In the present study, we analyzed retrospectively the DTI-ALPS index in patients with brain tumors, including both primary brain tumors and cerebral metastases, and compared them to healthy subjects. We compared the DTI-ALPS index of the tumor-affected hemisphere to that of the contralateral side and found statistically significant differences between the hemispheres. In comparison with HC, we found no difference between the patients' ipsilateral and matched DTI-ALPS index. In addition, the DTI-ALPS index of the contralateral side of the patients was larger than the one from HC, contradicting Toh et al. study on meningiomas [10].

On the other hand, we did not find significant group differences according to age, sex, or type of tumor. Furthermore, we did not find a linear correlation between DTI-ALPS index and age and tumor ADC. To our knowledge, only three publications so far have used the DTI-ALPS method to assess GS function in patients suffering from brain tumors [9,10,11]. However, only the study in meningioma [10] considered the tumor’s contralateral side and HCs. The DTI-ALPS index in brain tumors has been evaluated by Toh et al. [9,10,11]. First, they analyzed the formation of PTBE in meningiomas [10]. In that study, they compared the DTI-ALPS index in patients with meningiomas with and without PTBE and HC. They found that the DTI-ALPS index was larger in patients without PTBE than in both healthy subjects and patients with edema. There was no difference between the patients with edema and the HC. These findings disagree with our results, where there is no significant difference between the ipsilateral side and the matched HC and also that the contralateral DTI-ALPS index was larger in tumor patients than HC.

Previous publications have also observed no significant differences between patients and controls, such as in Parkinson's disease [16], focal epilepsy [25] and migraine [36]. However, to our knowledge, no publication has reported a higher index in patients than in healthy subjects. One could hypothesize that the diffusion along perivascular spaces in the contralateral side is abnormally enlarged in patients in an attempt to compensate for the increased amount of brain waste products due to the increased tumor's metabolic activity, tissue compression and general neuroinflammation. However, no definitive conclusions can be made yet and further research is required.

Besides, one needs to consider that Toh et al. [10] mainly studied PTBE in extra-axial tumors and its influence on the DTI-ALPS index, and we did not make a separation between patients with and without PTBE.

Additionally, contrary to our findings, they did not find a significant difference between ipsi- and contralateral hemispheres in meningioma patients, without further discussion. We speculate that for this reason, they did not investigate the DTI-ALPS index in the tumor’s non-affected contralateral side in further research. We could hypothesize that intra- and extra-axial tumors may influence the DTI-ALPS index using different mechanisms. However, we cannot draw definitive conclusions, and more research is needed [10].

Secondly, Toh and Siow [9] investigate the DTI-ALPS index in patients with gliomas in the hemisphere of the tumor. They compared the DTI-ALPS index in patients with different stages of glioma, the type of isocitrate dehydrogenase 1 (IDH1), and the sex of the patients. They found a significant difference in the DTI-ALPS index between patients with a lower glioma grade and those with a higher-grade glioma (II/III grade vs. IV grade glioma). Additionally, the index for the mutant IDH1 was higher than that for the wild-type mutation. They did not include HC in this research. It could have been interesting to compare our work with Toh et al. glioma study [9]. Unfortunately, it lacks ipsi—contralateral comparison and does not include HC. Consequently, we cannot compare our results.

Lastly, they studied perivascular edema in brain metastases [11]. They found that high ADC tumor values and low DTI-ALPS index were related to large edema volumes. This suggests that high volumes of edema may be related to intratumoral water diffusivity, disturbing glymphatic function. We did not find a significant relationship between the tumor's ADC and DTI-ALPS index. However, as the main focus of Toh et al. [11] is PTBE; therefore, making a comparison between our results is not possible.

In our opinion, the contralateral hemisphere should be analyzed in particular when considering gliomas as it must increasingly be assumed that higher-grade gliomas do not only show abnormalities in pathologically contrast-enhancing lesions on MRI. Cerebral glioma must increasingly be evaluated as diseases of larger parts of the brain, if not of the entire brain. Therefore, determining the DTI-ALPS index in the contralateral hemisphere, too, is significant.

As expected, we found that the DTI-ALPS index in the ipsilateral hemisphere was lower than that on the contralateral side. This might suggest impaired glymphatic function on the side of the brain with the tumor. However, the DTI-ALPS index does not solely represent diffusion in the perivascular space, but also other processes such as axonal degeneration [37]; and, therefore, cannot be directly linked to the GS function. Additionally, due to different tumor locations, the anatomy of the brain and periventricular veins may be altered and measuring the diffusivity along the perivascular spaces in the ipsilateral side may be not be totally free from unwanted influences due to anatomical displacements. Therefore, we excluded patients that due to the position of the tumor, the area of the periventricular veins was affected. Additionally, intra-axial tumors can directly disrupt the microstructure of the brain and directly affect the water diffusion in the brain due to e.g. cellular growth or tumor cell infiltration. On the other hand, extra-axial tumors, due to mass effect, can compress and exert pressure on adjacent structures but typically do not directly infiltrate brain tissue. This could suggest the existence of different brain waste clearance mechanisms. Also, as the DTI-ALPS index is measured in the deep white matter, it is not surprising that the apparent diffusion coefficients in the tumor-affected hemisphere will differ from the contralateral side in the case of intra-axial tumors. Nevertheless, both tumor types could compress the brain, or the extra-axial tumors could infiltrate the perivascular spaces, like the periventricular veins. Therefore, extra-axial tumor infiltration into brain tissue might have different influences on deep white matter than non-infiltrated ones, and this might need to be investigated separately.

In our cohort, we did not find significant gender differences regarding the DTI-ALPS index results. This is consistent with the findings of other DTI-ALPS tumor-related studies [9,10,11]. Previous studies investigating the DTI-ALPS index difference between males and females reached the same conclusion [8, 14, 16, 17, 35, 38,39,40,41], where there is no significant difference between sexes. However, contradictory results have been reported by Hsiao et al. [42] and Zhang et al. [43], where the DTI-ALPS index in women was larger than in men. This shows that sex differences might still need further investigation into different illnesses with a larger sample size.

Furthermore, we did not see a significant difference between the DTI-ALPS index and the type of tumor (primary or brain metastasis). In contrast to Toh et al. [11], we did not find a linear correlation between the tumor ADC and DTI-ALPS index. However, due to our small sample size, further studies are needed to evaluate these parameters. These contradicting results could be due to differences in calculating the ADC in the tumor. We selected a small area in the middle of the tumor, while they calculated the ADC by making an average of all slices where the tumor was present. However, we did not investigate PTBE and thus cannot directly compare this parameter.

Previous studies found a negative correlation between age and DTI-ALPS index [5, 15, 25,26,27, 42, 44, 45]. This is a common finding in healthy volunteers and in neurodegenerative diseases such as Parkinson’s and Alzheimer’s disease. However, in cases of brain tumors [9,10,11], REM sleep disorder [46], neuromyelitis optica [47], renal disease [48, 49], global amnesia [50], and migraine [36] did not find significant correlation was found between DTI-ALPS and age. In our research, we did also not find any correlation between patients' age and DTI-ALPS index. Therefore, the relationship between age and DTI-ALPS index should be further studied to determine which factors might be involved in the diffusion changes along perivascular spaces under other clinical conditions.

Here, we compare the DTI-ALPS index of the brain tumor hemisphere with the contralateral side and also differences between tumor type, age, and sex. Additionally, we included HC to compare the DTI-ALPS indexes. Nonetheless, there are several limitations that need to be addressed:

The sample size of this single-center, retrospective, consecutive study is small. Further studies should consist of more tumor patients and age- and sex-matched HC. Additionally, the study lacks data about the clinical course of the patients; thus, we cannot correlate GS dysfunction with subsequent, distinct psychiatric or neurological deficits. Further studies should therefore correlate the DTI-ALPS index with corresponding neuropsychological parameters.

The correct planning of the ROIs can be difficult and may require practice. In this study, a senior and junior neuroradiologist positioned the ROIs in consensus. Additionally, we planned the ROIs using only the fractional anisotropy image, as susceptibility-weighted images are not part of the clinical routine at our institution. Further studies should include susceptibility-weighted images for positioning of the ROIs to determine whether the course of the periventricular and peritumoral blood vessels has an effect here.

Additionally, the DTI-ALPS method makes an estimation of the diffusivity along the perivascular space in a small ROI parallel to the medullary veins. Therefore, makes an extrapolation that the whole brain glymphatic function could be derived from this small area within deep white matter, while the production and accumulation of amyloid β and tau proteins occurs in the brain cortex [51].

It is important to consider that the DTI-ALPS index does not solely reflect random water motion in the perivascular spaces, but it can be also influenced by factors such as axonal degeneration. Therefore, even if the DTI-ALPS appears as a technique to evaluate the GS, those indices might not reflect brain waste clearance directly. Hence, changes in the DTI-ALPS index should be interpreted with caution [37, 52, 53].