This was a retrospective analysis of 100 CT scans of adult individuals (56 males, 44 females; 200 sides total) obtained from Inkosi Albert Luthuli Central Hospital (IALCH) in Durban, KwaZulu-Natal, South Africa. The CT images were of 13 years of age and above, classified into age categories: <=29, 30-39, 40-49, 50-59, 60-69, 70-79, and >=80 years. Due to data distribution, the Indian and White sample groups were merged into one sample (29; 29%) and compared to the Black sample (71; 71%).

A total of 100 CT images from CT venograms and CT angiograms (with venous phase) were viewed and analyzed using syngo.plaza viewer software (version VB10) in axial and sagittal planes. Scanners used included Multi-Detector row CT (MDCT) Scanners (Lightspeed CT, GE Healthcare and SOMATOM Definition Flash CT, Siemens Healthineers) with 1mm and 2mm slice configurations. Inclusion criteria comprised scans of patients >13 years depicting both cerebral hemispheres with slice thickness 1-2 mm. Exclusion criteria comprised scans with observable venous vascular pathologies (AVMs, dAVF, brain tumours, veno-occlusive diseases). Ethical clearance was obtained from the Biomedical Research Ethics Committee (BREC) at UKZN (BREC/00005683/2023).

Four CS tributaries were evaluated: 1) Superior Ophthalmic Vein (SOV): classified as Type S (straight from anterior to posterior), Type T-A (tortuous anteriorly/distally), or Type T-P (tortuous posteriorly/proximally); 2) Inferior Ophthalmic Vein (IOV): Type A (drains into SOV before CS) or Type B (drains directly into CS); 3) Sphenoparietal Sinus (SPS): Type A (drained into CS only), Type B (drained into CS and pterygoid plexus via SEVs), or Type C (hypoplastic/poorly formed); 4) Superficial Middle Cerebral Vein (SMCV): Type A (drained into SPS before CS), Type B (drained laterally into CS), Type C (connected to SEVs and pterygoid plexus), or Type D (drained into superior petrosal/transverse sinus posteriorly).

Stata version 17 (StataCorp, College Station, Texas, USA) statistical software was used for analysis with significance level set at p < 0.05. Descriptive statistics, Chi-square tests, Mann-Whitney, and Kruskal-Wallis non-parametric tests were performed.

Three morphological patterns of the SOV were observed: Type S, Type T-A, and Type T-P (Figure 1). Bilaterally identical configurations occurred in 50% of the population: 38% Type S, 7% Type T-A, and 5% Type T-P (Table 1). Type S was the most prevalent across both Black SA (50.7% left, 59.2% right) and Indian/White SA (62.1% left, 65.5% right), as well as across sexes (females: 61.4% left, 65.9% right; males: 48.2% left, 57.1% right) (Table 2).

Two drainage patterns of the IOV were observed: Type A (draining into SOV) and Type B (draining directly into CS) (Figure 2). Forty-two percent of patients had Type A bilaterally, while 24% had Type B bilaterally (Table 3). Type A predominated on both left (58.0%) and right (60.0%) sides. A statistically significant association was observed on the right side for sex (p = 0.03), where males had higher Type A (69.6%) and females had higher Type B (52.3%) (Table 4).

Three drainage patterns of the SPS were observed: Type A (drained into CS only), Type B (drained into CS and pterygoid plexus), and Type C (hypoplastic) (Figure 3). Fifty-four percent of patients had bilaterally identical configurations (37% Type A, 15% Type B, 2% Type C) (Table 5). Type A predominated on both left (62.0%) and right (55.0%) sides, followed by Type B (29.0% left, 39.0% right), while Type C was least common (9.0% left, 6.0% right) (Table 6). Indian/White individuals exhibited no Type C cases bilaterally.

Four drainage patterns of the SMCV were observed: Type A (drained into SPS), Type B (drained laterally into CS), Type C (connected to SEVs and pterygoid plexus), and Type D (drained into superior petrosal/transverse sinus) (Figure 4). Thirty-six percent of patients exhibited bilaterally identical patterns (7% Type A, 3% Type B, 20% Type C, 6% Type D) (Table 7). Type C was the most prevalent overall on both left (47.0%) and right (38.0%) sides, followed by Type A and Type B (each 19.0% left, 23.0% right), while Type D was least frequent (15.0% left, 16.0% right) (Table 8).

Several studies have observed the tortuosity of the SOV, which sometimes becomes an obstruction in the access of the CS using the transvenous approach [12,14,22-24]. The study by Kurata et al. [22] observed the tortuosity of the SOV in three of the ten patients who were examined. The tortuosity of the SOV was observed distally on the left side in those three patients [22]. This is in line with the current study, as the left side had a high incidence of SOV with Type T-A (tortuous anteriorly) compared to the right side in sex, population, and age. Both sides can have different morphological patterns of the SOV; therefore, assessment of the SOV is crucial before performing transvenous embolization.

In a study by Wolfe et al. [23], out of the ten patients that were studied, one CT presented with tortuosity of the SOV on the left side, which resulted in the rupture of the vein during the direct surgical SOV approach as well as the retrograde transvenous embolization procedure. Despite this incident, they still recommend the use of the SOV as an alternative in accessing the CS by direct surgical cannulation if the IPS, superior petrosal sinus, and the intercavernous sinus are obstructed [23]. This agrees with the present study as the tortuosity of the SOV both anteriorly and posteriorly was less compared to the straight pattern of the SOV. This implies that the SOV can be used for transvenous embolisation in the SA population as this will limit the amount of venous rupture and intracranial bleeding during surgery.

Additionally, the study by Tsutsumi et al. [12] observed the whole course of the SOV, and its proximal part was found to be variable. Out of 46 patients examined, five (11%) patients had a tortuous proximal segment on the right side, and three (6.5%) patients presented with a tortuous proximal segment on the left side. This also agrees with the present study, as variations between the left and right sides were evident. This suggests that each CT may present with different patterns of the SOV on the left and right sides.

Venous sinuses are often assumed to facilitate the spread of infection from extracranial to intracranial by the SOV and IOV to the CS, which in turn leads to life-threatening septic cavernous thrombosis [26-28]. The IOV is said to converge with the SOV before draining into the CS [11,26,29]. Zhang and Stringer [26] observed a similar drainage pattern of the IOV in their study; however, the prevalence of these findings was not reported. Cheung and McNab [11] also observed a similar pattern in three out of ten orbital dissections.

This is in agreement with the current study as the prevalence of Type A was high when compared to Type B in both population groups and both sexes, with a statistical significance of p = 0.03 found on the right for sex. It is worth noting that Cheung and McNab (2003) observed the IOV draining directly into the CS, bypassing the SOV in two orbital dissections. Similar to the current study, low incidents of Type B were obtained for population, sex, and age. This implies that most individuals in SA are likely to present with Type A, and this can result in complications like embolisation of unintended areas as well as damage to surrounding structures like rupture of the SOV.

The SPS has been identified as a venous channel found under the lesser wing of the sphenoid bone, and it receives drainage from the SMCV [30; 31]. The relationship between the SPS and the SMCV in terms of their drainage has been studied previously [9,13,30]. However, these studies were more concerned about the drainage of the SMCV and less on the SPS.

Tanoue et al. [9] observed three different patterns of the SPS which were also observed in the current study. Out of the 37 patients that were assessed, Type A (72%) was the most prevalent, followed by Type C (14%), and Type B (4%) was the least prevalent in the Japanese study [9]. The current study agrees with the Tanoue et al., [9] as Type A was the most prevalent in all the parameters (population, sex, and age) studied on both the left and right sides. However, Type A was followed by Type B, and fewer incidences were recorded in the present study for Type C. It is important to note that the Indian/White sample group did not present with Type C on both the left and right sides. This assumes that it is possible to access the CS using the SPS in the Indian/White sample group, as the course of the SPS was observed in most patients coursing from the pterion along the lesser wing of the sphenoid to its termination in the CS (Type A).

The SMCV is one of the venous channels essential in the transvenous embolisation of the cavernous sinus dural arteriovenous fistula (CSDAF) [31]. It has been observed in previous studies, and it was said to drain either to the SPS before draining into the CS or just directly to the CS (9; 13; 31]. However, some authors have noted several drainage patterns of the SMCV to the CS. A study of 24 patients observed four drainage patterns where the SMCV drained into the anterolateral aspect of the CS in 35.4% of the patients; drained into the foramen ovale in 10.4% of the patients; 16.7% of the population had their SMCV draining into the lateral aspect of the CS and in 35.4% of the population they could not identify the SMCV [31]. Suzuki and Matsumoto [13] and Ikushima et al. [33] observed seven drainage patterns of the SMCV, which were classified according to their termination points as SPS, CS, Superior petrosal sinus, basal type, squamosal type, as well as the underdeveloped type.

The studies by Suzuki and Matsumoto [13] and Ikushima et al. [33] observed a high prevalence of the SPS type in their studies - 58% and 54% respectively. Tanoue et al. [9] observed four drainage patterns of the SMCV as observed by the present study and were classified into Type A - SPS, Type B - CS, Type C - SEV, and Type D - Superior petrosal sinus. Type A was the most prevalent in their study with Type D being the least recorded [9]. Contrary to the present study, Type C was the most prevalent, but Type D was the least recorded similar to previous studies, with notable differences on the left and right sides.

There were similar findings for Types A and B for females and males on the left side (18.2% and 19.6%, respectively). Similarly, in previous studies, patients with the same or mixed types on the left and right sides were observed as indicated in Table 7 [9]. This implies that the pattern of the SMCV may be challenging to use for transvenous access to the CS as it is small and variable. It can pose a high risk to surrounding structures like the middle meningeal artery, but it can be used as a potential access point for malformations and CSDAF. However, this will depend on individual patients' anatomy as the SMCV may not drain directly to the CS but drain to the pterygoid plexus through the SEVs and to the superior petrosal sinus/transverse sinus.