Impact of long-term storage on mid-infrared spectral patterns of serum and synovial fluid samples of dogs with osteoarthritis

Objective To evaluate impact of long-term storage on mid-infrared (MIR) spectral patterns of serum and synovial fluid (SF) of dogs with knee OA and controls. Design Serum (52 OA and 49 control) and SF (51 OA and 51 control) samples from dogs that had been in short-term (<3 years) frozen state (−80°C) had their MIR spectra obtained. The remaining aliquots were maintained in long-term (>5 years) frozen state before having MIR spectra acquired under the same testing conditions. Multi-level simultaneous component analysis was used to evaluate the effect of time. Partial least squares discriminant analysis was used to compare performance of predictive models built for discriminating OA from control spectra from each time point. Results Median interval of storage between sample measurements was 5.7 years. Spectra obtained at two time points were significantly different (P <0.0001), however, contribution of sample aging accounted for only 1.61% and 2.98% of serum and SF profiles’ variability, respectively. Predictive models for discriminating serum of OA from controls for short-term storage showed 87.3±3.7% sensitivity, 88.9±2.4% specificity and 88.1±2.3% accuracy, while, for long-term storage, values of the same figures of merit were 92.5±2.6%, 97.1±1.7% and 94.8±1.4%, respectively. Predictive models based on short-term stored SF spectra had 97.3±1.6% sensitivity, 89.4±2.6% specificity and 93.4±1.6% accuracy, while the values for long-term storage 95.7±2.1%, 95.7±0.8% and 95.8±1.1%, respectively. Conclusions Long-term storage of serum and SF results in significant differences in spectral variables, however, these changes do not significantly alter the performance of predictive algorithms for discriminating OA samples from controls.


64
Biobanking biological fluid samples (e.g., blood, urine, joint fluid) is an essential component of 65 prospective clinical studies and a means of maximizing efficiency of using obtained samples. 66 These biological samples and the contents therein (i.e., cellular, molecular) are at risk of being biological fluids to detect variability between disease states as a novel approach to disease 77 pattern recognition has been investigated with an increasing rate in the past few decades(4-12). 78 The spectral pattern of a given sample is the sum of all MIR light absorbance by the infrared-79 active molecular bonds within the sample that is displayed as a unique waveform (spectrum)(13). 80 Fourier-transform infrared spectroscopy (FTIR) is one of the techniques utilized in acquiring 81 MIR spectra of biological samples that is adjuvant free, simple and requires small volume of 82 samples(5-9). Use of the spectral variables based on this method has been successfully used in 83 differentiating various types of arthritis in humans(4). Research in experimental and clinical 84 models of osteoarthritis (OA) in various joints of animal models has also been able to 85 demonstrate ability to distinguish OA from control samples with high accuracy rates(5-9). Short-5 86 term (< one year) longitudinal studies using spectral variables in serum and joint fluid of animals 87 have shown reproducible spectral characteristics of OA that despite changes overtime, remain 88 distinguishable from control samples(7-9). Studies investigating long-term changes in clinical 89 OA in larger animal models (e.g., dogs) and human models require at least several years that 90 warrants biobanking of collected biological samples to allow for batch analysis. These biobanked 91 biological samples (e.g., blood, synovial fluid0 are typically stored in -80 degrees Celsius for 92 long-term storage to minimize the impact of time. However, the impact of long-term storage on 93 the spectral pattern of these biological samples has not been previously investigated. The first 94 aim of this study was to evaluate whether long-term storage results in changes in the MIR 95 spectral pattern of serum and synovial samples of dogs with and without knee OA as measured 96 by FTIR spectroscopy. The second aim of this study was to evaluate the impact of any observed 97 changes due to long-term storage on the ability to discriminate between the serum and synovial 98 fluid spectra patterns of dogs with and without knee OA. The hypothesis of the study was that 99 long-term storage will have minimal impact on MIR spectral pattern of serum and synovial fluid 100 of dogs with or without knee OA. owners for an earlier study (8,9). The sample size used in the current study was based on 106 available number of samples from the original project that had two arms investigating serum and 107 synovial fluid samples from client-owned dogs with OA secondary to naturally occurring 108 degenerative (non-traumatic) cranial cruciate ligament tear in one or both knees and controls(8,  The available sample inventory was reviewed for OA and Control samples with adequate volume 123 left for analysis. For the OA group the serum and synovial fluid samples were those obtained 124 prior to any surgical intervention. If more than one aliquoted sample was available, the clearest 125 was selected (i.e., non-to minimal blood contaminated for synovial fluid and none to minimal 126 hemolysis for serum samples). Only one serum sample per dog was selected. In the OA group, 127 even if both knees were affected, only one sample from one knee per dog was included in the 128 study. In the control group, both healthy knees of each dog had been sampled, but preferentially 129 only one sample from each dog was included.

130
The age of samples at the time of first spectral analysis was calculated based on the date sample 131 was obtained from the dog and the time when the sample was thawed, and the first spectral 7 132 analysis was performed. Samples from this initial storage period are referred to as "short-term   156 Analyses of non-spectral data were performed using SPSS software (IBM SPSS Statistics, v. 25).

157
Variables without a normal distribution were described by their median and interquartile range 158 (IQR). All acquired serum and synovial fluid spectra files from both time-points were imported In particular, the effect of sample aging on the spectroscopic signal is related to the within- found that only 1.61% of the total variability among the serum spectra was ascribable to the 216 effect of aging (between-samples differences accounted for 77.23% of the total variance, while 217 the remaining part is residual variation, associated to random error). The variability observed 218 between the two time points was not impacted by the type of sample (OA versus control).

219
Although the relatively low amount of spectral variance was associated with the differences 220 between short-term and long-term storage, permutation testing indicated that these differences

293
This is the first study evaluating impact of long-term sample storage on the quality of MIR 294 spectra of serum and synovial fluid samples used for discriminating OA from control samples.

295
Both serum and synovial fluid samples spectra had statistically significant differences in spectral 296 variables due to aging of the samples. However, the contributions of these differences to the overall 297 spectral variability for both serum and synovial fluid were relatively small (i.e., <3%). It is 298 important to note that despite the statistically significant differences between the spectra of 299 samples from short-term versus long-term storage, these differences were not a significant   Another limitation of this study is lack of additional, shorter intervals between 348 measurements, and longer storage period to assess possible limits for sample storage time and 349 trends as to when the sample variability due to degradation due to aging overwhelms the 350 differences based on disease state. Future studies, can prospectively store samples and sequentially 351 measure MIR spectra over shorter intervals (e.g. six months intervals) to document these changes.