The value of serum creatinine as biomarker of disease progression in spinal and bulbar muscular atrophy (SBMA)

Serum creatinine has been indicated as a potential marker of motor function in SBMA and results form previous longitudinal studies pointed to its decline over time. This is a longitudinal retrospective study investigating creatinine changes over a 36-month-period in 73 patients with SBMA. Severity and progression of the disease was assessed according to serum creatine kinase (CK) values, manual muscle testing (MMT), SBMA functional rating scale (SBMAFRS) score, 6-min-walk test (6MWT) value, and spirometry (forced vital capacity, fVC%) obtained at the baseline and at each of the annual follow-up visits. Baseline serum creatinine concentrations positively correlated with 6MWT, the MMT megascore score of both the upper (ULM) and lower (LLM) limbs and SBMAFRS. No correlation was found with CK or fVC% values. Similar correlation results were achieved at all the subsequent time points. Longitudinal assessments conducted by the generalized estimating equations (GEE) method returned significant changes for SBMAFRS (− 1.41 points per year, p < 0.001), ULM and LLM (− 0.69, p = 0.01; and − 1.07, p < 0.001, respectively), 6MWT (− 47 m, p < 0.001) but not for creatinine (− 0.82, p > 0.05). We also observed that creatinine levels at baseline did not correlate with changes in the other measures from baseline at each annual visit. Our data do not support a role for serum creatinine as sensitive biomarker of disease progression, and possibily prognosis, in SBMA.


Data collection
Patients' characteristics, including age at onset (described as subjective weakness in any part of the body, bulbar and/or spinal district) and at baseline visit, length of illness since onset of weakness, and number of CAG repeats, were collected.The severity and progression of the disease were assessed through the following measures obtained at the baseline visit and at each of the annual follow-up visit: 1. biochemical markers (serum creatine kinase, CK, and creatinine levels); 2. manual muscle testing (MMT) according to MRC score of the following muscles: deltoid, biceps brachii, triceps brachii, extensor carpi, opponens pollicis for upper limbs; iliopsoas, quadriceps femoris, anterior tibialis, and extensor hallucis longus for lower limbs; all muscles were tested bilaterally and cumulative scores for upper and lower limbs, namely upper limb megascore (ULM), range 0-50, and lower limb megascore (LLM), range 0-40, were used for statistical analysis; 3. SBMA functional rating scale (SBMAFRS) score 25,26 ; 4. 6MWT distance (meters); 5. respiratory muscle function according to the forced vital capacity (fVC, expressed as percentage of predicted value).For each patient, glomerular filtration rate (GFR) and blood urea nitrogen were also annotated to monitor renal function.

Statistical analysis
To verify any deviation from the normal distribution of the variables considered, the Shapiro-Wilk test was applied.Biochemical parameters were compared among data at different time-points (baseline, 12 months, 24 months, 36 months) using Wilcoxon Signed Rank Tests for repeated measurements on a single sample.Spearman's rho correlation coefficient were assessed to verify a possible correlation between creatinine serum levels and clinical parameters at different time-points.
For longitudinal assessments, Generalized Estimating Equations (GEE) were used to evaluate all measure progression over time (i.e. per year).Spearman regressions were also performed to evaluate the correlation between the baseline creatinine values and the delta for each outcome (calculated as "outcome evaluation at the specific time point-outcome evaluation at baseline") in order to evaluate the prognostic effect.Finally, at each time point, the delta of each parameter was also compared with the delta creatinine at the specific time point.Statistical analyses were performed in R (R Foundation, version 4.0.2), with statistical significance set at P < 0.05 for all tests.To graphically represent data tidyverse and beeswarm packages were used.

Ethics approval
This study was performed in line with the principles of the Declaration of Helsinki.Approval was granted by CESC (Comitato Etico per la Sperimentazione Clinica della Provincia di Padova), AOP1696.
At the baseline visit, 57 patients (78%) had mild muscle weakness in all four limbs, 12 (18.4%)had mild to moderate weakness and required walking support, and 4 (6.1%) were using a wheelchair, being therefore unable to complete the 6MWT.None of them complained of significant respiratory or swallowing deficits.
www.nature.com/scientificreports/Mean values of functional and biochemical measures at baseline and subsequent annual monitoring visits (12, 24 and 36 months), along with the number of individuals assessed at each time point, are reported in Table 1.
We did not observe any relation of creatinine with patients'age, disease duration, or CAG repeat number, whereas a significant correlation between disease duration and SBMAFRS (r = − 0.45; p = 0.00013), 6MWT (r = − 0.43; p = 0.0003), ULM and LLM (r = − 28; p = 0.023 and r = -0.50;p = 2.017 × 10 -05 , respectively) was noted.For each patient, renal function parameters were found within the normal range throughout the study period.

Discussion
SBMA is a slowly worsening neuromuscular disease 28 and a biomarker is not yet available to significantly track the disease progression in a period of time suitable for short-term trials 29 .Serum creatinine has been indicated as a potential marker of motor function in SBMA and results form previous longitudinal studies pointed to its decline over time in patients 13,28 .
In our retrospective study of 73 patients with SBMA, we confirmed a good and sustained correlation of creatinine with SBMAFRS, 6MWT and MMT although not with CK and FVC%.On the other hand, we also observed that creatinine values did not decrease significantly during the 36 months of observation, unlike SBMAFRS, 6MWT and MMT.Such a discrepancy between cross-sectional and longitudinal results of creatinine performance compared to the other outcome measures may possibily reflect the relative instability of creatinine concentrations due to mechanisms other than muscle mass/function.In fact, in addition to renal function, creatinine serum level is influenced by many variables including dietary intake or physical activity 19,[31][32][33][34] .Similarly, Dahlqvist et al. 13 observed that creatinine was stable among patients with protein levels below the reference range or increased in others during the 18-month observation period.Overall, these observations suggest a poor reliability of creatinine as a marker of short-term progression in SBMA.
In line with previous studies 28,29 , we confirmed that creatinine levels are unrelated to those of CK.Indeed, CK values are a marker of muscle injury rather than muscle function and they have also been reported not to correlate with functional parameters in SBMA 28,29 .Furthermore, CK values are vulnerable to a SBMA-specific impaired muscle metabolism of creatine 35 and patient's physical exercise prior to blood sampling 28 .Of interest, we reported a significant increase in CK levels in SBMA patients receiving beta2-agonist treatment who nevertheless showed improvement in motor performance 36 .
A relationship between creatinine and fVC values was also lacking, possibly because respiratory involvement may occur at advanced stages of the disease 37 and, in addition, no patients of our cohort complained of respiratory issues.
Further, we assessed whether creatinine measurement could have prognostic significance.However, creatinine levels at baseline failed to predict changes in other measures over the observation period, nor was there a clear correlation observed between changes in creatinine and other measures compared to baseline.
This study has limitations including the retrospective design and the drop of patient number at the 36-month visit.As regards the latter point, the missing data basically belong to those patients who were initially followed at our center and who then moved to a nearest center following the recognition of other reference clinics across the country in accordance with the Italian SBMA Registry 38 .Therefore, we are confident that we can rule out any bias related to the disease course as the reason for the decline in patient ratings at 36 months.
In conclusion, our data do not support a role for serum creatinine as sensitive biomarker of disease progression in SBMA.Further studies that will also consider more recent outcome measures such as muscle MRI are warranted.

Figure 1 .
Figure 1.Scatter-plot between creatinine serum levels and clinical parameters at baseline.Spearman's Rho coefficients (r) and p-values (p) from cross-sectional correlation are shown.6MWT 6-min-walk test, Fvc forced vital capacity, CPK creatine kinase.