| Research background and purpose:Rheumatoid arthritis(RA)is a chronic autoimmune disease with an incidence of about 0.5%–1.0%.The main pathological manifestation is the destruction of synovial tissue.In the long-term chronic injury caused by synovial damage,joint destruction gradually aggravates,and severe arthritis,joint swelling,and deformity may occur in the later stage of the disease,eventually leading to permanent disability.Although joint replacement or other surgery is an effective treatment measure at present,it also brings a heavy medical and economic burden to patients and society.Early detection and active treatment can reduce late complications and joint deformities,and avoid joint replacement and other operations and the resulting high surgical costs.Therefore,improving the screening rate or diagnostic rate of RA in the early stage is one of the clinical problems that need to be solved urgently.As an optical detection technique,Raman spectroscopy can show different spectral waveforms according to different substances,that is,each substance has its unique map.Blood is an important place of the body’s metabolic process,and most metabolic substances can be detected by blood.Therefore,special substances or biochemical changes produced by diseases can be monitored or identified by serum Raman spectroscopy,which provides a new idea for disease diagnosis or screening.Previous studies have demonstrated that Raman spectroscopy can be used to quantitatively measure serum anti-cyclic citrullinated peptide antibodies in RA patients with high specificity.On this basis,this study will be further deepened to establish serum Raman spectroscopy screening models for RA patients at different stages,explore the screening value of serum Raman spectroscopy for early RA patients,and provide an experimental basis for the clinical application of Raman spectroscopy technology.Materials and methods:A total of 216 RA patients and healthy adults admitted to the hospital from November 2021 to June 2022 were selected,including 88 patients with early RA,51 patients with non-early RA,and 77 healthy controls.Fasting elbow venous blood was collected from patients who met the inclusion criteria on admission,and about 5 ml of blood was drawn from each patient in a red anticoagulant tube and stored at 4 °C under low temperature.1.The age,gender,hemoglobin,platelets,albumin,serum creatinine,serum uric acid of all patients,rheumatoid factor,anti-CCP antibody,erythrocyte sedimentation rate,and C-reactive protein of RA patients were collected after routine biochemical tests were completed in the laboratory department.The values of rheumatoid factor,anti-CCP antibody,erythrocyte sedimentation rate,and C-reactive protein were defined as positive when they were higher than the upper limit of the reference value of the hospital laboratory.2.The remaining samples were separated at 3000 RPM for 5 minutes,and the upper serum was frozen and stored in an ultra-low temperature refrigerator at-80 °C for spectral detection.Raman spectrum detection and graphic data collection were performed once a month.After obtaining the original Raman spectra,Matlab 2017 b software was used for spectrum preprocessing,including eliminating cosmic rays,baseline correction,smoothing,normalization,etc.Then the variance analysis method was used to test the differences between groups and within groups.The wavenumber points with significant differences between groups were used as feature spectra and input into the support vector machine for learning and screening model construction.The k-fold(K = 5)cross-validation method was used to verify the performance of the model.Result:1.Basic information and biochemical test results(1)There were no significant differences in age,sex,platelets,serum creatinine,and the positive rate of RF,anti-CCP antibody,ESR,and CRP among all groups(P > 0.05).The ages of the normal control group,early RA group,and non-early RA group were54.43±8.04 years old,53.66±13.71 years old,and 56.27±11.24 years old,respectively,P=0.458.The number and proportion of females in the normal control group,early RA group,and non-early RA group were 56(72.73%),66(75.00%),and 43(84.31%),respectively,P=0.295.Platelet values in the normal control group,early RA group,and non-early RA group were 218.75±68.60,239.69±80.18,and 231.80±91.53,respectively,P=0.238.The creatinine levels of the normal control group,early RA group,and non-early RA group were 61.84±12.37μmol/L,59.96±13.30μmol/L and57.47±11.81μmol/L,respectively,P=0.162.The number and positive rate of RF in the early RA group and non-early RA group were 77(87.50%)and 43(84.31%),respectively,P=0.598.The number and positive rate of anti-CCP antibody in the early RA group and non-early RA group were 78(88.64%)and 44(86.27%),respectively,P=0.682.The number and positive rate of CRP in the early RA group and non-early RA group were 40(45.45%)and 20(39.22%),respectively,P=0.474.The number and positive rate of ESR in early RA and non-early RA groups were 54(61.36%)and 27(52.94%),respectively,P=0.332.(2)There were significant differences in hemoglobin,albumin,serum uric acid,and the course of RA patients among the groups(P < 0.05).The hemoglobin levels of the normal control group,early RA group,and non-early RA group were 137.68±17.79g/L,126.47±13.17g/L,and 127.16±13.99g/L,respectively,P < 0.001.The albumin levels of the normal control group,early RA group,and non-early RA group were 44.20±2.70g/L,41.48±4.30g/L,and 41.30±4.08g/L,respectively,P < 0.001.The serum uric acid levels of the normal control group,early RA group,and non-early RA group were318.84±68.93μmol/L,264.26±88.83μmol/L and 256.53±81.69μmol/L,respectively,P <0.001.The course of the disease was 3.5(1.0–12.0)months in the early RA group and45.0(28.0–60.0)months in the non-early RA group,P < 0.001.There were statistically significant differences in hemoglobin,albumin,and serum uric acid levels between the control group and the early RA group,and between the control group and the non-early RA group.The above indicators in RA patients were lower than those in the healthy control group,but there was no statistically significant difference between the early RA group and the non-early RA group.2.Spectral results(1)Raman spectra were processed using MATLAB 2017 b software,and each sample was averaged after 15 spectral measurements,showing that the average Raman spectral difference of the three groups was within 2 x standard deviation with no significant difference.(2)In order to further analyze the differences between groups,an analysis of variance method was used to test the differences between groups,and wavenumber points with significant statistical differences(P<0.05)between groups were selected as characteristic spectra,which were input into support vector machines for learning and modeling.The sensitivity,specificity,and accuracy of the model in screening early RA were 83%,78%,and 80%,respectively.The sensitivity,specificity,and accuracy of screening for non-early RA were 83%,86%,and 84%,respectively.The sensitivity,specificity,and accuracy of distinguishing early from non-early RA were 87%,85%,and 86%,respectively.The results showed that the AUC of the normal control group versus the early RA group was 0.860,the normal control group versus the non-early RA group was 0.903,and the early RA group versus the non-early RA group was 0.918.Conclusion:1.Serum Raman spectroscopy has a good screening ability for RA patients and can accurately stage the disease course,which can provide new ideas for the diagnosis or screening of early RA.2.The molecular information revealed by Raman spectroscopy may be consistent with known biomarkers of RA disease,or may be derived from unknown potential biomarkers.This technique opens up new possibilities for finding serum biomarkers of RA and studying the pathogenesis of RA.3.The levels of hemoglobin,albumin,and serum uric acid in RA patients are lower than those in normal controls,which may be related to inflammatory factors,malnutrition,and taking drugs. |
PDF Full Text Request