A prospective observational study of patient-reported functioning and quality of life in advanced and metastatic breast cancer utilizing a novel mobile application

Purpose To assess and describe patient-reported outcomes (PROs) in women with locally advanced/unresectable or metastatic breast cancer (aBC/mBC) with hormone receptor-positive/human epidermal growth factor receptor 2-negative (HR + /HER2 −) status receiving palbociclib combination therapy in a US real-world setting. Methods A prospective, noninterventional, multicenter longitudinal study was conducted in US patients initiating treatment with palbociclib combination therapy for HR + /HER2 − aBC/mBC. PRO data (SF-12; CES-D-10; mood; pain; fatigue; interference of aBC/mBC or its treatment on family life, social life, physical activity, energy, and productivity; overall health rating; and quality of life [QOL]) were collected via a custom-developed mobile application at daily, weekly, and cycle-based intervals. Patient medical information (demographics, clinical characteristics, treatment information, and adverse events) was collected from medical records at baseline and at the end of the 6-month follow-up period. Results Patients’ general health status (SF-12) remained consistent throughout treatment and was generally consistent with published norms for individuals diagnosed with cancer. The presence of depression (CES-D-10) was low and did not change substantially over time. Mean pain and fatigue scores using an 11-point numeric rating scale were low and remained stable. Patients, on average, reported neutral or positive moods. Patient-reported QOL and overall health was primarily “Good,” “Very good,” or “Excellent.” Findings were consistent regardless of patient experience with neutropenia. Conclusions Patients treated with palbociclib, on average, reported consistently low levels of pain and fatigue as well as good QOL and overall health that remained stable throughout the first 6 months of treatment regardless of episodes of neutropenia. Supplementary information The online version of this article (10.1007/s10549-020-06082-7) contains supplementary material, which is available to authorized users.


Introduction
Breast cancer is the most common cancer among women in the United States (US), with an estimated 276,480 new cases occurring in 2020. Although 5-year survival rates are high among patients diagnosed with early-stage breast cancer, only an estimated 28% of patients with advanced breast cancer (aBC) or metastatic breast cancer (mBC) survive for 5 years [1].
Hormone receptor-positive/human epidermal growth factor receptor 2-negative (HR + /HER2 −) breast cancer is the most common subtype, with an estimated annual incidence rate of 87.0/100,000 women during 2013-2017 in the US [1,2]. Treatment for patients with HR + /HER2 − aBC/mBC had been unchanged for almost two decades prior to the approval of palbociclib, a first-in-class oral cyclin-dependent kinase 4/6 (CDK4/6) inhibitor approved in 2015. Palbociclib is indicated for the treatment of HR + /HER2 − aBC/mBC in combination with an aromatase inhibitor (AI) as initial endocrine-based therapy and in combination with fulvestrant in patients with disease progression following endocrine therapy in postmenopausal women [3]. In the PALOMA trials, palbociclib combination therapy demonstrated significantly improved progression-free survival, tumor shrinkage, and overall clinical benefit compared with endocrine therapy plus placebo and was associated with primarily hematologic adverse events (AEs), including neutropenia, with a low frequency of febrile neutropenia (1.4%) [4][5][6][7][8].
Since approval, palbociclib has been broadly adopted for treating aBC/mBC, and two other CDK4/6 inhibitors, ribociclib and abemaciclib, have subsequently been approved. For HR + /HER2 − aBC/mBC, current indications for ribociclib are combination treatment with (1) an AI as initial endocrine-based therapy in premenopausal, perimenopausal, or postmenopausal women or (2) fulvestrant in postmenopausal women as either initial endocrine therapy or following disease progression on endocrine therapy [9]. Abemaciclib is currently approved for similar indications to those of palbociclib and as monotherapy in patients with HR + / HER2 − aBC/mBC with disease progression following endocrine therapy and prior chemotherapy in the metastatic setting [10]. CDK4/6 inhibitors have become widely used in practice and are considered standard or preferred options for the treatment of women with HR + /HER2 − aBC/mBC in first line as well as later lines of therapy [11,12].
There is an extensive body of evidence on the efficacy and safety of CDK4/6 inhibitors broadly and a growing body of real-world evidence focusing on palbociclib's effectiveness specifically, but little information to date is available regarding the day-to-day impact of aBC/mBC and its treatment on health-related quality of life (QOL) in a real-world setting. We sought to capture this information and the impact of neutropenia, the most common AE for palbociclib, on patient-reported, health-related QOL.

Study design
The study was a prospective, noninterventional, multicenter longitudinal study of US HR + /HER2-aBC/mBC patients initiating first-, second-, or third-line treatment with either palbociclib in combination with an AI (P + AI) as initial endocrine therapy 1 or palbociclib in combination with fulvestrant (P + Ful) after progression on prior endocrine therapy (according to the US Food and Drug Administration-approved indication for palbociclib) or any other approved therapies for aBC/mBC other than palbociclib [13]. This manuscript focuses on only those patients initiating palbociclib combination therapy.
Participating investigators screened patients for eligibility, obtained written informed consent, and enrolled patients. Investigators completed electronic case report forms (eCRFs) to capture demographic, medical history, and treatment information at enrollment and recorded interim changes in treatment, clinical outcomes, and AEs for 6 months after enrollment. Enrolled patients were provided access to and trained on the use of a custom-developed mobile application, downloaded to their smartphones, to complete baseline, daily, weekly, and cycle-based patientreported outcome (PRO) assessments for the 6-month period ( Fig. 1). PRO and eCRF data were combined at the patient level.
Data collection for this study occurred over a 33-month period (February 2017-October 2019). The data collection period for an individual patient was approximately 6 months, which could be truncated by treatment switching, patient withdrawal from the study, or death. Follow-up ended if patients discontinued palbociclib combination therapy.

Eligibility
Eligibility was assessed prior to enrollment during a scheduled visit; any woman who met the eligibility criteria was invited to participate. Eligible women were aged ≥ 18 years with a diagnosis of adenocarcinoma of the breast and evidence of aBC/mBC not amenable to resection or radiation therapy with curative intent, had evidence of HR + /HER2 − tumor based on the patient's most recent tumor biopsy, and owned or had regular access to a smartphone. Exclusion criteria were as follows: life expectancy < 3 months at the time of diagnosis of aBC/mBC, participation in any interventional clinical trial, and active treatment for other malignancies other than aBC/mBC.

Endpoints
Study endpoints included PRO assessments (to describe patients' symptoms, health status, mood, depression, ability to function, overall health, and QOL) and safety information (AEs).
Questionnaires assessing PROs included the 12-Item Short Form Health Survey (SF-12), the 10-Item Center for Epidemiological Studies Depression Scale (CES-D-10), and targeted questions, which were made available to patients via the mobile application downloaded onto their smartphones [14][15][16][17]. Patients completed a baseline questionnaire as well as a series of questions at daily, weekly, and cycle-based intervals. Not all measures collected are examined in this manuscript. Standard survey methodological principles were used to draft items [18].

Daily and weekly assessments
Pain and fatigue severity were measured using an 11-point numeric rating scale (NRS); 0 indicated no pain or fatigue, and 10 indicated the worst possible pain or fatigue.
Mood was rated daily on a 7-point scale ("Very sad," "Sad," "Discontent," "Neutral," "Content," "Happy," "Most happy") and summarized at the weekly level to determine the mean percentage of nonmissing days for each mood level across all patients.
Patients indicated weekly how breast cancer or its treatment interfered with family/social life, productivity, physical activities, and energy on a 5-point scale ("Not at all," "A little," "Moderately," "Quite a bit," "A great deal"). Additionally, patients rated their health and overall QOL during the past 7 days on a 5-point scale ("Poor," "Fair," "Good," "Very good," "Excellent").

SF-12
The SF-12 is a shortened version of the 36-Item Short Form Health Survey for measuring general health status in study populations [19]. Physical Component Summary (PCS-12) and Mental Component Summary (MCS-12) scale scores were calculated from responses to yield a mean score of 50 and a standard deviation (SD) of 10 in the US population; higher scores represented better health. The PCS-12 and MCS-12 summary measures provide a rough comparison

CES-D-10
The CES-D-10 is a 10-item self-administered questionnaire assessing depression during the last week. The 10 items are rated on a 4-point scale ranging from 0 (< 1 day) to 3 (5-7 days) and summed to produce a total score (range, 0-30). A total score of 10 or more is considered to indicate depression [14].

Safety endpoints
Safety endpoints included the incidence, severity, and duration of neutropenia events and other AEs occurring during the study. AEs were recorded in the eCRF and included selected AEs of special interest (nonfebrile neutropenia, febrile neutropenia, leukopenia, infection, fatigue, nausea, anemia, stomatitis, headache, diarrhea, thrombocytopenia, constipation, alopecia, vomiting, rash, and decreased appetite), dates of onset and resolution, National Cancer Institute Common Terminology Criteria for Adverse Events toxicity grade, outcome, action taken with treatment, and seriousness. AEs not listed on the eCRF were categorized as "other."

Statistical analysis
Analyses were conducted using the full analysis set (FAS), which comprised all patients in the study population with at least baseline mobile app data and data in both the screening and enrollment eCRFs. The SF-12 and CES-D-10 analysis set consists of all patients in the FAS who completed at least one baseline SF-12/CES-D-10 question and at least one postbaseline SF-12/CES-D-10 question.
Descriptive analyses were performed on categorical and continuous endpoints collected or derived from the eCRF and mobile application. Summary statistics were displayed by overall therapy and stratified by line of therapy.
At each cycle, the relationship between all PROs and episodes of neutropenia was explored by comparing patients who did not experience neutropenia during the study (for the cycle-based PROs) and patients who experienced neutropenia at the time point.
Linear mixed-effects models for repeated measures were used to summarize change from baseline values by neutropenia status for the cycle-based PCS-12, MCS-12, and CES-D-10 scores. Fixed effects included treatment, month, treatment by month interaction, neutropenia status at the time point, neutropenia status by month interaction, neutropenia status by treatment interaction, neutropenia status by month by treatment status interaction, and baseline score; a random effect was included for the patient. Least-squares mean estimates and 95% confidence intervals (CIs) for the change from baseline values by neutropenia status were reported for each stratum at each cycle and overall across cycles.
To examine the overall relationship between mood and CES-D-10 score, mixed models for repeated measures were conducted, with CES-D-10 score as the independent variable and a fixed effect for the respective mood question and a repeated effect to account for the correlation within patients. Although mood was available weekly, only the first week of each cycle was used in the analysis to match the recall period of the cycle-based CES-D-10.
As this was primarily a descriptive study, any statistical comparisons are presented at the 2-sided alpha = 0.05 level without adjustment for multiplicity. Any missing data were assumed missing at random. Analyses were conducted using SAS statistical software, version 9.4 (SAS Institute Inc, Cary, North Carolina).

Baseline characteristics
This study enrolled 139 evaluable patients from 25 participating sites. Demographic data are shown in Table 1, and clinical data are shown in Table 2. The median (range) age of patients was 60 (34-82) years; 83% were white, 9% were black/African American, 5% were Hispanic/Latino, and 1% were Asian. About half of patients (49%) were employed. Sixty-one percent of patients initiated P + AI and 39% initiated P + Ful. Almost all patients (96%) initiated palbociclib at a dose of 125 mg/day. Based on medical records, 22% of patients had Stage IV, de novo, disease at diagnosis. At enrollment, the median duration between mBC diagnosis and study enrollment was 1.0 month. Thirty-nine percent of patients had visceral metastases and 41% had bone-only disease. Most patients had an Eastern Cooperative Oncology Group performance status of 0 (63%) or 1 (24%).

Daily pain
For patients overall, mean (SD) daily level of pain was 2.2 (2.31) on the NRS (0-10, with 10 being the worst possible pain), averaged across week 1 of cycle 1. At week 1 of cycle 6, mean (SD) pain was slightly lower at 1.8 (2.19). There were no substantial changes through week 1 of cycle 6 among the two palbociclib treatment groups (Fig. 2a). For patients experiencing neutropenia at the timepoint, mean (SD) pain ranged from 1.8 (2.2) to 2.8 (2.0) across the first week of each cycle and was slightly higher at cycle 3 and cycle 6 than among those not experiencing neutropenia at any time during the study (Fig. 2b).

Daily fatigue
Overall, mean (SD) daily level of fatigue was 2.7 (2.3) on the NRS (0-10, with 10 being the worst possible fatigue), averaged across week 1 of cycle 1. No substantial changes in fatigue NRS were observed through week 1 of cycle 6 overall or between the two palbociclib combination treatment groups (Fig. 3a). Despite slightly higher mean fatigue NRS scores among patients with no neutropenia events relative to those who experienced neutropenia at week 1 of cycle 1, no substantial differences between these groups were observed through week 1 of cycles 2-6 ( Fig. 3b).

Weekly mood
At the first week of cycle 1, a majority (90%) of participants reported their mood ranged from "Neutral" to "Most happy" ( Figure S1). There were no substantial changes through week 1 of cycle 6 overall or among the two palbociclib treatment subgroups. For patients without an indication of depression (i.e., CES-D-10 score < 10) at cycle 1, week 1, the mean percentage of nonmissing days where mood was ranked as "Most Happy," "Happy," or "Content" was 62% versus 24% in patients with depression. For patients without depression, aBC advanced breast cancer, ECOG, Eastern Cooperative Oncology Group; mBC metastatic breast cancer, NSAID nonsteroidal anti-inflammatory drug; P + AI palbociclib in combination with an aromatase inhibitor, P + Ful palbociclib in combination with fulvestrant, SD standard deviation a Liver, brain, lung/pleura, and ovary metastases were categorized as visceral sites. Bone, lymph node, and skin/soft tissue metastases were categorized as nonvisceral sites. Other sites were categorized as visceral or nonvisceral by clinical review. Patients who had metastasis at any visceral site were classified as having visceral metastases; those without metastases at any visceral site were classified as having nonvisceral metastases b No patients with an ECOG performance status of 3 or 4 were reported the mean percentage of nonmissing days with a positive mood was stable across the first week of each cycle, while it ranged from 15 to 38% for patients with depression ( Figure  S2). The use of a mixed model for repeated measures, with fixed effects for negative impact on mood or emotions, suggests patients with mood or emotions not adversely impacted by breast cancer reported lower scores on the CES-D-10 (i.e., less indicative of depression). Specifically, the CES-D-10 least-squares mean (95% CI) for patients reporting "Not at all" was 4.41 (3.2-5.6), whereas the CES-D-10 least-squares mean for patients reporting "A great deal" of impact was 13.6 (11.0-16.3) (Fig. 4).
The use of a mixed model for repeated measures, with fixed effects for percentage of days with content, happy, or most happy mood, indicated that happy mood is associated with lower scores on the CES-D-10 (i.e., less indicative of depression): β = -0.04 (P < 0.001) (Fig. 5).
Descriptive analysis revealed no substantial difference in weekly mean percentage of nonmissing days for instances in which mood was ranked as either "Most  happy," "Happy," or "Content" between patients who did or did not experience neutropenia.

Interference from breast cancer or its treatment
Most patients across the first week of all cycles indicated aBC/mBC or its treatment interfered "Not at all" or "A little" with family life ( Figure S3a) or social life ( Figure S3b). Similarly, most patients indicated aBC/mBC or its treatment limited physical activity ( Figure S4a), energy or stamina ( Figure  S4b), and productivity ( Figure S4c) "Not at all" or "A little." The percentage of patients who indicated aBC/mBC or its treatment limited these items "Quite a bit" or "A great deal" did not exceed 20% across the first week of all cycles and,
Descriptive examination of the mean (SD) PCS-12 scores by neutropenia status indicated patients with neutropenia at cycle 6 had a score of 39.0 (11.3) compared with 41.8 (10.63) for those who did not experience neutropenia during the study (Fig. 6c). For the MCS-12, the mean (SD) scores at cycle 6 were 50.0 (12.1) and 52.5 (9.1), respectively, for patients with neutropenia at cycle 6 and those without a neutropenia event during the study (Fig. 6d). Based on the mixed models for repeated measures, the least-squares mean (95% CI) change from baseline across all cycles for patients with and without neutropenia was 0.
Based on the mixed models for repeated measures, the least-squares mean (95% CI) change from baseline across all cycles was 0.0 (− 0.7 to 0.6). Least-squares mean changes from baseline were consistently lower in P + AI patients than in P + Ful patients. At cycle 6, least-squares mean changes from baseline were − 0.4 (− 1.2 to 0.5) and 0.5 (− 0.6 to 1.6), respectively, for P + AI and P + Ful. Descriptive examination of the mean (SD) CES-D-10 scores by neutropenia status indicated patients with neutropenia at cycle 6 had a score of 9.2 (6.9) compared with 6.2 (5.6) for those who did not experience neutropenia during the study (Fig. 7b). Based on the mixed models for repeated measures, the least-squares mean (95% CI) change from baseline across all cycles for patients with neutropenia and those without neutropenia was 0.8 (− 0.1 to 1.8) and − 0.3 (− 1.1 to 0.4), respectively.

Overall weekly health rating and quality of life
Most patients across the first week of all cycles indicated their overall health rating and QOL was "Good," "Very good," or "Excellent" (Figure S5a and S5b). In the first week across all cycles, weekly overall health and QOL was ranked as "Poor" by less than 8% and 4% of patients, respectively. These findings were generally similar regardless of neutropenia status, where ratings remained stable across cycles.

Discussion
Treatment for HR + /HER2 − aBC/mBC has evolved with the approval of the CDK4/6 inhibitor class combined with endocrine agents, beginning with palbociclib in February 2015 [20]. Real-world data regarding patient QOL and experiences while receiving palbociclib, however, are limited.
To our knowledge, our study is one of the first to evaluate the day-to-day impacts of aBC/mBC and its treatment on patients and the effects of treatment-induced neutropenia on patients' daily functioning outside the context of clinical trials. We found that, on the whole, patients treated with palbociclib combination therapy experienced low levels of baseline pain and fatigue, which remained stable across cycles; general health status as measured by the SF-12 remained consistent throughout treatment and was generally consistent with published norms for individuals diagnosed with cancer (excluding skin cancer), and the presence of depression, as indicated by the patient-reported CES-D-10, was low and did not change substantially over time [18]. Patients, on average, reported a neutral or positive mood. Patient-reported QOL and overall health was primarily "Good," "Very good," or "Excellent." Patients with neutropenia, on the whole, did not show decreases in QOL during the study, and findings were consistent with those who did not experience neutropenia.
With limited real-world information available regarding patient experiences with aBC/mBC and treatment-induced neutropenia, this study provides valuable data to inform patient and physician treatment discussions and health care decision-making.

Limitations
The sample of participating sites was a convenience sample and may not be representative of all US centers treating patients with aBC/mBC. Only 139 evaluable patients of a targeted 300 were enrolled. The patient selection and monitoring procedures are those applied per the treating physician's usual treatment paradigm and not dictated by the protocol. Heterogeneous patient populations could make the interpretation of the outcomes difficult.
As with all studies requiring patients to self-report outcomes and behavior, completeness and accuracy of reporting can be a concern. AE collection was limited to what was spontaneously disclosed during standard of care visits and retrospective medical record review. Patients without a smartphone were not eligible for participation and may have different outcomes. Some patients may have discontinued the study early because of progression or declining health, which is associated with QOL. Early withdrawals decreased sample sizes at later cycles but examination of PRO assessments by completion status found no substantial differences. Correlation of disease burden with severity of pain was not examined in this study. Results of this study are subject to potential selection bias and responder bias, and whether individuals who elected not to participate would have reported different outcomes is unknown.

Conclusion
This study demonstrated that palbociclib-treated patients, on average, reported consistently low levels of pain and fatigue as well as good QOL, and overall health remained stable during 6 months of treatment. Episodes of neutropenia did not impact QOL.
Acknowledgements The authors thank all of the patients and study investigators.
Funding This study was conducted by RTI Health Solutions under the direction of Pfizer Inc. and was funded by Pfizer Inc.