Understanding Pain Related to Adjuvant Endocrine Therapy after Breast Cancer: A Qualitative Report

Abstract

Objectives:

Most patients report pain while taking adjuvant endocrine therapy (AET) for the treatment of breast cancer. While studies have examined patients’ experiences with side effects, none solely capture patients’ experiences with AET-related pain, a troubling symptom which reduces quality of life and impairs treatment adherence. This study explored themes of AET-related pain to inform future intervention development.

Methods:

Between 11/2017–11/2018, female patients (n=30) with early-stage breast cancer enrolled between 3–36 months post-initiation of AET. Purposeful sampling was stratified by adherence level, age, distress level, and time taking AET. Study staff conducted, transcribed, and coded semi-structured interviews via inductive thematic coding to identify pain-related themes and achieved high inter-coded reliability (Kappa = .96).

Results:

Several pain-related themes were observed. Attitudes around pain are generally negative and management needs are largely unmet. Patients reported preferences for non-pharmacological management strategies and cited AET pain as a reason for medication breaks but not discontinuation. Patients within 19 months of starting AET and low adherers reported more intense and disruptive pain.

Conclusions:

Patients’ experiences varied by patient attributes and revealed modifiable factors that may be targeted through behavioral interventions. AET-related pain is a complex side effect for which psychosocial support may be beneficial.

Introduction

Adjuvant endocrine therapy (AET) is a critical treatment for hormone-sensitive breast cancer and provides up to 50% reduction in recurrence and cancer-related mortality (Burstein et al., 2019). AET operates by blocking endogenous estrogen and progesterone from binding with hormone receptors on cancer cells (e.g., aromatase inhibitors (AIs)) or through reduction in levels of estrogen in the body (e.g., tamoxifen) (Ramchand et al., 2019). Practice guidelines recommend AET for up to 10 years (Burstein et al., 2019). However, such medications are associated with impairing side effects, such as pain, hot flashes, sexual dysfunction, and sleep disruption (Condorelli & Vaz-Luis, 2018).

In particular, AET-related pain, including diffuse musculoskeletal and joint complaints, arthralgias, stiffness, cramps, tingling or numbness, and headaches, is reported by 61–74% of patients and poses significant challenges to daily routines and functioning (Moscetti et al., 2015; Haidinger & Bauerfeind, 2019; Boonstra et al., 2013; van Londen et al., 2014). Reasons for AET-related pain are not well understood but may result from physiological changes that contribute to bone and cartilage breakdown, increased inflammation, and decreased pain tolerance (Niravath, 2013). Additionally, medical-related factors exist that place women at an increased risk for AET-related pain, including low vitamin D levels and chemotherapy prior to AET initiation (Niravath, 2013; Younus & Kligman, 2010). Furthermore, the natural aging process, menopause, and existing co-morbidities, such as fibromyalgia, contribute to the risk of new or worsened pain for patients receiving AET (Winters et al., 2007; Presant et al., 2006; Andrews et al., 2018; Rief et al., 2011).AET-related pain has been associated with lower medication adherence (Brett, Boulton, et al., 2018), decreased motor functioning (Boonstra et al., 2013), impaired sexual and social functioning (Boonstra et al., 2013; Ribi et al., 2020), inability to care for oneself or family, greater disability (Boonstra et al., 2013), and worse quality of life (Ferreira et al., 2019). Ultimately, strategies for addressing pain are essential for patients on these therapies.

Some strategies to manage AET-related pain exist and consist of medication (e.g., nonsteroidal anti-inflammatories, duloxetine), non-pharmacologic interventions (acupuncture, yoga, exercise, massage), switching AET regimens or type, prescribing a lower dose or break, and discontinuing AET (Winters et al., 2007; Presant et al., 2006; Hershman et al., 2018). While some evidence-based behavioral therapies exist for breast cancer-related pain or for management of common AET side effects (e.g., hot flashes; Hershman et al., 2018; Johannsen et al., 2013;), few studies have explored these interventions for AET-specific pain, which is qualitatively different due to the physiological underpinnings and primary musculoskeletal presentation of pain (Shelby et al., 2019; Fenlon et al., 2018; Dorfman et al., 2019; Warth et al., 2020). Furthermore, while qualitative studies have explored the general experiences of patients taking AET (Jacobs et al., 2020; Harrow et al., 2014; AlOmeir et al., 2020; Wells et al., 2016; Moon et al., 2017; Brett, Fenlon, et al., 2018; Cahir et al., 2015), none focus solely on understanding AET-related pain experiences and consequences. Therefore, this qualitative study aims to understand nuances of the pain experience, if experiences vary by patient characteristics, and to guide the development of AET pain-specific resources and support for patients with early-stage breast cancer.

Methods

Study Design and Procedures

Between November 2017 and November 2018, study staff recruited patients for a qualitative study about AET experiences and adherence. Main study results are reported elsewhere (Jacobs et al., 2020). This single-site study was performed at Massachusetts General Hospital, a large academic medical center in the northeast region of the United States. The study was approved for human subjects research by the Dana Farber/Harvard Cancer Center institutional review board (DF/HCC IRB #17–315), and patients provided informed consent prior to enrolling.

Eligibility criteria included females diagnosed with stage 0-IIIB breast cancer who had completed treatment, were between 3–36 months post-initiation of AET, and had an Eastern Cooperative Oncology Group performance status ≤ 2. While musculoskeletal issues are more prevalent for women on AIs, this is a symptom experienced by women on both AIs and tamoxifen, and therefore we enrolled both. 3 months post-initiation of AET has commonly been selected in prior literature as an early follow-up timepoint to assess preliminary AET-related symptoms (Cuzick, Sestak, Cella, & Fallowfield, 2008). 36 months post-initiation is commonly used as follow-up endpoint for studies assessing AET symptoms and aftereffects (Garcia et al., 2021; Galvano et al., 2019; Wagner et al., 2020). As such, we included participants within this timeframe after AET initiation to contextualize experiences of pain of patients compared to existing literature on other experiences in this timeframe. Additionally, patients had to be ≥ 21 years old and fluent in English. Patients were ineligible if they were diagnosed with a condition that would interfere with study procedures, such as a severe, untreated psychiatric or cognitive disorder, or if they had discontinued AET in the prior month. Study staff identified patients with upcoming clinic appointments via the electronic health record (EHR). Staff received permission from oncology clinicians to approach eligible patients and introduce the study.

Study staff presented the study to participants and provided information that the purpose of the interviews was to better understand experiences related to adjuvant endocrine therapy. Interested and eligible patients provided informed consent and completed a stratification questionnaire to ensure adequate distribution of patients by age, distress level, adherence level, and time taking AET (3–19 versus 20–36 months). Patients did not screen into the study if their corresponding strata goal had been met or qualitative thematic saturation already achieved. Enrolled patient stratification breakdown may be found in Table 1 and method of stratification was as follows.

Table 1.

Participant stratification

	Low Adherence	High Adherence
Low Distress	14	7
High Distress	6	3

Patients reported the amount of AET taken over the past 30 days on a Visual Analog Scale (Kalichman et al., 2009) (0–100%) and if they experienced any challenges taking it, which was used to categorize high versus low adherers. Distress levels were assessed using the nine-item Patient Health Questionnaire (PHQ-9) (Kroenke et al., 2001) and seven-item Generalized Anxiety Disorder (GAD-7) (Spitzer et al., 2006). High distress was classified by scores ≥ 8 on the PHQ-9 and/or ≥ 10 on the GAD-7, as supported by existing literature (Thekkumpurath et al., 2011; Trevino, Saracino, & Roth, 2020). Study staff collected disease information from the EHR, and patients self-reported sociodemographic characteristics.

Upon enrollment, trained study staff conducted and audio recorded individual, one-time, semi-structured interviews in person in a private clinic room or by telephone, and patients were remunerated $20. A private, HIPAA-compliant transcription company transcribed the interviews verbatim, and study staff reviewed interviews to ensure accuracy of transcripts and complete deidentification. Interviewers took notes throughout interviews to assist in following the semi-structured interview; however, these notes were securely discarded and not used for data analysis. Participants did not review study transcripts or feedback on study finding.

The study team (JJ, EW) created and refined the interview guide with a qualitative expert (EP) and feedback from cancer center psychologists and breast oncologists. Interviews inquired about experiences with AET and preferences regarding a psychosocial intervention. The study team (JJ, EW) attended institutional qualitative interview trainings and completed interview role plays prior to conducting all interviews. Both interviewers were female, held the roles of principle investigator of the study and clinical research coordinator, and had obtained a Doctor of Philosophy in clinical psychology and bachelor’s degree, respectively. Both interviewers maintained interests in psycho-oncology and symptom management and have numerous years of experience conducting related clinical research with women with breast cancer. Grounded in prior research and clinical experience, interviewers held the assumption that women may experience difficult side effects and symptoms related to adjuvant endocrine therapy, which they objectively probed through open-ended interview questions and supervision.

Patient Characteristics

In total, 333 patients were approached, 212 consented, and 30 participated in semi-structured interviews. Consented patients were excluded or did not interview for the following reasons: screened out of stratification measures (n=174); ineligible (n=1); lost to follow-up (n=5); withdrew (n=2). Patients were primarily partnered (76.7%), middle-aged (M=55.13 years, SD=12.37), White (83.3%), non-Hispanic (93.3%) women. Of these, 30% (9/30) reported high distress, and approximately 40% (12/30) were taking tamoxifen with the rest on AIs. Patient characteristics are detailed in Table 2. Semi-structured interviews lasted on average 27 minutes (range=12–43).

Table 2.

Patient characteristics

	N (%)
Gender: Woman	100 (30)
Race
White	25 (83.3)
Asian	3 (10)
Black or African American	1 (3.3)
Other	1 (3.3)
Ethnicity
Hispanic or Latino/a	2 (6.7)
Education
Doctoral/medical/law degree	2 (6.7)
Master’s degree	10 (33.3)
College graduate	11 (36.7)
2 years of college/associates degree	6 (20.)
High school graduate/GED	1 (3.3)
Relationship Status
Married/living as if married	21 (70)
Single, never married	4 (13.3)
Non-cohabitating relationship	2 (6.7)
Divorced/separated	3 (10)
Employment Status
Full-time or part-time	21 (70)
Caring for home or family	1 (3.3)
Unable to work due to illness/disability	1 (3.3)
Retired	7 (23.3)
Income
$25,000-$49,999	3 (10)
$50,000-$99,999	5 (16.7)
$100,000-$149,999	14 (46.7)
> $150,000	6 (20)
Missing	2 (6.7)
Type of AET
Anastrozole	8 (26.7)
Exemestane	4 (13.3)
Letrozole	6 (20)
Tamoxifen	12 (40)
Breast Cancer Stage
I	17 (56.7)
II	10 (33.3)
III	3 (10.0)

	M (SD)	Range
Age (years)	55.13 (12.37)	27–76
Time from AET start date to enrollment (years)	1.76 (.75)	.28–2.9
Adherence to AET on VAS (% taken)	89.27 (17.71)	5–100
Total PHQ-9 (Sum)	4.47 (4.64)	0–16
Total GAD-7 (Sum)	4.40 (4.19)	0–16

Data Analysis

Two trained study staff (EW, RC) created a comprehensive pain-related thematic coding framework for relevant interview codes. The framework utilized inductive thematic content analysis with a six-step process including data familiarization, initial coding, generating themes, reviewing themes, defining, and naming themes, and report writing (Table 3; Braun & Clarke, 2006). Using NVivo 11 software, study staff (EW, RC) independently coded each transcript, compared coding schemes, saved the audit trail throughout, and resolved discrepancies with PI oversight (JJ). Data saturation was achieved once no new codes were identified. A high level of coding reliability was achieved (Kappa=.96). Staff (EW, RC) reviewed specific codes to identify and compare themes associated with AET-related pain and across strata.

Table 3.

Evolution of the Coding Framework and Generated Themes

	Addition to Coding Framework	Generated Theme
Specific location of pain	Version 2	Pain Experiences, Attitudes, and Attributions
Strong language about pain	Version 1
Pain as a reason to not take AET	Version 1
Length/acuteness of pain	Version 1
Attribution of pain to AET	Version 1
Impact of pain on daily living	Version 1	Pain Interference
Social support related to pain management	Version 3	External Support for Pain
Desire for a pain intervention or program	Version 1
Patient/provider communication	Version 1
Methods for managing the pain	Version 1	Individual Strategies and Barriers for Pain Management
Barriers to managing the pain	Version 1

Results

Patient-Reported Themes

The following themes were identified with illustrative quotes in Table 4.

Table 4.

Qualitative themes

Pain Experiences, Attitudes, and Attributions “It’s miserable. Being in pain is just miserable, and that’s what tamoxifen makes you.” “It’s very nasty. I don’t like it one bit, but I don’t feel I have a choice. It causes a lot of muscle and bone pain. I’ve already sort of not been a great sleeper since going through menopause, and it’s just made that worse. I feel I’m crabbier, probably because I don’t sleep as well, and I’m often uncomfortable. I don’t really feel like I have a choice, so on we go.” “I think that the most outstanding side effect for me is joint, bone and muscle pain.” Possible worsening of pain “I called the hospital and then I finally started taking it. I was really frightened, I did not want any more pain. I’m very sensitive to pain in my hands and my feet and I just didn’t want that sort of thing.” “The muscle weakness and joint pain were one of the things that were common side effects, and when I started out, I didn’t have a big problem with it, but now that my MS has kicked up, I do.” “Well, I’ve always had joint pain. I’ve had increasing joint pain since taking it. I’ve been almost three years now, and I never know whether it’s just the natural aging process or whether it’s due to the anastrozole, the increased joint pain.” Attribution “I had a little bit of joint pain, or my feet hurt a little bit every once in a while, but I didn’t really have [pain]. I can’t tell if the side effects are from the medicine or if it’s from menopause itself.” “It’s hard to quantify the association of whether that’s to the exemestane or whether it’s just typical post-treatment experience. I think the same would go for the tiredness. The joint pain is clearly related to the medication, but the others, I’m less clear as to whether the correlation is directly to the drug or not.” Pain Interference “No, because I feel like I’m good to remember taking my medication. I just haven’t because I don’t want to have pain.” “I had really bad headaches. Every day, I used to get headaches. That was because of the stroke. So, on those days, I used to skip. I used to skip for, continuously, two, three days, depending on the level of headaches I had. They tend to give me headaches, because then I would skip like that.” “I was finding was I was getting in the middle of the day and having so much pain, I’d end up taking the pain medication, but I wouldn’t take the blood pressure medication, or the cholesterol medication, or the acid reflux drug.”

External Support for Pain “He [oncologist] said, “If this doesn’t work then we could switch to the letrozole,” and it’s been helping. It was mostly the leg pain that I had a little bit of issue with all through my treatment, especially with the Neulasta and that whole thing. So, he said, “All right, well, let’s try it.” And honestly, it was pretty quick that I realized it was a much better fit for me.” “I would say the joint pain more so has been frustrating, and it hasn’t impacted my way of thinking about taking it. My husband has said, ‘You should think about whether you still want to deal with this with all the issues that you’re having with it.’ But I feel like that is not really a viable option, meaning that I do not think I should stop taking it because, again, the benefits outweigh the challenges.” P30 “It was really painful and causing cramping when I’m sleeping and when I’m at work. He [oncologist] said, “Okay. Let’s just see if it is the medication.” I went off of it for a couple of weeks. Maybe it was three weeks or something like that. And then if it went away, then the deal was I would just send him an email and say, “It is the medication.” And I would go back on it at that time just to see that that what was causing it. And sure enough it was. It had gone away when I was off of it. So, I emailed him and then I went back on it again. And it has all the side effects and all of that.” “I saw my primary care provider and was telling her everything I was experiencing. And she said, ‘Well, I think we should try you on something to see if we can help with your discomfort and help with the sleep.’ And as I’ve said, I feel better. And I don’t know if that’s from adding the Paxil into the mix or just kind of that’s what happens after a while with the medication, and then we’ll see.” Individual Strategies and Barriers for Pain Management “I used to have a lot of body aches, but then I went for alternative medicine.” “Because of the nausea, I took all my other pills in the morning, and then I had to increase my pain medication because I take Tylenol, and I take ibuprofen, and I take naproxen. So, I’ve had to increase that. And at night, I take the exemestane with those pain medications again so that I’m able to sleep. I take the pill with the pills. So I’ve been able to juggle it, and that way, the nausea’s not so bad if I take it at night because I’m sleeping, and it doesn’t impact me not taking my other pills.” “I can kind of tolerate the bone pain because it does-- It is when I wake up or sit for a period of time and then it does get better as I move.” P28 “I got a massage once or twice to help with the feet, but that’s just temporary. That’s just a temporary relief.”

Pain Experiences, Attitudes, and Attributions

Patients primarily reported negative attitudes about pain, often using strong language. Pain was an unwanted, feared side effect, which contributed to poor attitudes about AET. Patients commented on the location and experience of pain, commonly mentioning body aches, headaches, bone and joint pain, and cramps. Duration and acuity of pain was highly variable and often compounded preexisting pain. Some patients experienced pain briefly and others for the entirety of their AET regimen. Attribution of pain also varied, and some reported ambivalence or suspected other underlying causes, such as comorbid conditions, other breast cancer treatments, aging, or menopause-related changes. Patients expressed hopelessness and helplessness related to pain, expectations that AET would induce pain, and associated pain with long-term health consequences.

Pain Interference

Patients reported several ways in which pain interferes with their daily living and functioning, including hindering AET adherence. Reducing pain was often a reason to skip or avoid taking AET. They also discussed providers using pain as a reason to prescribe an AET break. Patients identified pain as a cause of sleep disruption, increased nausea, and unstable mood. Notably, patients described a nuanced understanding that the benefits of AET outweighed the pain and did not ascribe pain as a reason to consider discontinuing AET entirely.

External Support for Pain

Patients reported discussing decisions about AET due to their pain with family members and receiving emotional support from other cancer survivors. Patients reported receiving practical and informational support from oncologists through new or revised prescriptions, information on AET, and referrals. Patients also reported receiving useful informational support from primary care physicians and pharmacists. While noting that their care team occasionally had differing opinions on the underlying cause of their pain, patients expressed high levels of trust in their providers. Some patients wished they had received more information before starting AET while others felt they had too much information about the possibility of pain as a side effect, contributing to hypervigilance about bodily sensations. No participants mentioned referrals specifically for psychological pain treatment or support.

Individual Strategies and Barriers for Pain Management

Patients reported feeling little control over their ability to manage pain despite high levels of AET self-directed research or knowledge. Strategies to manage pain were non-pharmacological (e.g., changing time of medication, exercise, massage), pharmacological (e.g., pain medication, switching AET medication), and psychological (e.g., acceptance). Patients reported strong preferences to manage pain without medication. Difficulties with management strategies were reported, including fears of overmedicating and logistical barriers, such as physician access.

Pain Experiences by Patient Attributes

Experiences with pain varied by distress level, AET type, age, time on AET, and adherence level. Overall, high adherers reported primarily mild joint and muscle pain, while low adherers described pain with stronger descriptors (e.g., “terrible” and “miserable”) and minimal success in management other than medication breaks. Patients with high distress reported pain as one of several persistent side effects and often attributed pain to AET. Alternatively, for patients reporting low distress, pain improved over time and management strategies were effective. Pain was particularly disruptive among patients who had been on AET between 3–19 months. For patients within 20–36-months, pain improved over time after several management attempts. These patients expressed concern that pain indicated long-term health consequences. Younger patients spoke often of pain in conjunction with menopause or their menstrual cycle, whereas older patients mentioned other medical conditions and comorbidities. While younger patients expressed intense discomfort and hopelessness, older patients spoke with acceptance of pain. Lastly, some nuances were observed per AET medication type. Patients taking tamoxifen reported several failed pain management attempts and often described pain as one of many symptoms. Patients taking AIs often reported intense bone and joint pain, and some reported satisfaction with pain reduction after having switched from a different AET.

Discussion

Four broad themes capture the pain experiences of patients taking AET after early-stage breast cancer: pain experiences, attitudes, and attributions; pain interference; external support for pain; and individual strategies and barriers for pain management. This study provides a novel qualitative and thematic context for understanding the AET-related pain experience, especially for certain subgroups of women with breast cancer, and may reveal modifiable targets for a pain-management intervention.

Patients who had been on AET longer reported less pain, concordant with evidence that suggests AET-related pain peaks about 6 months after initiation (Boonstra et al., 2013). As supported in existing literature, patients, especially those older than 50, mentioned comorbidities or preexisting conditions as pain contributors (Winters et al., 2007). While literature suggests that pain is a source of AET discontinuation (Winters et al., 2007; Presant et al., 2006), patients in our study articulated the benefits of AET and solely articulated thoughts of AET breaks, not discontinuation, due to pain. Patients also reported high levels of trust and support from healthcare providers, which may be a nuance of our sample that makes self-guided discontinuation less likely. Social desirability related to self-reporting nonadherence must also be considered in interpreting this finding. Women taking an AI report higher rates of pain than those taking tamoxifen (Choo et al., 2019; Burstein, 2007). Findings of this qualitative study are consistent with this; however, patients in this sample taking tamoxifen also reported many failed attempts to manage pain. The complexity within this finding may be a result that pain was reported as one of several side effects on tamoxifen, so these patients are more burdened with side effect management and not solely with pain management overall.

Patients in this sample did not report social interference due to AET-related pain, which is somewhat surprising given our previous findings that women desire more support than they currently receive (Jacobs et al., 2020). In addition, general medical populations of females with pain report high social isolation (Mellado, et al., 2016). Healthcare providers, primarily oncologists, provided a large amount of support for patients, which is reflective of the context in which this study was conducted at an urban academic medical center.

A wide variety of pain management strategies contributed to mixed success of mitigating pain. Feelings of helplessness and hopelessness resulted from inadequately managed or untreated pain, which is consistent with general pain experiences of patients with cancer (Erol et al., 2018). Consistent with existing literature, patients overall expressed preferences for non-pharmacological pain management strategies (Liou et al., 2021), emphasizing the opportunity for an AET-related pain management intervention in this population.

These recurrent themes underscore areas of need and several modifiable factors for improving AET pain management. Interventions for this population may draw upon evidence-based cognitive behavioral therapy for pain management (Dorfman et al., 2019) while addressing the nuances of the AET and breast cancer experience, such as the potential to affect adherence and daily functioning as noted here. Interventions can encompass strategies to decrease negative emotions about pain and AET, enhance skills for coping with pain, and improve self-efficacy for managing pain. For example, patients often expressed worries about pain or negative beliefs due to pain, which may be addressed with cognitive behavioral strategies, such as thought reframing. Interference in daily functioning due to pain can be mitigated with behavioral techniques such as activity pacing and scheduling and behavioral activation, which may be especially pertinent after breast cancer treatment (Klinger & Levenson, 1989, Sharp, 2001). Relaxation training and acceptance-based skills may bolster patients’ ability to tolerate pain as well as uncertainty about breast cancer recurrence (McCracken & Eccleston, 2003). Intervening on these modifiable targets, in tandem with existing pharmacological and non-pharmacological strategies, may ultimately enhance AET pain management.

Key considerations and limitations of this study include the homogenous sample of primarily White, educated, upper-middle class women. Additionally, this sample was selectively recruited to represent low adherers or high adherers with notable side effect burdens, so pain experiences may be different for other subgroups. We did not selectively screen women for pain; therefore, women on any AET, regardless of whether they experienced pain, were included in the study. Pain experiences in the current sample were not limited to AET-specific pain and results may capture the overlap of AET-related pain among other pain experiences. Previous research indicates differences in pain experiences between early-stage or advanced breast cancer (Hamood et al., 2018). Our findings are limited to those with early-stage breast cancer and may not generalize to later stage disease nor to patients who have since discontinued AET altogether. Furthermore, other forms of breast cancer treatment and preexisting pain confer greater risk of pain after breast cancer (Sheridan et al., 2012). This information was not collected or included for consideration in the current study. Additional research in this area must clarify the role of AET-related pain within more diverse populations, at more advanced stages, for those who have discontinued AET entirely, and in other settings (e.g., rural, community). Moreover, understanding of the role of social support, particularly from healthcare providers, in the management of AET-related pain is warranted to clarify patient preferences for information and role of the care team in pain management. Considerations for risk factors of pain, such as patients closer to time of AET initiation and those with comorbidities, are necessary in treating this population.

Ultimately, this qualitative study explored patients’ experiences with pain while taking AET and identified key themes to understanding these perspectives. Given non-pharmacological pain management preferences, there is an opportunity to tailor evidence-based pain management strategies to meet the unique needs of patients on AET after early-stage breast cancer.

Data Availability:

Deidentified data is available upon request.