Arachnoid cysts do not contain cerebrospinal fluid: A comparative chemical analysis of arachnoid cyst fluid and cerebrospinal fluid in adults

Arachnoid cysts (AC) are relatively common benign lesions of the arachnoid, with a reported prevalence as high as 1.1% in the adult population [1]. Clinical presentations of AC include headache, dizziness, seizures [2] and dyscognition [3]. They can be found all along the cranio-spinal axis, but have a marked predisposition for the temporal fossa [4]. The mechanisms underlying the formation and filling of arachnoid cysts are not well understood, but clinical, epidemiological, and laboratory data indicate that genetic mechanisms are involved in the formation of arachnoid cysts [5, 6]. Three prevailing theories exist for the filling of the cyst: 1) active secretion of fluid by cells in the cyst wall [7, 8], 2) fluid influx due to an oncotic pressure gradient [9], and 3) trapping of fluid by a valve mechanism [10]. It is conceivable that the chemical composition of the AC fluid relative to the cerebrospinal fluid CSF reflects the mechanism by which the fluid enters the cyst. If the composition is identical to CSF a valve mechanism appears likely, whereas if the filling is caused by oncotic pressure, a higher concentration of proteins in the cyst fluid compared with CSF would be expected. Likewise, cyst fluid composition could reflect the mechanism of transport across the cyst wall, if such a mechanism is involved.

In their study of clinical chemical analysis of cyst fluid in pediatric patients, Sandberg et al. [11] described a similar chemical composition to that of reference CSF in the majority of patients investigated, but in 14 of 41 (34%) the protein concentrations were elevated above 0.50 g/L in the cyst fluid. Based on these findings, the authors hypothesized that higher protein content could contribute to the expansion of the cysts by an oncotic pressure gradient. We have recently described the up-regulation of the CSF-secreting cation chloride co-transporter NKCC1 in AC membranes compared with normal arachnoid [8]. This finding supports fluid secretion as the main mechanism of fluid accumulation in AC. The objective of the present study was to analyze the chemical parameters of AC fluid and compare with CSF from the same patient, to gain further knowledge of AC concerning the mechanisms of filling and sustaining of such cysts.

Chemical analysis of AC fluid and CSF, obtained during elective surgery for arachnoid cysts from 15 patients, was performed in a routine hospital laboratory. The results from the measurements are presented in table 2. There was no significance difference in osmolarity or concentrations of sodium, potassium, chloride, calcium, magnesium or glucose between AC fluid and CSF. The concentration of phosphate was higher in AC fluid relative to CSF (0.39 versus 0.35 mmol/L; p = 0.02), while the concentration of total protein (0.30 versus 0.41 g/L; p = 0.004), lactate dehydrogenase (17.9 versus 35.6 U/L; p = 0.002) and ferritin (7.8 versus 25.5 ug/L; p = 0.001) was significantly lower in AC relative to CSF. Bilirubin, iron, triglycerides and immunoglobulins were below quantification limit for analysis setup. Patient 12 was identified as an extreme outlier for protein measurement and was excluded from this analysis.

Correlations were calculated between possible blood contamination parameters ferritin, lactate dehydrogenase and protein on the CSF samples without observed blood contamination (n = 11). The correlation between ferritin and LD was 0.77 (p = 0.006), ferritin and protein was 0.36 (p = 0.312) and lactate dehydrogenase and protein was 0.34 (p = 0.34). The protein concentrations in the cyst fluid from the two clinically different patients (patient 3 and 4, respectively 6.16 g/L (hematoma) and 3.54 g/L (previous operation)) were significantly elevated relative to the others (p < 10^-16).

The aim of the present study was to collect information that may contribute to the understanding of the mechanism for arachnoid cyst filling. If the cyst were filled by a simple valve mechanism, which has been observed in suprasellar cysts, it would be expected that the composition of the cyst fluid would be identical to that of CSF. On the other hand, if the filling was caused by oncotic pressure, one would expect a difference in osmolarity between the AC fluid and CSF as well as a higher protein content in the AC. If the underlying mechanism were active secretion or transport, this would probably cause a different concentration of certain molecules or ions depending on the underlying transport mechanisms. We found an isotonic AC fluid with a lower protein concentration than in CSF; this is not consistent with an oncotic pressure filling mechanism.

Macromolecules such as albumin (67 kDa), ferritin (24 subunits between 19 and 21 kDa, total weight around 450 kDa) and lactate dehydrogenase (tetramer of about 37 kDa, total weight around 140 kDa) would be expected to pass freely through a slit in the cyst membrane. However, the ratio between cyst fluid and CSF for these three protein complexes was reduced (dependent on size), from 0.73 for protein (of which 2/3 albumin) to 0.31 for ferritin. Our findings are therefore contradictive of a slit valve mechanism underlying the filling of temporal AC.

The skewed distribution of phosphate could imply a selective or active transport of fluid and solutes over the AC membranes. This is consistent with previous findings of morphological and enzyme ultracytochemical structures in the wall of arachnoid cysts assumed to be capable of fluid secretion, as reported by Go et al [7]. Furthermore, our group has recently published evidence that there are differences between arachnoid cysts and normal arachnoid tissue in that the Na-K-Cl cotransporter NKCC1 is up-regulated in arachnoid cysts compared with normal arachnoid [8], and a small subset genes are differentially expressed in arachnoid cysts compared with normal arachnoid tissue [5]. The phosphate level in the CSF is kept lower than in the blood [16], due to active transport mechanisms in the choroid plexus epithelia. Higher phosphate concentration in the cyst fluid could imply that the cyst epithelium is either not, or is differentially equipped with transport mechanisms relative to the choroid plexus. There are several lines of evidence suggesting that co-transporters such as GLUT1, MCT1 and NKCC1 have the ability to transport water along with their respective substrates, regardless of the osmotic gradients [17].

As we have only studied temporal cysts, we cannot generalize the assumption of an active transport mechanism to all ACs. For other locations, other fillings mechanisms may well exist, such as a slit-valve in supracellar cysts [18, 19]. In our study, two patients differed from the others - one with a previous hematoma in close proximity to the cyst and one being a reoperation. The protein levels in the cyst fluid from these patients were significantly higher than in the rest of the study group. In their study Sandberg et al [11] found that 14 of the 41 patients had markedly higher protein concentrations (above 0.5 g/L) in the cysts; four of those had extreme values such as the ones we observed. The cause for some cysts to have a marked elevated protein level is not well understood, but products from previous bleeding in the cyst may be one explanation. Correlation analysis between protein, ferritin and lactate dehydrogenase was performed as a statistical control to rule out blood contamination as a possible explanation of the protein content in the CSF. The protein concentration of CSF relative to plasma is about 0.5-1%; a leak of plasma to the CSF would introduce a large source of error to the analysis. If the difference in concentration of protein, lactate dehydrogenase and ferritin was caused by blood contamination, the correlation should be expected to be significant and close to 100%. These correlation results do not show a strong association between assumed blood contaminants and the findings, thus not supporting blood contamination as explaining variable.

A limitation for this study is the small number of samples. Applied on such a dataset, the T-test is vulnerable to outliers such as the excluded patient 12 in protein concentration (outlier plots not shown). Nonetheless, our results can support reflections around the filling mechanisms for arachnoid cysts. As far as the authors know, this is the first publication of AC fluid analysis matched with CSF from the same patient.

The chemical composition of AC fluid found in this study does not support an oncotic pressure or valve mechanism as responsible for filling an AC. Due to the pattern of differences, we postulate that the filling mechanism for temporal AC is by either a selective or active transport mechanism or a secretion from the cyst-lining cells.