OPEN Research Support

PhD student
Kim Michéle Feder
Physiotherapy, Lillebaelt Hospital - Vejle & UC SYD

Projekt styring
Projekt status    Open
Data indsamlingsdatoer
Start 01.03.2019  
Slut 26.02.2023  

Late-term shoulder pain and shoulder function after breast cancer surgery. Clinicoepidemiological and socioeconomic aspects of physiotherapeutic rehabilitation, and the effect of an individualized rehabilitation program.

Short summary

Breast cancer is associated with late-term upper limb impairments. The most frequent are lymphedema, sensory disturbances, pain and impaired shoulder function, with up to 70% reporting at least one symptom three to five years after surgery. To date a substantial knowledge gap exists in how to help these women.


Worldwide, breast cancer is the second most common cancer (1). The most frequently used surgical treatments are breast conserving surgery (BCS) and mastectomy (2). In parts due to early diagnosis and optimized treatment methods (3), 5-year survival has improved to a current 87% (4). But still late term upper limb impairments remain common (5,6,7). The most frequent are lymphedema, sensory disturbances, pain and impaired shoulder function (3,6-33), with up to 70% reporting at least one symptom three to five years after surgery (6,19-27,34). These impairments lead to difficulties in activities of normal daily living (28,35) and decreased quality of life (QoL) (36,37).

Physiotherapy research has focused on prevention and treatment of lymphedema (6,19-27,34). However, other upper limb impairments need focus as prevalences of up to 50% is reported for impaired shoulder function and pain 6 years after surgery (6,34,36,38,39,40), and a substantial knowledge gap exists in how to help these women.

Socioeconomic factors has been shown to influence patients´ participation in postoperative rehabilitation (41,42), but other potential risk factors (e.g. type of surgery and co-morbidity) combined with socioeconomic factors influence on physiotherapeutic rehabilitation and long-term shoulder pain and -function need further elucidation.

A report from the Danish Cancer Society in 2018 showed a decreasing proportion of breast cancer patients referred to physiotherapeutic rehabilitation since 2007 (44), but the reason and the distributions of referrals among breast cancer patients are unknown.

The results of this Ph.D project are expected to have an impact on the quality of physiotherapeutic rehabilitation offered to patients with breast cancer, by providing evidence-based knowledge to physiotherapists and patients about the importance of postoperative physiotherapeutic interventions, management of shoulder pain and impaired shoulder function, and the importance of potential risk factors and inequality in use of physiotherapeutic rehabilitation.

Aim in this PhD-study:

Study 1. To evaluate if socioeconomic factors, at the time of diagnosis for breast cancer, predicts initial referral and amount of physiotherapy received.

Study 2. To identify potential risk factors (e.g. type of surgery, oncological treatments, co-morbidities,) for shoulder impairments 3-5 years after breast cancer treatment.

Study 3. To compare the effect of a structured intervention consisting of an expert assessment followed by an individualized physiotherapeutic rehabilitation program, relative to a basic physiotherapeutic rehabilitation program delivered in a pamphlet, on changes in patient reported shoulder pain and function from baseline to 12 weeks after initiating the treatment among women with late-term shoulder impairments 3-5 years after primary breast cancer surgery.

Description of the cohort

Study 1

Design: A national register based cohort study following the STROBE statement (45).

Setting: Nation-wide register data from patients treated for breast cancer in Denmark in the years 2002-17, using primary data extraction from the Danish Health Data Board (DHDB).

Participants: The study population will be defined by Danish Breast Cancer Cooperative Group (DBCG), and crossed checked with the Patobank and the Danish National Patient Registry (LPR) to accommodate possible bias.

Data source and data variables: Data extraction performed per patient receiving a BCS or mastectomy +/- ALND since 2002 will be made from the DNPR. Referral to physiotherapeutic rehabilitation and number of sessions with a physiotherapist, education, marital status and level of income among other will be extracted and cross checked from the "Danish Act of Health §140" (GES) and Educational register (HFAUDD), Statistics Denmark (DST) and DBCG.

Study size: Annually 3.726 patients undergo surgery due to breast cancer (2). The sample size of the study is 63.943 patients.

Statistical method: Demographics will be reported as mean (SD), median (quartiles) and percentages. In study 1, breast cancer patients who underwent surgery and registered with a DC50 code in LPRR, will for example be divided into their level of education (short, medium and long), and differences in initial referral and amount of physiotherapeutic rehabilitation (referral to physiotherapy in LPR and registration pr. visit in GES) will be estimated.

Study 2 and 3

Design: Cross-sectional trial (study 2) with a subsequent prospective two parallel group randomized controlled and assessor blinded superiority trial (study 3), following the STROBE (45) (study 2) and the CONSORT guideline (46) (study 3).

Participants: Women who underwent BCS or mastectomy +/- ALND for breast cancer in 2016-19 in Denmark will be recruited through a letter inquiry including a questionnaire. The questionnaire will be used to specify the participants according to the potential risk factors (e.g. demographics, co-morbidity and socioeconomic status) and evaluate late term upper limb impairments; only focusing on shoulder pain and impaired shoulder function, 3-5 years after primary surgery (study 3). In study 3 women will be divided as case/non-case based on the following criteria: Self-reported shoulder pain, self-reported pain by flexion/rotation/adduction of the shoulder and a score of 20 or above on the Oxford Shoulder Score (OSS) (47). Women characterized as case will be offered to be part of study 3, after which they will be randomized to Intervention Group (IG) or Control Group (CG).

Exclusion criteria: Cancer relapse, disseminated cancer cells and cancer fixed to the thoracic wall, primary- or secondary reconstruction, bilateral surgery due to breast cancer, surgery in the affected shoulder prior to inclusion, previous shoulder or upper extremity fractures (left/right), currently receiving chemo, immune- or radiotherapy, co-morbidity expected to influence shoulder function (rheumatoid arthritis, previous stroke, multiple sclerosis), unable to understand or read Danish.

Inclusion procedures: Patients reporting shoulder discomfort in study 2, and providing informed and written consent will be referred to the primary investigator for final inclusion (Study 3).

Sample size calculation: In the period 2013-16, on average 1.177 breast cancer surgeries were performed per year in the included hospitals (2,48-50). A 40% participation rate is expected in study 2 (51), 1/3 of the patients are expected to be classified as cases (36,8,9), 1/3 is expected to decline participation in study 3, resulting in a population quantity at approximately 200 eligible women. Giving a power of at least 90%, two-sided significance level α=0.05, and baseline standard deviation (SD)=15.41 DASH units (52), the estimated total sample size is n=102 (~ n=51 in each group), to detect minimal clinical relevant difference, defined as 10 points on the DASH-score (53).

Randomization: After baseline assessment, the patients will be randomized in a ratio 1:1 to either IG or CG as per a computer-generated randomization schedule through OPEN (54) using permuted blocks of random sizes (two to six). A secretary, otherwise not affiliated with the project, will conduct the randomization procedure.

Interventions: Women randomized to CG will be handed a pamphlet with standard exercise recommendations for the shoulder (55,56). Patients randomized to IG will be referred to an examination in the shoulder sector at Lillebaelt Hospital - Vejle, Orthopedic Department to be diagnosed specifically for their individual shoulder impairments. The diagnosis for the IG group will be used to guide the individual physiotherapeutic rehabilitation program. The therapist in collaboration with the patient will decide the amount and type of intervention. Outcomes: An assessor, blinded to group allocation, will immediately before randomization and after intervention conduct baseline- and 12 weeks follow-up measurements.

Primary outcome measure: The patient reported "Disability of the Arm, Shoulder and Hand (DASH) questionnaire (0-100 (0 = no impairment)) (57), will be used as the primary outcome at baseline and 12 weeks follow-up. It consists of a primary dimension measuring function and pain, and two subscales measuring impairments in relation to work and sport/hobby (53).

Secondary outcome measure: Numeric pain intensity Rating Scale (NRS) (58); Global Rating of Change scale (GRC) (59,60), Hospital Anxiety and Depression score (HADS) (61), and adverse events/harms.

Demographic data: CPR-number; BMI; co-morbidities; type of surgery for breast cancer (BCS or mastectomy +/- ALND); amount of radiation and chemo treatment.

Statistical analysis: In study 3 a descriptive analysis will be used as mean (SD), median (quartiles) and percentages, to specify patients according to type of late term upper limb impairments and subgroups of type of surgery, radiotherapy etc. In study 3, primary efficacy analysis will be performed in the intention-to-treat population, with baseline observation carried forward used for missing data. Analysis of covariance will be used to compare mean changes from baseline in DASH scores, as well as the secondary continuous outcomes. Results are expressed as the difference between group means (95% CIs) with associated P-values.

Data and biological material

We hypothesize that in breast cancer patients:

1a) Highly educated women are referred and participate more physiotherapeutic rehabilitation compared to low educated women.

1b) Women with a high "family income" are referred and participate more physiotherapeutic rehabilitation compared to women with a low "family income".

1c) Married/"living with partner" women are referred and participate more physiotherapeutic rehabilitation compared to unmarried/"living alone" women.

2) High-risk breast cancer patients are more likely to develop late term upper limb impairments, such as shoulder pain and impaired shoulder function, compared to the rest breast cancer survivors.

3) An expert assessment followed by an individualized physiotherapeutic rehabilitation program is superior to the basic (control) intervention, measured on patient reported outcome measure "Disability of the Arm, Shoulder and Hand questionnaire" (DASH), in reducing shoulder pain and improve shoulder function in women with late-term shoulder impairments 3-5 years after primary breast cancer surgery.

Collaborating researchers and departments

The project is strategic organized and approved at Lillebaelt Hospital - Vejle, in collaboration between Department of Physio- and Occupational Therapy and Department of Surgery.