria and the Medication Appropriateness Index (MAI) described below (Table 3).

Table 3.

MAI and Beers criteria

Abbreviation: MAI, Medication Appropriateness Index.

Beers Criteria and Its Derivations

The Beers criteria consist of a list of medications deemed inappropriate for use in older adults, divided into 2 components: (a) specific drugs or drug classes that are considered inappropriate for use in older adults because they are either ineffective or pose unnecessarily high risk where a safer alternative exists; and (b) drugs that may be inappropriate for use in older adults based on the presence of coexisting diseases or conditions [93]. The Beers criteria have been updated twice since 1991 [93–95]. The most recently updated first and second components of the Beers criteria are outlined below (Tables 4 and 5, respectively). Polypharmacy studies using the most recent Beers criteria have been reported in a variety of clinical settings across several countries [8–10, 12, 14–25, 28–30, 32, 34–39, 46, 77, 96–103].

Table 4.

Comparison of the Beers criteria (first component), Zhan criteria, and HEDIS DAE list

Table 4.

(continued)

^aThe Beers criteria do NOT consider methylphenidate as inappropriate.

^bBoth the Zhan criteria and the HEDIS DAE list consider atropine and combination products as inappropriate in addition to the other anticholinergic agents.

^cBoth the Zhan criteria and the HEDIS DAE list consider phenobarbital as inappropriate.

Abbreviations: CNS, central nervous system; COX, cyclooxygenase; GI, gastrointestinal; HEDIS DAE, Healthcare Effectiveness and Data Information Set Drugs to be Avoided in the Elderly; NSAID, nonsteroidal anti-inflammatory drug; SIADH, syndrome of inappropriate antidiuretic hormone secretion.

Adapted with permission from National Committee for Quality Assurance. HEDIS 2009 National Drug Code (NDC) List.

Available online at http://www.ncqa.org/tabid/891/Default.aspx, accessed October 5, 2009; and from Fick DM, Cooper JW, Wade WE et al. Updating the Beers criteria for potentially inappropriate medication use in older adults: Results of a US consensus panel of experts. Arch Intern Med 2003;163:2716–2724, copyright ©2003 American Medical Association. All rights reserved.

Table 5.

The 2002 Beers criteria for potentially inappropriate medication use in older adults: Considering diagnoses or conditions

Abbreviations: CCB, calcium channel blockers; CNS, central nervous system; COPD, chronic obstructive pulmonary disease; INR, international normalized ratio; MAOI, monoamine oxidase inhibitor; NSAID, nonsteroidal anti-inflammatory drug; SIADH, syndrome of inappropriate antidiuretic hormone secretion; SSRI, selective serotonin release inhibitor; TCA, tricyclic antidepressant.

Reprinted with permission from Fick DM, Cooper JW, Wade WE et al. Updating the Beers criteria for potentially inappropriate medication use in older adults: Results of a US consensus panel of experts. Arch Intern Med 2003;163:2716–2724, copyright ©2003 American Medical Association. All rights reserved.

In addition to the number of medications, older adults receiving medications identified as potentially inappropriate by the Beers criteria have an increased risk for polypharmacy-related adverse outcomes [1, 2, 104], including increased rates of adverse drug reactions [17, 45, 50, 105], hospitalization [30, 39, 45–47, 53, 59, 60], emergency room visits [18, 45, 60], falls [40, 51, 56–58], fractures [52, 58], and lower scores on measures of health-related quality of life [19, 106].

The Beers criteria were further modified by Zhan et al. to develop a streamlined list of potentially inappropriate medications and to delineate medications that carry a higher risk for side effects than others [107]. The drugs listed by the Zhan criteria are compared with the Beers criteria in Table 4. The Zhan criteria identify fewer at-risk medications ultimately deemed inappropriate by expert consensus [81]. Some studies have used both the Zhan and Beers criteria concurrently [25, 81, 99, 108, 109]. The simplicity of the Zhan criteria, however, makes them an easier screening tool in population-based evaluations of polypharmacy [14, 25, 107, 109, 110].

The Beers criteria and their derivations have also been used as potential quality indicators. For example, The Healthcare Effectiveness and Data Information Set (HEDIS) is a program designed by the National Committee for Quality Assurance to identify standards of care for 71 clinical measures across 8 domains; it is used by >90% of health care insurance providers, including Medicare and Medicaid. Recently added to the HEDIS measures in 2007 and revised in 2008 is the list of “Drugs to be Avoided in the Elderly” (DAE) [111]. The DAE lists potentially inappropriate medications for older adult patients, incorporating a curtailed list of high-risk medications similar but not identical to those identified by the Beers criteria (Table 4), and can be found online as well (http://www.ncqa.org/tabid/892/Default.aspx).

In addition to evaluating for potentially inappropriate medications, the potential for clinically significant drug-drug interactions is also being evaluated by insurance plans for older adults. These plans are now incorporating Beers and “Beers-like” indices. As per an initial insurance-based analysis in 2007, almost 25% of approximately 30,000 beneficiaries received at least one prescription for a medication considered inappropriate by the Beers criteria, and up to 6% were reported as having had an adverse drug event [112]. As a result, the HEDIS program has adopted such quality measures to curtail potentially inappropriate medication use and thus adverse drug events among older adults.

Medication Appropriateness Index

The MAI uses 10 items to assess the degree of appropriateness of a particular medication along a 3-point Likert rating scale (Table 6) [113]. If a medication receives at least one “inappropriate” score on any item, it is deemed inappropriate overall. Several modifications have since been applied to the original MAI. Some studies have incorporated the following modifications: (a) taking into account that some items may be more suitable in particular clinical contexts than others [114, 115]; (b) summating the item scores to create an overall single score of medication appropriateness [60, 115, 116]; and (c) condensing the parameters to just three (indication, efficacy, therapeutic duplication) [26, 116]. The MAI has been applied in several clinical scenarios, including in hospitalized [26, 27, 96, 113, 114, 117] and ambulatory patients [3, 13, 60, 77, 97, 115, 116, 118], as well as used in evaluating medications taken as-needed in addition to regularly scheduled medications [117].

Table 6.

Medication Appropriateness Index (MAI)

Abbreviation: DK, Don't know.

Reprinted from Hanlon JT, Schmader KE, Samsa GP et al. A method for assessing drug therapy appropriateness. J Clin Epidemiol 1992;45:1045–1051, copyright 1992, with permission from Elsevier.

Unlike the Beers criteria, the MAI has not been extensively evaluated in outcomes-based studies. However, higher MAI scores have also been associated with higher rates of hospitalization and emergency room visits as well as a higher risk of adverse drug reactions [97, 116]. Moreover, higher MAI scores are associated with lower self-reported health scores among older adults [27].

Comparison of the Beers Criteria (and Derivations) Versus MAI

Shortcomings of all approaches derived from the Beers criteria include the following: (a) the list is not entirely exhaustive and needs to be updated periodically as new drugs are introduced; (b) it does not assess specific aspects of polypharmacy such as inappropriate dosing, which the MAI assesses; (c) it does not take into account that some “inappropriate” medications may prove beneficial for a particular patient under specific circumstances [1, 2, 77, 119].

Some weaknesses of the MAI include the following: (a) not enough data may be present to apply all 10 of the items; (b) it takes approximately 10 minutes to apply the MAI to each medication; (c) many studies have used more than one evaluator to ensure consistent scoring; and (c) it has been studied primarily in older veterans [26, 27, 60, 113, 116, 120].

Medication Duplication or Underuse

Medication duplicity, which can lead to unnecessary medication use and thus increase the risk of adverse drug events and potential drug interactions, is a criterion evaluated by the MAI, but this component of polypharmacy may still be overlooked. The Unnecessary Drug Use Measure is a modified form of the MAI developed specifically to incorporate these properties in the assessment of polypharmacy [121].

Neither the Beers criteria nor the MAI addresses the full scope for potential drug interactions as well as medication underuse, which refers to the situation in which the addition of a particular agent may actually prove beneficial for a patient with a specific disease [2, 77, 119, 122–128]. Such medication underuse has been associated with adverse effects in older adults and can contribute to drug-related hospitalizations beyond those attributable to adverse drug reactions or nonadherence [129–131]. Separate measures have been developed to address this component specifically [117, 132, 133]. Furthermore, studies have shown discordant results among these different approaches in evaluating the prevalence of polypharmacy. As a result, a combined and/or more comprehensive approach using one or more criteria should be considered [81, 119]. For example, the Screening Tool of Older Persons' Prescriptions (STOPP) and Screening Tool to Alert doctors to Right Treatment (START) criteria have recently been formulated to address multiple components of potentially inappropriate medication use/duplicity and medication underuse, respectively [134–136].

Herbal and Complementary/Alternative Medications

Herbal or complementary/alternative medication (CAM) is becoming increasingly prevalent among adults in the U.S. [89, 137–139]. Studies from the 1990s and early 2000s demonstrated that herbal/CAM use among older adults ranged from 6% to 15% [137, 138]. More recent studies report prevalence rates of 26% to 36% [89, 139]. However, this number may underestimate the true prevalence as demonstrated by a study that reported more than half of older adults do not disclose such use to their physicians [137].

An evaluation of herbal/CAM use is not typically included in standard definitions of polypharmacy described above. However, herbal/CAM use can increase the risk for drug interactions [90, 138]. Many of these interactions pertain to herbal agents such as garlic, ginkgo, and ginseng, which increase the bleeding risk associated with antiplatelet and anticoagulant agents such as aspirin and warfarin [138]. As such, herbal/CAM use should be incorporated as part of any assessment of polypharmacy.

Intervention Studies to Decrease Polypharmacy

Most intervention studies have been limited to the implementation of a pharmacist or interdisciplinary team to review medication usage, leading to reduction in the number of medications and/or use of potentially inappropriate medications among older adults in a variety of clinical settings [96, 140–147]. Overall, these studies have shown that these approaches have led to a significant reduction in suboptimal prescribing, and thus potential for adverse drug events in otherwise susceptible older adult patients. However, the impact on such intervention on clinical outcomes may depend upon the particular population. For example, the use of pharmacist-led review of prescription drug appropriateness and subsequent modification translated into less frequent adverse drug events in an outpatient setting but not in those older adults going from hospital discharge to long-term care facilities [144, 145, 147].

However, in many settings, a review of the medication list by a pharmacist or interdisciplinary team may not be available. In these situations, the implementation of electronic drug databases may be useful to help identify at-risk drugs, drug classes, dosages, and schedules [97, 148, 149]. Several electronic drug databases are available to clinicians. One study reviewed several databases and suggested that LexiComp (http://www.lexi.com), Clinical Pharmacology (http://www.clinicalpharmacology.com), and Micromedex (http://www.micromedex.com) had overall high-quality scores based on a composite evaluation of their scope, completeness, and ease-of-use in ability to answer several clinical questions [150]. However, their comparison specifically in a geriatric- or geriatric oncology–based setting has not been reported. Furthermore, the clinical significance and/or relevance of potential drug interactions and thus the “risk” of a certain drug or drug combination require clinician interpretation.

Polypharmacy in Older Adults with Cancer

Prevalence and Clinical Significance

Older adults with cancer are potentially vulnerable to the adverse effects of polypharmacy because cancer treatment often involves exposure to chemotherapy and other adjunctive or supportive medications that may increase the risk of drug interactions. Furthermore, the majority of adults with cancer are ≥65 years, with pre-existing medical conditions requiring pharmacotherapy [151]. A recent workshop sponsored by the National Institute of Aging and the National Cancer Institute reported that the prevalence and impact of medication use in the management of older adults with cancer is an unexplored area that mandates further investigation [152].

Only a few studies have evaluated the prevalence of polypharmacy specifically in geriatric oncology patients. One study reported that 63% in this group had a potential adverse drug interaction, with the majority of such patients receiving at least eight medications on average, and more than half of these interactions classified as moderate-to-severe risk [73]. When applied to outpatients receiving chemotherapy compared with supportive care only, this prevalence decreased to 27% and 31%, respectively [68, 69].

Another study evaluated the number of both prescription and nonprescription medications in older outpatients receiving chemotherapy for a variety of cancer types [74]. These patients were ≥65 years, had three comorbid conditions on average, and were receiving nine medications and at least three chemotherapeutic and/or supportive medications (mainly antiemetics). Several potential chemotherapy-drug interactions were identified; however, the frequency of adverse drug events or chemotherapy toxicity was not reported.

Most recently, the 2003 Beers criteria have been used to evaluate polypharmacy in older adults with cancer, one in an oncology-specific Acute Care for the Elderly unit and another in an outpatient setting [71, 72]. The mean age of patients evaluated was 73.5 and 74 years, respectively. Beers criteria–based prevalence of polypharmacy was 21% and 11%, respectively. Both studies were coupled with pharmacist-based interventions in medication review and subsequent modification, with 53% and 50% of patients, respectively, leading to reduction in the number of at-risk medications.

Herbal/CAM Use in Older Adults with Cancer

Herbal/CAM use can pose a significant risk in older adults with cancer. Its use in adults with cancer in the U.S. has been evaluated in several studies, with a prevalence ranging from 25% to 91%, depending on the study population and the definition of CAM used [153–189]. Only one study focused on older cancer patients, reporting a CAM prevalence of 33%, but limited the cancer type to breast, colorectal, prostate, and lung [189]. Predictors for herbal/CAM use have included the following: (a) female sex [153, 155–157, 160, 161, 164, 167, 169, 172, 181, 189]; (b) younger age [154, 156, 159, 160, 162, 172, 175, 176, 185, 186, 189]; (c) higher education levels [153, 158, 162–165, 169, 172, 176, 177, 185–187]; (d) higher income levels [165, 177, 179, 186]; (e) higher scores on measures of cancer-related physical and/or mental symptoms [153–155, 167, 170, 175, 184, 189]; and (f) advanced disease [155, 157, 162, 165]. However, none of these studies focused on herbal/CAM use in the context of polypharmacy in older adults with all cancer types or associated herbal/CAM use with outcomes.

A study evaluating outpatients undergoing chemotherapy demonstrated that almost a quarter to a half reported taking an herbal supplement or vitamin, respectively [182]. In evaluating all supplements, the 5 most frequently used supplementary agents were vitamin C (47%), a multivitamin (46%), vitamin E (42%), coenzyme Q10 (23%), and selenium (22%). When excluding vitamins, the 5 most commonly used supplementary agents were coenzyme Q10 (23%), selenium (22%), eicosapentaenoic acid (fish oil) (20%), garlic (18%), and zinc (17%).

The potential interactions of such herbal agents with chemotherapy have been reviewed [190]. For example, irinotecan has augmented gastrointestinal toxicity in patients concomitantly taking St. John's wort [191]. Those agents deemed a higher risk are those with competing cytochrome P450 and/or P-glycoprotein interactions such as garlic, ginseng, Echinacea, St. John's wort, ginkgo, and kava. This finding is of concern because garlic, ginseng, and gingko remain in the top 10 of the most frequently used herbal agents among adults nationally as of 2006 [89]. Although the prevalence of herbal/CAM use has not been fully elucidated specifically in older adults with cancer, the importance of evaluating herbal/CAM use has led some centers to provide resources for both cancer patients and their providers to evaluate an individual agent's potential benefits as well as toxicities [192].

Knowledge Gaps

Several gaps remain in our knowledge of polypharmacy in the geriatric oncology population (Table 7). First, to our knowledge, no prospective, longitudinal studies have reported the association of polypharmacy with cancer therapy toxicity or other adverse drug events. Second, the risk of drug-chemotherapy and drug-drug interactions in this target population needs to be further explored. Third, prior studies have used only single methods in identifying and/or measuring polypharmacy, such as number of medications or the Beers criteria; however, multiple methods of evaluating polypharmacy may provide greater insight into the associated risk of adverse drug events and determine which approaches are more closely linked to that risk. Specific attention to over-the-counter medication or herbal/CAM use in these evaluations of polypharmacy in older adults with cancer is also needed.

Table 7.

Knowledge gaps in evaluating polypharmacy in older adults with cancer

Future Directions

Polypharmacy in its various guises is a common problem facing older adults. In this article, we describe several common definitions of polypharmacy in the geriatric population, but, regardless of definition, polypharmacy has been clearly linked with several adverse outcomes, including increased risk of adverse drug reactions [17, 45, 50, 97, 105]; medication nonadherence [34, 84–86]; hospitalization [30, 39, 45–47, 53, 59, 60]; emergency room visits [18, 45, 60, 97, 116]; falls and/or fractures [40, 51, 56–60]; and lower self-reported health scores [19, 27, 106].

Given the added degree of pharmacologic complexity that chemotherapy and cancer-specific supportive care may engender, older adults with cancer are more vulnerable to the risks associated with polypharmacy. Studies directed toward prevalence and associated outcomes in this unique group of older adults are under way (ClinicalTrials.gov Identifier: NCT00477958). These studies will allow better evaluation of potential drug interactions, herbal/CAM use, and predictors of polypharmacy in addition to chemotherapy toxicity in this vulnerable patient population.

Meanwhile, based on what we already know about polypharmacy in older adults with cancer, we would recommend these steps to hematologists and oncologists treating these vulnerable patients:

1. Perform a careful review of the patient's list of medications, including indications and dosages.

2. Directly inquire about over-the-counter and herbal/complementary agents.

3. Evaluate in advance the potential interactions between the chemotherapy regimen and other medications to minimize drug interactions and subsequent toxicity; discuss with pharmacy staff where appropriate.

4. Consider use of electronic drug databases that may help identify at-risk drugs, drug classes, dosages, and schedules, bearing in mind the limitations of such tools, especially if pharmacy-based support is not readily available or accessible.

5. Maintain an open and active line of communication with the patient's other medical providers regarding changes or additions to medication lists.

6. Continue to perform routine medication reconciliation at every clinical visit in conjunction with pharmacy and/or nursing staff where appropriate.

The knowledge that we have gained thus far from the geriatrics literature can facilitate oncologists in developing more effective strategies to assess, monitor, and ultimately prevent polypharmacy in older adults with cancer.

Author contributions

Conception/Design: Ronald J. Maggiore, Arti Hurria, Cary P. Gross

Provision of study material or patients: Ronald J. Maggiore

Collection/Assembly of data: Ronald J. Maggiore

Data analysis and interpretation: Ronald J. Maggiore

Manuscript writing: Ronald J. Maggiore, Arti Hurria, Cary P. Gross

Final approval of manuscript: Arti Hurria, Cary P. Gross