Optimizing hand-function patient outcome measures for inclusion body myositis

Abstract

Inclusion body myositis is the most commonly acquired myopathy after the age of 45. The slowly progressive and heterogeneous disorder is a challenge for measuring clinical trial efficacy. One current method for measuring progression utilizes the Inclusion Body Myositis-Functional Rating Scale. We have found that the upper extremity domain scores in the Inclusion Body Myositis-Functional Rating Scale do not consistently change until there is extreme loss of grip and finger flexor strength. Therefore, we performed a cross-sectional observational study of 83 inclusion body myositis patients and 38 controls recruited at the 2019 Annual Patient Conference of The Myositis Association. We evaluated new Inclusion Body Myositis Patient-Reported Outcome measures for upper extremity function modified from the NIH Patient-Reported Outcomes Measurement Information System as well as pinch and grip strength. We found that Patient-Reported Outcome measures hand-function have a higher correlation with pinch and grip strength than the Inclusion Body Myositis-Functional Rating Scale.

1. Introduction

Sporadic inclusion body myositis (IBM) is the most commonly acquired myopathy after the age of 45. The clinical presentation of IBM is quite distinctive, with a classic pattern of focal weakness of the deep finger flexors and quadriceps. There is selective weakness of the flexor digitorum profundi and flexor pollicis longus occurs with sparing of long finger extensors and intrinsic hand muscles. The Inclusion Body Myositis Functional Rating Scale (IBM-FRS) has been used as an outcome measure to track disease progression and response to treatment in clinical trials [1–3]. In our prior paper, we found that the IBM-FRS upper extremity function items had lower than expected correlation with finger flexor strength and even less with grip strength [4]. We expand on our prior findings to examine pinch strength (as a representative to finger flexor strength) in addition to grip strength.

Hand grip involves multiple forearm and intrinsic hand muscles, including the flexor digitorum superficialis, abductor pollicis brevis, flexor pollicis, adductor pollicis, flexor digiti minimi brevis, and the lumbricals, with less involvement of the opponens pollicis and opponens digiti minimi. As a particular hallmark of IBM compared to other neuromuscular disorders, IBM classically targets flexor digitorum profundi leading to deep finger flexor weakness with relative sparing of the hand intrinsic muscles [5–8]. Longitudinal measurement of pinch strength in IBM has only recently been characterized, and a subgroup of rapid deteriorators have shown to have a detectable yearly decline [9]. We specifically characterize pinch strength via dynamometer performed by the flexor pollicis longus and flexor digitorum profundus II of the dominant hand with the hope of finding another biomarker of IBM severity.

The hand function domain of the IBM-FRS queries handwriting, the ability to cut food and handle utensils, and fine motor tasks such as opening doors, using keys, and picking up small objects [1]. Patients are asked to choose the best description that fits their current ability in handwriting, with a subsequent score ranging from 0 to 4 based on degree of function (0 = unable to grip pen, 1 = able to grip pen but unable to write, 2 = not all words are legible, 3 = slow or sloppy; all words are legible, 4 = normal function). In our prior study, patients reported more functional deficit (a lower score) only when they had minimal or no finger flexor strength [4].

The smoldering progression of IBM raises the importance of appropriate outcome measures capable of detecting changes within a set clinical trial timeline. In particular, there is a need to find optimal biomarkers that have the sensitivity to detect small clinical changes in IBM over a 1–2-year period. Our prior study showed a less than ideal correlation between the IBM-FRS upper extremity items and degree of grip and finger flexor strength [4]. Often there is no change in the IBM-FRS score until there is significant loss of strength. A common critique of the IBM-FRS is its absence of patient-reported outcome contribution. Patient-reported outcome (PRO) measures are accepted by the Federal Drug Administration (FDA) as an important outcome measure taking into account the patient’s experience with the disease [10,11]. Our study evaluates this hypothesis by adapting the upper extremity function items from the Patient-Reported Outcomes Measurement Information System (PROMIS) program of the National Institute of Health (Table 1). We assessed whether these PROs have improved correlation with pinch and grip strength as well as duration of disease compared to the current upper extremity function items in IBM-FRS.

Table 1.

PRO Upper Extremity Function scale for IBM.

1. Are you able to turn a key in a lock?

2. Are you able to brush your teeth?

3. Are you able to pick up coins from a table top?

4. Are you able to write with a pen or pencil?

5. Are you able to open or close a zipper?

6. Are you able to open previously opened jars?

7. Are you able to hold a plate full of food?

8. Are you able to button your shirt?

9. Are you able to trim your fingernails?

10. Are you able to cut your toenails?

11. How much difficulty do you currently have removing wrappings from small objects?

12. How much difficulty do you currently have opening medications or vitamin containers (e.g., childproof containers, small bottles)?

We present a cross-sectional study to address the following: (1) The distribution of pinch and grip strength and their association to upper extremity function items of PROs and FRS in a population with IBM; and (2) how these variables (pinch strength, grip strength, PROs and FRS) change with duration of symptoms.

2. Patients and methods

We performed a cross-sectional observational study of 83 IBM patients recruited at the 2019 Annual Patient Conference of The Myositis Association. Diagnosis of IBM was based on patient self-report. Concurrently, 38 control patients, defined as patients with no significant past medical history nor any history of neuromuscular disease, were recruited.

This study was approved by the Washington University institutional review board.

Grip strength of the dominant hand was measured three times consecutively by a Jamar hydraulic hand dynamometer. We limited to three trials to prevent fatigue of the hand confounding our data. The patients were seated with elbow at 90° of flexion with forearm and wrist in a neutral position. They were instructed to squeeze the dynamometer as hard as possible for at least three seconds. The average grip dynamometer reading was normalized by age and sex according to data collected from the NIH Toolbox project [12].

Quantitative tip pinch strength was measured with CITEC CT 3002/30 handheld dynamometer (Haren, The Netherlands) with the pinch grip applicator [13,14]. Pinch was measured with the elbow at 90° of flexion and forearm in neutral position. The patients placed their thumb on one side of the pinch pad and their index finger on the other. It was emphasized to apply pressure through the distal phalanx of both fingers as hard as possible for at least three seconds. At least two measurements were made to ensure consistency. Additionally, manual assessment of finger flexor strength was performed by a board-eligible or board-certified neuromuscular physician and described as complete inability to flex fingers (graded 0/5), severe weakness (no resistance, graded 1/5), moderate weakness (very little resistance, graded 2/5), mild weakness (breakable resistance, graded 3/5), subtle weakness (resistance that is near unbreakable, graded 4/5), and full strength (graded 5/5).

Patients with IBM were asked to fill out both the IBM-FRS and upper extremity function items of PROs from the NIH PROMIS. Each patient answered the ten IBM-FRS questions qualitatively and answers were subsequently converted to a 0–4 point scale. Patients were asked to rate each of the adapted twelve hand function PROs (see Table 1) on a similar 0–4 point scale: 0 (unable to do), 1 (with much difficulty), 2 (with some difficulty), 3 (with little difficulty), and 4 (without any difficulty).

Demographic and clinical variables were summarized as number with percent for categorical variables and as median with interquartile range for continuous variables. Scatterplots and histograms were used to visually assess the distribution of grip and pinch strength between patients with IBM and control patients. Box plots were created to examine the relationship between disease duration and strength. Heat maps were created to assess the magnitude of correlation between variables of IBM severity and individual IBM-FRS and PRO upper extremity function items. Spearman correlations were calculated to assess the magnitude of correlation between measures of strength, IBM-FRS and PRO upper extremity function items, and disease duration. Correlation tests for two independent correlations were used to assess for differences in correlations between patients with IBM and control patients, and correlation tests for two dependent correlations were used to assess for differences in correlations within IBM patients. A two-sided test p-value less than 0.05 was considered significant, and all analyses were completed using R version 3.5.2.

3. Results

3.1. Demographics

We compared 83 IBM patients to 38 controls. The IBM patients had a male predominance (57%), while the control subjects had female predominance (75%) as befitting a disease that has a male predominance with the controls most likely caregivers or relatives of the patients. The IBM cohort had a median age of 68 with interquartile range [IQR] spanning 62–72.5 years, while the control cohort had a median [IQR] age of 64.5 [58.5–71.5] years. Our IBM cohort reported a duration of symptoms ranging between 1 and 41 years. Almost all patients self-identified as non-Hispanic white.

3.2. Both grip and pinch strength differentiate IBM and control patients

Normalized to age- and sex-norms, the median [IQR] grip strength was 34% [20–48] in the IBM cohort versus 117% [104–134] in the control cohort (Fig 1B). These findings are similar to our 2018 cohort [4], where patients with IBM (squares) and control patients (circles) had different strengths (Fig 1A and B).

Fig. 1.

(A) Grip strength (non-normalized) averaged over three trials of control patients (blue circles) and patients with IBM (red squares). (B) Histogram of frequency by decile bins of grip strength normalized to age and sex of control patients (blue) and patients with IBM (red). (C) Mean pinch strength by CITEC dynamometer with the pinch grip applicator differentiated by age. As there is no normalized data, (D) presents histogram of frequency by one pound of pinch strength.

Pinch strength by CITEC dynamometer attached to pinch grip applicator also differentiated the patients with IBM from the control patients. The median [IQR] pinch strength was 1.7 [1.1–3.0] lbs for the IBM cohort versus 5.7 [4.6–6.8] lbs for the control cohort (Fig 1C and D).

3.3. Pinch strength by dynamometer is correlated with manual measurements of finger flexor strength, but is an independent measure to grip strength

Pinch strength by CITEC dynamometer correlated with manually measured qualitative finger flexor strength graded 0–5 (r = 0.79, Fig 2A) indicating that the pinch dynamometer is representative of our neuromuscular examination. Overall, it appeared that pinch strength was associated with normalized grip strength . However, when we analyzed the IBM and control patients separately, pinch and grip strength had a lower degree of association for control versus IBM cohort, though this difference was not statistically significant (Control: r = 0.47, IBM: r = 0.61, p = 0.33; Fig 2B). After normalizing grip strength by age and sex, a significant difference was found between the correlation of pinch and grip strength in control patients versus IBM patients (Control: r = −0.05, IBM: r = 0.56, p < 0.001, Fig 2C) [4,12].

Fig. 2.

(A) Comparison of pinch strength (by CITEC dynamometer) versus finger flexor strength measured semi-quantitatively (r = 0.79) and grip strength measured by dynamometer (r = 0.77). Patients with IBM in squares and controls in circles. (B) Comparison of pinch versus hand grip strength among patients with IBM (square) and control (circle). Dotted lines denote median pinch and median hand grip strength of IBM cohort. Solid lines and shaded gray areas correspond to localized smoothed means and 95% confidence intervals of smoothed means, respectively. (C) Normalized grip strength can be calculated given published baseline for age and sex, but this is not available for pinch. Normalization more clearly shows that pinch and grip are independent measures.

We further divided our IBM cohort (Fig 2B squares) into four quadrants based on the cohort’s median pinch and hand grip strength. Patients with IBM in the upper right quadrant are patients with grip and pinch strength greater than median of the cohort—and have more preserved hand function. These patients have an increase in variability when compared to patients in the lower left quadrant, suggesting that pinch and grip are independent measures for our patients with IBM, especially for those with more preserved hand function; this is also seen when grip strength is normalized to age and sex (Fig 2C) [12].

These findings further suggest that pinch utilizes a subset of muscles involved in grip (flexor pollicis longus and flexor digitorum profundus II out of the multiple forearm and intrinsic hand muscles).

3.4. Cross-sectional analysis of pinch and hand grip strength revealed disease progression early in the course

There appears to be a ceiling effect of disease on both maximal weakness of pinch and handgrip strength in our patients with IBM, particularly when disease duration is greater than 15 years (Fig S1). The ranges of pinch and grip strength show a similar pattern over time—more variable at earlier stages of the disease, with a precipitous drop in mean strength in the 6–10-year disease duration range with subsequent slow progression to a more similar level of weakness over the course of about 20 years.

3.5. PRO upper extremity function items correlated better with pinch and grip strength than the FRS upper extremity function items

Similar to our prior findings [4], upper extremity function items in IBM-FRS only moderately correlated with pinch and grip strength (r < 0.45, Fig S2). In contrast, the IBM-PRO Upper Extremity Function Scale items correlated better with pinch and grip strength (Fig S3), especially with regards to ability to trim fingernails (r = 0.58 vs. pinch strength, 0.50 vs grip strength), open/close a zipper (r = 0.57 vs. pinch strength, 0.54 vs grip strength), button a shirt (r = 0.55 vs. pinch strength, 0.48 vs grip strength), and turn a key (r = 0.51 vs. pinch strength, 0.53 vs grip strength). The sum of IBM-PRO Upper Extremity Function Scale also had higher correlation with grip strength (r = 0.62 vs. 0.34, p < 0.01, Fig 5B and E) and pinch strength (r = 0.66 vs. 0.39, p < 0.01, Fig 5C and F).

Fig. 5.

IBM-FRS hand item sum (A) and IBM Upper Extremity Function Scale sum (D) plotted against patient-reported disease duration. No clear decline with time is seen, which is consistent with the heterogeneity of the patients in a cross-sectional study. However, the IBM-PRO Upper Extremity Function Scale have a higher correlation with grip (E) and pinch (F) strength than FRS hand item sum (B and C, respectively).

IBM-FRS upper extremity items have insufficient capability to detect graded strength loss in our prior report [4]. We found a steep drop off in measured hand strength before patients change the rating from “no difficulty” to “little, some or much difficulty,” reducing a 5-point scale to more of a binary function. In this study, the IBM-PRO Upper Extremity Function Scale demonstrates better ability to detect changes in hand strength. This allows it to function more reliably as a 5-point scale, which also explains its superior correlation to measured grip and pinch strength. (Figs. 3 and 4).

Fig. 3.

Box plot of grip strength and patient’s rating of individual 12-item IBM-PRO Upper Extremity Function Scale.

Fig. 4.

Box plot of pinch strength and patient’s rating of individual 12- item IBM-PRO Upper Extremity Function Scale.

The adapted IBM-PRO Upper Extremity Function Scale sum does not follow an appreciable trend with disease duration (Fig 5D). Neither does IBM-FRS upper extremity sum (Fig 5A), which was reported in our 2018 cohort as well [4].

4. Discussion

Our study furthers our first cross-sectional study of hand/upper extremity function in patients with IBM [4]. That paper described grip strength in a cross-sectional cohort of 81 patients with IBM and found that the IBM-FRS does not follow disease duration or correlate well with grip strength. This current study further expands on how to best quantify hand function and investigated the usefulness of patient-reported outcome measures.

We characterize pinch strength in our cohort using a similar dynamometer that was used in the longitudinal study by Oldroyd et al. [9]. In our cross-sectional study, distribution of both grip and pinch strength differs between our IBM and control cohorts (Fig 1). We were able to show that the CITEC pinch measure is appropriately correlated with our manual finger flexor strength testing (Fig 2A). Similar to Oldroyd, et al. [9]., we find that patients with IBM had about half the grip strength compared to the reference population [12]. Both our studies found variability in patient grip and pinch strength as consistent with heterogeneity of IBM in cross-sectional studies.

Our data supports following grip strength and pinch strength independently in a clinical trial. While there is some overlap in the muscles used between these two movements, they are not redundant enough for one to fully represent the other. Grip and pinch is particularly in control patients as well IBM patients with more preserved hand function (Fig 2B and C). Oldroyd et al. found a yearly 10% decline in pinch strength compared to 2.4% decline in grip [9], which further supports grip and pinch strength as independent measures in IBM. Cortese et al. also found a ~10% yearly change in manual measurement of finger flexor strength previously [2].

Pinch and grip strength drop the most over the first 5–10 years of reported disease duration, after which there is a relative plateau (Fig S1). This is consistent with recent report of weakness manifesting predominantly in pinch and grip strength in younger patients, while the older population had more notable quadricep/ambulation changes [9]. The subsequent plateau (Fig S1) reflects a ceiling effect, which may be due to clinical development of muscle fibrosis and contractures. We suggest that selecting patients who are early in the disease process (< 10 years since symptom onset) for clinical trials would maximize the chance of seeing meaningful stabilization, and if fortunate, improvement with emerging therapeutics.

We previously found that the IBM-FRS has a surprisingly low correlation with measured hand strength [4]. Patients needed to lose significant grip strength to have a lower score in IBM-FRS upper extremity items, raising the risk of this scale being insensitive to clinical changes over a 1–2 year trial period. A common criticism of the IBM-FRS is that it was modified from the amyotrophic lateral sclerosis (ALS)-FRS without factoring in IBM patient experience. Patient-reported outcomes (PROs) are recommended by the FDA for use in a clinical trial [10,11] as they are regarded as a multidimensional framework that prioritizes the experiences, attitudes, and beliefs of patients. PROs give weight to issues important to the patient that may not be recognized as important for health care providers or family. PROs that assess upper extremity functions, such as the ability to use utensils and perform self-care, allow us to assess functions in IBM patients that support longer period of independence.

Our study found that the IBM-PRO Upper Extremity Function Scale items had a higher correlation with objective measures of grip and pinch strength by dynamometry compared to the upper extremity items in the current FRS (Figs. S2 and S3). Unfortunately, decreases in the hand items of the IBM-FRS or the IBM-PRO Upper Extremity Function Scale were not associated with disease duration, as we continue to see the inherent heterogeneity of IBM in this cross-sectional study.

Our study proposes separate outcome measures for different functional domains (eg. hand function, ambulation, etc.) in IBM clinical trials. This makes sense for a heterogeneous disease such as IBM and is supported by the study by Ramdharry et al. who performed a Rasch analysis on the IBM-FRS [15]. They found different rates of change in the different domains and also proposed separating out the IBM-FRS into an IBM-FRS upper limb scale which emphasized hand-function.

Our study is limited by the cross-sectional nature of assessing 83 patients with self-reported duration of symptoms anywhere from 2 years to 39 years as opposed to a truly longitudinal study. The diagnosis of IBM and duration of symptoms was dependent on patient report, and the length of disease may be subject to recall bias. While the strength in our study is the large number of IBM patients, we had a smaller, predominantly female control cohort. Additionally, similar to Oldroyd et al., we did not have the normalized data for pinch strength [9]. We did not have enough control patients at each age and sex to establish reference values. We previously showed that the IBM cohort is better distinguished from the control cohort when grip strength is normalized to published population reference values—taking into account the possible difference in the decline in strength in each sex [4]. Therefore, we hypothesize that normalizing pinch strength will help differentiate between patients with weaker pinch and between our IBM and control cohort.

Despite its limitations, our study demonstrate that the IBM-PRO Upper Extremity Function Scale items better reflect measured function by dynamometer in IBM patients compared to the IBM-FRS. This is a unique cross-sectional look at pinch and grip strength and patient-reported outcome measures in IBM patients.