Secondary Logo

Journal Logo


Interventions for the Treatment of Aromatase Inhibitor–Associated Arthralgia in Breast Cancer Survivors

A Systematic Review and Meta-analysis

Yang, Gee Su MSN, RN; Kim, Hee Jun PhD, RN; Griffith, Kathleen A. PhD, MPH, CRNP; Zhu, Shijun PhD; Dorsey, Susan G. PhD, RN, FAAN; Renn, Cynthia L. PhD, RN

Author Information
doi: 10.1097/NCC.0000000000000409
  • Free


Breast cancer is the most frequently diagnosed cancer (231 841 estimated cases in 2015) and ranks second as a cause of cancer death among women in the United States.1 Third-generation aromatase inhibitors (AIs) are widely used and are the most effective, standard adjuvant endocrine treatment for postmenopausal women with breast cancer.2,3 Aromatase inhibitors play a critical role in decreasing the estrogen that primarily stimulates breast cancer progression by 90% via covalent (eg, exemestane) or noncovalent (eg, letrozole, anastrozole) binding, thereby aiding in preventing breast cancer recurrence.4 Treatment with AIs improves disease-free survival, lowers rates of endometrial malignancy, and has a more favorable toxicity profile compared with tamoxifen.5,6 Since AI therapy has become more common, clinical experience has identified a unique musculoskeletal adverse effect of AI therapy.5 Nearly half of women who take AIs report AI-related joint pain or stiffness, and 20% to 30% discontinue the treatment because of joint pain.5,7–9 Patients report symmetrical joint pain or stiffness in hands, knees, feet, hip, and back, with a mild thickening of the soft tissues.5 Neuropathic, diffused, mixed, or persistent joint pain is the most common characteristic of this condition.8 Results from the IDEAL (Investigation on the Duration of Extended Adjuvant Letrozole) trial demonstrated that a quarter of the enrolled women discontinued initial AI therapy within 2 years specifically because of musculoskeletal symptoms; in addition, the overall noncompliance probability was 18.4% at 2.5 years.10

As arthralgias are quite common in breast cancer patients taking AIs, and such symptoms can continue years into survivorship, there is a growing need to develop and evaluate AI-associated arthralgia (AIA) interventional strategies. Still, few forms of pain management have been established in clinical settings because of the lack of a clear consensus on the underlying biological mechanisms of development of AIA. Although it has been proposed that estrogen deprivation may contribute to the development of AIA, this has not been shown in all studies on estrogen: whether the effect results from localized estrogen deprivation or a systemic problem is also not clear.11,12

If strong evidence on the effectiveness of specific interventions is provided, clinicians, policy makers, and patients can make the best decision based on the information to get the desirable treatment effect.13 Aromatase inhibitor–associated arthralgia is an extremely common problem among these women and negatively impacts their day-to-day well-being. Because AI therapy should be continued for a minimum of 5 years,14 AIA may be a persistent problem for women, resulting in lack of adherence to AI therapy. To our knowledge, no systematic review and meta-analysis has been conducted on all intervention types for AIA. Recently, Chien and colleagues15 studied the effect of acupuncture on AIA via a systematic review and meta-analysis and reported that acupuncture may alleviate AIA in breast cancer survivors. Bae et al16 also evaluated the effects of acupuncture on AIA in 4 randomized controlled trials (RCTs), and their systematic review suggested that acupuncture may potentially improve AIA. However, they focused only on acupuncture and did not assess other types of interventions for AIA. Therefore, the purpose of this study was to identify current various types of pain management for AIA and evaluate the study quality and effects of interventions on AIA in breast cancer survivors.


Search Strategy and Data Sources

A systematic review and meta-analysis were guided by the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA), which consists of a 27-item checklist and a 4-phase flow diagram.17 The PRISMA approach helps authors to formulate relevant and precise questions by using the acronym “PICOS,” which refers to 5 components: the information about population (P), the exposure or interventions (I), the comparator group intervention (C), the outcomes of intervention being assessed (O), and the type of study design (S).

The following databases were searched from 2000 to August 2015: PubMed, CINAHL (EBSCOHost), PsycINFO (OVID), and Web of Science (Web of Knowledge). We also identified eligible studies through a search of additional sources (eg, Google Scholar, ProQuest for dissertations and theses, and manual search in citations of original publications). Unpublished dissertation studies were included to avoid publication bias. The keywords “Intervention,” “Management,” “Treatment” were combined with “Arthralgia,” “Arthritis,” “Joint pain,” “Joint disease,” “Musculoskeletal disease,” “Musculoskeletal pain,” “Musculoskeletal symptom,” and “Aromatase Inhibitors,” including “Anastrozole (Arimidex),” “Letrozole (Femara),” and “Exemestane (Aromasin),” as well as “Breast cancer,” and “Breast neoplasm.” The search was limited to only the English language, publication dates from 2000 to August 2015, human subjects, and clinical experimental or quasi-experimental study design.

Each abstract and subsequent full-text article considered as qualitative synthesis were reviewed by 2 authors (G.S.Y. and H.J.K.) independently based on the PICOS criteria. At each step, 2 reviewers compared the results and discussed discrepancies to reach consensus. Eligible studies were sorted according to type of pain management, including pharmacological approach, acupuncture, nutritional supplementation, relaxation techniques, and physical exercise. Data of each study were assessed by 1 reviewer (G.S.Y.) with a self-developed structured codebook and confirmed by the other reviewer (H.J.K.) (Figure 1).

Figure 1
Figure 1:
Flow diagram of interventions for the treatment on aromatase inhibitor–associated arthralgia in breast cancer survivors.

Inclusion and Exclusion Criteria

The inclusion criteria and exclusion criteria were assessed according to the PICOS framework. The population of interest (P) was postmenopausal women with a history of stage I, II, or III breast cancer and current use of AIs. Also, women who had joint pain associated with AIs and reported persistent pain or stiffness in their joints after initiation of AI therapy were also appropriate for the inclusion criteria. Women with metastatic breast cancer (stage IV) were excluded because of possible visceral/bone pain. For the intervention of interest (I), any intervention defined as treatment or activity strategy to decrease AIA symptoms was included. Trials with designs that did not allow for an evaluation of the effectiveness of intervention were excluded. As for comparators (C), any comparator or none was included, for example, baseline, placebo, usual care, another active treatment, or no control treatment. In terms of the outcome of interest (O), joint pain assessed by validated pain measurement tools was included. Regarding study design (S), all RCTs and quasi-experimental designs (eg, 1-group pretest-posttest studies) were included. Prospective cohort studies and longitudinal studies were also included if the study was designed to compare a treatment group and control group, although no direct intervention was applied in patients. Case studies, case series, qualitative studies, systematic literature reviews, meta-analyses, and poster abstracts were excluded (Table 1).

Table 1
Table 1:
Study Eligibility Criteria From PICOSa Framework

Quality Assessment

Once all full-text articles were collected, each article was assessed to evaluate quality using the Quality Assessment Tool for Quantitative Studies (QATQS) with a standardized guideline and dictionary for reference.18 The QATQS tool is frequently used to assess studies by grading 8 components including selection bias, study design, confounders, blinding, data collection methods, withdrawals and dropouts, intervention integrity, and analysis. The first 6 components were rated as strong, moderate, or weak, whereas the last 2 components were not rated; after which overall study quality was determined. The 2 authors compared the 6 components and the overall quality of each article and then discussed discrepancies so that they could arrive at agreement (Table 2).

Table 2
Table 2:
Criteria for Assessing Study Quality

Data Analysis

We primarily used the pain mean score as assessed by Brief Pain Inventory (BPI) at the end timepoint of intervention because this questionnaire was the most frequently used instrument across the studies. We calculated pooled standardized mean differences (SMDs) and 95% confidence intervals (CIs) to determine the magnitude of effect sizes of interventions on pain. If standard deviations (SDs) were not provided, we obtained SDs from standard errors (SEs), CIs, and sample sizes. Publication bias was assessed with Funnel plot and Egger test. Heterogeneity across studies was identified by I2 statistic, which was regarded as high (I2 = 81.52%). Hence, the random-effects model was used to get pooled SMDs.19 Using the random-effects model, subgroup analyses were performed by testing pooled SMDs for a significant difference in pain between categorized groups, such as study type, study design, study quality, adherence to intervention protocols, and the number of participants. P < .05 was considered statistically significant (2-tailed). Comprehensive Meta-analysis version 3 (Englewood, New Jersey) was used to analyze the data.


The systematic search through PubMed, CINAHL, PsycINFO, and Web of Science yielded a total of 509 titles, and 24 were identified from additional sources (eg, Google Scholar, ProQuest for dissertations and theses, and manual search in citations of original publications). An abstract review reduced the potential number of studies to 58, and full-text review further reduced 8 RCTs and 11 cohort 1-group pretest-posttests, for a total of 1076 postmenopausal women with a history of breast cancer. Average age in the studies ranged from 56 to 71 years. Participants were recruited from a variety of settings, such as university hospitals, clinics, breast cancer centers, and community health clubs. In addition, AIA intervention studies were conducted in United States, Spain, France, Japan, United Kingdom, China, and Australia. The overall adherence to intervention protocols was high (85.4%), ranging from 66.7% to 100% (Table 3). Joint pain symptom outcomes of the trials were measured by validated self-reported questionnaires, including BPI, Western Ontario and McMaster Universities Osteoarthritis Index, visual analog scale, Fibromyalgia Impact Questionnaire, Modified Score for the Assessment and Quantification of Chronic Rheumatoid Affections of the Hands, and Health Assessment Questionnaire. The selected 19 studies were included for qualitative synthesis, among which 15 were available for meta-analysis. As for the quality Zassessment, they had overall moderate quality according to the QATQS tool, with a substantial agreement (κ = 0.640, P < .001). Studies were categorized into 5 intervention types: pharmacological approaches, (electro)acupuncture, nutritional supplementation, relaxation techniques, and physical exercise.

Table 3
Table 3:
Characteristics and Quality of the Included Studies

Characteristics and Quality of the Included Studies


Four cohort 1-group pretest-posttest studies examined the effects of pharmacological interventions in a total of 251 participants with AIA. Pharmacological approaches include taking duloxetine (Cymbalta) as a serotonin and norepinephrine reuptake inhibitor,20 prednisolone,21 zadaxin (Thymalfasin) as immunotherapy22 and switching between AI drugs.23 Those studies reported no severe adverse events and demonstrated overall improvement in joint pain and stiffness. In the duloxetine trial (30 mg daily for 1 week and then 60 mg daily for 7 weeks), 72.4% of participants experienced at least 30% reduction in the average pain level compared with baseline (P < .001), and approximately 80% completed the protocol.20 A short course of prednisolone (5 mg of oral prednisolone once a day in the morning for 1 week) provided immediate relief in joint pain in nearly 70% of participants, with 63% still reporting improvement at 1 month.21 Also, the participants who started letrozole after a 1-month washout period of anastrozole reported improvement in pain (P < .001) and physical and mental quality of life (P < .001 and P = .01, respectively), indicating that letrozole was better tolerated compared with anastrozole. However, 74% of participants still complain of arthralgia at 6 months of letrozole therapy.23 Lastly, zadaxin therapy (subcutaneous injection of thymosin α1 1.6 mg twice a week for 4 weeks) had evidence of improvement in pain severity (P = .014), pain-related functional interference (P = .001), and physical well-being (P = .001) compared with baseline.22

Studies using duloxetine, immunotherapy, and switching between AIs were evaluated as having moderate quality, whereas the prednisolone study was weak because of lack of blinding and no use of valid and reliable measurement tools.


Four RCTs and 2 cohort 1-group pretest-posttest studies investigated the effectiveness of traditional acupuncture and electro-acupuncture in a total of 226 women with AIA. Overall, the acupuncture interventions were well tolerated, and infection or development of lymphedema was not observed in any participants. However, all of the studies did not demonstrate effectiveness of acupuncture. In an RCT study by Crew et al,24 the true acupuncture (TA) group received full-body and auricular acupuncture as well as joint-specific acupuncture points (acupoints), whereas the sham acupuncture group had superficial needle insertion at nonacupoint locations. The TA group reported significant decrease in pain severity (P = .003), pain-related interference (P = .002), and stiffness (P = .01). Mao and colleagues25 also conducted an RCT of electrostimulation delivered by a transcutaneous electrical nerve stimulation unit. Pain severity and pain-related interference improved more in the TA group compared with the waiting list control group (P < .001 and P = .003, respectively). Prior to those RCT studies, these research groups had conducted pilot studies using the same protocols to test efficacy and safety and had found that pain was significantly decreased.6,26

On the contrary, there are another 2 studies reporting the opposite results. In Bao and colleagues’27 study, a significant difference was not observed in decrease in pain between experimental and sham groups (P = .31) when comparing effects of stimulating 15 real acupoints called major Qi with sham acupoints that are midpoints of the line between 2 real acupoints. Oh and colleagues28 applied real acupuncture by connecting needles through an electrode with bilateral rotation at various acupoints until de qi sensation (eg, tingling, numbness) occurred, whereas the sham group received acupuncture that did not penetrate the skin. There were no significant differences in joint pain, stiffness, and physical function between the sham and real acupuncture groups.

In acupuncture intervention, 4 studies were evaluated as having strong quality,24,25,27,28 whereas the others were assessed as having moderate26 and weak6 quality because of issues of participants’ representativeness and lack of blinding.


Nutritional supplementation included 3 studies evaluating the effects of taking a high dose of vitamin D, glucosamine plus chondroitin, and omega-3 fatty acids (O3-FAs) in 362 breast cancer survivors with AIA. Two studies were designed as RCTs,29,30 and 1 study was a cohort 1-group pretest-posttest design.31 Participants taking high-dose vitamin D2 (50 000 IU) in Rastelli and colleagues’30 study showed a greater reduction in average pain (P = .007), pain severity (P = .04), and pain interference (P = .034) compared with the placebo group at 2 months, although there were no differences in pain at 6 months. A daily dose of 1500 mg glucosamine plus 1200 mg chondroitin over 24 weeks also produced improvement in pain (P = .05) and stiffness (P = .03), as well as hand grip and pinch strength with minimal adverse effects at 12 weeks.31 In the RCT study by Hershman et al,29 the intervention group received daily 3.3 g O3-FAs (ie, eicosapentaenoic acid plus dehydroepiandrosterone in a 40:20 ratio), whereas the control group took the placebo, a blend of soybean and corn oil, for 24 weeks. Interestingly, both placebo treatment and O3-FAs improved joint pain in participants; therefore, 2 groups showed no difference (P = .52). In addition, no changes in cholesterol (P = .66), C-reactive protein (P = .71), and lipid profiles except for triglycerides were observed. In nutritional supplementation study, 2 studies had strong quality,29,30 whereas the other one was moderate.31


There were 2 studies examining the effectiveness of relaxation techniques, including yoga and tai chi, in 22 women taking AIs. The studies were designed as a 1-group cohort with pretest-posttest. A feasibility study was conducted to evaluate the impact of yoga on pain, functional outcomes, and health-related quality of life.32 Iyengar yoga, composed of precise postures (asanas), breathing exercise (pranayama), and meditation, was done for 2 hours per week over 8 weeks. The participants reported improvement in pain (P = .016), flexibility (functional reach: P = .048, sit and reach: P = .009), and physical function (P = .015) following yoga. In contrast, another study using tai chi was shown to be ineffective in decreasing pain (P = .058) and functional well-being (P = .223).9 This technique is focused on body awareness, deep breathing, and weight bearing to address the symptoms for 1 hour twice per week over 8 weeks. In this study, anxiety, depression, emotional well-being, fatigue, and sit-and-reach flexibility were significantly improved. Both studies were evaluated as weak quality because of selection bias and lack of blinding of outcome assessors and participants.


Four studies (n = 215) were conducted to examine the effectiveness of physical exercise in participants treated with AIs. Two studies were designed as 1-group pretest-posttest,33,34 and 2 were designed as RCTs.35,36 DeNysschen et al33 investigated an 8-week, home-based program that combined upper- and lower-body resistance exercise with self-selected aerobic exercise by assessing pain, functional performance, and cardiovascular endurance. Participants reported a significantly lower number of painful joints (P = .01) and an improvement in physical activity (P = .01). Significant improvements in grip strength, biceps curl, and sit-to-stand were also observed, whereas cardiovascular endurance showed no significant changes. In Fields’s35 study, participants carried out Nordic walking, which is a form of walking with gripping poles moved in an opposite manner to lower-limb motion, to evaluate the effectiveness on pain and psychosocial components over 12 weeks. The intervention group participated in supervised Nordic walking training followed by self-managed Nordic walking session, whereas the control group received usual care. The findings suggested that both groups experienced improvement on pain, depression, and self-efficacy; interestingly, the control group showed a marginally greater decrease in pain compared with the experimental group (P = .10). In Irwin and colleagues’36 study, participants were randomized into exercise (a combination of supervised resistance training program at a local health club and a home-based aerobic exercise program) or usual care group. At 12 months, the worst joint pain scores had decreased by 29% in the intervention group versus a 3% increase in the usual care group. Pain severity (P < .001) and pain interference (P < .001) were significantly lower in the experimental group. Lastly, Nyrop et al34 investigated a 6-week self-directed walking program in elderly breast cancer survivors taking AIs. The women completed walking at least 5 days a week for at least 30 accumulated minutes a day. As a result, a total time of walking per week increased over a 6-week period. Joint pain, stiffness, and fatigue also decreased by 10%, 32%, and 19%, respectively, although not significantly. In the physical exercise intervention, 3 studies had moderate quality,33,34,36 and 1 study was evaluated as having weak quality because of selection bias, lack of blinding, and significant confounders.35


Figure 2 and Table 4 show the results of the meta-analysis from 15 selected studies. Four of 19 studies were excluded in this analysis because of the lack of sufficient raw data from the original publication. Studies included were 3 with pharmacological approaches,20,22,23 4 using acupuncture,6,24–26 3 using nutritional supplementation,29–31 2 with relaxation techniques,9,32 and 3 with physical exercise interventions.33,35,36

Figure 2
Figure 2:
Forest plot for evaluating effects of interventions on pain in breast cancer survivors. Standardized mean differences and weights were obtained from a random-effects model. The square size is proportional to the weight of each study in this analysis. In the meta-analysis, there are 5 types of interventions: (1) pharmacological approaches,20,22,23 (2) acupuncture,6,24–26 (3) nutritional supplementation,29–31 (4) relaxation techniques,9,32 and (5) physical exercise.33,35,36
Table 4
Table 4:
Subgroup Analyses of the Selected Studies on Pain


The funnel plot for measurement of pain is presented in Figure 3. In the presence of publication bias, the bottom of the plot commonly shows a high concentration of studies on one side of the mean, indicating that smaller studies are more likely to be published if they can report a large effect size.37 Although the funnel plot looked asymmetrically distributed around the pooled SMD, it did not show a high concentration on one side of the bottom. Egger test for asymmetry was not significant (t = 1.765, P = .101).These results indicate that there was no publication bias for pain.

Figure 3
Figure 3:
Funnel plot for examining publication bias on pain across studies (n = 15). Evidence of negative publication bias is shown in the funnel plot (assessed with Egger test).


Statistical evidence for between-study heterogeneity in the effects of interventions on pain is shown (I2 = 81.52%, P < .001). Across the studies, the overall effect size of the interventions used for improving AIA was shown to be large (SMD, −0.744; 95% CI, −1.061 to −0.428). Seven of 15 studies demonstrated a significant effect in reducing AIA, among which the duloxetine therapy used in Henry and colleagues’20 study had the greatest effect size (SMD, −2.513; 95% CI, −3.243 to −1.783) (Figure 2).


Subgroup analyses were conducted on study type, study design, study quality, the number of participants, and adherence to intervention protocols, as shown in Table 4. There was a significant difference between subgroups of study type (P = .005). Pharmacological approaches (SMD, −1.186; 95% CI, −2.312 to −0.060) and acupuncture (SMD, −1.150; 95% CI, −1.729 to −0.571) showed a very large effect in improving pain, and relaxation technique (SMD, −0.667; 95% CI, −1.274 to −0.060) had a moderate effect on pain; nutritional supplementation (SMD, −0.124; 95% CI, −0.341 to 0.092) and physical exercise (SMD, −0.562; 95% CI, −1.499 to 0.375) showed no significant effect on pain, although they had tendency of decreasing joint pain. In addition, study design (1-group pretest-posttest vs RCT, P = .124), number of participants (<50 vs ≥50, P = .238), adherence to the intervention protocol (high adherence [80%–100%] vs moderate adherence [60%–79%], P = .103), and study quality (weak vs moderate vs strong, P = .111) had no differences between subgroups.


This systematic review was the first to evaluate the effects of current pain management on AIA in breast cancer survivors and to identify the quality and strength of the studies. We found that the overall effect size of AIA interventions for pain improvement across the studies was large. In particular, pharmacological approaches and acupuncture showed a large effect on pain, and relaxation techniques had a moderate effect, whereas effects of nutritional supplementation and physical exercise were shown to be insignificant. This may be in part due to inconsistency of intervention outcomes on joint pain in nutritional supplementation and physical exercise studies, although their study quality was evaluated moderate to high. However, we should be cautious when interpreting reported findings because 1-group pretest-posttest studies may inflate mean effect sizes in meta-analysis.38 Because interventions of musculoskeletal adverse symptoms have been studied only since the late 2000s, two-thirds of current studies were designed as a 1-group pretest-posttest with a small sample size to evaluate feasibility and safety, indicating that there may be potential risk of biases in findings. In particular, pharmacological approaches and relaxation techniques showed a wide range of 95% CIs of SMDs, although benefits of these interventions seem to be significant. Evaluation of each intervention effect might be changed if additional RCTs are conducted. Further investigation into the influence of nutrition and physical exercise is needed to better discern their potential for pain management.

In this review, more than half of all studies described a procedure for an appropriate power analysis of sample size adequacy to detect 80% to 90% power.6,24–26,29,30,32,36 These studies calculated the sample size based on their pilot study findings or the assumption that the SD is equal to or greater than half of the mean and considering a dropout rate. According to Turner et al,39 when there are at least 2 adequately powered studies included in the meta-analysis, the underpowered studies will contribute a trivial amount to the overall power. Therefore, our results from the meta-analysis may provide sufficient power to detect the large effects of intervention on pain.

As for the effect of acupuncture, the result from this review was different from that of the previous meta-analysis of acupuncture in the treatment of AIA.15 According to that study, there was a trend of decreased pain in participants following acupuncture; however, the effect was not significant. In contrast, we concluded that acupuncture significantly reduces pain. In Chien and colleagues’15 analysis, only 2 RCTs24,25 were selected in meta-analysis, whereas we included 4 acupuncture studies regardless of study design, which might lead to results that are more statistically significant in terms of the intervention effect. Despite the positive impact of acupuncture on pain, its effect may be biased in out meta-analysis because the other 2 studies showing no observed change in pain27,28 required exclusion because reported findings included median values or percentages only and were therefore incompatible with methods used in other studies for the meta-analysis.

Although most studies offered promising modalities because of high efficacy and safety in treating AIA, several limitations remain to be addressed in this analysis. With the exception of acupuncture, each intervention type was discussed in only 2 or 3 studies at most, which may be too small to determine their precise treatment effects on joint pain. To evaluate the individual intervention type appropriately, investigators should consider conducting more studies on pain management strategies that are underexamined or have not yet been examined but are frequently used by AIA patients in daily life. For example, pharmacological approaches may include acetaminophen, cyclooxygenase 2 inhibitors, or nonsteroidal anti-inflammatory drugs, for which causal mechanisms have not been identified.40,41

Aromatase inhibitor–associated arthralgia is a complex symptom affected by the interrelation among biological, psychological, and sociocultural aspects of patients; however, we were not able to include secondary outcomes (eg, depression, sleep disturbance, functional well-being, menopausal symptoms, etc) in the meta-analysis. A small number of studies evaluated these aspects, and those secondary outcomes were assessed using a variety of measurement tools. Quantitatively analyzing these important secondary outcomes regarding AIA intervention effects would help researchers better understand the symptom if additional studies are conducted on secondary outcomes.

Lastly, 4 of the studies that used sham/placebo treatment failed to find significant differences between experimental and control groups.27–29,35 The sham or placebo treatment may have had potential effects in the control group. Sham acupuncture may trigger physiological effects or closely correlate with real acupoints that are diffused in the body.27 Also, the placebo drug of O3-FA may have caused the natural improvement over time, contamination in the placebo group, or the possibility that the ingredients (soy/corn oil) were active in decreasing AIA.29,42 These unexpected outcomes might have had negative impacts on evaluation of study quality and the meta-analysis. To support interpretation of placebo data, objective evidence, such as musculoskeletal (eg, hand/knee joints and tendons) sonography, may be considered in addition to self-reported questionnaires.

Implications for Practice

Oncology nurses are in an excellent position to improve AIA by directly providing patients with high-quality care. With optimal intensity of acupuncture, physical exercise or relaxation techniques, and precise dosage of medication or nutritional supplementation, nurses may be able to implement such validated interventions as pain management modalities to mitigate the symptoms. Especially, some interventions, including acupuncture and relaxation techniques, can be safely and rapidly applied to women in clinical settings as potential therapies to improve AIA symptoms according to findings from this review. Enabling oncology nurses to offer safe therapeutic pain interventions can reduce burden for the patients with AIA.

In addition, nurse researchers may consider conducting larger trials to evaluate the therapeutic efficacy of AIA interventions, especially pharmacological approaches and relaxation techniques that have been proven effective in only studies with a small sample size, because large RCT studies may strengthen these bodies of evidence. The use of cost-benefit analysis may potentially play an important role in improving decision making on which intervention types would be appropriate for further RCTs or developing new interventions by identifying their benefits and deficits.43


This review suggests the overall effect size of current interventions on AIA is large, and the evidence is based on a body of research with moderate study quality. Pharmacological approaches, acupuncture, and relaxation techniques demonstrated a significant effect on pain. Based on our findings, nutritional supplementation and physical exercise do not seem to contribute to control of joint pain. Additional studies with more methodological rigor and diversity may be required to fully evaluate precise effects of these interventional approaches.

Suboptimal adherence to AI drugs may lead to unsatisfactory therapeutic efficacy, ultimately reducing rates of disease-free survival in postmenopausal breast cancer patients. Therefore, it is important for healthcare providers to implement effective strategies for patients so that they remain compliant with the AIA therapy, which has considerable clinical benefit. Promising interventions to reduce AIA, including acupuncture, relaxation, and pharmacotherapy, are indicated for use based on findings reported here.


The authors thank Drs Sue A. Thomas and Jeanne M. Geiger-Brown for teaching PRISMA reporting guideline. They also thank Ms Jane Sellman for proofreading.


1. American Cancer Society. Cancer Facts & Figures 2015. Accessed June 1, 2015.
2. Cuzick J, Sestak I, Baum M, et al. Effect of anastrozole and tamoxifen as adjuvant treatment for early-stage breast cancer: 10-year analysis of the ATAC trial. Lancet Oncol. 2010;11(12):1135–1141.
3. Burstein HJ, Prestrud AA, Seidenfeld J, et al. American Society of Clinical Oncology clinical practice guideline: update on adjuvant endocrine therapy for women with hormone receptor-positive breast cancer. J Clin Oncol. 2010;28(23):3784–3796.
4. Nabholtz JM. Long-term safety of aromatase inhibitors in the treatment of breast cancer. Ther Clin Risk Manag. 2008;4(1):189–204.
5. Burstein HJ, Winer EP. Aromatase inhibitors and arthralgias: a new frontier in symptom management for breast cancer survivors. J Clin Oncol. 2007;25(25):3797–3799.
6. Crew KD, Capodice JL, Greenlee H, et al. Pilot study of acupuncture for the treatment of joint symptoms related to adjuvant aromatase inhibitor therapy in postmenopausal breast cancer patients. J Cancer Surviv. 2007;1(4):283–291.
7. Henry NL, Giles JT, Ang D, et al. Prospective characterization of musculoskeletal symptoms in early stage breast cancer patients treated with aromatase inhibitors. Breast Cancer Res Treat. 2008;111(2):365–372.
8. Laroche F, Coste J, Medkour T, et al. Classification of and risk factors for estrogen deprivation pain syndromes related to aromatase Inhibitor treatments in women with breast cancer: a prospective multicenter cohort study. J Pain. 2014;15(3):293–303.
9. Galantino ML, Callens ML, Cardena GJ, Piela NL, Mao JJ. Tai chi for well-being of breast cancer survivors with aromatase inhibitor–associated arthralgias: a feasibility study. Altern Ther Health Med. 2013;19(6):38–44.
10. Fontein DBY, Nortier JWR, Liefers GJ, et al. High non-compliance in the use of letrozole after 2.5 years of extended adjuvant endocrine therapy. Results from the IDEAL randomized trial. Eur J Surg Oncol. 2012;38(2):110–117.
11. Gaillard S, Stearns V. Aromatase inhibitor–associated bone and musculoskeletal effects: new evidence defining etiology and strategies for management. Breast Cancer Res. 2011;13(2):1–11.
12. Nevitt MC, Felson DT, Williams EN, Grady D. The effect of estrogen plus progestin on knee symptoms and related disability in postmenopausal women: the Heart and Estrogen/Progestin Replacement Study, a randomized, double-blind, placebo-controlled trial. Arthritis Rheum. 2001;44(4):811–818.
13. Atkins D, Best D, Briss PA, et al. Grading quality of evidence and strength of recommendations. Br Med J. 2004;328(7454):1490.
14. Burstein HJ, Griggs JJ, Prestrud AA, Temin S. American Society of Clinical Oncology clinical practice guideline: update on adjuvant endocrine therapy for women with hormone receptor-positive breast cancer. J Clin Oncol. 2010;6(5):243–246.
15. Chien T-J, Liu C-Y, Chang Y-F, Fang C-J, Hsu C-H. Acupuncture for treating aromatase inhibitor–related arthralgia in breast cancer: a systematic review and meta-analysis. J Altern Complem Med. 2015;21(5):251–260.
16. Bae K, Yoo HS, Lamoury G, Boyle F, Rosenthal DS, Oh B. Acupuncture for aromatase inhibitor–induced arthralgia: a systematic review. Integr Cancer Ther. Nov 2015;14(6):496–502.
17. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol. 2009;62(10):e1–e34.
18. Thomas BH, Ciliska D, Dobbins M, Micucci S. A process for systematically reviewing the literature: providing the research evidence for public health nursing interventions. Worldviews Evid Based Nurs. 2004;1(3):176–184.
19. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Introduction to Meta-analysis. Chichester, UK: John Wiley & Sons, Ltd; 2009.
20. Henry NL, Banerjee M, Wicha M, et al. Pilot study of duloxetine for treatment of aromatase inhibitor–associated musculoskeletal symptoms. Cancer. 2011;117(24):5469–5475.
21. Kubo M, Onishi H, Kuroki S, et al. Short-term and low-dose prednisolone administration reduces aromatase inhibitor–induced arthralgia in patients with breast cancer. Anticancer Res. 2012;32(6):2331–2336.
22. Zhang Q, Tang D, Zhao H. Immunological therapies can relieve aromatase inhibitor–related joint symptoms in breast cancer survivors. Am J Clin Oncol. 2010;33(6):557–560.
23. Briot K, Tubiana-Hulin M, Bastit L, Kloos I, Roux C. Effect of a switch of aromatase inhibitors on musculoskeletal symptoms in postmenopausal women with hormone-receptor–positive breast cancer: the ATOLL (articular tolerance of letrozole) study. Breast Cancer Res Treat. 2010;120(1):127–134.
24. Crew KD, Capodice JL, Greenlee H, et al. Randomized, blinded, sham-controlled trial of acupuncture for the management of aromatase inhibitor–associated joint symptoms in women with early-stage breast cancer. J Clin Oncol. 2010;28(7):1154–1160.
25. Mao JJ, Xie SX, Farrar JT, et al. A randomised trial of electro-acupuncture for arthralgia related to aromatase inhibitor use. Eur J Cancer. 2014;50(2):267–276.
26. Mao JJ, Bruner DW, Stricker C, et al. Feasibility trial of electroacupuncture for aromatase inhibitor–related arthralgia in breast cancer survivors. Integr Cancer Ther. 2009;8(2):123–129.
27. Bao T, Cai L, Giles JT, et al. A dual-center randomized controlled double blind trial assessing the effect of acupuncture in reducing musculoskeletal symptoms in breast cancer patients taking aromatase inhibitors. Breast Cancer Res Treat. 2013;138(1):167–174.
28. Oh B, Kimble B, Costa DSJ, et al. Acupuncture for treatment of arthralgia secondary to aromatase inhibitor therapy in women with early breast cancer: pilot study. Acupunct Med. 2013;31(3):264–271.
29. Hershman DL, Unger JM, Crew KD, et al. Randomized multicenter placebo-controlled trial of omega-3 fatty acids for the control of aromatase inhibitor–induced musculoskeletal pain: SWOG S0927. J Clin Oncol. 2015;33(17):1910–1917.
30. Rastelli AL, Taylor ME, Gao F, et al. Vitamin D and aromatase inhibitor–induced musculoskeletal symptoms (AIMSS): a phase II, double-blind, placebo-controlled, randomized trial. Breast Cancer Res Treat. 2011;129(1):107–116.
31. Greenlee H, Crew KD, Shao T, et al. Phase II study of glucosamine with chondroitin on aromatase inhibitor–associated joint symptoms in women with breast cancer. Support Care Cancer. 2013;21(4):1077–1087.
32. Galantino ML, Desai K, Greene L, DeMichele A, Stricker CT, Mao JJ. Impact of yoga on functional outcomes in breast cancer survivors with aromatase inhibitor–associated arthralgias. Integr Cancer Ther. 2012;11(4):313–320.
33. DeNysschen CA, Burton H, Ademuyiwa F, Levine E, Tetewsky S, O’Connor T. Exercise intervention in breast cancer patients with aromatase inhibitor–associated arthralgia: a pilot study. Eur J Cancer Care. 2014;23(4):493–501.
34. Nyrop KA, Muss HB, Hackney B, Cleveland R, Altpeter M, Callahan LF. Feasibility and promise of a 6-week program to encourage physical activity and reduce joint symptoms among elderly breast cancer survivors on aromatase inhibitor therapy. J Geriatr Oncol. 2014;5(2):148–155.
35. Fields J. A Feasibility Study of a Nordic Walking Intervention for Women Experiencing Aromatase Inhibitor Associated Arthralgia [dissertation]. Southampton, UK: University of Southampton; 2015.
36. Irwin ML, Cartmel B, Gross C, et al. Randomized exercise trial of aromatase inhibitor–induced arthralgia in breast cancer survivors. J Clin Oncol. 2015;33(10):1104–1111.
37. Rothstein HR, Sutton AJ, Borenstein M. Publication bias in meta-analysis: prevention, assessment and adjustments. Chichester, UK: John Wiley & Sons, Ltd; 2005.
38. Carlson KD, Schmidt FL. Impact of experimental design on effect size: findings from the research literature on training. J Appl Psychol. 1999;84(6):851–862.
39. Turner RM, Bird SM, Higgins JPT. The impact of study size on meta-analyses: examination of underpowered studies in Cochrane reviews. PLoS One. 2013;8(3):e59202.
40. Lintermans A, Neven P. Pharmacology of arthralgia with estrogen deprivation. Steroids. 2011;76(8):781–785.
41. Menas P, Merkel D, Hui W, Lawton J, Harper A, Carro G. Incidence and management of arthralgias in breast cancer patients treated with aromatase inhibitors in an outpatient oncology clinic. J Oncol Pharm Pract. 2012;18(4):387–393.
42. Henry NL, Griggs JJ. The power of the placebo in symptom management. J Clin Oncol. 2015;33(17):1870–1872.
43. Arrow KJ, Cropper ML, Eads GC, et al. Is there a role for benefit-cost analysis in environmental, health, and safety regulation? Science. 1996;272(5259):221–222.

Aromatase inhibitor; Arthralgia; Breast cancer; Intervention; Meta-analysis; Systematic review

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.