Background
Breast cancer (BC) is the most common cancer among women. Advances in treatments have led to a growing population of BC survivors,1 many of whom experience distress, including anxiety, depression, and stress-related disorders.2–5 Compared with healthy controls, BC survivors report higher levels of distress.6,7 Due to the prevalence of cancer distress, the American College of Surgeon’s Commission on Cancer mandates screening for cancer center accreditation.8 Various guidelines and tools are available to help screen for and measure distress both during and after cancer treatment.9,10
Although it is clearly established that psychological distress is prevalent among survivors, further population-based and longitudinal studies with longer follow-ups are needed to better characterize its nature and trajectory. Such insights are essential for refining cancer distress management strategies. In this study, we examined the trajectory of distress in BC survivors by evaluating annual global mental health (GMH) scores over time since cancer diagnosis using the Patient-Reported Outcomes Measurement Information System Global-10 (PROMIS-10). Additionally, we assessed co-occurring global physical health (GPH) scores, posttraumatic stress symptoms (PTSS; using the Impact of Event Scale-Revised [IES-R]), and depressive symptoms (using the Patient Health Questionnaire-2 [PHQ-2]) in association with GMH.
Patients and Methods
Study Design
Clinical and survey data were abstracted from the Mayo Clinic Breast Disease Registry, a prospective longitudinal cohort of patients seen at least once at Mayo Clinic Rochester for newly diagnosed BC. Patients who provide informed consent complete a baseline survey within 1 year of diagnosis, and annually thereafter. This study uses survey data through year 4. Surveys incorporate the following instruments: PROMIS-10 (assessed annually), and IES-R and PHQ-2 (assessed at years 2 and 4). Clinical data pertaining to tumor and treatment characteristics are abstracted by nurses from electronic health records. Patients aged ≥18 years with stage 0–III BC who completed the PROMIS-10 distress measure on the baseline survey and at least one follow-up survey were included in this analysis (n=2,140), and individuals with stage IV disease or BC recurrence documented at any time were excluded. This study was approved by the Mayo Clinic Institutional Review Board.
Measures
PROMIS-10
The PROMIS-10 system is a 10-item measure that assesses general physical function as well as social and emotional well-being. It generates a GPH score based on 4 items and a GMH score based on another 4 items. PROMIS measures use a T-score metric that is converted from raw sum scores. The T-score of 50 is standardized to the mean United States general population with a standard deviation of 10 points.11,12 Higher T-scores demonstrate more of what is being measured, such as greater global physical or mental health than the general population. In cancer populations, prior validation studies have established that a 3-point difference in T-scores is clinically meaningful.12–14
IES-R
The IES-R is a 22-item measure that evaluates the level of PTSS. The IES-R generates a total score ranging from 0 to 88, with subscale scores for intrusion, avoidance, and hyperarousal symptoms.15 In this study, we analyzed the complete IES-R and used a cutoff score of ≥20 as a marker of clinical significance. Additionally, subscale scores were examined to assess changes in the IES-R components. This measure is commonly used to evaluate PTSS severity in BC survivors.16
PHQ-2
The PHQ-2 consists of the first 2 questions of the PHQ-9 and is used to screen for and assess depressive symptoms, including major depressive disorder. Scores range from 0 to 6, with a recommended cutoff of ≥3 indicating possible depression.17 This measure is commonly used in individuals with cancer to screen for depressive symptoms and major depression.18
Statistical Analysis
Data were summarized using frequencies and percentages for categorical variables and means with standard deviations for continuous variables. Associations of baseline characteristics with whether any PROMIS-10 score values were missing in subsequent surveys, primarily due to failure to return the surveys, were assessed using 2-sample t tests. Changes in PROMIS-10 scores over the 4-year study period were first evaluated using linear regression models. Up to 5 observations were included for each individual: one for the baseline score and one for each completed survey from years 1 through 4. Separate analyses were conducted for GMH and GPH T-scores. For each, the PROMIS-10 score of interest was modeled as the outcome variable, survey year was modeled as the categorical exposure variable, and subject ID was included as a blocking term to account for within-subject effects. These global tests were then followed by a series of 4 paired t tests assessing changes in scores from baseline to each of years 1, 2, 3, and 4, subset to individuals who returned both the baseline survey and the survey from the year being analyzed. Changes in the individual components of the PROMIS-10 scores from baseline to each of years 1, 2, 3, and 4 were also examined using paired t tests. Within-subject changes from year 2 to year 4 in IES-R intrusion, avoidance, and hyperarousal scores and in PHQ-2 scores were also examined using paired t tests, subset to individuals who had nonmissing scores at both survey years. Secondary analyses examined changes in scores over time among individuals who returned all 5 surveys from baseline through year 4.
Results
Study Population
A total of 2,140 individuals completed a baseline PROMIS-10 survey plus at least one follow-up survey and were included in the study (Figure 1). At year 1, 2,137 (99.9%) individuals completed a follow-up survey including the PROMIS-10 measure; at year 2, 1,515 (70.8%) completed a follow-up survey including the PROMIS-10, IES-R, and PHQ-2; at year 3, 1,155 (54.0%) completed a follow-up survey including the PROMIS-10; and at year 4, 779 (36.4%) completed a follow-up survey including the PROMIS-10, IES-R, and PHQ-2. Baseline demographic and clinical characteristics are described in Table 1. More than 99% of individuals were biologically female, 95.6% were White, and 7.4% were aged <40 years at diagnosis. Most were diagnosed with stage I or II disease. More than half of participants had lumpectomy, with about two-thirds receiving endocrine therapy, 31% chemotherapy, and 57% radiotherapy.
Flow diagram of inclusion criteria.
Abbreviations: IES-R, Impact of Event Scale-Revised; PHQ-2, Patient Health Questionnaire-2; PROMIS-10, Patient-Reported Outcomes Measurement Information System Global-10.
Citation: Journal of the National Comprehensive Cancer Network 23, 5; 10.6004/jnccn.2024.7353
Baseline Characteristics
| Characteristic | n (%) |
|---|---|
| Total, n | 2,140 |
| Gender | |
| Female | 2,125 (99.3) |
| Male | 15 (0.7) |
| Age at cancer diagnosis | |
| n | 2,140 |
| Mean [SD] | 58.9 [12.49] |
| Median (IQR) | 59.5 (49.9–68.3) |
| Age at cancer diagnosis | |
| <40 y | 158 (7.4) |
| 40–49 y | 378 (17.7) |
| 50–59 y | 556 (26.0) |
| 60–69 y | 597 (27.9) |
| ≥70 y | 451 (21.1) |
| Race | |
| White | 2,046 (95.6) |
| Black or African American | 24 (1.1) |
| American Indian/Alaska Native | 9 (0.4) |
| Asian | 24 (1.1) |
| Other/Unknown | 30 (1.4) |
| Not disclosed | 7 (0.3) |
| Education | |
| High school graduate or less | 249 (11.7) |
| Vocational education beyond high school | 184 (8.7) |
| Associate’s degree or some college | 493 (23.2) |
| Bachelor’s degree | 617 (29.0) |
| Graduate school | 581 (27.4) |
| Missing | 16 |
| Household finance situation | |
| Difficulty paying bills or need to cut back | 135 (6.4) |
| Enough money to pay bills, but little spare money for extra things | 383 (18.1) |
| After paying bills, still enough money for extra things | 1,599 (75.5) |
| Missing | 23 |
| Tumor stage | |
| 0 | 359 (16.8) |
| I | 1,052 (49.3) |
| II | 558 (26.2) |
| III | 163 (7.6) |
| Missing | 8 |
| Type of surgery | |
| Lumpectomy only | 1,073 (52.3) |
| Mastectomy only | 925 (45.1) |
| Mastectomy and lumpectomy | 53 (2.6) |
| Missing | 89 |
| Patient underwent endocrine therapy | |
| No | 712 (33.3) |
| Yes | 1,427 (66.7) |
| Missing | 1 |
| Patient underwent chemotherapy | |
| No | 1,477 (69.1) |
| Yes | 662 (30.9) |
| Missing | 1 |
| Patient underwent radiotherapy | |
| No | 925 (43.2) |
| Yes | 1,214 (56.8) |
| Missing | 1 |
| ER status of first breast cancer | |
| Negative | 316 (15.3) |
| Positive | 1,754 (84.7) |
| Missing | 70 |
| PR status of first breast cancer | |
| Negative | 458 (22.3) |
| Positive | 1,599 (77.7) |
| Missing | 83 |
| HER2 status of first breast cancer | |
| Negative | 1,480 (87.8) |
| Positive | 205 (12.2) |
| Missing | 455 |
| Pathologic BRCA1/2 mutation | |
| No | 2,090 (97.7) |
| Yes | 50 (2.3) |
| Vital status | |
| Alive | 2,089 (97.6) |
| Deceased | 51 (2.4) |
Abbreviations: ER, estrogen receptor; PR, progesterone receptor.
Trajectory of PROMIS-10 Scores
The mean [SD] PROMIS-10 mental and physical health T-scores at baseline were 52.3 [7.7] and 51.3 [7.4], respectively. Individuals who provided physical and mental PROMIS-10 score values for all follow-up surveys had higher mean baseline PROMIS-10 scores, were more likely to present with lower tumor stage, and were less likely to have received chemotherapy or radiotherapy than those not providing values for all surveys (Supplementary Table S1, available online in the supplementary material). The mean GMH T-scores decreased significantly over time with follow-up (Table 2). This difference was most notable in year 2, when the GMH T-score demonstrated a mean difference of 0.9 units. Although most changes were statistically significant, none met the 3-point difference threshold of clinical significance.12–14 Changes in the individual components of the mental health T-score are provided in Supplementary Table S2. Over time, 3 of the 4 components (quality of life, rating of mental health, and satisfaction with social relationships) decreased, whereas self-reported emotional health increased.
Trajectory of PROMIS-10 Scores
| Assessment | Mean Baseline Score [SD] |
Mean Follow-Up Score [SD] |
Mean Difference in Score [SD] |
Test Statistica | P Valuea | P Valueb |
|---|---|---|---|---|---|---|
| Global mental health T-score | <.001 | |||||
| Year 1 (n=2,100) | 52.3 [7.7] | 51.9 [8.1] | 0.4 [6.4] | 2.73 | .007 | |
| Year 2 (n=1,489) | 52.8 [7.6] | 51.9 [8.2] | 0.9 [6.7] | 5.36 | <.001 | |
| Year 3 (n=1,143) | 52.8 [7.4] | 52.3 [8.1] | 0.5 [7.0] | 2.44 | .01 | |
| Year 4 (n=766) | 52.9 [7.2] | 52.4 [8.1] | 0.5 [7.4] | 1.79 | .07 | |
| Global physical health T-score | <.001 | |||||
| Year 1 (n=2,072) | 51.3 [7.5] | 51.9 [7.6] | −0.6 [6.1] | −4.51 | <.001 | |
| Year 2 (n=1,466) | 51.7 [7.4] | 52.0 [7.5] | −0.3 [6.3] | −1.97 | .05 | |
| Year 3 (n=1,124) | 51.9 [7.3] | 52.4 [7.8] | −0.5 [6.9] | −2.18 | .03 | |
| Year 4 (n=749) | 51.9 [7.3] | 52.3 [7.8] | −0.4 [7.2] | −1.69 | .09 | |
Sample sizes indicate number of individuals providing PROMIS-10 score values at both the baseline survey and the survey from the year being investigated.
Abbreviation: PROMIS-10, Patient-Reported Outcomes Measurement Information System Global-10.
Paired t test.
Linear regression model assessing association between PROMIS-10 score and time since survey return, accounting for individual subject effects by modeling subject ID as a blocking term.
In contrast to the GMH T-scores, the GPH T-scores increased significantly over time, with changes compared with baseline ranging from 0.3 to 0.6 units. Although most changes were statistically significant, none met the 3-point difference threshold of clinical significance.12–14 Regarding the individual components of the physical health T-score, patients reported that their physical health declined slightly but their level of fatigue improved (Supplementary Table S2). Ability to carry out every-day physical activities and pain levels were not consistently associated with time since BC diagnosis. A subsequent analysis of the GMH and GPH T-scores subsetting to participants who returned the baseline survey and all surveys from year 1 through 4 yielded similar results to those in Table 2, though with higher P values due to smaller sample sizes (Supplementary Table S3).
Posttraumatic Stress–Related and Depressive Symptoms
The IES-R and PHQ-2 were completed at years 2 and 4. At year 2, the total IES-R score was clinically significant at ≥20 among 12.7% of participants (Table 3) and the median total score was 4. We observed significant decreases in scores between year 2 and year 4 (12.7% to 7.7% prevalence) for the total scale, as well as the intrusion, avoidance, and hyperarousal subscales (P<.001 for each; Figure 2, Supplementary Table S4). PHQ-2 score distributions were not significantly different across survey years (P=.78). At year 2 and 4, >95% of participants had a PHQ-2 score of ≤2 with respective prevalence rates of 4.7% and 4.8%.
Year 2 Follow-up Survey Results: IES-R and PHQ-2 (N=2,140)
| n (%) | |
|---|---|
| IES-R intrusion subscale, year 2 | |
| 0 | 523 (35.6) |
| 1 | 210 (14.3) |
| 2 | 161 (11.0) |
| 3 | 129 (8.8) |
| 4 | 96 (6.5) |
| 5 | 76 (5.2) |
| ≥6 | 274 (18.7) |
| Missing | 671 |
| IES-R avoidance subscale, year 2 | |
| 0 | 620 (42.6) |
| 1 | 160 (11.0) |
| 2 | 108 (7.4) |
| 3 | 103 (7.1) |
| 4 | 80 (5.5) |
| 5 | 64 (4.4) |
| ≥6 | 320 (22.0) |
| Missing | 685 |
| IES-R hyperarousal subscale, year 2 | |
| 0 | 814 (55.2) |
| 1 | 228 (15.5) |
| 2 | 133 (9.0) |
| 3 | 83 (5.6) |
| 4 | 50 (3.4) |
| 5 | 39 (2.6) |
| ≥6 | 128 (8.7) |
| Missing | 665 |
| Categorized IES-R total scale, year 2 | |
| 0–19 | 1,236 (87.3) |
| 20–23 | 52 (3.7) |
| 24–31 | 65 (4.6) |
| 32–36 | 23 (1.6) |
| ≥37 | 39 (2.8) |
| Missing | 725 |
| PHQ-2 scale, year 2 | |
| 0 | 1,057 (71.0) |
| 1 | 171 (11.5) |
| 2 | 190 (12.8) |
| 3 | 32 (2.2) |
| 4 | 22 (1.5) |
| 5 | 6 (0.4) |
| 6 | 10 (0.7) |
| Missing | 652 |
Abbreviations: IES-R, Impact of Event Scale-Revised; PHQ-2, Patient Health Questionnaire-2.
Longitudinal changes in IES-R scores from year 2 to year 4 among patients who returned both surveys. (A) Intrusion (n=718): mean [SD], 3.0 [4.1] to 2.2 [3.4], respectively; P<.001. (B) Avoidance (n=713): mean [SD], 3.0 [4.2] to 2.3 [3.8], respectively; P<.001. (C) Hyperarousal (n=729): mean [SD], 1.4 [2.6] to 1.0 [2.3], respectively; P<.001. (D) Total scale (n=683): mean [SD], 7.3 [9.6] to 5.5 [8.4], respectively; P<.001.
Abbreviation: IES-R, Impact of Event Scale-Revised.
Citation: Journal of the National Comprehensive Cancer Network 23, 5; 10.6004/jnccn.2024.7353
Discussion
Using a large, established BC survey registry, we assessed the prevalence of distress among BC survivors with stage 0–III disease through 3 different measures of psychological distress derived from annual survey follow-up data over approximately 4 years after diagnosis. The PROMIS-10 measure revealed small but statistically significant decreases in mental health and small but statistically significant increases in physical health over time. Continued distress screening and interventions are needed for the minority of survivors experiencing significant symptom. Additionally, further studies should specifically assess distress subtypes that might be unique to the postchemotherapy, postsurgery, and postradiation phases of BC survivorship. Further analysis is also necessary to better understand factors contributing to distress, such as age, disease stage, and adjuvant treatments received, such as chemotherapy, radiation therapy, and endocrine therapy.
Although we observed a small decrease in GMH over time, more specific symptom measures remained the same or improved. Depression symptoms, as measured by the PHQ-2, did not change, and depression was uncommon, with a prevalence of 4.7% and 4.8% at years 2 and 4, respectively (defined by a cutoff point of ≥3). Additionally, the more specific PTSS, measured by the IES-R, improved over time, suggesting that survivors experience less-intrusive thoughts about cancer as treatment becomes more distant. The prevalence of posttraumatic distress (defined by a cutoff point of ≥20) was 12.7% at year 2 and 7.7% at year 4. Taken together, these measure comparisons highlight that tool selection directly impacts distress prevalence and determines whether distress is detected in the population. Many different tools are currently used to screen for cancer-related distress, with some being more specific to particular disorders such as depression, anxiety, or stress-related conditions.
There is currently no consensus on the optimal tool for distress screening among cancer survivors. The ASCO guidelines for managing anxiety and depression in cancer survivors recommend using the 2-item PHQ-2 to screen for depression in adults with cancer. If a patient reports a score of ≥2, completing the full PHQ-9 is advised to guide subsequent interventions accordingly.19 In contrast, the NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for Distress Management9 provide a single-item measure of general distress using the NCCN Distress Thermometer and Problem List, and recommend screening for distress at every medical visit or, at a minimum, during the initial visit, at appropriate intervals, and as clinically indicated (such as with changes in disease status). However, neither of these guidelines is specifically tailored to cancer survivors or the unique concerns that arise during survivorship, such as fear of recurrence.
Barriers to implementing consistent distress screening programs include the challenge of incorporating measures into routine clinical practice. Moreover, research has shown that screening alone is ineffective without an established system in place for appropriate referrals and interventions for individuals identified as experiencing increased distress.10 Neal et al20 demonstrated the feasibility of using the PROMIS measure in a large cancer center.
This study provides important information from a large clinical database with long follow-up, high survey response rates, and curated clinical data. However, limitations include that most of our population was White, the potential for referral bias due to the study being conducted at a tertiary care center, and the potential for survival bias inherent in the longitudinal design. Additionally, our cohort consisted solely of BC survivors, which may limit the generalizability of the findings to individuals with other types of cancer. Future studies are needed to examine distress trajectories in more diverse population cohorts and community settings.
Conclusions
This study evaluated a large cohort of BC survivors who completed 3 different measures of psychological distress up to 4 years after diagnosis. Although there was a small decrease in general mental health over time among respondents, this decrease was not clinically significant, symptom-specific measures of depression did not change, and PTSS improved. Although most patients do not report clinically significant distress, further studies are needed to optimize detection of distress among the subgroup of patients with clinically significant symptoms, to better understand whether more symptomatic patients are more or less likely to respond to serial questionnaires in later survivorship, and to inform optimal measure selection for distress screening and management at different phases of survivorship.
References
- 1.↑
Siegel RL, Miller KD, Wagle NS, Jemal A. Cancer statistics, 2023. CA Cancer J Clin 2023;73:17–48.
- 2.↑
Carreira H, Williams R, Müller M, et al. Associations between breast cancer survivorship and adverse mental health outcomes: a systematic review. J Natl Cancer Inst 2018;110:1311–1327.
- 3.↑
Cordova MJ, Riba MB, Spiegel D. Post-traumatic stress disorder and cancer. Lancet Psychiatry 2017;4:330–338.
- 4.↑
Abbey G, Thompson SB, Hickish T, Heathcote D. A meta-analysis of prevalence rates and moderating factors for cancer-related post-traumatic stress disorder. Psychooncology 2015;24:371–381.
- 5.↑
Shelby RA, Golden-Kreutz DM, Andersen BL. PTSD diagnoses, subsyndromal symptoms, and comorbidities contribute to impairments for breast cancer survivors. J Trauma Stress 2008;21:165–172.
- 6.↑
Avis NE, Levine BJ, Case LD, et al. Trajectories of depressive symptoms following breast cancer diagnosis. Cancer Epidemiol Biomarkers Prev 2015;24:1789–1795.
- 7.↑
Maass SWMC, Boerman LM, Verhaak PFM, et al. Long-term psychological distress in breast cancer survivors and their matched controls: a cross-sectional study. Maturitas 2019;130:6–12.
- 8.↑
American College of Surgeons Commission on Cancer. Cancer Program Standards: Ensuring Patient-Centered Care, 2016 Edition. American College of Surgeons; 2015:56–57.
- 9.↑
Riba MB, Vanderlan J, Anderson B, et al. NCCN Clinical Practice Guidelines in Oncology: Distress Management. Version 1.2025. Accessed December 6, 2023. To view the most recent version, visit https://www.nccn.org
- 10.↑
Smith SK, Loscalzo M, Mayer C, Rosenstein DL. Best practices in oncology distress management: beyond the screen. Am Soc Clin Oncol Educ Book 2018;38:813–821.
- 11.↑
Hays RD, Bjorner JB, Revicki DA, et al. Development of physical and mental health summary scores from the patient-reported outcomes measurement information system (PROMIS) global items. Qual Life Res 2009;18:873–880.
- 12.↑
Jensen RE, Potosky AL, Moinpour CM, et al. United States population-based estimates of patient-reported outcomes measurement information system symptom and functional status reference values for individuals with cancer. J Clin Oncol 2017;35:1913–1920.
- 13.↑
Yost KJ, Eton DT, Garcia SF, Cella D. Minimally important differences were estimated for six Patient-Reported Outcomes Measurement Information System-Cancer scales in advanced-stage cancer patients. J Clin Epidemiol 2011;64:507–516.
- 14.↑
Seneviratne MG, Bozkurt S, Patel MI, et al. Distribution of global health measures from routinely collected PROMIS surveys in patients with breast cancer or prostate cancer. Cancer 2019;125:943–951.
- 15.↑
Weiss DS, Marmar CR. The impact of event scale – revised. In: Wilson JP, Keane TM, eds. Assessing Psychological Trauma and PTSD. The Guilford Press; 1997:399–411.
- 16.↑
Koopman C, Butler LD, Classen C, et al. Traumatic stress symptoms among women with recently diagnosed primary breast cancer. J Trauma Stress 2002;15:277–287.
- 17.↑
Kroenke K, Spitzer RL, Williams JB. The Patient Health Questionnaire-2: validity of a two-item depression screener. Med Care 2003;41:1284–1292.
- 18.↑
Thekkumpurath P, Walker J, Butcher I, et al. Screening for major depression in cancer outpatients: the diagnostic accuracy of the 9-item Patient Health Questionnaire. Cancer 2011;117:218–227.
- 19.↑
Andersen BL, Lacchetti C, Ashing K, et al. Management of anxiety and depression in adult survivors of cancer: ASCO guideline update. J Clin Oncol 2023;4:3426–3453.
- 20.↑
Neal JW, Roy M, Bugos K, et al. Distress screening through Patient-Reported Outcomes Measurement Information System (PROMIS) at an academic cancer center and network site: implementation of a hybrid model. JCO Oncol Pract 2021;17:e1688–1697.
