Cancer and Leukemia Group B Cancer Control and Health Outcomes Committee: Origins and Accomplishments

Origins of the Cancer Control Committee

Cancer Control Sciences Committee. Under the leadership of Dr. Vincent Vinciguerra, the Cancer Control Sciences Committee (CSCC) focused on smoking cessation, cancer screening and prevention of colorectal, prostate, ovarian, and lung cancers, supportive care research, and research that recognized the needs of minority and elderly populations. The first protocols focused on breast cancer screening, symptom control, and smoking cessation. In 1992, the first phase III randomized prevention trial, CALGB 9270, “Aspirin for the Prevention of Colorectal Adenomas in Patients with Previous Colorectal Cancer,” was proposed. In 1996, Dr. Electra D. Paskett, an epidemiologist from Wake Forest University School of Medicine, replaced Dr. Vinciguerra as chair of the committee. From 1996 to 2001, the Cancer Control Sciences Committee continued to focus on two areas defined as cancer control-cancer prevention and symptom control.

Psycho-oncology Committee. In 1976, CALGB formed the Psychiatry Committee under the leadership of Dr. Jimmie Holland. Through the Committee's collaboration with the disease committees, the CALGB became the first cooperative group to report quality of life (QOL) as an outcome variable in clinical trials. The advantages in studying psychosocial variables within a cooperative group were several large patient populations, greater diversity due to different geographic locations, and control of medical and treatment variables by the treatment protocols. From 1976 to 2001, Dr. Holland, with the support of Dr. Alice Kornblith, led a robust committee that conducted QOL studies as part of CALGB treatment trials, evaluated psychological distress in cancer patients and their caregivers, developed interventions to reduce such distress, and examined the medical, psychological, and social consequences of cancer survivorship.

Clinical economics. Responding to changes in the nature and financing of health care in the 1970s, social scientists developed a series of related approaches to quantifying cost per unit health benefit (1). They realized that the existing clinical trial infrastructure could be used to collect cancer therapy cost data in parallel with collection of the traditional measures of clinical benefit in clinical trials (e.g., time to progression and overall survival). Active participation of the NCI cooperative groups in economic studies dates from 1994 when NCI provided funding for trials of minimally invasive surgery and stipulated that these trials had to include assessments of costs in addition to QOL. In this context, the CALGB created the Clinical Economics Working Group in August 1994 under the leadership of Dr. Jane C. Weeks, a health economist and medical oncologist, and in Spring 1996, the Working Group achieved full committee status.

CCHO Committee. To better develop and coordinate the research activities of the three committees (Cancer Control, Psycho-oncology, and Clinical Economics), a Health Outcomes Research Council, comprising the chairs of the three committees, was formed within CALGB in March 1997. When Drs. Holland and Weeks stepped down as chairs of their respective committees in 2001, CALGB consolidated all health outcomes research under one umbrella committee, CCHO, with Dr. Paskett as the chair.

To accomplish the goals of the CCHO, four subcommittees were formed: Prevention, Symptom Control, Clinical Economics, and Quality of Life. Since 2004, Dr. Elizabeth B. Lamont has lead the Clinical Economics Subcommittee, whose name was then changed to Health Services to reflect the Subcommittee's increasing methodologic diversity in evaluating the translation of clinical trial efficacy to population-level effectiveness and its associated costs. Dr. Kornblith leads the Quality of Life Subcommittee. In 2005, the goals of the subcommittees were further refined and the Symptom Control Subcommittee was refocused and renamed the Symptom Intervention Subcommittee with Dr. Charles Shapiro as the chair. Dr. James R. Marshall took over from Dr. Judy Garber as head of the Prevention Subcommittee in 2005.

Significant Achievements

Psycho-oncology/Quality of Life Subcommittee

The CALGB Psycho-oncology Committee and the successor, Quality of Life Subcommittee, focus on studies related to the long-term adjustment of cancer survivors, effect of treatment on patient's QOL, methodologic issues in behavioral research, interventions to improve adaptation to illness, and psychological and sociodemographic predictors of survival.

Long-term adjustment of cancer survivors. The rich registry of patients treated on CALGB protocols made it possible for the Psychiatry Committee to undertake the first studies of long-term survivors and the effect of treatments on their lives. Rowland et al. studied one of the earliest cohorts of children with acute lymphoblastic leukemia treated in a randomized trial. The results showed the negative effect of cranial radiation on intelligence and achievement tests for those treated with cranial radiation. Continuing this line of research from 1984 to 2003, three studies of other cohorts of survivors were conducted to determine the long-term treatment effects: Hodgkin's disease survivors, adult leukemia cancer survivors, and breast cancer survivors. A core of identical measures was used, enabling us to compare the results of the three groups (2–5).

Twenty-two percent of Hodgkin's disease survivors, a mean of 6 years after treatment, were found to report distress at a level suggestive of a psychiatric disorder along with a range of sexual problems, conditioned nausea and vomiting, and insurance problems (CALGB 8561 and 8562; ref. 6). When compared with leukemia survivors, 5 years post-treatment, Hodgkin's disease survivors' adjustment was found to be worse, with greater distress reported than by leukemia survivors (22% versus 14%; 2, 4, 6). Hodgkin's disease survivors were also found to have significantly worse adaptation than leukemia survivors in terms of the effect of cancer on their family life, their sexual relationships, and their energy level. In addition, 39% of Hodgkin's disease survivors versus 25% of leukemia survivors reported experiencing conditioned nausea and vomiting when triggered by sights, smells, or tastes that were reminders of their treatment. These studies of the long-term issues associated with cancer treatment have influenced the scientific community's research agenda.

Effect of treatment on patient's adjustment. By assessing psychosocial factors in the context of CALGB phase III clinical trials, we have been able to compare the effect of different treatments on the lives of patients, controlling for medical and sociodemographic factors. Without this information, neither oncologists nor patients can make informed decisions about which treatment to choose. Whereas there were many studies in which the QOL of patients on phase III trials were examined over the years, two were particularly significant from the standpoint of research supporting the successful treatment. In CALGB 8864, women with advanced-stage breast cancer were randomly assigned to 160, 800, and 1,600 mg/d megestrol acetate. One of the previously recognized major side effects of high-dose megestrol acetate was weight gain. The QOL question was whether high doses of megestrol acetate would create greater psychological distress than lower doses. Women taking the highest doses had significantly more severe side effects, worse physical functioning, worse psychological distress, and worse overall QOL than those taking the 160 mg/d dose. In conjunction with the findings that there were no significant differences between treatment arms in time to disease progression and survival (7), the QOL studies led to the clear conclusion that high-dose megestrol acetate was not beneficial.

QOL was also a particularly important component of CALGB 9221, a phase III trial in which patients with myelodysplastic syndrome were randomized to 5-azacytidine or supportive care. The QOL assessments of patients receiving 5-azacytidine showed greater improvement in fatigue, physical functioning, positive affect, and psychological distress compared with supportive care (8). These QOL findings contributed to the Food and Drug Administration approval of 5-azacytidine as treatment for myelodysplastic syndrome.

A third study was unique in that it tested the clinical value of hydrazine sulfate as an alternative therapy for lung cancer. A QOL evaluation was done to determine whether there would be a significant difference in QOL due to the addition of hydrazine sulfate to cisplatin and vinblastine compared with a placebo plus chemotherapy in patients with advanced non–small cell lung cancer (CALGB 8931; ref. 9). The placebo group was found to have superior QOL at month 2 with respect to physical functioning, fatigue, and overall QOL. Given that the clinical trial found no differences in overall survival between the two groups, the QOL findings further supported the conclusion that hydrazine sulfate provided no benefit to the lung cancer patient.

Clinical Economics/Health Services Subcommittee

The central scientific contributions of the Clinical Economics/Health Services Subcommittee over its 10-year history can be divided into two broad categories, methodologic innovation and empirical research. In its early years, a high priority was establishing a methodologically sound and feasible approach to economic analysis in conjunction with clinical trials within the cooperative group setting. The committee established criteria for prioritizing trials for economic companions (i.e., those studying treatments for relatively common diseases and those comparing two strategies that differ substantially in cost but only modestly in effectiveness). A standardized approach to the collection and analysis of resource use, cost, and utility data was developed, and infrastructure to support this effort was built within the Group.

As the Subcommittee matured, the portfolio expanded to include primary health services research studies. Key examples include both empirical (10–13) and methodologic (14, 15) studies. At first, CALGB sought to estimate the additional cost of treating cancer patients in the NCI-sponsored clinical trials in the United States. A retrospective cost study was conducted on a random sample of patients enrolled in 1 of 35 active phase III trials or phase I or any phase II trials between October 1, 1998 and December 31, 1999 and a group of nonparticipants with a similar stage of disease who were treated at comparable clinical research institutions across the United States. Direct treatment costs were measured by using medical records, telephone survey, and Medicare claims data from ∼2.5 years. Adjusted costs were 6.5% higher for trial participants than nonparticipants; therefore, the additional treatment costs of an open reimbursement policy for government-sponsored cancer clinical trials seemed minimal (16). This study provided important information to policymakers at the local and federal levels regarding reimbursement for clinical care costs incurred by patients enrolled on NCI-sponsored trials.

In the second study, an empirical companion study to a randomized controlled trial of laparoscopic-assisted colectomy versus open colectomy for colon cancer (CALGB 9396), investigators determined short QOL outcomes (10). The Symptoms Distress Scale, QOL index, and a single-item global rating scale at 2 days, 2 weeks, and 2 months postoperative, duration of postoperative in-hospital analgesic use, and length of stay were recorded prospectively. In an intention-to-treat analysis comparing Symptoms Distress Scale pain intensity, Symptoms Distress Scale summary, QOL index summary, and global rating scale scores at each time point, the only statistically significant difference observed between groups was mean (median) global rating scale score for 2 weeks postsurgery [76.9 (80) for laparoscopic-assisted colectomy versus 74.4 (75) for open colectomy (P = 0.009)]. In the hospital, patients assigned to laparoscopic-assisted colectomy required fewer days of both parenteral and oral analgesics compared with patients assigned to open colectomy [mean (median), 3.2 (3) versus 4.0 (4) days for parenteral (P < 0.001) and 1.9 (1) versus 2.2 (2) days for oral (P = 0.03)]. Thus, only minimal short-term QOL benefits were found with laparoscopic-assisted colectomy for colon cancer compared with standard open colectomy.

The decision analysis of health care costs and effectiveness of multiple staging options for patients with esophageal cancer (12) involved comparison of the staging modalities of computed tomographic scan, endoscopic ultrasound with fine-needle aspiration biopsy, positron emission tomography, thoracoscopy/laparoscopy, and combinations of these techniques. The inputs to the model included cost estimates from Medicare claims data, life expectancies from the NCI Surveillance, Epidemiology, and End Results data, and radiologic test characteristics from the published literature and from results of CALGB 9380 (17). The investigators found that under baseline assumptions computed tomographic scan plus endoscopic ultrasound with fine-needle aspiration biopsy was the most inexpensive strategy and offered more quality-adjusted life-years, on average, than all other strategies with the exception of positron emission tomography plus endoscopic ultrasound with fine-needle aspiration.

Cancer control

The CSCC focused on both prevention and symptom control studies. A pivotal trial for the CCHO Committee was CALGB 9270, “Aspirin for the Prevention of Colorectal Adenomas in Patients with Previous Colorectal Cancer,” which examined the effect of 325 mg aspirin on preventing colorectal adenomas among patients (n = 635) treated for early-stage colorectal cancer in a randomized, placebo-controlled, phase III trial. Results indicated that at least one adenoma was found in 17% of the participants in the aspirin group and 27% of those in the placebo group (hazard ratio, 0.64; P = 0.022) after a median of 12.8 months following randomization (18).

This trial was important because it (a) was the first large-scale prevention study CALGB sponsored, (b) added to the scientific data on the efficacy of aspirin in colorectal cancer prevention, and (c) laid the foundation for a proposed Southwest Oncology Group/CALGB colorectal cancer prevention study. The results of this study created some controversy as a similar study by Baron et al. (19) found efficacy for aspirin at 81 mg but not 325 mg among patients with a history of an adenomatous polyp of the colon or rectum. With the more recent safety issues associated with other nonsteroidal anti-inflammatory drugs, the results of the aspirin studies have become increasingly important.

In February 2006, accrual was completed to CALGB 79809, a phase III trial of zoledronic acid in the prevention of bone loss in early-stage breast cancer patients with chemotherapy-induced ovarian failure. This is the first trial to evaluate a highly potent third-generation bisphosphonate in women with chemotherapy-induced ovarian failure, a setting in which bone loss is expected to high.

Future Plans

Quality of life

The future plans for the Quality of Life Subcommittee are to develop and expand protocols in the areas of established expertise.

Effect of treatment on cancer patients' QOL. The types of studies important to this Subcommittee stem from three major paradigms of research: (a) one cancer treatment is expected to improve patients' overall survival or disease-free survival compared with another, at potentially greater toxicity; (b) equivalence trials in which no treatment differences are expected but with one arm expected to be significantly less toxic, resulting in an improved QOL; and (c) marginal differences between treatments in disease-free, progression-free, or overall survival are expected, but differences in QOL between treatment arms might influence treatment recommendations.

An example of an active equivalence trial is CALGB 49907, “A Randomized Trial of Adjuvant Chemotherapy with Standard Regimens, Cyclophosphamide, Methotrexate and Fluorouracil (CMF) or Doxorubicin and Cyclophosphamide (AC) versus Capecitabine in Women 65 Years and Older with Early-Stage Breast Cancer.” The hypothesis is that standard adjuvant chemotherapy and capecitabine will be equally effective but that capecitabine will be less toxic than either cyclophosphamide, methotrexate, and 5-fluorouracil or doxorubicin and cyclophosphamide resulting in a better emotional state both during treatment and in the initial post-treatment period. As the side effects of chemotherapy are largely reversible, by 1 year post-treatment, no significant differences in patients' physical and emotional state are expected between the two treatment arms. This study also examines older cancer patients' adherence to an oral medication, capecitabine. Poor adherence could significantly compromise treatment outcome as well as the validity of the clinical trial.

Cancer survivors. An ongoing study is designed to examine the adjustment of African American cancer survivors across three disease sites (colon, prostate, and breast cancer) compared with African American population controls (CALGB 11990). There are no studies of African American survivors of this magnitude. Through CALGB registry of all patients treated on CALGB clinical trials over many years and the use of tumor registries, a large cohort of African American cancer survivors, at least 3 years post-treatment and free of disease, is being identified. The degree to which a cancer diagnosis and treatment significantly worsens African American survivors' adjustment will be compared with African American noncancer population controls.

Health Services Subcommittee

In a climate of decreasing research funding, research agendas that leverage existing data sources and infrastructures to ask and answer novel questions related to cancer care are increasingly attractive. With this in mind, the future plans of the Subcommittee are to develop an adjunctive research infrastructure within CALGB that will use the standard clinical trial data that are already being collected to facilitate the development of new projects. The Subcommittee has proposed the establishment of a CALGB clinical trial registry containing a minimum amount of data (i.e., basic demographic, disease, treatment, and outcome information) on trial participants for CALGB researchers to carry out hypothesis generating secondary analyses that might serve as stepping stones for larger, clinical trial companions or free-standing projects (with or without de novo data collection). Ideally, this effort will aggregate prior years of data into a common data file format that will be flexible enough for regular updates of survival information, addition of data from new CALGB studies (i.e., prospectively maintained), and easy de-identification. Such a minimum data set may serve as a starting point for areas of inquiry that would require linkage to other data types, either de novo (e.g., patient survey data) or administrative (e.g., health insurance claims data). For example, with this type of data architecture and appropriate patient consent, patients could be surveyed at the time of trial entry to establish the types and amount of certain factors (e.g., social, medical, and psychological) potentially relevant to traditional clinical trial outcomes of survival and toxicity but not routinely evaluated at baseline. Following study completion, evaluation of associations between such factors and classic clinical trial end points (e.g., survival and toxicity) as well as nonclassic end points (e.g., health service utilization as measured through health insurance claims) could be undertaken. Such an emblematic study might ascertain degree of instrumental social support at trial enrollment using standard measures (e.g., Berkman et al.) and then evaluate the extent to which social support (in addition to more traditional demographic and cancer-related covariates) predicts survival, toxicity, and resource use. Further, if such a study is to be carried out within the framework of phase III studies, evaluation of differential effects of social support according to treatment arm would be possible. In addition to this future direction, the Subcommittee will continue to develop scientifically rigorous and clinically relevant companion trials in economics, health services, and clinical epidemiology that support and enhance CALGB therapeutic trials.

Prevention Subcommittee

The present foci of the Prevention Subcommittee include three strategies: tobacco prevention, screening, and chemoprevention. Although there is little disputing the importance of tobacco exposure to cancer risk, progress in convincing Americans not to smoke has been painfully slow. In spite of the success of the cooperative oncology groups with such prevention studies as trials of tamoxifen for the prevention of breast cancer and finasteride for the prevention of prostate cancer, the cooperative groups have not to date exhibited great success with tobacco cessation. A challenge for the coming decade will be for CALGB to develop more effective strategies for encouraging tobacco cessation and for assisting those who choose to quit smoking.

A second promising prevention strategy involves screening. There is debate about the efficiency with which prostate and breast cancer screening will decrease the cancer burden because, there being no precancerous lesions that signal substantial risk, screening is able only to increase the diagnosis of prostate and breast cancer. Screening, however, diagnoses these cancers at a more curable stage. Nonetheless, increases in diagnosis of breast and prostate cancer have been accompanied by only modest decreases in breast and prostate cancer mortality (20). Similarly, colorectal cancer screening is widely regarded as having the potential to significantly decrease the toll of colorectal cancer mortality. There is very strong, clinical trial-based evidence that the use of fecal occult blood testing decreases colorectal cancer mortality (21) as does sigmoidoscopic screening (22). A cohort study suggests that colonoscopic screening decreases the risk of colorectal cancer (23). In spite of this evidence, there is considerable evidence that people are screened at less than optimal intervals (24). Given the considerable potential of breast, prostate, and colorectal cancer screening to decrease the toll of these three common cancers, an important challenge for CALGB and the other cooperative oncology groups may well be to increase screening in average-risk, or at least high-risk, individuals.

The third promising focus is chemoprevention. Some substantial triumphs have been recorded in chemoprevention; these include the trials of tamoxifen to prevent breast cancer (25) and trials of finasteride to prevent prostate cancer (26). A pilot trial currently under way tests whether diet change is possible among men with prostate cancer; this trial will be in preparation for a trial of diet change to prevent the progression of early-stage prostate cancer. CALGB is presently also supporting a trial of selenium to prevent the development of prostate cancer among men with high-grade prostatic intraepithelial neoplasia.⁶ Additional pilot chemoprevention trials are testing statin drugs as means for preventing breast cancer; one of these, supported by CALGB pilot funds, tests statins against a biomarker of breast cancer risk: mammographic density. A new study tests a novel angiogenesis inhibitor, sorafenib, against the progress of dysplastic Barrett's esophagus to esophageal cancer.

Symptom Intervention Subcommittee

The mission statement of the Symptom Intervention Subcommittee is to test interventions (drug and nondrug) to alleviate symptoms and side effects of cancer or its treatment. Three protocols are currently in development stages within CALGB. The first is CALGB 70205, a phase II study of methylnaltrexone for opioid-induced constipation. Opioid-induced constipation is a prevalent problem in ambulatory cancer patients who are receiving narcotic analgesics. Methylnaltrexone is an opioid antagonist that does not cross the blood-brain barrier and is effective in inducing laxation in a hospice population. This phase II trial is testing whether methylnaltrexone is effective in an ambulatory cancer population and whether nonresponders will respond to a second dose.

CALGB 170301 is a proposed phase III randomized, placebo-controlled trial of duloxetine, a selective serotonin reuptake inhibitor that is approved for use in painful diabetic neuropathy, for treatment of chemotherapy-induced neuropathy. Chemotherapy-induced peripheral neuropathy is common in cancer patients receiving paclitaxel or oxaliplatin, and there are no effective measures to treat or prevent it. This trial tests whether patients with moderate to severe chemotherapy-induced peripheral neuropathy report improvements on validated instruments designed to evaluate chemotherapy-induced peripheral neuropathy pain scores and overall QOL.

The third protocol is CALGB 70305, a phase III trial of lymphedema prevention. This trial randomizes breast cancer patients who have had an axillary node dissection to lymphedema education or lymphedema education plus a tailored physical therapy program.

Conclusion

The range of studies completed, under way, and planned for the future attest to the breadth of the CCHO Committee mission. CALGB's ongoing support has enabled us to be one of the premier QOL groups in the cooperative clinical trials groups and a leader in the area of Clinical Economics/Health Services. Going forward, the CCHO Committee will further the goals of each Subcommittee while promoting cross-fertilization of protocols with a multidisciplinary focus. Through such multidisciplinary approaches, the burden of cancer can be alleviated.