Physical function in patients newly diagnosed with multiple myeloma; a Danish cohort study

Prior to start of the treatment in an outpatient setting, all eligible patients were tested with the following physical function measurements: Six-Minute-Walk-Test (6MWT) [24] as a functional measure of aerobic capacity, Sit-to-Stand-Test (SST) [25] as a functional measure of lower body strength, grip strength [26, 27] as a measure of upper body strength and a direct measure of isometric knee extension strength [27, 28]. Prior to testing, the haematologist performed a systematic assessment of the impact of the radiologically assessed bone disease to determine restrictions regarding the physical tests (and exercise as well, to be used in the randomised controlled trial). In relation to testing, our focus was on the femoral bone. The assessment captured size of osteolytic lesions, fractures, and if applicable, estimated the time of fractures, and the haematologist assessed the degree of pain. Based on Mirel’s scoring system [29], this combined information of location, fractures/size of lesions and pain were used to assess whether the fractures and/or bone destructions should restrict certain tests. That was the case if an osteolytic lesion in the femoral bone involved between one third and up to two thirds of the diameter and caused pain, or if an osteolytic lesion involved more than two thirds of the diameter or involved the cortical bone (cortical thinning), even without associated pain. In these cases we only tested the unaffected side and omitted SST.

The physical function data (6MWT, SST, grip strength and knee extension strength) used in the current analysis are data from the baseline measures in the randomised controlled trial (ClinicalTrials.gov, ID NCT02439112), conducted by a project team of trained physiotherapists. Patient demographic and medical characteristics were collected from the patients’ medical records. The testing of 6MWT, SST and grip strength followed guidelines [25, 27, 30‐33], while the knee extension strength was measured by a standardised protocol developed for the randomised controlled trial. It was measured by a dynamometer (Lafayette Manual Muscle Tester), which was perpendicularly fixated to a bench by a strap. The participant was sitting on the bench with hip and knee flexion of 90° and arms resting on the side. Then the strap with the dynamometer was placed around the participant’s lower leg. The lower border of the dynamometer was placed five centimeters from the top of the lateral malleolus. The patient had three measures of right and left side, respectively regarding grip strength (until maximum value) and knee extension strength (each try was 5 s). The highest value was used for the analysis.

We included normative data of physical function outcomes from different healthy populations [34‐36] and published data from other cancer disease populations; malignant lymphoma before starting chemotherapy and without bone metastasis or elevated risk of fracture [37], prostate cancer after surgery or radiotherapy [38, 39] and breast cancer post-treatment [40, 41]. These cancers were chosen to compare MM data to other haematological cancers, both malignant lymphoma without bone destructions, and solid cancers where bone destructions are common. Comparison with other haematological cancers is relevant, because typically patients with MM represent a minor part of the included patients in exercise studies in haematological cancer populations, and therefore it is unknown if they differ and should be approached separately or with a special focus. The two non-haematological cancer diagnoses (prostate cancer and breast cancer) are the most common gender specific diagnoses and both cancer types share an issue of bone health, bone destructions and bone pain [42]. In the following our study population is called the EMMY population (Exercise in Multiple MYeloma).

Characteristics of the cohort are reported as counts and proportions and stratified by gender. The physical outcome measures 6MWT, SST, grip strength and knee extension strength are reported as mean and standard deviation (SD) and stratified by gender and age groups. Data are compared by z-test (after standardisation to mean = 0 and SD = 1) to reference values from normative populations and furthermore, to published data from patients with malignant lymphoma, prostate cancer and breast cancer, respectively. Moreover, we present outcome measures as box plots stratified by bone involvement and fractures and compare the standardised measurements by Wilcoxon rank sum test. Sample size calculation showed that by inclusion of 100 patients, differences of 0.33 SD with 90% power in the age- and gender-standardised outcome measures, compared to the reference populations, could be detected.

All mean scores, regardless of gender, were lower than for the normal population [34] and furthermore, all mean scores were below the lower SD-reference line for the normal population (Fig. 1a and b). The difference between EMMY and the reference population was statistically significant (p < 0.0001, z-score − 1.25). The 6MWT measurement was neither modified by the presence of vertebral fracture (p = 0.061), bone disease (p = 0.657) nor fracture (p = 0.758) (Fig. 3). Compared to lymphoma cancer (mixed genders) aged 55–59 years [37], the EMMY population had a shorter walking distance with a mean difference of 73 m and 171 m for males and females, respectively (Fig. 1a and b). Males with prostate cancer aged 70–74 years [38] achieved a longer walking distance than the EMMY population (Fig. 1a). Females with breast cancer aged 55–60 years [40] had a shorter walking distance than females from the EMMY population (Fig. 1b, Breast B), but younger females with breast cancer (approximately 47 years) [41] had almost the same walking distance as females from the EMMY population (Fig. 1b, Breast A).

Compared to the normal population [34], males between 60 and 80 years (Fig. 1c) and females between 60 and 75 years (Fig. 1d) had a lower number of mean raises. The total EMMY population (males and females) had statistically significantly lower mean raises than the reference group (p < 0.0001, z-score − 0.55), and number of mean raises was modified by the presence of bone involvement (p = 0.033) or fracture (p = 0.044), but not by vertebral fracture (p = 0.058) (Fig. 3). Comparing SST scores for males from the EMMY population to males with prostate cancer within the age group 65–70 years [39] or to females with breast cancer [40] the number of raises was almost identical.

Grip strength (mean (SD)) in the total group was statistically significantly higher than in the normal population [35] (Fig. 2a and b) (p < 0.00001, z-score 0.49) and modified by the presence of fracture (p = 0.025) or vertebral fracture (p = 0.005), but not bone involvement (p = 0.261) (Fig. 3). Compared to the population with lymphoma (mixed group of gender), the females from the EMMY population scored lower than the population with lymphoma cancer, while males had almost the same grip strength [37]. Though, this must be with reservations of comparing a mixed group of gender with females and males, respectively. For males with prostate cancer [38] there was no difference in mean grip strength compared to the EMMY population. Females with breast cancer [40] had a lower grip strength than the EMMY population.

Within the different age groups, the EMMY population (both genders) generally had lower strength compared to the normal population [36] (Fig. 2c and d). For the total group this difference was statistically significant (p = 0.0005, z-score − 0.39) and not modified by the presence of bone involvement (p = 0.246), fracture (p = 0.792) or vertebral fracture (p = 0.543) (Fig. 3). The lymphoma population [37] had much lower strength than the EMMY population. Females with breast cancer [41] and the patients from the EMMY population had almost the same strength in the age span 40–50 years.

The Danish test procedure for 6MWT (used in our study) [45] is in accordance with the American Thoracic Society test procedure [24], but Rikli et al. [34] deviated from that procedure regarding instruction to the patient. In the ATS test procedure patients are encouraged to walk as far as possible and are told that they will experience exertion [24], while Rikli et al. [34] told them to walk the best they could, but to avoid pushing themselves to overexertion or beyond what they thought would be safe for them.

Potentially, this could have the consequence that the reference values could be higher, if Rikli et al. [34] had followed the ATS procedure. Thus, the 6MWT difference between the patients with MM compared to reference values may be underestimated. Overall, the test position in the knee extension strength measure does not differ from the one used in the EMMY population. There is a difference regarding grip strength (using sitting or standing position) in the review [35], but the authors conclude that the different positions do not affect grip strength. We assume, that SST is very standardised, and thus does not differ between studies.

In the field of MM and physical function, the size of our cohort is quite large, and essential characteristics such as age, gender and bone disease are in accordance with the expected in the general MM population [3]. The R-ISS scores were according to the expected [46], although a bit higher proportion (30%) of our participants had R-ISS 3, which probably reflects that our cohort is population based and thereby included more patients with high Beta-2-microglobulin due to renal insufficiency than are observed in randomized clinical trials, because these patients do not fulfil inclusion/exclusion criteria [46]. Comorbidities were rare and mild, probably reflecting that included participants should be able to perform exercise training without being hampered by comorbid condition. Exclusion of patients with comorbidities may have been the physician’s decision as well as the patient’s choice of non-participation. We are not able to provide exact data on that. However, as a consequence, our data reflect the impact of MM and not comorbidities. Though, we have missing data, only a minor part was due to bone disease in the femoral bone (n = 3 for the SST). Thus, we believe our study and findings are representative for patients with MM in everyday clinical practice and thereby heighten the external validity of our study.

The associations between physical function and bone disease or fracture, indicate that these subgroups need special attention in a physical function perspective.

It is a strength that we have age specific data from normal samples, but regarding age-specific comparisons, when divided into age groups we are hampered by a rather small number of participants, especially in the lower and upper age groups.

There are some shortcomings in the comparisons, since we were unable to cover the total age span of the EMMY population in the comparisons with the normal population as well as comparisons with other cancer disease populations. We do not have data from citizens under the age of 60 years for 6MWT and SST, and under the age of 55 years for the knee extension strength. However, we assume that the association between age and physical performance will follow the same pattern for the younger age groups (< 60 years) [47], at least, according to the literature, for the walking distance [48, 49] and grip strength [47]. Furthermore, we did not have data on all the needed physical outcomes in the cancer disease populations. Finally, we need to address that the EMMY data are at time of diagnosis, which is different from the time points in the other cancer population studies, except for the lymphoma population. The differences in time points, and settings as well could influence the external validity.

Generally, our results indicate, that patients with MM have lower physical function at time of diagnosis and that this particularly is the case for patients with bone involvement. After start of anti-myeloma treatment, physical function may worsen, but we lack strong data on this. Bone studies in MM have shown that early bone fractures are common within the first weeks and observed in about 15% within 3 months [50]. This is assumed to cause deterioration of physical function. Patients undergoing HDT-SCT can experience loss in function during treatment. Potentially, such loss can be prevented or minimised by exercise [51] as shown in exercise studies among other cancer populations [19] and among elderly [20], although little is known about exercise interventions at time of diagnosis [52, 53] . Other treatments than HDT-SCT, typically offered to patients over the age of 65–70 years, are less intensive, but still may affect the physical function as well. Since the patients are older and may be frail [54] early detection of physical decline and subsequent early prevention by providing exercise interventions is of importance.

Our study accommodates the gap of knowledge of physical function in newly diagnosed patients with MM. Although our cohort is relatively large, further research is needed if we want to establish evidence of the physical function limitations. This could have implications for clinical practice, either by identifying patients at risk at group or individual level, and then establish an exercise regimen aiming at preventing physical decline and thereby importantly maintaining independence and quality of life.