This study investigated the relationship between the mRS scores of patients after thrombolytic therapy and imaging biomarkers, including APT, ASL, and their combination. The results indicate the feasibility of APT, ASL, and their combination in predicting the therapy efficacy of subacute phase ischemic stroke patients after thrombolytic therapy. Both rASL and rAPT in the good prognosis group were significantly different from those in the poor prognosis group, indicating their clinical value in predicting the effectiveness of the therapy. The combination of rASL and rAPT had the highest AUC in predicting outcomes and might be a potential imaging biomarker for predicting stroke prognosis after thrombolytic therapy.
The rAPT was significantly different between the two groups, indicating that it could be used to predict the therapeutic efficacy of subacute phase ischemic stroke. This is consistent with previous studies which have shown that APT MRI can detect tissue pH [
31,
32]. When the cerebral artery is occluded, the corresponding blood perfusion pressure will gradually decrease, a series of pathophysiological changes will occur in the human body, the lactic acid concentration in the lesion area will gradually increase, the pH value will decrease, and the proton exchange rate of saturated amide and water will also be significantly reduced. In addition, studies using animal models have found that the study has found that the emergence of APT effect, the reduction of pH value will cause the slowdown of the exchange rate of amide protons and hydrogen protons in water, while the MTR in the brain tissue of dead rats will gradually decrease, and the calibration of APT in cerebral ischemia models has found that there will be obvious differences between PH values and normal brain tissues, which is basically consistent with the results of tissue staining behavior [
18]. This may make rAPT an effective marker for understanding the outcome of subacute phase ischemic stroke. The mean rAPT value in the good prognosis group was approximately zero, and the standard error was low. These findings are also similar to those of previous studies [
33]. Sui, H. J found that the ADC value of the cerebral infarction lesion area will gradually change from low to high with time [
34], and in the subacute phase (days 2–10), as the acid-base balance of the perfusion area continues to alkalinize, the APTw reduction area shrinks, and the perfusion area improves correspondingly with the opening of the lateral branch circulation and reconstruction, and “false normalization” can occur. The average course of infarction cases in this group was 115 h and 50 min (about 4.8 days), from 49 h and 30 min to 238 h and 47 min (about 2–10 days), in the stage of acidosis to alkalinization change, resulting in microenvironments that affect APT values, such as PH concentration, which do not appear to change significantly. As a result, APT imaging cannot detect a significant difference between the stroke area and the contralateral normal area after thrombolysis. The mean value in the good prognosis group was lower than that in the poor prognosis group, which is inconsistent with previous research [
35]. Studies have found that the APT signal of patients with different periods of ischemic stroke will be reduced to varying degrees, and the signal intensity in the hyperacute phase will be the lowest, which is basically consistent with the most severe cases of acidosis in the hyperacute ischemic area. With the prolongation of the onset time, acidosis will gradually decrease, and the signal intensity of APT can reflect the characteristics of PH value of ischemic tissues at different stages and different time points, which has a more advanced predictive effect. PH imaging may be a crucial substitute imaging biomarker for acute ischemic stroke, as it reflects the metabolic health of the tissue. The initial pH adjustment results in a slightly higher APT imaging contrast change because the amide proton exchange rate is predominately base-catalyzed [
36]. Besides, several things will interfere with APT in vivo imaging: The direct saturation impact of the water proton, also known as the spillover effect, will occur when the saturation pulse excites the amide proton because its magnetic resonance frequency is very close to the water center frequency. The field strength and equipment settings have a significant impact on the direct saturation effect, and low field strength will make this effect more pronounced. Between two protons that are spatially very close to one another, there occurs a relaxation effect. The nuclear Euclidean effect occurs when one proton is energized and saturated and transfers energy to the other proton, increasing their resonance signal (Nuclearoverhauserenhancement, NOE) [
37,
38]. These will deliver upsetting polarization move outcomes, which will influence the precision of Well-suited imaging results. Because the overflow impact changes evenly with the water community recurrence, MTRaysm can eliminate the impact of this impact on the Able sign. Nonetheless, the charge move impact of strong macromolecules is not evenly circulated at the middle recurrence of water. Likewise, the reverberation recurrence of a few aliphatic protons is right at −3.5 ppm, which will likewise create the NOE outcome, which will disrupt the exactness of MTRaysm boundaries [
39]. There was also a statistically significant difference in rASL between the groups with good prognoses and those with poor prognoses. This is similar to previously described results [
40]. In normal brain tissues, intracellular pH is regulated by active and passive mechanisms and kept at 7.2 [
41]. When ischemia occurs, the blood supply to the brain decreases. Such a change in the microenvironment can be detected in the form of CBF using ASL. In the case of poor cerebral perfusion, the extrusion of CO2 from the cell is limited, and the glucose and oxygen supplies are reduced [
42], causing a decrease in bicarbonate buffering capacity and the depletion of glycogen and phosphocreatine. In general, intracellular pH relies on cerebral perfusion, intracellular energy reserves time. Affected by ischemia, brain tissue undergoes anaerobic metabolism, leading to lactic acid accumulation and decreased pH [
5,
36]. Cerebral hypoperfusion and reduced bicarbonate buffering capacity worsen tissue acidification, which frequently results in cell death and tissue damage. In this study, patients with lower mRS scores after thrombolysis generally presented with hyperperfusion, and the mean rASL value in the poor prognosis group was significantly lower than that in the good prognosis group. When an ischemic stroke occurs, the blood supply to a part of the brain decreases, leading to brain damage. Such damage can be alleviated if thrombolytic therapy can establish reperfusion on time. Hyperperfusion is believed to occur after effective vascular recanalization and tissue reperfusion because local autoregulation may require a few days to compensate for the extra blood volume that tissues receive as a result of their initial maximal vasodilatation due to acute ischemia [
43]. In contrast, patients with higher mRS scores after thrombolysis typically showed hypoperfusion in the ASL. A low rASL value after thrombolytic therapy may indicate ineffective reperfusion. This may be due to the failure of the thrombolytic or no-reflow effect. According to previous research [
44,
45], multiple factors, including microvascular obstruction, edema, and occlusion via the endothelium, could cause the no-reflow effect, even though the primary occlusion was resolved. Owing to this phenomenon, the tissue was unable to receive the nutritional support needed to recover sustainably, which might lead to more severe brain damage and a higher mRS score.
As mentioned above, the CBF value detected by ASL reflects the perfusion microenvironment, and the
acquired by APT imaging indicates the metabolic health of the tissue. Combining ASL and APT may provide comprehensive information and further understanding of the stroke lesion microenvironment and pathophysiology. In this study, the combination of ASL and APT led to further improvements in the AUC. Prior studies used DWI, ASL, and APT mismatching to identify the ischemic core and penumbra [
32]. The perfusion deficit area might recover spontaneously, and ischemic cores are potentially reversible; however, further information about the metabolic milieu is needed to make a diagnosis [
46]. Clinically, head MR is usually performed after thrombolysis to understand post-thrombolytic recanalization. However, in this study, the combination of ASL and APT led to further improvements in AUC, and the combination of APT and ASL may be a potential imaging biomarker that reflects the outcome of thrombolytic therapy in patients with subacute ischemic stroke, thereby helping to guide treatment and identify high-risk patients. It is possible to apply the MRI techniques in this study for pre-thrombolytic selection of the stroke patients who will gain clinical benefit from thrombolytic therapy.
The general data analysis of patients in the infarction group showed that there was no clear correlation between the clinical and laboratory test results and the APTw value, which may be because the measurement of clinical indicators and laboratory tests of patients were one-time measurement results, and did not coincide with the MRI examination time, and there was a certain error, which required further large-sample epidemiological research.
This study has several limitations. First, the patients did not have a baseline MRI scan before thrombolytic therapy because such reperfusion therapy is time-critical, whereas an MRI scan is time-consuming. Patients were not divided into anterior circulation and posterior circulation strokes for study and it will be our next research plan. Moreover, when patients with cerebral infarction were admitted for examination, some patients were in serious condition, and patients or their parents were more worried and nervous and could not accurately describe the last known well time. Because most cerebral infarction patients are elderly, they are less sensitive to changes in their bodies, so it is difficult to detect their own accurate onset time. Secondly, the sample size of the study was relatively small. Further studies with larger sample sizes should be conducted. Finally, APT imaging was acquired with only one slice using a 2D SSFSE sequence, whereas ASL imaging was performed using a 3D sequence.