Developmental venous anomalies: appearance on whole-brain CT digital subtraction angiography and CT perfusion

This study demonstrates that CTP may permit assessment of the hemodynamic alterations associated with DVA. For each DVA in this study, absolute CBV and CBF values were significantly increased in the affected hemisphere as compared to the unaffected hemisphere. On MR perfusion, Camacho et al. [12] also identified increased CBF, CBV, and MTT in four uncomplicated DVA cases. The authors concluded that hemodynamic alterations were not associated with the presenting symptoms of the patients and that the alterations did not imply a more ominous future outcome for the DVA. The results of this study suggested that increased regional CBV and CBF on CTP may indicate the presence of a DVA on CTP of the brain and case 1 presented with facial numbness symptoms that may have been influenced by the presence of the DVA. Previous studies have identified decreased CBF values in the parenchyma near two large DVA on xenon-enhanced CT [21] and in one patient on brain perfusion SPECT that the authors believed was correlated with the SPECT abnormalities, DVA location, and an electroencephalograph focus [22]. However, differentiation of the intravascular hemodynamic changes and the surrounding parenchymal perfusion changes using the ROI approach in this study is limited due to the averaging of the CTP values within the ROI and future studies using a pixel-by-pixel evaluation of the DVA and the surrounding parenchyma could provide an answer.

Although MTT was consistently elevated and suggestive of delayed or slow flow in this small group of five patients, further study is required to determine how the whole-brain hemodynamic (i.e., CBV and CBF) and temporal CTP values (i.e., MTT, TTP, and delay) within the DVA can be utilized to further delineate the hemodynamic parameters for a larger collection of DVA cases. Because MTT is an indicator of perfusion pressure, the higher MTT values in the DVA indicated lower flow or perfusion and increased MTT values have previously been associated with DVA in four patients on MR perfusion [12]. The unaffected hemispheres of each case appeared to be influencing the relative percentage change in MTT values, but because of insignificant p values and the small sample size, evaluation of trends in MTT values in patients with a DVA over time will require further research.

TTP results for DVA cases must be interpreted in the context of each individual patient. TTP was not indicative of DVA in this study. This was possibly due to the wide range of DVA types that can both increase and decrease the time for blood to get to a region. TTP is a temporal parameter that can be falsely prolonged depending on technical factors of the scanning process and by the size and shape of the arterial input function. Cases 1, 3, and 4 had higher relative TTPs in the affected hemisphere (range 1.99-6.45%) that suggested slower arrival of contrast to the DVA. This may be anticipated in slow flow venous structures with lower proportions of arterial influence on TTP values. For cases 2 and 5, rapid flow and lower relative TTP (−4.09% in case 5 to −7.43% in case 2) in the affected hemisphere could be interpreted as contribution from a possible arteriovenous shunting component but this could not be confirmed and no arterial abnormalities were noted on CT DSA for cases 2 and 5. Further DVA case analyses will be required to evaluate if TTP may serve as a stratifier for CTP pressures in DVA and a possible CTP parameter that may, in combination with the CBV, CBF, and MTT results, help to delineate an individual’s risk of future complications.

Flow-related complications of DVA that may lead to a symptomatic presentation include an increase of inflow because of an arteriovenous shunt, a decrease of outflow, or a remote shunt with increased venous pressure [10]. In our study, case 1 presented with symptoms that was possibly secondary to a DVA (Fig. 5). Pereira et al. suggested characteristics of an increased risk of symptomatic DVA presentation: large and complex DVA with changes on MRI suggesting venous congestion, acute or subacute ischemic changes, asymmetrical medullary zone appearance, and association with AVMs [10]. Mechanical compression of intracranial structures around the DVA produced symptoms in 20.3% (14 of 69 cases). Unexplained symptomatology occurred in 8.7% (six of 69 cases) while the remaining 71% (49 of 69 cases) of symptomatic DVA were attributed to flow-related pathomechanisms (19 with increased inflow, 26 with decreased outflow, and four with a shunt that increased venous pressure) [10]. The findings of the present study suggest that anatomical variations in DVA that can influence hemodynamic flow may be characterized by the combination of assessment by CTA-V (Fig. 4d), whole-brain dynamic volume CTP, and 4D DSA.

As reviewed by Hammoud et al. [13], nonhemorrhagic ischemia associated with DVA is uncommon and rarely reported in the literature. A review by Hon et al. [23] of 15 studies that included 714 patients with DVA found that less than 1% were associated with infarction on the initial presentation. The etiology of nonhemorrhagic infarcts associated with DVA has been proposed as thrombosis of the collecting or draining vein [13] and has been identified on angiography [24]. Stenosis at the dural opening of DVA is also frequently noted and this has been proposed as a mechanism for initiating a DVA thrombosis [4, 25]. In this study, no case had clinically discernible stenosis, but case 3 had multiple terminal draining veins that lead into a small dural opening (Fig. 4a, d). DVA that demonstrate thrombotic complications are treated the same as cortical venous or venous sinus thromboses. An identical laboratory workup is required and consideration for treatment with anticoagulation is appropriate. Investigation of procoagulating factors or prothrombotic conditions is also important [10]. Regardless of the treatment method implemented, it is important that the integrity of the DVA be preserved to avoid the substantial neurologic complications previously documented when it is excised either inadvertently or due to clinical necessity [10, 13, 26]. DVA-associated hemorrhagic and nonhemorrhagic infarctions are rare, but implementation of prevention measures upon identification, such as hydration in warm climates, smoking cessation, or anticoagulant medications for individuals at increased risk of thrombosis as indicated may reduce risk of adverse events.

Although DVA complications are rare, the complications of DVA are more frequently hemorrhagic than ischemic in nature [25, 27, 28]. The Hon et al. [23] review found that 6% with DVA presented with symptomatic hemorrhage and the hemorrhage rate following the first event was 0-1.28% per year for 422 of these 714 patients. In a prospective study by the same authors on 93 adults, one patient presented with symptomatic hemorrhage and there were no symptomatic hemorrhages following the first presentation for a total of 492 person-years of follow-up [23]. Formation of a thrombus within a DVA has also been proposed as a cause of DVA-associated hemorrhage secondary to increased venous pressure [27]. Additionally, DVA hemorrhages may not result from the DVA itself but may be associated with co-occurring or developing anomalies including CM, capillary telangiectasias, and AVMs. Hemorrhage of DVA with an associated CM has been well-documented [1, 7, 9, 16, 29, 30] and Hong et al. [14] described angioarchitectural factors of DVA that may cause concurrent sporadic CM to form. Further evaluation is required to determine if DVA progression to capillary telangiectasias and CMs or AVM proximity to DVAs can be longitudinally monitored with CTP.

Previous CT technology has not allowed whole-brain visualization and as demonstrated in this study, comprehensive DVA analysis requires whole-brain CTP for visualization of the peripheral or superior cerebral cortices (case 1 in Fig. 5; cases 2, 4, and 5) or in the posterior fossa (case 3 in Fig. 4).

With the recent scientific advances in understanding of the pathophysiology [10] and angioarchitectural factors [14] of DVA coupled with the increasing clinical availability of whole-brain CTP, this preliminary study data indicates that the incremental diagnostic information for DVA can be provided by whole-brain CTP studies. Further investigations are required to use CTP as a method for identifying individuals at greatest risk for adverse clinical outcomes related to DVA. Additionally, it is possible that assessment of the CTP parameters associated with DVA may help to avoid incorrect diagnoses such as arterial infarcts, focal hemorrhage, demyelinating disease [31], neoplasm, or similar appearing abnormalities on CT [13]. Misdiagnosis of DVA and DVA-associated infarcts has been documented and may be secondary to the heterogeneous imaging characteristics of cerebral venous infarctions [32]. Additionally, DVA are frequently not identified on other imaging studies including non-contrast CT and structural MRI but may be detected with CTP, as was true of case 5 in this study (Table 1).

In order to assess the risk associated with diagnostic procedures, both physical risk and radiation dose must be considered. Due to its interventional nature, conventional angiography has a 0.5% risk of permanent neurologic complication [33] and a typical effective radiation dose of 10.6 mSv [34]. The whole-brain CT imaging protocol used in this study, that included CT DSA and CTP, had an effective dose of 4.3-4.4 mSv (Table 2). Effective radiation dose for 64-slice CTP has been reported as 7.5-11.4 mSv [34]. The whole-brain CTP protocol used in this study is lower than conventional angiography or 64-slice CT and is less than one third of the level reported as acceptable by the FDA for brain perfusion imaging [35]. As with any imaging study, the benefit of the diagnostic information acquired must always be weighed versus the radiation risk.

There were limitations of this study. CTP value variation using ROIs and placement of the ROIs near vessels or artifacts can adversely impact the clinical interpretation of results. Variation with the 8 mm ROIs was increased for two reasons: (1) a HU densities in the ROI are averaged in the vicinity of each voxel together and (2) when the ROI is placed in areas with vessels, the flow and volumes are significantly higher. CTP values obtained within the area of the DVA and in the contralateral hemisphere have different compositions of parenchymal and vascular components and this variability could not be quantified. Vascular and tissue differentiation could be improved by pixel-by-pixel CTP value evaluations and this could increase precision in defining the range of CTP values at important DVA structures, such as the terminal vein opening. Additionally, no cases had angiographical confirmation of associated pathologies. Finally, expanded CTP studies with higher numbers of patients using standardized system configurations (e.g., hardware, software, and protocols) with human variables (e.g., estimates of cardiac output and hematocrit) incorporated into the DICOM files may help to facilitate individual CTP analyses, as well as comparisons between individuals.

Most DVA are incidental findings on imaging studies and little is known about which individuals will progress to hemorrhagic or ischemic complications. The cases presented suggest that the physiologic appearance of DVA on CT DSA and whole-brain CTP can provide incremental clinical information to practitioners for diagnostic purposes and with future research may help to provide some of the monitoring and prognostic data required to identify the rare subset of patients with DVA that are at increased risk of adverse outcomes.