Old Wine in a New Bottle: Vasograde Scale and Biomarkers, Composite Model in Subarachnoid Hemorrhage Authors: Balkrishna D Nimavat, Kapil G Zirpe Published: 2025 ABSTRACT AbstractHow to cite this article: Nimavat BD, Zirpe KG. Old Wine in a New Bottle: Vasograde Scale and Biomarkers, Composite Model in Subarachnoid Hemorrhage. Indian J Crit Care Med 2025;29(12):979–982.Keywords: Biomarkers, Delayed cerebral ischemia, Inflammation, Prediction tools, Subarachnoid hemorrhage, VASOGRADE, Delayed cerebral ischemia In Search of a “Pitch-perfect” Scoring System Achieving a pitch-perfect scoring system – one with excellent discrimination, objectivity, ease of use, and clinical decision-making utility is central in health care and critical care medicine. Despite significant medical advancements, the mortality and morbidity associated with subarachnoid hemorrhage (SAH) remain high. Delayed cerebral ischemia (DCI) is the major contributor to poor outcomes in SAH. There is now a paradigm shift in understanding the pathogenesis of DCI: From large vessel vasospasm (macrocirculation) to a broader multifactorial process involving cerebral autoregulation failure, microthrombosis, neuroinflammation, and neurovascular uncoupling. Several scoring systems have been developed to predict DCI risk and functional outcome in SAH, based on either clinical or radiological findings. Among them, Hunt and Hess (H-H), World Federation of Neurosurgeons (WFNSs), and modified Fisher score (mFS) are widely used.1 Hunt-Hess and WFNS purely depend on clinical findings, whereas mFS depends on computed tomography (CT) findings (imaging-based). Area under the receiver operating characteristic curve (AUC) of H-H and WFNS scoring system is around 0.77–0.78 (considered good but not excellent) for predicting in-hospital mortality, while mFS performs lower at approximately 0.67.1 Table 1 shows the classification of the scoring system in SAH. Table 1Classification of scoring system in SAH1–7 Based on clinical parameters Based on radiological parameters Based on clinical + radiological parameters H-H1Modified Fisher scale (mFS)1VASOGRADE1WFNSs1Hijdra score2Subarachnoid Hemorrhage International Trialists (SAHIT) score3Yasargil Grading System4SAH early brain edema score (SEBES)1SAH score or HAIR score (H-H, age, intraventricular hemorrhage IVH and re-bleeding within 24 hours)1Johns Hopkins GCS Grading Scale4Early DCI (EDCI) score (WFNS, mFS, SEBES, and Intraventricular hemorrhage IVH)5And many more!Ogilvy and Carter Combination Scale6Shunt-dependency in Symptomatic Aneurysmal SAH (SDASH) Score (to predict shunt-dependency)7EDCI score, early DCI score; GCS, Glasgow Coma Scale; HAIR, Hunt and Hess+ age+ intraventricular hemorrhage+ re-bleeding in 24 hours; H-H, Hunt and Hess; IVH, intraventricular hemorrhage; mFS, modified Fisher scale; SAH, subarachnoid hemorrhage; SDASH, shunt-dependency in symptomatic aneurysmal subarachnoid hemorrhage; SEBES, subarachnoid hemorrhage early brain edema score; WFNS, World Federation of Neurosurgeons Classification of scoring system in SAH1–7 EDCI score, early DCI score; GCS, Glasgow Coma Scale; HAIR, Hunt and Hess+ age+ intraventricular hemorrhage+ re-bleeding in 24 hours; H-H, Hunt and Hess; IVH, intraventricular hemorrhage; mFS, modified Fisher scale; SAH, subarachnoid hemorrhage; SDASH, shunt-dependency in symptomatic aneurysmal subarachnoid hemorrhage; SEBES, subarachnoid hemorrhage early brain edema score; WFNS, World Federation of Neurosurgeons “More is Better” – Merger of Clinical and Radiological scores In the quest for a more accurate prediction of DCI, Dupont SA et al. combined social, clinical, and radiological information to predict vasospasm in SAH. They observed that the Hijdra sum score and history of smoking were strong predictors of cerebral vasospasm and were superior to the mFS.2 Crobeddu et al. identified a model that could predict SAH patients who are at low risk of DCI, facilitating early transfer to the ward. Their model consisted of age, WFNS at different timings, mFS, presence or absence of hematoma, and site of aneurysm.8 de Rooij NK et al. combined clinical and radiological findings. They categorize the possibility of DCI as high (>40%) and low risk (<20%) based on clinical condition at time of admission, CT blood burden (cisternal and intraventricular), and age. The discriminatory ability of this approach was fair, with an AUC of 0.69.9 de Oliveira Manoel et al. developed a simple and reliable grading scale (labeled as VASOGRADE) for SAH to predict DCI and risk stratification. A summary of that article is projected in Table 2.10 They further validated the VASOGRADE scale for prediction of DCI with an AUC of 0.69 with good calibration.11 de Oliveira Souza NV et al., in their retrospective cohort study, linked the VASOGRADE score for predicting functional outcome. They proposed practical utilities of the risk-chart/VASOGRADE scale, including tailoring of monitoring, aggressiveness, and decision of disposition. VASOGRADE-green was associated with a good clinical outcome, while yellow/red were associated with a high risk of DCI, and transfer to high-volume aSAH centers was recommended.12 Table 2VASOGRADE grading scale for predicting DCI in SAH VASOGRADE scale Combined WFNS + mFS Risk of DCI Outcome length of stay (LOS) ICU LOS Hospital LOS Green1–2 + 1–210.5%Less (7.95 days)Less (12.27 days)Yellow1–3 + 3–431.1%More (12.97–13.09 days)More (19.03–20.39 days)Red4–5 + ANY52.8%WFNS, World Federation of Neurosurgeons VASOGRADE grading scale for predicting DCI in SAH WFNS, World Federation of Neurosurgeons Other scoring systems in SAH where they used combined clinical+ radiological criteria are HAIR (prediction of in-hospital mortality), SAHIT, SAH Score, and EDCI. VASOGRADE Scale: Simple Composite Bedside Scale VASOGRADE scale is a composite score incorporating clinical and radiological criteria to estimate the risk of DCI and functional outcome in SAH. It employs the red, amber, green (RAG) rating (simplest example is traffic light: Red, Amber, and Green) to categorize disease severity. It helps to triage urgent priority/high-risk vs low-risk groups. VASOGRADE scale is easy to use and provides a homogeneous platform to communicate among healthcare workers. Area under the AUC of the VASOGRADE scale for predicting poor outcome was 0.63 in the original study (>0.7 in recent literature).1 Oversimplification of Complex Issue! Although composite scoring systems are often perceived as superior, this is not always true. There are lots of downsides to the VASOGRADE scale, including: (1) Oversimplification of a complex issue – RAG scale seems easy to use as a trichotomous scale but does not account for dynamic changes in clinical condition and spectrum of disease process, similar to the difference between organ failure (dichotomous) vs organ dysfunction (spectrum). (2) It is the sugar coating/repackaging of a basic clinical approach that most clinicians already know and do by clinical assessment (like focal deficit and GCS/WFNS score) and radiological findings (intraventricular involvement/ mFS score). (3) Prediction/AUC varies depending on the intervention (post-coiling vs post-clipping). In clipping patients, the VASOGRADE score has one of the best predictive values compared to other scoring systems (mFS, H-H, WFNS, SEBES), and while in the coiling group, no score showed predictive accuracy/statistical significance. (4) Clinical score and combined score have better prediction for poor outcome compared to the radiological score. Surprisingly, the combined score does not show superiority over the clinical score.1 It is also noteworthy that the value of AUC of any scoring system for predicting poor outcome is different from that of AUC for predicting DCI. Composite Model with Biomarkers The complexity of DCI suggests the need for multimodal approaches. With a better understanding of DCI pathogenesis, the role of neuroinflammation in the pathophysiology of SAH/DCI is now increasingly recognized. It is worth exploring the role of inflammatory biomarkers in SAH for the diagnosis of DCI, prognosis, and to identify the phenotype with the right modalities/interventions. The concept of merging clinical, radiological, imaging, and biomarkers as a composite model is not new in critical care. It has been tried in many other medical illnesses like sepsis, acute respiratory distress syndrome (ARDS), pancreatitis, and acute coronary syndrome (ACS). Many systemic biomarkers, like high glucose, high lactate, low albumin, high PCT, and high D-dimer have been associated with poor outcomes in SAH.13 Considering they are non-specific in acute-illness and sepsis settings, search for more specific neurological biomarkers emerged. At least 50 such biomarkers have been studied in SAH for the prediction of DCI and poor outcome. Copeptin, neurofilament proteins (neurofilament light NfL), S100B, microRNAs, kallikrein, glial fibrillary acidic protein (GFAP), and IL-6 are some such potential neuro-biomarkers in SAH.14 Table 3 summarizes sets of biomarkers and their role in SAH. Adding biomarkers to a composite model results in another set of concerns-Insufficient specificity and sensitivity of the biomarkers, lack of standardization, optimal cut-off to discriminate outcome, availability/feasibility, and single biomarker vs sets of biomarkers. Yi et al. showed an association of high systemic inflammatory response index (SIRI) and systemic immune inflammation (SII) Index at day 5–7 with the occurrence of DCI.15 Hao Y et al in their meta-analysis showed that SIRI and SII are associated with unfavorable outcomes, but surprisingly, SIRI was not a strong predictor of DCI (while SII demonstrated an association with DCI).16 Sharma et al. took a set of biomarkers (Indices like SIRI and systemic immune-inflammatory index SII) + VASOGRADE scale to predict DCI.17 They proposed VASS DCI scoring (VASOGRADE, age, SIRI, SII for prediction of DCI) for prediction of DCI. Area under the AUC observed by VASS DCI seems modest in their findings. Strong suits of such an approach are as follows: (1) Easy to use and (2) Bedside feasibility (as no fancy biomarker/test needed) with comparable AUC with the available scoring system. But again, it is associated with a fundamental downside of biomarkers, i.e., optimal cut-off SIRI/SII, other confounding factors in the ICU, like sepsis, pneumonia, and surgery. SIRI and SII need more validation on a prospective trial with an optimal cut-off. Table 3Sets of biomarkers and role in prognostication of SAH Sets of biomarkers Component of a set Relation with DCI/SAH functional outcome Low-grade inflammation score (LGIS)C-reactive protein, white blood cell count, platelet count, neutrophil to lymphocyte ratio (NLR)Higher LGIS – poor outcome Higher LGIS – higher chances of symptomatic cerebral vasospasm (SCVS)18Systemic inflammatory response index (SIRI)Neutrophil, monocyte, and lymphocyte countsValue of SIRI and SII at 5–7 days is significantly higher in the DCI group compared to the non-DCI group15Systemic immune inflammation indexNeutrophil, platelet, and lymphocyte countsHigh SIRI and SII are associated with poor outcomes in SAH, but only SII is predictor for DCI (Not SIRI)16Pan-immune-inflammation value (PIV)Neutrophil, monocyte, platelet, and lymphocyte countsHigh-PIV independent predictor of DCI19 Sets of biomarkers and role in prognostication of SAH VASS DCI SAH: A Curate's Egg The concept of composite model (here VASS DCI SAH) is conceptually appealing – holistic, simple, and feasible at the bedside. Yet it remains in an early developmental phase, requiring biomarker validation, standardized thresholds, and evaluation in a prospective trial setting. In future, with clear cut-offs and phenotypic therapeutic implications, they may become transformative. Currently, biomarker-augmented scores risk diluting the strengths of established tools like VASOGRADE. Take-home Message: Bounded Rationality – Satisfice Rather than Optimize! Level of recommendation and making guidelines or protocols depends on the feasibility of tests, cost, wider applicability, apart from vigorous griding of available evidence. VASOGRADE/VASS DCI SAH scale passes this test with flying colors, but simplification comes with the cost of losing granular data and a lack of trajectory of illness over time. This criticism is not directed solely at VASOGRADE (and other composite models/scores), as this is a problem with bounded rationality associated with our knowledge about SAH/DCI. Derek Angus, in his article, gave an example of the “sepsis” construct and bounded rationality.20 Here, in SAH/DCI, we have a considerable knowledge gap, and finding simple solutions (like VASS DCI SAH/VASOGRADE) seem like bounded rationality (personal/editorial opinion). Orcid Balkrishna D Nimavat https://orcid.org/0000-0001-8648-9289 Kapil G Zirpe https://orcid.org/0000-0002-8140-727X Footnotes Source of support: Nil Conflict of interest: None References