SEARCH

SEARCH BY CITATION

Keywords:

  • survivorship;
  • secondary cancers;
  • health status;
  • late effects;
  • quality of life

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

BACKGROUND:

Little research has identified the physical and mental health status of survivors of multiple primary cancer diagnoses.

METHODS:

By using data from the population-based 2009 National Health Information Survey, 154 survivors of multiple primary cancer diagnoses, 1427 survivors of a single cancer diagnosis, and 25,004 individuals without a history of cancer diagnosis were identified. The multiple cancer group was compared with the single cancer and no cancer groups with regard to physical and mental health status using analysis of covariance and binary logistic regression.

RESULTS:

Relative to the no cancer group, the multiple cancer group reported significantly poorer mental health status, greater lifetime, recent, and current prevalence of a variety of medical conditions and comorbidities, and more health-related disability. Although observed group differences between the multiple cancer and single cancer groups were less pronounced than those between the multiple cancer and no cancer groups, a consistent pattern was also evident; the multiple cancer group reported significantly poorer status relative to the single cancer group across a range of mental and physical health and illness-related disability indices.

CONCLUSIONS:

Diagnosis of 2 or more primary cancers (excluding nonmelanoma skin cancers) is associated with increased risk for poorer physical and mental health status over and above that associated with diagnosis of a single primary cancer. Survivors of multiple and single primary cancer diagnoses should be considered as distinct subgroups, and increased attention should be devoted to the unique status and needs of survivors of multiple primary cancer diagnoses. Cancer 2012;3645–3653. © 2011 American Cancer Society.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

The number of cancer survivors now residing in the United States is estimated to exceed 11.7 million.1 This is more than 3× the number of cancer survivors residing in the United States in 1970.2 This growth in the sheer number of cancer survivors residing in the United States has been accompanied by a corresponding growth in research devoted to examination of the physical, social, and psychological status of these survivors. Despite the proliferation of research on cancer survivors, however, several gaps in this survivorship literature continue to exist.

In particular, little research has focused upon the current status and needs of individuals who have experienced 2 or more primary cancer diagnoses. The number of such survivors of multiple cancers residing in the United States was estimated to be nearly 900,000 in 2005.2 This represented about 8% of all cancer survivors in the United States.3 Furthermore, second (or more) primary cancer diagnoses represent about 16% of all incident cancer diagnoses in the United States.4 Despite this large and growing number of survivors of multiple cancers, only 2 studies have examined the current physical and mental health status of these individuals.5, 6 In a study limited to Hawaii residents, survivors of multiple primary cancer diagnoses reported poorer global quality of life, less vitality, greater cancer-specific distress, and poorer existential well-being than matched survivors of a single primary cancer diagnosis.6 In an analysis of data from the 2009 national Behavioral Risk Factor Surveillance Survey, survivors of multiple primary cancer diagnoses reported poorer current status across a variety of physical and mental health indices than survivors of a single primary cancer diagnosis.5 The latter, in turn, reported poorer current physical and mental health status than noncancer controls. Considered together, the results of these 2 studies suggest the impact of additional primary cancer diagnoses beyond an initial primary cancer diagnosis is additive. This suggests the importance of distinguishing between survivors of multiple and single cancers and highlights the need for additional research to identify and consider the unique needs of survivors of multiple cancers.

In its landmark report, From Cancer Patient to Cancer Survivor: Lost in Transition, the Institute of Medicine recommended cancer survivorship research make greater use of population-based, national health surveys and increase focus on understudied cancer survivors, such as survivors of multiple primary cancer diagnoses.7 The present study reflects these recommendations. Data from the population-based, 2009 National Health Interview Survey (NHIS) were used to identify the current physical and mental health status of survivors of multiple primary cancer diagnoses. The current status of survivors of multiple cancer diagnoses was directly compared with both survivors of a single cancer diagnosis and healthy, noncancer controls. Based upon previous research,5, 6 it was hypothesized survivors of multiple cancer diagnoses would report poorer physical and mental health status than both survivors of a single cancer diagnosis and noncancer controls.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

NHIS

Data were obtained from the 2009 NHIS survey. Sponsored by the National Center for Health Statistics, Centers for Disease Control and Prevention, the NHIS is an annual, national survey that tracks health trends in the United States.8 NHIS data are collected through a personal household interview. The NHIS uses a multistage sample designed to represent the civilian, noninstitutionalized population of the United States, with oversampling of black, Hispanic, and Asian individuals. For the 2009 NHIS survey, 33,856 households were surveyed and interviews were completed with 27,383 adults ≥18 years. (An additional 348 interviews were completed by an adult proxy when the target adult was physically or mentally unable to complete the interview. These proxy interviews were excluded in the current study.) In the 2009 survey, the household response rate was 82.2%, with 80.1% of these households yielding a completed adult interview.8 As NHIS survey data are in the public domain, approval of the data analyses reported here was not required by a local institutional review board.

NHIS Cancer Survivor Questions

The 2009 NHIS survey included questions to identify cancer survivors and to identify the type of cancer and age at diagnosis for each respondent reporting 1 or more cancer diagnoses. Respondents were asked, “Have you ever been told by a physician or other health professional that you had cancer or a malignancy of any kind?” Those who answered “Yes” were asked, “What kind of cancer was it?” Responses were categorized into 29 different types of cancer plus an “other” category. Respondents were also asked, “How old were you when this cancer was first diagnosed?” Identical, additional questions were asked to identify the type and age of diagnosis for up to 2 additional lifetime cancer diagnoses. Specific wording and response alternatives for all 2009 NHIS survey questions can be found at www.cdc.gov/nchs/nhis/nhisquestionnaires.htm. Time since initial cancer diagnosis was calculated as the difference between age at initial cancer diagnosis and age at survey participation.

NHIS Demographic and Physical and Mental Health Status Indices

Demographic indices

Information regarding age, sex, race, ethnicity, education, and marital status was obtained. Information regarding race and ethnicity was used to categorize respondents as either nonminority (white, non-Hispanic) or minority (all others). Marital status information was used to categorize respondents as either partnered or nonpartnered.

Mental health indices

Six parallel questions assessed feelings experienced during the past 30 days, including sad, nervous, restless or fidgety, hopeless, everything was an effort, and worthless. Responses were recorded on a 5-point scale ranging from “none of the time” to “all of the time.” The 6 items can be summed to create an index of psychological distress (Total Distress; range, 0-24).9 Scores ≥13 were considered indicative of serious psychological distress.9 Participants responding at least “a little of the time” for ≥1 of the 6 “feelings” items were asked, “How much did these feelings interfere with your life or activities?” Responses were recorded on a 4-point scale ranging from “not at all” to “a lot.” Higher scores indicated greater interference.

Physical health indices

Current height and weight were obtained and used to calculate body mass index (BMI). General health change in the past year was assessed by asking respondents to compare their current health to their health 1 year previously. Response options included “worse,” “about the same,” and “better.“ The number of work days lost during the past year because of illness or injury (excluding maternity leave) was assessed, as was the number of days during the past year that illness or injury kept the respondent in bed more than half the day.

Lifetime history of 9 medical conditions was assessed: hypertension, coronary heart disease, heart attack, stroke, emphysema, asthma, ulcer, diabetes, and arthritis. For each, respondents were asked, “Have you EVER been told by a physician or other health professional that you had_____?” Respondents indicated “Yes” or “No.” The number of conditions for which a respondent answered “Yes” was summed to create an index of number of lifetime comorbidities.

More recent and current history of various medical conditions was also assessed. Two questions asked respondents whether they had been told by a physician or health professional in the past 12 months they had weak or failing kidneys or a liver condition (yes vs no). Four questions assessed the presence of pain (yes vs no) lasting ≥1 day during the past 3 months of different types: lower back, neck, face/jaw, severe migraine/headache. The 3 questions for neck, face/jaw, and severe migraine/headache were combined into a single dichotomous index of any head/neck pain (yes vs no). Two questions asked respondents whether they had experienced a head or chest cold or stomach or intestinal illness with vomiting or diarrhea within the past 2 weeks (yes vs no). Two questions asked respondents whether they currently used a hearing aid or had trouble seeing, even with glasses or contact lenses (yes vs no). Respondents rated their current hearing ability on a 6-point scale ranging from “excellent” to “deaf,” with higher scores indicating poorer hearing. Finally, respondents were asked whether they had experienced symptoms of pain, aching, or stiffness in or around a joint (other than back or neck) within the past 30 days (yes vs no). Those who responded “Yes” were asked to rate this joint pain, on average, using an 11-point scale ranging from 0 to 11 (higher scores indicating greater pain) and to indicate whether their current usual activities were limited because of arthritis or joint symptoms (yes vs no).

Lifestyle behavior indices

Lifetime smoking and alcohol use were assessed. Respondents were asked, “Have you smoked at least 100 cigarettes in your entire life” (yes vs no) and “In your entire life, have you had at least 12 drinks of any type of alcoholic beverage” (yes vs no).

Identification of Single Cancer, Multiple Cancer, and No Cancer Control Groups

In the 2009 NHIS survey, 2227 respondents indicated they had been told by a physician or health professional they had cancer. Further eligibility criteria for the single cancer group included: 1) report of single lifetime cancer diagnosis other than nonmelanoma skin cancer, 2) ≥18 years of age at diagnosis, and 3) time since diagnosis ≥1 year. Eligibility criteria for the multiple cancer group included: 1)report of ≥2 lifetime cancer diagnoses other than nonmelanoma skin cancers, 2) ≥18 years of age at initial diagnosis, and 3) time since initial diagnosis ≥1 year. By using these criteria, 1427 respondents constituted the single cancer group, and 154 respondents constituted the multiple cancer group. In the multiple cancer group, 134 individuals reported 2 lifetime cancer diagnoses, 16 reported 3 diagnoses, and 4 reported 4 diagnoses. The 25,004 adult NHIS respondents who indicated they had never been told by a health professional they had cancer constituted the no cancer control group.

Statistical Analyses

The multiple cancer, single cancer, and no cancer groups were compared on current age, marital status (partnered vs nonpartnered), minority status (white, non-Hispanic vs all others), and education using chi-square and t test. Variables for which significant differences were found were used as covariates in subsequent analyses. Our primary analyses involved 2 sets of comparisons; the multiple cancer group was compared with the no cancer group followed by comparison of the multiple cancer group with the single cancer group. Analysis of covariance (ANCOVA) was used to compare groups on continuous outcome variables. Cohen's d was used as the measure of effect size.10, 11 For categorical outcome variables, binary logistic regression, including appropriate covariates, was used to identify group differences. For all multiple cancer versus no cancer comparisons, the no cancer group was used as the reference group. For all multiple cancer versus single cancer comparisons, the single cancer group was used as the reference group. For all odds ratios (ORs), 95% confidence intervals were calculated.

Supplementary analyses examined whether a lifetime history of smoking or alcohol use moderated comparisons of the multiple cancer and single cancer groups and multiple cancer and no cancer groups performed in our primary analyses. For continuous outcome variables, all 1-way ANCOVA analyses were repeated using lifetime smoking history (yes vs no) as a factor in 2 (group) × 2 (smoking) ANCOVA analyses followed by use of lifetime alcohol use history as a factor in 2 (group) × 2 (alcohol use) ANCOVA analyses. For dichotomous outcome variables, all binary logistic regression analyses were repeated including a group × smoking interaction term in the analyses followed by repeat of these analyses including a group × alcohol use interaction term in the analyses. For all supplementary analyses, a significant group × smoking or group × alcohol use interaction was considered evidence of a moderating effect of lifetime smoking or alcohol use.

Analyses were performed using SPSS (version 17.0; SPSS Inc., Chicago, Ill). To minimize type I error, the criterion for statistical significance was set at P < .02.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

Demographic characteristics for the 3 study groups are shown in Table 1. Significant group differences were found for age, sex, and minority status. The no cancer group was younger and less likely to be female and white, non-Hispanic. Consequently, these 3 variables were used as covariates in all analyses. Clinical characteristics for the 2 cancer survivor groups are also shown in Table 1. As expected, the multiple cancer and single cancer groups differed with respect to age at initial cancer diagnosis and time since initial diagnosis. The multiple cancer group was younger at initial diagnosis relative to the single cancer group (48.8 years vs 52.5 years; P < .01) and also had a longer interval from initial diagnosis to the time of NHIS participation relative to the single cancer group (16.1 years vs 11.2 years; P < .001). Consequently, time since initial cancer diagnosis was added as a covariate to all analyses comparing the multiple cancer and single cancer groups. Table 2 shows cancer diagnoses represented in the single cancer and multiple cancer groups.

Table 1. Demographic and Clinical Characteristics of the MC, SC, and NCC Groups
CharacteristicMC GroupSC GroupNCC GroupPa
  • Abbreviations: Dx, diagnosis; GED, general equivalency diploma; MC, multiple cancer; NCC, no cancer control; SC, single cancer; SD, standard deviation.

  • a

    For analysis of variance, t test or chi-square.

Education   .066
 ≤Grade 8n = 11; 7.2%n = 108; 7.6%n = 1668; 6.7% 
 Grades 9-11n = 16; 10.5%n = 154; 10.8%n = 2561; 10.3% 
 Grade 12 or GEDn = 43; 28.1%n = 424; 29.8%n = 6678; 26.8% 
 Some college/technical schooln = 46; 30.1%n = 377; 26.5%n = 7501; 30.1% 
 College graduaten = 24; 15.7%n = 218; 15.3%n = 4289; 17.2% 
 Postbaccalaureate study/degreen = 13; 8.5%n = 141; 9.9%n = 2190; 8.8% 
Race/ethnicity white, non-Hispanicn = 119; 77.3%n = 1096; 77.1%n = 14,206; 57.1%<.001
Marital status married/partneredn = 80; 51.9%n = 702; 49.2%n = 12,271; 49.2%.792
Womenn = 94; 61.0%n = 904; 63.3%n = 13,876; 55.5%<.001
Current age, yearsMean = 64.9; SD = 14.6Mean = 63.7; SD = 14.6Mean = 46.1; SD = 17.4<.001
Age at initial cancer Dx, yearsMean = 48.8; SD = 16.0Mean = 52.5; SD = 16.6 .009
Time since initial cancer Dx, yearsMean = 16.1; SD = 11.0Mean = 11.2; SD = 10.4 <.001
Table 2. Cancer Diagnoses in the MC and SC Groups
DiagnosisMC GroupSC Group
  1. Abbreviations: MC, multiple cancer; SC, single cancer.

  2. Percentages within columns do not add to 100% due to rounding in SC group and multiple cancer diagnoses in MC group.

Bladdern = 10; 6.5%n = 40; 2.8%
Breastn = 42; 27.3%n = 334; 23.4%
Cervixn = 24; 15.6%n = 143; 10.0%
Colonn = 31; 20.1%n = 103; 7.2%
Kidneyn = 19; 5.8%n = 20; 1.4%
Lungn = 23; 14.9%n = 45; 3.2%
Lymphoman = 8; 5.2%n = 43; 3.0%
Melanoman = 22; 14.3%n = 104; 7.3%
Ovaryn = 13; 8.4%n = 42; 2.9%
Prostaten = 33; 21.4%n = 198; 13.9%
Skin (nonmelanoma)n = 0; 0.0%n = 0; 0.0%
Skin (kind unknown)n = 19; 12.3%n = 81; 5.7%
Thyroidn = 8; 5.2%n = 34; 2.4%
Uterusn = 18; 11.7%n = 77; 5.4%
Othern = 68; 44.2%n = 163; 11.4%

Comparison of the multiple cancer and no cancer groups on mental health outcomes revealed the multiple cancer group reported significantly greater Total Distress scores (P < .001; effect size, 0.45 standard deviation [SD]; Table 3) and were more likely to meet criteria for serious psychological distress (OR, 1.79; P < .001; Table 4). No difference was found for the extent to which feelings interfered with activities. This suggests when distress is present, the extent of activity interference caused by distress is similar between the multiple cancer and no cancer groups (Table 3). Comparison of the multiple cancer and single cancer groups on these same mental health indices again revealed the multiple cancer group reported significantly greater Total Distress scores (P < .02; effect size, 0.21 SD; Table 3). No significant differences were found with regard to the presence of serious psychological distress (Table 4) or extent to which feelings interfered with activities (Table 3).

Table 3. Mental and Physical Health Outcomes for the MC, SC, and NCC Groups
OutcomesMC vs NCCMC vs SC
MCNCCESbPcMCSCESbPc
MeanaSDMeanaSDMeandSDMeandSD
  • Abbreviations: BMI, body mass index; ES, effect size; MC, multiple cancer; NCC, no cancer control; SC, single cancer; SD, standard deviation.

  • a

    Means shown are adjusted for age, sex, and minority status.

  • b

    Cohen's d.

  • c

    F test for 2-group analysis of covariance.

  • d

    Means shown are adjusted for age, sex, minority status, and time since initial diagnosis.

  • e

    Statistically significant.

  • f

    Asked of respondents acknowledging ≥1 symptom in past 30 days.

  • g

    Higher scores indicate current health better than 1 year previously.

  • h

    Asked of respondents reporting employment during past year.

  • i

    Higher scores represent poorer hearing.

  • j

    Asked of respondents reporting joint pain in past 30 days.

Mental health outcomes            
 Distress total10.255.498.513.820.45<.001e10.105.499.134.540.21.013e
 Feelings interfere with activitiesf2.291.092.051.030.23.0632.301.092.271.080.03.830
Physical health outcomes            
 Current health vs health 1 year previouslyg2.030.662.100.510.14.0731.980.662.040.600.10.256
 No. of health-related bed days in past year15.1151.494.6025.820.40<.001e16.7451.499.3434.100.20.018e
 No. of health-related lost work days in past yearh10.3225.263.6915.620.42.001e11.2025.267.9328.760.12.422
 Current hearingi2.021.221.710.890.35<.001e2.341.222.191.080.14.087
 Joint  pain severity in past 30 daysj5.572.905.172.450.16.1245.552.905.352.430.08.487
 BMI27.096.0427.726.180.10.21727.466.0427.605.720.02.791
 No. of lifetime comorbidities1.741.720.921.200.68<.001e2.321.721.821.470.28<.001e
Table 4. Prevalence of Mental and Physical Health Conditions for the MC, SC, and NCC Groups
ConditionPrevalenceaGroup Comparison
MC vs NCCbMC vs SCc
MCSCNCCOR95% CIPOR95% CIP
  • Abbreviations: CI, confidence interval; MC, multiple cancer; NCC, no cancer control; OR, odds ratio; SC, single cancer.

  • a

    Raw prevalence rates.

  • b

    NCC group is the reference group; analyses adjusted for age, minority status, and sex.

  • c

    SC group is the reference group; analyses adjusted for age, minority status, sex, and time since initial diagnosis.

  • d

    Statistically significant.

  • e

    Asked of respondents reporting joint pain/arthritis in the past 30 days.

Mental health conditions         
 Serious psychological disorder9.8%5.5%3.3%1.791.36-2.35<.001d1.871.03-3.41.040
Physical health conditions         
 Weak/failing kidneys, past 12 months9.1%5.0%1.8%1.741.31-2.32<.001d1.730.94-3.18.078
 Liver condition, past 12 months7.1%2.6%1.4%1.931.41-2.64<.001d3.011.48-6.15.002d
 Joint symptoms, past 30 days55.2%52.2%31.4%1.201.02-1.41.0321.040.74-1.47.810
 Limited by joint symptoms/arthritise46.8%37.6%32.3%1.261.04-1.53.017d1.490.99-2.23.057
 Head/neck pain, past 3 months40.9%30.5%26.0%1.451.23-1.71<.001d1.651.15-2.36.006d
 Low back pain, past 3 months38.3%37.3%27.9%1.130.95-1.33.1611.020.72-1.45.898
 Head/chest cold, past 2 weeks16.2%10.2%11.5%1.321.06-1.64.013d1.641.03-2.63.038
 Stomach/intestinal illness, past 2 weeks11.0%5.8%4.5%1.641.27-2.11<.001d2.061.18-3.63.012d
 Use hearing aid13.0%8.2%2.2%1.431.11-1.86.005d1.640.94-2.87.080
 Trouble seeing even with glasses20.1%15.8%8.1%1.391.14-1.70.001d1.340.87-2.05.181

Comparison of the multiple cancer and no cancer groups on 7 continuous physical health outcomes using ANCOVA (Table 3) revealed the multiple cancer group reported more health-related bed days (P < .001; effect size, 0.40 SD) and more health-related lost work days in the past year (P < .001; effect size, 0.42 SD). The multiple cancer group also reported poorer current hearing (P < .001; effect size, 0.35 SD) and a greater number of lifetime comorbidities (P < .001; effect size, 0.68 SD). Overall, the multiple cancer group reported poorer status on 6 of the 7 indices, the lone exception being reports of a lower BMI by the multiple cancer group. Comparison of the multiple cancer and single cancer groups on these same continuous physical health outcomes (Table 3) revealed the multiple cancer group reported significantly more health-related lost work days in the past year (P < .02; effect size, 0.20 SD) and a greater number of lifetime comorbidities (P < .001; effect size, 0.28 SD). Overall, the multiple cancer group reported poorer status on 6 of the 7 indices, the lone exception again being BMI.

Comparison of the multiple cancer and no cancer groups with regard to reports of prevalence of recent and current physical health outcomes using binary logistic regression (Table 4) revealed the multiple cancer group reported a significantly higher prevalence for 8 of the 10 outcomes examined. The lone exceptions were the prevalence of lower back pain in the past 3 months and joint symptoms in the past 30 days. Odds ratios for the 8 physical health outcomes for which significant group differences were found ranged from 1.26 (limitations by joint symptom/arthritis) to 1.93 (liver condition in past 12 months), with a mean odds ratio of 1.52. Comparison of the multiple cancer and single cancer groups with regard to prevalence of these same recent or current physical health outcomes revealed significant differences for 3 of the 10 outcomes. The multiple cancer group was more likely to report a liver condition within the past year (OR, 3.01; P < .01), head/neck pain within the past 3 months (OR, 1.65; P < .01), and a stomach/intestinal illness within the past 2 weeks (OR, 2.06; P < .02). Although significant differences emerged only for these 3 outcomes, ORs exceeded 1.0 for all 10 outcomes, indicating greater risk in the multiple cancer group (P = .002; binomial test, 2-tailed). The mean OR for comparison of the multiple cancer and single cancer groups across all 10 physical health outcomes was 1.66.

Supplementary analyses examined whether lifetime history of smoking or alcohol use moderated comparisons of the multiple cancer and single cancer groups shown in Tables 3 and 4. A significant group × alcohol interaction was found for 3 of 20 outcome variables examined (liver condition in past 12 months [P = .018], head/neck pain in past 3 months [P = .02], number of lifetime comorbidities [P = .003]), and a significant group × smoking interaction was found for 3 of 20 outcome variables (liver condition in past 12 months [P = .002], head/neck pain in past 3 months [P = .007], head/chest cold in past 2 weeks [P = .018]). Inspection of cell proportions for the multiple cancer versus single cancer comparisons for reports of a liver condition in the past year indicated differences between the multiple cancer and single cancer groups were more pronounced in survivors without a lifetime history of alcohol use (8.1% vs 1.9%) than those with a history of alcohol use (6.9% vs 2.7%). In contrast, differences between the multiple cancer and single cancer groups in reports of a liver condition were more pronounced in survivors with a smoking history (9.7% vs 3.0%) relative to those without a smoking history (3.3% vs 2.0%). An identical pattern of results was present for reports of head/neck pain in the past 3 months; no history of alcohol use was associated with larger differences between the multiple cancer and single cancer groups in reports of head/neck pain, whereas a history of smoking was associated with larger differences in reports of head/neck pain between the multiple cancer and single cancer groups. Finally, differences between the multiple cancer and single cancer groups in reports of a head/chest cold in the past 2 weeks were more pronounced in survivors with a smoking history (19.4% vs 10.7%) relative to those without a smoking history (11.5% vs 9.5%).

Similar supplementary analyses for the comparison of the multiple cancer and no cancer groups found a significant group × alcohol interaction for 1 of 20 outcome variables (No. of lifetime comorbidities [P = .001]) and a significant group × smoking interaction for 0 of 20 outcome variables. Inspection of group means for both the multiple cancer vs single cancer and multiple cancer vs no cancer comparisons suggested reports of a greater No. of lifetime comorbidities were most pronounced in multiple cancer survivors with no history of alcohol consumption.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES

As anticipated, results of the present study indicate survivors of multiple cancer diagnoses report poorer physical and mental health status than individuals without a history of cancer diagnosis. Relative to individuals without a history of cancer diagnosis, survivors of multiple cancers reported consistently poorer mental health status, a greater lifetime, recent, and current prevalence of a variety of medical conditions and comorbidities, and more health-related disability. No evidence suggested a lifetime history of smoking or alcohol use moderated these comparisons of the multiple cancer and no cancer groups. The magnitude of differences between the multiple cancer and no cancer groups was in the clinically significant range,12, 13 with a mean effect size of 0.35 SD for mental health indices, mean ORs in the 1.5 range for prevalence of various medical conditions and comorbidities, and effect sizes >0.35 SD for health-related disability indices.

More notably, however, the present study indicates survivors of multiple cancer diagnoses generally report poorer mental and physical health status than survivors of a single cancer diagnosis. Granted, observed group differences are less pronounced than those between survivors of multiple cancer diagnoses and no cancer controls. However, a consistent pattern of survivors of multiple cancer diagnoses reporting poorer status relative to survivors of a single cancer diagnosis across a variety of mental and physical health indices is apparent. Although a lifetime history of smoking or alcohol use moderated comparison of the multiple cancer and single cancer groups for a few individual outcome variables, no general pattern of moderating impact was present for these lifestyle behaviors. When considered together with findings for the no cancer control group, the overall pattern of results is quite clear; survivors of multiple cancer diagnoses report poorer status on a variety of mental and physical health indices relative to survivors of single cancer diagnoses, who in turn report poorer status than individuals without a history of cancer.

Results of the present study are consistent with the only 2 previous studies of survivors of multiple cancer diagnoses in suggesting subsequent primary cancer diagnoses beyond an initial cancer diagnosis have an additive negative impact upon mental and physical health status.5, 6 Notably, each of the extant studies of multiple cancer survivors has used somewhat different methodologies. Specifically, different methods of sample recruitment have been used: statewide Surveillance, Epidemiology, and End Results cancer registry6 versus national face-to-face interview (present study) versus national telephone survey.5 Furthermore, whereas Burris and Andrykowski5 included survivors of nonmelanoma skin cancer in their multiple and single cancer survivor groups, both Gotay et al6 and the current study excluded individuals with nonmelanoma skin cancer from the survivor groups. Finally, each of the extant studies examined a somewhat different set of mental and physical health status indices. Gotay et al6 focused on mental health and quality of life indices, whereas this study and Burris and Andrykowski5 included indices of physical health and disability. The consistency of results across the extant studies of multiple cancer survivors, despite methodological differences, suggests observed differences in physical and mental health status between survivors of multiple and single cancer diagnoses are robust.

Although it is well known a cancer diagnosis can yield decrements in mental and physical health status in both the short and long term,14 the added risk for negative late and long-term mental and physical health effects conferred by additional, subsequent primary cancer diagnoses is only now being realized. This has both methodological and clinical implications. Methodologically, survivors of multiple and single primary cancer diagnoses should be considered as distinct subgroups of cancer survivors for research purposes. When conducting survivorship research, the number of participants reporting multiple and single primary cancer diagnoses in the study sample should be identified in the sample description. Data should be presented separately for these 2 groups, and subgroup analyses should be conducted when possible. Clinically, our data suggest survivors of multiple cancers may be at risk for greater illness burden than survivors of a single diagnosis. Consequently, screening for prior cancer diagnoses is important when assessing risk for poor mental and physical health outcomes in a cancer survivor. However, it is not known whether survivors of multiple cancer diagnoses experience different clinical rehabilitation needs and require different clinical management and supportive care approaches than survivors of a single cancer diagnosis. Research is needed to examine these important clinical questions further.

The limitations of this study should be acknowledged. As is the case in large population-based surveys (eg, Coups and Ostroff,15 Hewitt and Rowland16), personal cancer history was assessed by self-report and not ascertained from medical records. Fortunately, self-reports of the prevalence and type of lifetime cancer diagnosis are generally accurate.17, 18 Nevertheless, some degree of misclassification in our 3 study groups is possible. The cross-sectional nature of the data reported here also limits the ability to attribute observed group differences between multiple and single cancer survivors to the impact of additional primary cancer diagnoses. Although case-control designs similar to that used here are often used in cancer survivor research to identify the impact of a cancer diagnosis, the possibility that any observed group differences might have existed before cancer diagnosis, or be associated with risk for multiple primary cancer diagnoses, cannot be ruled out. In addition, as we excluded individuals diagnosed with cancer during childhood (<18 years) from our survivor groups, our findings may not generalize to all multiple cancer survivors. Furthermore, as only 35 of 154 respondents in the multiple cancer group were minority respondents, we were unable to examine whether our general pattern of findings was also evident within different racial/ethnic subgroups. Finally, although the Institute of Medicine has recommended national household and health care surveys be analyzed to capture information on cancer survivorship,7 this approach has limitations. Specifically, we were limited in the health status variables we could examine by the questions in the NHIS survey. Had we set out to prospectively examine the physical and mental health status of survivors of multiple cancers, we would certainly have asked additional questions and included standardized multi-item measures often used in survivorship research. Furthermore, the lack of clinical information regarding cytotoxic treatment and disease staging, inherent in the analysis of cancer survivor data from large, multipurpose health surveys such as the NHIS, precludes more fine-grained analyses of how these factors might affect physical and mental health status among both our multiple cancer and single cancer survivor groups. This lack of clinical information prevents the drawing of conclusions regarding how disease and treatment factors might be associated with better or poorer physical and mental health status among multiple cancer survivors. This important question remains to be addressed in future research.

In conclusion, our results suggest additional primary cancer diagnoses are associated with a significant increase in illness burden beyond that associated with an initial cancer diagnosis. Coupled with similar results from previous studies of multiple cancer survivors, the need for additional research is evident. Indeed, survivors of multiple primary cancer diagnoses represent a remarkably understudied group of cancer survivors. One of 6 incident cancer diagnoses in the United States is a diagnosis of a multiple primary cancer.4 This would suggest about 266,000 of the estimated 1.6 million incident cancer diagnoses in the United States in 20111 would involve an individual with a prior primary cancer diagnosis, a number comparable to the estimated 288,000 new cases of invasive or in situ female breast cancer diagnosed in the United States in 2011.1 Given the hundreds (if not thousands) of extant studies examining the experience of female breast cancer survivors (eg, Cormier et al,19 Ewertz and Jensen,20 Yanez et al21), efforts to examine the impact and survivorship experience associated with multiple primary cancer diagnoses are likely overdue. Future research should focus on obtaining a more detailed picture of specific deficits in physical and mental health status experienced by survivors of multiple cancers, identifying subgroups of survivors at greater risk for these deficits. As this information emerges, strategies for enhancing physical and mental health should be developed, implemented, and disseminated to this underserved and potentially unique group of cancer survivors.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. CONFLICT OF INTEREST DISCLOSURES
  9. REFERENCES
  • 1
    American Cancer Society. Cancer Facts and Figures 2011. Atlanta, GA: American Cancer Society; 2011.
  • 2
    American Cancer Society. Cancer Facts and Figures 2009. Atlanta, GA: American Cancer Society; 2009.
  • 3
    Marriotto AB, Rowland JH, Ries LAG, Scoppa S, Feuer EJ. Multiple cancer prevalence: a growing challenge in long-term survivorship. Cancer Epidemiol Biomarkers Prev. 2007; 16: 566-571.
  • 4
    Travis LB. The epidemiology of second primary cancers. Cancer Epidemiol Biomarkers Prev. 2006; 15: 2020-2026.
  • 5
    Burris JL, Andrykowski MA. Physical and mental health status and health behaviors of survivors of multiple cancers: a national, population-based study. Ann Behav Med. [published online ahead of print July 6, 2011.]
  • 6
    Gotay CC, Ransom S, Pagona IS. Quality of life in survivors of multiple primary cancers compared with cancer survivor controls. Cancer. 2007; 110: 2101-2109.
  • 7
    Institute of Medicine. From Cancer Patient to Cancer Survivor: Lost in Transition. Washington, DC: National Academies Press; 2006.
  • 8
    National Center for Health Statistics. 2009 National Health Interview Survey (NHIS) Public Use Data Release: NHIS Survey Description. 2010. Available at: www.cds.gov/nchs/nhis Accessed on July 1, 2011.
  • 9
    Kessler RC, Barker PR, Colpe LJ, et al. Screening for serious mental illness in the general population. Arch Gen Psychiatry. 2003; 60: 184-189.
  • 10
    Cohen J. A power primer. Psychol Bull. 1992; 112: 155-159.
  • 11
    Zakzanis KK. Statistics to tell the truth, the whole truth, and nothing but the truth: formulae, illustrative numerical examples, and heuristic interpretation of effect size analyses for neuropsychological researchers. Arch Clin Neuropsychol. 2001; 16: 653-667.
  • 12
    Hays RD, Wooley JM. The concept of clinically meaningful differences in health-related quality-of-life research: how meaningful is it? Pharmacoeconomics. 2000; 18: 419-423.
  • 13
    Norman GR, Sloan JA, Wyrwich KW. Interpretation of changes in health-related quality of life: the remarkable universality of half a standard deviation. Med Care. 2003; 41: 582-592.
  • 14
    Stein KD, Syrjala KL, Andrykowski MA. Physical and psychological long-term and late effects of cancer. Cancer. 2008; 112( 11 suppl): 2577-2592.
  • 15
    Coups EJ, Ostroff JS. A population-based estimate of the prevalence of behavioral risk factors among adult cancer survivors and noncancer controls. Prev Med. 2005; 40: 702-711.
  • 16
    Hewitt M, Rowland JH. Mental health service use among adult cancer survivors: analyses of the National Health Interview Survey. J Clin Oncol. 2002; 20: 4581-4590.
  • 17
    Abraham L, Geller BM, Yankaskas BC, et al. Accuracy of self-reported breast cancer among women undergoing mammography. Breast Cancer Res Treat. 2009; 118: 583-592.
  • 18
    Bergmann MM, Calle EE, Mervia CA, et al. Validity of self-reported cancers in a prospective cohort study in comparison with data from state cancer registries. Am J Epidemiol. 1998; 147: 556-562.
  • 19
    Cormier JN, Askew RL, Mungovan KS, Xing Y, Ross MI, Armer JM. Lymphedema beyond breast cancer: a systematic review and meta-analysis of cancer-related secondary lymphedema. Cancer. 2010; 116: 138-149.
  • 20
    Ewertz M, Jensen AB. Late effects of breast cancer treatment and potentials for rehabilitation. Acta Oncol. 2011; 50: 187-193.
  • 21
    Yanez B, Thompson EH, Stanton AL. Quality of life among Latina breast cancer patients: a systematic review of the literature. J Cancer Surviv. 2011; 5: 191-207.