http://ict.sagepub.com/Integrative Cancer Therapies

http://ict.sagepub.com/content/12/2/126The online version of this article can be found at:

 DOI: 10.1177/1534735412447582

2013 12: 126 originally published online 27 June 2012Integr Cancer TherLeanna J. Standish, Erin Sweet, Eleonora Naydis and M.Robyn Andersen

the Effectiveness of Integrative Oncology Offered in Community ClinicsCan We Demonstrate That Breast Cancer ''Integrative Oncology'' Is Effective? A Methodology to Evaluate

  

Published by:

http://www.sagepublications.com

can be found at:Integrative Cancer TherapiesAdditional services and information for    

  http://ict.sagepub.com/cgi/alertsEmail Alerts:

 

http://ict.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

http://ict.sagepub.com/content/12/2/126.refs.htmlCitations:  

What is This? 

- Jun 27, 2012OnlineFirst Version of Record  

- Mar 4, 2013Version of Record >>

at UNIV WASHINGTON LIBRARIES on May 21, 2013ict.sagepub.comDownloaded from

Integrative Cancer Therapies12(2) 126 –135© The Author(s) 2012Reprints and permission: sagepub.com/journalsPermissions.navDOI: 10.1177/1534735412447582http://ict.sagepub.com

Introduction

The use of complementary and alternative medicine (CAM)1 among breast cancer patients is significant and growing.2,3 It is estimated that between 50% and 80% of women in the United States with breast cancer supplement their conventional medical treatment regimen with some form of CAM therapy.1,4-9 Treatment is influenced by the growth of integrative oncology (IO) clinics that are operat-ing especially in large metropolitan areas of the United States. In spite of the growing number of cancer centers offering IO care, there has not yet been an adequate con-trolled study of clinical, laboratory, or quality of life out-comes associated with IO care.10,11

IO care can improve patient outcomes through several possible mechanisms. These include the potential of improving the effectiveness of conventional treatments, treating side effects, and improving compliance. CAM treatments may facilitate cancer rehabilitation in the post-treatment setting, prevent cancer relapse, improve survival, and provide palliative end-of-life care. There are a number

447582 ICTXXX10.1177/1534735412447582Standish et alIntegrative Cancer Therapies2012© The Author(s) 2010

Reprints and permission: http://www.sagepub.com/journalsPermissions.nav

1Bastyr University, Kenmore, WA, USA2Fred Hutchinson Cancer Research Center, Seattle, WA, USA

Corresponding Author:Leanna J. Standish, Bastyr Integrative Oncology Research Center, Bastyr University, 14500 Juanita Dr NE, Kenmore, WA 98028, USA. Email: ljs@bastyr.edu

Can We Demonstrate That Breast Cancer “Integrative Oncology” Is Effective? A Methodology to Evaluate the Effectiveness of Integrative Oncology Offered in Community Clinics

Leanna J. Standish, ND, PhD, FABNO1, Erin Sweet, ND1, Eleonora Naydis, ND, LAc1, and M. Robyn Andersen, PhD, MPH2

Abstract

Background. Many women diagnosed with breast cancer receive both standard cancer treatment and care from providers trained in the emerging field of medicine called integrative oncology (IO) in which science-based complementary and alternative medical therapies are prescribed by physicians. The effectiveness of IO services has not been fully studied, so is yet unknown. Purpose. Determine if a matched, case-controlled prospective outcomes study evaluating the efficacy and safety of breast cancer IO care is feasible. Methods. Methodological proof of principle requires demonstration that (1) it is possible to find matched control breast cancer patients using the Surveillance, Epidemiology and End Results (SEER) Program’s western Washington Cancer Surveillance System (CSS) and (2) an IO clinic can recruit breast cancer patients into a matched controlled study. Results. A pilot study was conducted in 2008 (n = 14) to determine if matched controlled women could be identified in the western Washington SEER database. All 14 women who were approached agreed to participate. The cases were matched to the CSS along 5 variables: age and stage at diagnosis, race, and marital and Estrogen Receptors/Progesterone Receptors (ER/PR) status. Multiple matches were found for 12 of the 14 participants. Conclusion: A prospective cohort study with a matched comparison group is a feasible and potentially rigorous study design with high patient acceptability. It may provide valuable data for the evaluation of the effectiveness of IO care on patient health, relapse rate, and health-related quality of life (HRQOL). A federally funded matched case controlled outcomes study is currently under way at Bastyr University and the Fred Hutchinson Cancer Research Center.

Keywords

breast cancer research methods, integrative oncology, naturopathic medicine, traditional Chinese medicine

at UNIV WASHINGTON LIBRARIES on May 21, 2013ict.sagepub.comDownloaded from

Standish et al 127

of botanical substances with demonstrated potent anticancer activity in vitro and in vivo.12,13 Consultation with IO pro-viders is also likely to improve health behaviors through patient education on nutrition, exercise, and stress reduction techniques. IO providers also counsel patients to avoid inap-propriate CAM use that might lead to drug-herb or drug-nutrient interactions, potentially leading to reduced efficacy of standard care (SC). CAM therapy may also provide patients with an opportunity to participate in decision mak-ing about their treatment, leading them to feel that they have some control over their health. This may increase patients’ satisfaction with their care and improve their health-related quality of life (HRQOL). The placebo effect of CAM is likely high in the IO setting because of the emphasis placed on communication and development of a therapeutic alli-ance with each patient and his or her family to address all levels of health (body, mind, and spirit). For all these rea-sons, it would be valuable to assess the effects of IO holisti-cally and to assess a variety of outcomes, including patient HRQOL, and not just disease status or survival.

Integrative Oncology Studies to DateMany IO therapies involve the use of disease and immune-modifying botanicals and nutrients that have varying levels of in vivo evidence for anticancer activity. The anticancer mechanism of action of CAM therapies used in IO includes modulation of immunological and molecular mechanisms of cancer initiation, promotion, and metastasis. The basis of IO research has to date been of 2 types: (1) systematic reviews of randomized clinical trials (RCTs) and nonran-domized clinical trials of single natural product CAM treat-ments, such as mistletoe14,15 and Trametes versicolor mushroom,16-19 and (2) whole-person, health services–oriented retrospective and prospective studies of IO survival out-comes compared with various nonrandomized controls.

RCTs and Nonrandomized Clinical Trials of Single Natural Product CAM Treatments

The NIH has funded trials of both T versicolor and mistle-toe in Germany (NCT00006477), France (NCT00006354), and the United States (NCT00680667, NCT00079794, NCT0005235) to evaluate the safety and immunological effects of T versicolor and mistletoe in phase I and II trials since 2001.

Based on these data, many IO physicians prescribe both T versicolor and mistletoe to their cancer patients, and they are often included in core protocols in IO clinics in the United States. Other specific targeted IO therapies, such as curcumin, silymarin, resveratrol, garlic, omega-3

fatty acids, co-enzyme Q10, artemisinin, green tea, and others, although in common use in IO clinics, are now being evaluated in clinical trials as single agents. These products are available over the counter and used by breast cancer patients and, at some IO clinics, prescribed as tar-geted therapies. As patients and clinicians await the results of these trials, studies using other methodologies, as they are commonly used in practice, would be helpful.

Use of Historic Data as Controls in Retrospective StudiesRetrospective studies on survival outcomes in metastatic breast cancer patients receiving IO care have provided promising data regarding the possible survival benefit of IO care. Block et al20 were among the first to compare results in breast cancer patients treated at IO clinics with historical data published in breast cancer clinical trials. Median sur-vival in the IO-treated stage IV patients was 38 months, and 5-year survival was 27%; this compared favorably to his-toric clinical trial results in stage IV breast cancer patients published in the 1990s. Median survival reported in stage IV breast cancer ranged from 12 to 24 months and 5-year survival was 17% among comparison patient data. These results may indicate a clinically significant treatment effect of IO care. However, the possibility of differences between the IO cases and the women with whom they were com-pared make the interpretation of the results uncertain. CAM-seeking cancer patients tend to be white, educated, anxious, depressed, and employed, though this is also often true of clinical trial participants. Any differences in these demographics between the 2 groups might explain the dif-ference in survival.

Use of “Single-Visit” Controls to Evaluate IO Outcomes in Lung Cancer Patients

McCulloch et al21-23 have shown improved survival curves for stage I to IV lung and colon cancer patients receiving IO care at Pine Street Clinic located in the San Francisco Bay area. In 2006, this group presented retrospective outcomes data of survival in 239 non–small-cell lung carcinoma patients following treatment with an adjunctive natural medicine protocol that included science-based herbal and nutraceutical treatments similar to those used by other natu-ropathic and Chinese medicine IO physicians.21 The control group consisted of non–small-cell lung carcinoma patients who presented at the IO clinic but did not proceed with treatment. Treated patients showed an overall 44% reduced risk of death over 10 years compared with untreated con-trols, even when baseline treatment and health status adjustments were made. IO combined with conventional

at UNIV WASHINGTON LIBRARIES on May 21, 2013ict.sagepub.comDownloaded from

128 Integrative Cancer Therapies 12(2)

therapy, compared with conventional therapy alone, reduced the risk of death in stage I by 95%, stage II by 64%, stage III by 29%, and stage IV by 75%. Long-term IO care combined with conventional therapy improved sur-vival compared with conventional therapy alone. These data add indirect support for the hypothesis that adjunctive science-based natural medicine may have a positive sur-vival benefit in lung cancer. McCullough et al21-23 work is important because it demonstrates a significant clinical dif-ference between lung cancer patients who were exposed to IO care versus those who were motivated to make an appointment at the IO clinic but did not continue care after the first appointment. Results from the well-designed retro-spective studies of IO with conventional therapy by McCulloch et al indicate that prospective studies of IO care combined with conventional treatment are warranted.

Methodological and Ethical Challenges Associated With RCTs of IO

Designing a RCT of IO care’s safety and efficacy, and not just a single-agent treatment, is fraught with several meth-odological challenges. First, federally funded trials of IO care would require FDA approval to conduct research Investigational New Drug Applications (INDs) for each element of IO care that involves a therapeutic substance (T versicolor mushroom extract, mistletoe, curcumin, green tea, resveratrol, artemisinin, silymarin, and melatonin to mention just a few commonly used IO treatments). Second, an IO RCT would require the development of a single stan-dardized treatment algorithm that could be used as a strict protocol. Most IO clinics do not have strict standards of care or protocol guidelines of the sort needed for a defini-tive RCT, instead allowing providers to customize care to each individual patient. Third, once such a protocol is developed, an estimated effect size for the protocol as a whole will be needed. If IO treatments are synergistic and their combination is more than the sum of their individual effects, this effect size could not be computed based on the effects found in single-treatment studies. Estimates for con-ventional medicine RCTs examining behavioral and nutri-tional interventions commonly come from epidemiological research studies often conducted using nonrandomized study designs and matched comparison groups.

There is also a potentially significant problem for an IO care RCT in terms of patient recruitment. Patients already seeking IO care would be unlikely to enroll in a study with possible randomization to a nontreatment group. Similarly, those not already seeking IO care may be reluctant to sign up for randomization to IO activities that they might find burdensome and that are not demonstrated to improve health.

Building strong evidence for efficacy of IO care may likely require RCTs. However, until we know more about the likely effects of IO care, we cannot proceed with RCTs. Therefore, it is appropriate to use other designs to create evidence to support an RCT proposal. For now, prospective outcomes studies using matched comparison women to assess the effects of IO care holistically via a health services approach may be most appropriate. The ethics of using clin-ical data collected from patients receiving reimbursed health care to evaluate those health services has been evalu-ated favorably.24

Use of Matched Controlled Studies to Measure the Effect of CAM Treatments in Cancer Patients

Matched controlled studies of CAM in breast cancer patients are rare. One, however, has been done. Lesperance et al25 analyzed survival and recurrence outcomes in 90 women with unilateral, nonmetastatic breast cancer diag-nosed between 1989 and 1998. All women were patients registered in the British Columbia Cancer Agency, Vancouver Island Centre. Women, who were prescribed mega-doses of vitamins and minerals for at least 2 months by a single IO physician in addition to standard therapies, were enrolled as cases. Control women were matched based on data available in hospital records, including age, year of diagnosis, stage, TNM status, tumor grade, ER/PR status, and conventional oncology treatment history. Breast cancer–specific survival and disease-free survival times were not improved for the vitamin/mineral-treated group over those for the controls. These data support the use of a matched controlled design to evaluate IO effectiveness in breast cancer but are of limited value in evaluating these current IO practices for 2 reasons: (1) they did not control for CAM use in the control breast cancer patients and (2) current-day IO physicians rarely prescribe oral high-dose vitamins and minerals, except to correct deficiencies or for targeted short-term treatment.

Although matching designs have added value and valid-ity over historic literature controls, only clinics associated with major hospitals offering conventional treatment could use the methods used by Lesperance et al25 to examine the outcomes of their care. Even then the size of sample a clinic associated with a single hospital can collect and examine is limited. We also know that CAM use and cancer outcomes are predicted by additional variables that the Lesperance study did not use for matching and that are commonly not available in routine hospital records. CAM use among breast cancer patients has been associated with higher lev-els of depression and anxiety, and these are associated with poorer outcomes in cancer patients. Ideally, education and ethnicity, which predict both CAM use and cancer outcomes,

at UNIV WASHINGTON LIBRARIES on May 21, 2013ict.sagepub.comDownloaded from

Standish et al 129

need to be included as matching variables in future matched controlled outcomes studies. A well-designed IO prospec-tive outcomes study would also ideally include baseline measures of depression, anxiety, and HRQOL to be able to control post hoc for baseline status differences between IO cases and their matched controls.

A prospective matched controlled research design may provide adequate methodology to determine the safety, effectiveness, and cost-effectiveness of physician-provided IO care. Such designs do not restrict patient choice and do not require a FDA IND for each element of the protocol. The difficulty of such studies for most clinics is in identify-ing appropriate comparison patients and collecting appro-priate data on these women recruited for comparison. Using cancer registries to identify comparison patients who do not use IO could solve these problems. Such a study can be designed to include freestanding community IO clinics individually or networks of such clinics broadening the patient population studied and allowing for faster recruit-ment and enrollment, which could allow for more timely evaluation.

We feel that the IO field is now ready for well-designed matched, controlled, prospective epidemiological longitu-dinal studies of this type. But to whom exactly should IO clinic patients be compared? Clearly, a matched compari-son group should match the IO clinic cohort under study on as many disease status variables as possible. This is the advantage of working with a cancer registry. Matching vari-ables that are collected by a SEER cancer registry at patients’ diagnosis that could be included in IO matched controlled studies are shown in Table 1. Thus, patients in IO clinics in a cancer registry’s area of coverage could be iden-tified and comparison patients with similar demographic

characteristics and who were diagnosed with similar dis-ease states could be identified.

To examine the possibility that a cancer registry could provide the needed data to allow for the identification of a group of breast cancer patients suitable for study as com-parison patients for a future study, we worked with the local site for the National Surveillance, Epidemiology, and End Results (SEER) program of the National Cancer Institute (http://seer.cancer.gov) called the Washington State Cancer Surveillance System (CSS; www.fhcrc.org/content/public/en/labs/phs/projects/cancer-surveillance-system/about.html). The CSS is housed at the Fred Hutchinson Cancer Research Center (FHCRC) in Seattle. Like many cancer registries, it routinely provides data for epidemio-logical research studies.

This cancer registry has been found to reliably identify 99.8% of women diagnosed with breast cancer in its 13 county recruitment areas, This provides to a comprehensive listing of cancer patients and the possibility of identification of a representative sample of similar patients in a compari-son group formed via the registry. Demonstration of feasi-bility requires that (1) an IO clinic can recruit breast cancer patients into a matched controlled outcomes study and (2) it is possible to find matched control breast cancer patients using a cancer registry.

The Bastyr Integrative Oncology Research Center (BIORC) is an outpatient research center located on the Bastyr University (BU) campus in Kenmore, Washington. Established in 2009, BIORC offers state-of-the-science CAM therapies to cancer patients, which are provided by CAM physicians, that is, naturopathic physicians (NDs) who are board certified in naturopathic oncology and doc-tors of acupuncture and Oriental medicine (DAOM) who have specialty training in traditional Chinese medicine oncology. IO medical providers are supported by a shared nursing staff, and providers share patients’ medical charts. Complex cases are discussed among traditional Chinese medicine (TCM) and ND providers in a shared open office space. Progress notes are sent directly to each patient’s medical oncologist and primary care physician. Often, treatment decisions in complex stage IV cancer cases will be made jointly by written, phone, or e-mail communica-tions between IO and medical oncology providers.

At their first BIORC appointment, patients are invited to participate in a prospective outcomes study and, if agree-able, then consent to participate. After signing an informed consent form and a medical records release form, they are then asked to complete a 30-minute questionnaire, which is offered again at 6 months and then annually for 5 years or more. Baseline demographic data are acquired at the first appointment and used to identify each cancer patient in the CSS registry and to find potential matched comparison patients. The questionnaire also includes validated instru-ments to assess HRQOL; the Medical Outcomes Survey (SF-36), worry about cancer scale, Perceived Stress Scale,

Table 1. SEER Cancer Surveillance System Variable List

Patient personal identification data Name DOB Zip code at time of diagnosis SSNPatient demographic identification Race/Ethnicity Marital statusDisease markers at FOC Diagnosis date TNM classification Histopathology Tumor markers (ER/PR, Her-2/neu)Descriptive data Conventional therapeutic regimen Treatment timeline

Abbreviations: SEER, Surveillance, Epidemiology and End Results; DOB, date of birth; SSN, Social Security Number; FOC, first office visit; ER/PR, Estrogen Receptors/Progesterone Receptors.

at UNIV WASHINGTON LIBRARIES on May 21, 2013ict.sagepub.comDownloaded from

130 Integrative Cancer Therapies 12(2)

Meaning in Cancer Experience Survey, and Cancer Treatment Decision Making. Each patient is asked about their current marital and financial status and CAM use prior to their use of the clinic. Follow-up questionnaires assess general CAM use and adherence to IO prescriptions and treatments recommended at the clinic because these are potential predictors of cancer outcomes.

Matching IO Cases to the Cancer RegistryPrior to opening BIORC, we needed to ensure that the CSS contained adequate numbers of potential matches for IO breast cancer patients. In 2008, we designed a pilot study to demonstrate the acceptability to patients of such proce-dures and the feasibility of using the registry to identify women matched for demographic and disease characteris-tics at the time of their diagnosis, thus allowing us to determine whether there were a sufficient number of cases with similar prognostic characteristics in the registry to allow matching. This pilot study was approved by both the internal review boards of BU and of the FHCRC. We asked the CSS to find matched control women for 14 (n = 14) women with breast cancer cases receiving IO treatment under the care of the first author. These 14 women, matched in terms of age, stage at diagnosis, estrogen and progesterone receptor status, ethnicity, and marital status, agreed to participate. They signed a consent form and

completed a questionnaire, and their names were provided to the cancer registry.

ResultsWhen the CSS registry was queried about the existence of other breast cancer cases matching the demographic and disease characteristics of the IO breast cancer case women, the results shown in Table 2 were obtained.

Using the criteria that the case and the control must each have been diagnosed within 2 years of each other, we found eligible matched controls for 12 of the 14 participants (87%), ranging from 1 to 81 matches each, with an average of 20 matches per IO patient. However, the CSS was unable to find matches for 2 of the study patients: 1 woman of Hispanic ethnicity (there is only a small Hispanic popula-tion in the greater Seattle metropolitan area) and 1 relatively rare breast cancer case of a premenopausal woman with newly diagnosed stage IV lobular breast cancer. With 2 exceptions, our pilot matching protocol demonstrates that the CSS registry contains an adequate number of potential matched controls for most IO breast cancer cases. When the age at diagnosis criteria is relaxed from within 2 years to within 5 years, many more matches were obtained (see Table 2). These pilot data provided methodological proof of principle and evidence of feasibility to proceed with a matched controlled prospective outcomes study of IO care in breast cancer patients.

Table 2. Results of Matching Pilot Study Demonstrating the Ability to Match BIORC Breast Cancer Patients Who Receive Integrative Oncology Care Using the SEER CSS Registrya

ID Year of Diagnosis Race Ethnicity Marital Status Stage ER PR

No. of Matches Within 2 Years

of Age

No. of Matches Within 5 Years

of Age

B-014 2000 White 0 2 Localized + + 81 155B-012 2003 White 0 2 Regional + + 77 137B-002 2006 White 0 2 Localized + + 16 41B-001 2007 White 0 2 Metastatic + + 4 7B-021 2007 n/a n/a n/a Localized − − 3 6B-019 2007 White 0 1 Localized + − 1 2B-020 2007 White 0 1 Localized − − 6 6B-010 2007 White 0 1 Localized + + 10 34B-011 2008 White 0 1 Localized + + 20 33B-003 2008 White Hispanic n/a Regional + + 0 0B-015 2009 n/ab n/ab n/ab n/a + + 12 29B-016 2009 n/ab n/ab n/ab Metastatic + + 0 0B-017 2009 n/ab n/ab n/ab In situ + + 5 15B-018 2009 n/ab n/ab n/ab n/a + + 8 13

Abbreviations: BIORC, Bastyr Integrative Oncology Research Center; SEER, Surveillance, Epidemiology and End Results; CSS, Cancer Surveillance System.a0 = No ethnicity; 1 = single; 2 = married.bData on demographics in registry incomplete at time of match for women who had been diagnosed in 2009. It may take up to 6 months for Western Washington CSS to collect and enter data into the registry.

at UNIV WASHINGTON LIBRARIES on May 21, 2013ict.sagepub.comDownloaded from

Standish et al 131

Cohort 2AThose reporting no CAM other than

multivitamins and prayer.

Cohort 2BCAM treatments but no IO specialist.

Cohort 2CCAM and IO specialist

Cohort 1A (n=300)

Sub-group of Cohort 1 patients eligible for CSS matching.Additional Inclusion Criteria:

• ICD9 174.× or 233.0 only• Diagnosis ≤2 years from IO FOC• Minimum of two IO clinic visits within 7 months• Completed Baseline Questionnaire

Aims: Describe IO treatments received, cost of IO care, & HRQOL at enrollment, at 6 months post enrollment, at one year and then yearly for as long as funding is available

• Describe HRQOL compared to CSS matched control cohort

• Collect pilot data on survival & disease free survival compared to CSS matched control cohort

Table 3. Cohort Descriptions for a Prospective Matched Controlled Outcomes Study of IO Care

Cohort 1: Bastyr, IO Clinic Patients (n = 600) Cohort 2: FHCRC, CSS Matched Controls (n = 900)

All IO clinic patients with breast cancer being treated at one of the participating IO clinics.

Inclusion criteria• Confirmed breast cancer (ICD9 174.x & 233.0 & V10.3)• Female• ≥18 years of age• Signed informed consent• First IO clinic visit ≤3 months prior to enrollmentAims• Describe IO treatments received, cost of IO care and HRQOL at

enrollment and 6 and 12 months postenrollment

Consists of up to 4 patients from the CSS database for each cohort 1A patient.

Inclusion criteria• Confirmed breast cancer• ≥18 years of age• Female• Diagnosis ≤2 years prior to enrollment• Matches a patient in the 1A cohort for age at diagnosis

(±2 years), stage at diagnosis, and time since diagnosis (±6 months)

• Signed informed consentAims• Describe use of CAM by 2B and 2C cohorts and compare

the HRQOL of women in cohort 2 (A and B) with those in cohort 1A at enrollment, at 6 months postenrollment, at 1 year, and then yearly for as long as funding is available

• Cohort 2 participants will each be classified after enrollment into 1 of 3 subcohorts:

Abbreviations: IO, integrative oncology; FHCRC, Fred Hutchinson Cancer Research Center; CSS, Cancer Surveillance System; ICD, International Clas-sification of Diseases; CAM, complementary and alternative medicine; HRQOL, health-related quality of life.

Discussion

Based on pilot data showing that identifying breast cancer patients who match IO patients on key demographic and disease characteristics is feasible, we proceeded to design and propose a prospective matched controlled outcomes study, which is now funded and under way. We hypothe-sized that a matched case-control prospective outcomes study of breast cancer IO was feasible and could provide valuable data for key questions regarding whether and how well IO and CAM treatments work. We selected breast cancer because it is among the most common cancers and one that IO clinic providers see frequently. Also, the wide-spread use of self-prescribed CAM and IO services by this group of patients2 and the burgeoning body of scientific

data that elucidate the mechanism of action of some com-monly used modalities1,17,26,27 makes this group of particu-lar interest scientifically.

Prospective Outcomes Monitoring Study Design ConsiderationsMost breast cancer patients who seek out IO care also com-plete standard treatments, including surgery, chemother-apy, radiation, and hormone treatments. Some of these patients do so before seeking IO care and others do not. They may also seek CAM treatments either on their own or seek care from CAM professionals (chiropractors, massage therapists, etc) who do not have specialty training or board certification in IO. Because self-prescribed CAM use and

at UNIV WASHINGTON LIBRARIES on May 21, 2013ict.sagepub.comDownloaded from

132 Integrative Cancer Therapies 12(2)

care from nonexpert licensed care CAM professionals is widespread among breast cancer patients, we felt that a matched controlled outcomes study should include an assessment of both CAM and conventional care received. Thus, a matched controlled outcomes study requires sepa-rate examination of 5 cohorts nested within IO patients and CSS-matched controls.

CohortsCohort 1: IO cases (breast cancer cases receiving

integrated oncology treatment in addition to SC)Cohort 1a: a subset of IO cases in which the women

have been diagnosed within 2 years of the first office visit to an IO clinic

Comparison GroupsCohort 2a: SC alone (SC alone, no reported CAM use)Cohort 2b: CAM + SC (breast cancer patients who

use CAM treatments, most of which they would self-prescribe in addition to SC). This group includes breast cancer patients who receive care from licensed CAM providers who do not have advanced oncology training. Such a group could be examined separately depending on the avail-ability of enough patients for each group.

Cohort 2c: CAM + non–research site IO. This group includes breast cancer patients who receive care

from licensed IO providers who are not part of any of the participating IO clinical research sites. We expect this cohort to be small in number because we have recruited the most well-known IO clinics with the highest census. We will compare baseline and changes in HRQOL in breast cancer patients who self-prescribe CAM (cohort 2B) matched with comparison women who do not use CAM (cohort 2A). See Table 3 for definitions of the cohorts.

For this study, FHCRC and BU have to date recruited the participation of 4 additional western Washington State IO clinics (NCT01366248; http://clinicaltrials.gov); see Table 4 for the 5 western Washington IO clinic sites cur-rently recruiting.

Care at each IO clinic includes evaluation and manage-ment by a ND with board certification in naturopathic oncology (Fellows of the American Board of Naturopathic Oncology, http://www.oncanp.org/nat_onc.html). Although specific treatments differ among the IO clinics, the approach and modalities are similar and include botanical and nutri-tional medicine (oral and parenteral), mind-body medicine, and exercise programs. Acupuncture is often recommended by NDs for symptom management and palliative care.28 TCM is provided by a co-licensed ND/LAc or by referral to a DAOM who specializes in oncology (http://www.bastyr.edu/education/acupuncture/degree-curriculum/DAOM.asp). Most IO clinics include TCM evaluation and manage-ment that includes Chinese herbal formulas, moxibustion, qi gong, and dietary recommendations. One aim of the IO outcomes study is to describe variations across IO clinics in type, dose, and frequency of IO treatments prescribed. With time and sufficient number of enrolled patients, we will be able to associate IO treatments with specific outcomes.

The BU/FHCRC Breast Cancer Outcomes Study tracks each patient’s conventional oncological treatment as well as IO therapies prescribed. Patient adherence and ability to tol-erate both standard of care and IO treatments are tracked. Data on all conventional adjunctive treatments that are pre-scribed and completed are obtained from oncology medical records. Data on all IO care treatment recommendations are collected from the attending IO physician’s medical record. Adherence to IO treatment recommendations are assessed in the questionnaire that each patient fills out at baseline, 6 months, and then yearly.

Each IO clinic uses a standard set of recruitment and enrollment procedures and forwards data to BU and the FHCRC who coordinate work with the CSS registry. At baseline, each participant patient completes informed con-sent procedures and signs a medical records release. Women classified as IO cases fill out a 30-minute baseline questionnaire either at their first IO clinic visit or by mail

Table 4. Five Western Washington IO Clinic Recruitment Sites for Breast Cancer IO Prospective Matched Controlled Outcomes Study

IO Clinic Address Study Clinician(s)

Bastyr Integrative Oncology Research Center

Kenmore, WA 98028

Leanna J. Standish, ND, PhD, LAc, FABNO

Institute of Complementary Medicine

Seattle, WA 98122

Chad Aschtgen, ND, FABNO

Red Cedar Wellness Center

Bellevue, WA 98004

Janile Martin, ND, FABNO; Laura A. James, ND, FABNO

Seattle Integrative Oncology at Providence Integrative Cancer Care

Lacey, WA 98503

Chad Aschtgen, ND, FABNO

Seattle Cancer Treatment & Wellness Center

Renton, WA 98057

Paul Reilly, ND, FABNO; Mark Gignac, ND, FABNO; Letitia Cain, ND, FABNO

at UNIV WASHINGTON LIBRARIES on May 21, 2013ict.sagepub.comDownloaded from

Standish et al 133

shortly after. FHCRC mails questionnaires again at 6 months, then annually for 5 years or more. Each patient is asked about their current marital and financial status as well as adherence to IO prescriptions and treatments recommended because these are potential predictors of cancer outcomes.

Medical data are collected from each IO patient’s chart and includes the following variables: relapse status; stage; histology; tumor markers; ER, PR, and Her-2/neu recent scan reports; IO and conventional treatments, including sur-gery, chemotherapy, hormone therapy, monoclonal anti-body therapy, and radiotherapy (and doses) received; comorbidities; date of relapse; and date of death. Laboratory data that are extracted include tumor marker values, 25-OH-vitamin D, and natural killer cell activity. The REDCap (Research Electronic Data Capture) clinical data-base is used for data entry. REDCap is a secure, Web-based application designed to support data capture for research studies.29

Data on adverse effects of IO treatment will be collected using the Monitoring of Side Effects Scale. These data are categorized for type and rated for severity using the NCI Common Terminology Criteria for Adverse Events (CTCAE version 4.D). They will be used to assess safety of IO care. To evaluate cost-effectiveness of IO, we collect data on the cost of consultations and procedures (mostly reimbursed by third-party payers) and IO pharmacy that is not covered by medical insurance.

Matched controls are eligible if they were diagnosed within 2 years of the IO case and invited to participate by FHCRC staff. Controls will be queried about their use of CAM, both self- and physician prescribed, so that each can be assigned to 1 of the following categories: (1) breast can-cer cases receiving integrated oncology treatment in addi-tion to SC, (2) those that receive SC alone, (3) breast cancer patients who self-prescribe CAM in addition to SC, or (4) breast cancer patients who receive care from licensed CAM providers who are not part of the selected IO clinics. All follow-up questionnaires are distributed and collected by study staff working at FHCRC who will ensure that all follow-up data are collected using identical procedures in both the IO case and for both comparison groups.

Using a matched control comparison cohort as a study design, it may be possible to determine if IO care improves the breast cancer survivor’s HRQOL at 6, 12, 24, 36, 48, and 60 months and forward in time after the initiation of IO care. The BU/FHCRC partnership has taken the necessary first steps to be able to compare outcomes, including relapse-free survival and changes in HRQOL, with that of women from the comparison cohort matched for prognosis at diagnosis who have not used IO care but have, and have not, used CAM treatments on their own. To ensure that the clinical outcome measures of this IO study are commensu-rate with NCI guidelines, we will use the definition of

“invasive disease free survival” given by a panel of experts in breast cancer clinical trials representing medical oncol-ogy, biostatistics, and correlative science, as noted in Hudis et al.30

A prospective matched case-comparison cohort will pro-vide preliminary data where none are currently available. It may allow us to describe a standard protocol and provide an effect size estimate necessary before a RCT of IO care can be proposed. Collaboration with the western Washington State CSS registry allows for identification of an appropri-ate comparison group of women who have similar progno-ses at diagnosis and for whom similar data on medical care use and outcomes can be collected.

Measuring IO Treatment AdherenceThe 6-month and yearly questionnaire that is mailed to each study participant includes 2 questions to capture compli-ance with IO recommendations and prescriptions for IO substances, procedures, activities, and referrals:

Please provide information for each provider listed below. Check the box indicating the provider you have seen, provide the date of your last appointment. [The list contains] oncologist, primary care physi-cian, naturopathic doctor, Chinese medicine pro-vider, chiropractor, massage therapist, or other.

We are interested in your use of the following foods, herbs, and other substances. If you used these foods, herbs or other substances, please indicate with check marks those you used and answer the questions about that substance.

The list contains 70 CAM substances (herbs, nutriceuticals, etc) and 33 CAM activities (acupuncture, yoga, biofeed-back, etc).

Sample Size ConsiderationsWe anticipate that enrollment of 4 women with breast cancer from the CSS registry per IO clinic–identified case receiving IO care will yield sufficient matched control women per IO case (1 case: up to 4 controls) for comparison purposes. This includes both matched women who do not use CAM treat-ments and matched women who use self-prescribed CAM treatments independent of a licensed CAM provider or MD with specialty training in IO. Such a cohort will allow us to provide a comprehensive description of CAM use and CAM users in a community setting and prospective examination of the influence of IO care and, perhaps, self-prescribed CAM treatment on the HRQOL of users. Collection of outcomes data on the patient cohorts will include survival, disease-free survival, and a comprehensive multidimensional assessment

at UNIV WASHINGTON LIBRARIES on May 21, 2013ict.sagepub.comDownloaded from

134 Integrative Cancer Therapies 12(2)

of HRQOL. Short-term end points are changes in HRQOL. We calculate that a sample size of 300 IO cases and 1200 matched controls would provide adequate power for a study that compares clinical outcomes of IO plus SC com-pared with SC alone. Long-term outcomes of disease-free survival and relapse rates will require at least 5 years of data.

Our results thus far demonstrate that matching a cohort of IO patients to a matched group of women with breast cancer using a cancer registry is possible and provides a methodology by which the effectiveness of IO as routinely practiced in a community could be studied.

ConclusionBased on the arguments and pilot data presented in this article, the National Center for Complementary and Alternative Medicine is supporting a prospective matched-controlled breast cancer IO outcomes study (2010-2015). This study, currently under way at Bastyr University and FHCRC, is collecting quality-of-life and clinical outcome data from breast cancer patients receiving IO care in west-ern Washington State. As of May 2012, the Bastyr/FHCRC study has enrolled 185 IO breast cancer cases and 78 matched controls.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Supported by NIH-NCCAM grants U19 AT1998, R01 AT5873, U19 AT6028 and Cleavage Creeks Cellars. This work was funded by Cleavage Creek Winery and NCCAM grant R01 AT5873.1`

References

1. Danhauer SC, Mihalko SL, Russell GB, et al. Restorative yoga for women with breast cancer: findings from a random-ized pilot study. Psychooncology. 2009;18:360-368.

2. Greenlee H, Kwan ML, Ergas IJ, et al. Complementary and alternative therapy use before and after breast cancer diagnosis: the Pathways Study. Breast Cancer Res Treat. 2009;117:653-665.

3. Nahleh Z, Tabbara IA. Complementary and alternative medi-cine in breast cancer patients. Palliat Support Care. 2003;1: 267-273.

4. Navo MA, Phan J, Vaughan C, et al. An assessment of the utilization of complementary and alternative medication in women with gynecologic or breast malignancies. J Clin Oncol. 2004;22:671-677.

5. DiGianni LM, Garber JE, Winer EP. Complementary and alternative medicine use among women with breast cancer. J Clin Oncol. 2002;20(18, suppl):34S-38S.

6. Gray RE, Fitch M, Goel V, Franssen E, Labrecque M. Utili-zation of complementary/alternative services by women with breast cancer. J Health Soc Policy. 2003;16:75-84.

7. Henderson JW, Donatelle RJ. Complementary and alterna-tive medicine use by women after completion of allopathic treatment for breast cancer. Altern Ther Health Med. 2004;10: 52-57.

8. Ashikaga T, Bosompra K, O’Brien P, Nelson L. Use of com-plimentary and alternative medicine by breast cancer patients: prevalence, patterns and communication with physicians. Support Care Cancer. 2002;10:542-548.

9. Shen J, Andersen R, Albert PS, et al. Use of complementary/alternative therapies by women with advanced-stage breast cancer. BMC Complement Altern Med. 2002;2:8.

10. Tagliaferri M, Cohen I, Tripathy D. Complementary and alternative medicine in early-stage breast cancer. Semin Oncol. 2001;28:121-134.

11. Gerber B, Scholz C, Reimer T, Briese V, Janni W. Comple-mentary and alternative therapeutic approaches in patients with early breast cancer: a systematic review. Breast Cancer Res Treat. 2006;95:199-209.

12. Anand P, Kunnumakkara AB, Sundaram C, et al. Cancer is a preventable disease that requires major lifestyle changes. Pharm Res. 2008;25:2097-2116.

13. Standish LJAL, Ready A, Sivam G, Torkelson C, Wenner C. Textbook of Integrated Oncology: Botanical Oncology. Oxford, UK: Oxford University Press; 2008.

14. Kienle GS, Glockmann A, Schink M, Kiene H. Viscum album L. extracts in breast and gynaecological cancers: a systematic review of clinical and preclinical research. J Exp Clin Cancer Res. 2009;28:79.

15. Ostermann T, Raak C, Bussing A. Survival of cancer patients treated with mistletoe extract (Iscador): a systematic literature review. BMC Cancer. 2009;9:451.

16. Hsieh TC, Wu JM. Cell growth and gene modulatory activi-ties of Yunzhi (Windsor Wunxi) from mushroom Trametes versicolor in androgen-dependent and androgen-insensitive human prostate cancer cells. Int J Oncol. 2001;18:81-88.

17. Standish LJ, Wenner CA, Sweet ES, et al. Trametes ver-sicolor mushroom immune therapy in breast cancer. J Soc Integr Oncol. 2008;6:122-128.

18. Wasser SP, Weis AL. Therapeutic effects of substances occurring in higher Basidiomycetes mushrooms: a modern perspective. Crit Rev Immunol. 1999;19:65-96.

19. Zjawiony JK. Biologically active compounds from Aphyl-lophorales (polypore) fungi. J Nat Prod. 2004;67:300-310.

20. Block KI, Gyllenhaal C, Tripathy D, et al. Survival impact of integrative cancer care in advanced metastatic breast cancer. Breast J. 2009;15:357-366.

at UNIV WASHINGTON LIBRARIES on May 21, 2013ict.sagepub.comDownloaded from

Standish et al 135

21. McCulloch MBM, Gao J, Colford J. Survival in non-small cell lung cancer patients following treatment with an adjunc-tive Chinese herbal/vitamin protocol. Paper presented at: National Cancer Institute Conference on Traditional Chinese Medicine and Cancer Research; April 10-12, 2006; Bethesda, MD.

22. McCulloch M, Broffman M, van der Laan M, et al. Lung cancer survival with herbal medicine and vitamins in a whole-systems approach: ten-year follow-up data analyzed with marginal struc-tural models and propensity score methods. Integr Cancer Ther. 2011;10:260-279.

23. McCulloch M, Broffman M, van der Laan M, et al. Colon cancer survival with herbal medicine and vitamins combined with standard therapy in a whole-systems approach. Integr Cancer Ther. 2011;10:240-259.

24. Evans S, Block JB. Ethical issues regarding fee-for-service funded research within a complementary medicine context. J Altern Complement Med. 2001;7:697-702.

25. Lesperance ML, Olivotto IA, Forde N, et al. Mega-dose vita-mins and minerals in the treatment of non-metastatic breast

cancer: an historical cohort study. Breast Cancer Res Treat. 2002;76:137-143.

26. Schneider C, Segre T. Green tea: potential health benefits. Am Fam Physician. 2009;79:591-594.

27. Hirpara KV, Aggarwal P, Mukherjee AJ, Joshi N, Burman AC. Quercetin and its derivatives: synthesis, pharmacological uses with special emphasis on anti-tumor properties and prodrug with enhanced bio-availability. Anticancer Agents Med Chem. 2009;9:138-161.

28. Standish LJ, Kozak L, Congdon S. Acupuncture is underuti-lized in hospice and palliative medicine. Am J Hosp Palliat Care. 2008;25:298-308.

29. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap): a metadata-driven methodology and workflow process for pro-viding translational research informatics support. J Biomed Inform. 2009;42:377-381.

30. Hudis CA, Barlow WE, Costantino JP, et al. Proposal for stan-dardized definitions for efficacy end points in adjuvant breast can-cer trials: the STEEP system. J Clin Oncol. 2007;25:2127-2132.

at UNIV WASHINGTON LIBRARIES on May 21, 2013ict.sagepub.comDownloaded from