Critical shift phase in deep vein thrombosis development after major sellar region tumor surgeries: a systematic review.

Introduction
Major sellar region tumors, such as pituitary adenomas, craniopharyngiomas, tuberculum sellae meningiomas, chordomas, and cystic lesions, present considerable surgical difficulties due to their close association with vital neurovascular structures, endocrine systems, and cerebrospinal fluid (CSF) pathways [1].
Although modern microsurgical and endoscopic methods have enhanced resection results, these interventions cause significant physiological disruptions, particularly related to coagulation homeostasis. A significant postoperative concern after major sellar tumour surgeries is venous thromboembolism (VTE), which includes deep vein thrombosis (DVT) and pulmonary embolism (PE). Recent investigations have revealed a heightened occurrence of venous thromboembolism in these patients, with a notably elevated risk manifesting during the early postoperative phase [1]. Some studies have indicated that the incidence of patients with brain tumours experiencing venous thrombosis following surgery, in the absence of anticoagulant prophylaxis, varied between 3% and 30% [2–4]. This hypercoagulable state is shaped by various factors, such as extended operative duration, tumour biology, surgical trauma, and postoperative endocrine dysfunction [5].
The pathophysiology of VTE following the resection of sellar tumours is complex and influenced by multiple factors. The surgical approach to the sellar region can significantly impact the hypothalamic-pituitary axis, resulting in notable endocrine disturbances that affect coagulation pathways. The shift from central diabetes insipidus (CDI) to the syndrome of inappropriate antidiuretic hormone secretion (SIADH) is a documented postoperative occurrence that can disrupt fluid-electrolyte equilibrium, leading to endothelial dysfunction and a heightened risk of thrombosis [5].
Furthermore, the inflammatory and systemic stress responses that occur after tumour resection exacerbate a prothrombotic condition, especially in scenarios that involve significant surgical dissection or CSF leakage. Although the risks are well-recognized, the exact duration and mechanisms that contribute to the shift to a hypercoagulable state remain unclear. Moreover, the specific impact of perioperative sodium fluctuations and their direct influence on coagulation dynamics also arise as an issue that warrants further investigation [6].
Therefore, this study seeks to answer the research question: ‘During which postoperative phase are patients undergoing major sellar region tumor resection most vulnerable to developing deep vein thrombosis, and what pathophysiologic mechanisms contribute to this risk?’ Accodingly, this study seeks to conduct a systematic review of how fluid-electrolyte imbalances, specifically during the transition from CDI to SIADH, influence postoperative deep vein thrombosis (DVT) outcomes in patients undergoing surgery for major sellar region tumors. Considering the critical link among perioperative sodium imbalances, vascular stability, and the activation of coagulation, it is necessary to gain a more profound insight into this phenomenon to enhance thromboprophylactic approaches.

Methods
A thorough and systematic literature search was performed across three notable scientific databases including PubMed, Cochrane Library, and ScienceDirect in October 2024 until February 2025, adhering to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. No restrictions were applied regarding publication year to ensure comprehensive inclusion of relevant studies. The search strategy was carefully planned to pinpoint relevant research examining the risk of DVT occurrence in patients who are undergoing surgical resection of significant tumours located in the sellar region. Terms and keywords pertaining to ‘sellar tumors,’ ‘deep vein thrombosis,’ ‘venous thromboembolism,’ ‘craniopharyngioma surgery,’ ‘pituitary surgery,’ and ‘postoperative complications’ were utilised in broad combinations to guarantee the comprehensive inclusion of all pertinent articles. Boolean operators (‘AND,’ ‘OR,’ and ‘NOT’) were used to combine Medical Subject Headings (MeSH) and free-text terms, ensuring both precision and breadth. The final search strategy included the following key combinations: (‘Pituitary Neoplasms’ OR ‘Craniopharyngioma’ OR ‘Skull Base Neoplasms’ OR ‘Sella Turcica’) AND (‘Deep Vein Thrombosis’ OR ‘Venous Thromboembolism’) AND (‘Surgery’ OR ‘Postoperative Complications’). To improve the sensitivity of the search, reference lists of eligible studies and previously published reviews were carefully reviewed for supplementary literature. This systematic review has been registered in the PROSPERO database (CRD42024616487) to maintain methodological transparency.
Inclusion criteria were defined to include prospective, retrospective cohort, and randomized controlled trials (RCTs) studies that specifically evaluated the risk of DVT in patients undergoing surgical procedures for significant sellar region tumours. Research examining pituitary adenomas, craniopharyngiomas, chordomas, or other substantial sellar-based neoplasms considered suitable, as long as it detailed postoperative thrombotic outcomes, encompassing DVT incidence, onset timing, and related risk factors. Exclusions were made for case reports, case series, non-human studies, editorials, and review articles. Research focussing on the non-invasive treatment of sellar tumours, as well as studies lacking specific thromboembolic outcome data, were also excluded from consideration. Both adult and pediatric populations were included to capture the full spectrum of postoperative neuroendocrine and hemostatic changes across age groups.
Systematic reviews, narrative reviews, and case reports were excluded, as they do not provide primary data relevant to the study objectives. Exclusions were also made for non-human studies and editorials. Research focusing on the non-invasive treatment of sellar tumours, as well as studies lacking specific thromboembolic outcome data, were excluded from consideration.
The Newcastle-Ottawa Scale (NOS) was utilised to evaluate the risk of bias in the cohort studies included in the analysis. The NOS assesses three key domains: the selection of study participants, the comparability of study groups, and the evaluation of outcomes, with a maximum score of 9 reflecting high methodological rigour. The bias assessment was carried out by two independent reviewers, with any discrepancies addressed through a discussion involving a third reviewer. Research studies that achieved NOS scores of 7 or above were categorised as presenting a low risk of bias. On the other hand, studies with scores falling below 5 were identified as having a high risk and were subsequently subjected to more rigorous examination in sensitivity analyses.
Data extraction was conducted using a standardized data collection form, and the extracted information was compiled into a structured database. Two independent reviewers (CTS and ABSS) performed data extraction, documenting essential study characteristics such as sample size, DVT incidence, perioperative fluid-electrolyte disturbances, timing of DVT onset, and other risk or predictive factors. A third reviewer (DPWW) verified the accuracy of all extracted data, and any discrepancies were resolved by discussion and consensus. This multi-step verification ensured the reliability and reproducibility of data handling.
A comprehensive synthesis was performed to encapsulate trends in the incidence of DVT across various surgical and perioperative contexts, emphasising the critical shift phase in coagulation homeostasis during the transition from CDI to SIADH.

Results

Search result and study selection
A thorough literature search was conducted across PubMed, Cochrane, and Science Direct, resulting in 68 records that were relevant to the study objective. A total of 18 studies underwent initial screening after the removal of duplicate records (n = 8) and ineligible studies (n = 42). The records marked as ineligible were excluded primarily because they did not meet the predefined inclusion criteria. Most were excluded due to the absence of postoperative thrombotic outcome data, being review or editorial papers, or because they focused on non-surgical management of sellar region tumors such as medical or radiotherapeutic treatment. Case reports and small case series without adequate outcome description were also excluded.
Six studies were excluded after the title and abstract screening, resulting in 12 reports that were requested for retrieval. However, the number of studies eligible for full-text evaluation was reduced to eight, as four reports were inaccessible. Five studies were excluded during the eligibility assessment, predominantly because they did not specify the timeline of deep vein thrombosis (DVT) development (n = 3) or included other types of brain tumours (n = 2). In the end, three of the highest quality studies were chosen for the systematic review and met the inclusion criteria as shown in Figure 1.

DVT development after major sellar tumor surgeries
All studies included in this review indicated a notable postoperative thrombotic risk, with the incidence of DVT varying from 3.2% in a large multi-center cohort to 15.8% in two single-center investigations, as shown in Table 1 [1,5,7]. Although the variation in incidence can be linked to factors such as sample size, surgical technique, and perioperative care, a clear pattern has been observed: the greatest risk of DVT manifests within the initial 10 days following surgery, aligning with significant physiological changes in coagulation and fluid-electrolyte equilibrium.
The surgical approach appeared to be a critical factor in the risk of DVT, with varying techniques being applied across studies. In all three, craniotomies were performed [1,5,6]. Castelli et al. [5] included additional surgical techniques such as endoscopic resection and Ommaya reservoir placements (both closed and open), while Shi et al. [6] utilized pterional, unifrontal basal interhemispheric approach (UBIA), subfrontal, transcallosal, and combined pterional-UBIA approaches. It is important to highlight that one of the studies identified craniotomy as a significant independent risk factor for VTE development (OR 2.78, p = 0.017) and prolonged surgical duration (OR 1.78, p = 0.029) [1]. These findings suggest that more invasive procedures and longer operative duration may contribute to endothelial injury and postoperative hypercoagulability.
Fluid-electrolyte imbalances, particularly fluctuations in sodium levels and the transition from CDI to the SIADH, were significantly associated with DVT development following sellar tumor surgery. Castelli et al. [5] stated that the majority of DVT cases in their cohort occurred during periods of sodium dysregulation, with the maximum incidence reported between postoperative days 7 and 10. Shi et al. [6] also discovered that all patients who encountered vascular events had sodium concentrations that were outside the normal range. In that study, patients with pre-existing CDI that was deteriorating experienced 70.4% of thrombotic events, while 29.7% developed new-onset CDI. Additionally, postoperative ADH levels were statistically significantly increased (p < 0.005), indicating that neuroendocrine dysregulation may be a contributing factor to thrombotic risk.
The pathophysiological interaction between dysnatremia and postoperative hypercoagulability is emphasised by the correlation between sodium imbalance and DVT. Castelli et al. [5] identified the CDI-to-SIADH transition as a critical period for thrombotic events, emphasising the potential involvement of abrupt serum sodium changes in the activation of the coagulation cascade and venous stasis. This association is further supported by the results of Shi et al. [6], which show that patients who experienced vascular complications consistently displayed significant fluctuations in serum sodium levels.
Biochemical markers that were predictive of thrombotic risk were consistently elevated accross studies. D-dimer was validated as a predictive biomarker for thrombotic complications by Qiao et al. [1], who identified a strong association between postoperative D-dimer elevation and DVT incidence. Similarly, Castelli et al. [5] reported a protracted increase in C-reactive protein (CRP) in patients with thrombosis, indicating a correlation between hypercoagulability and postoperative inflammation.
In addition to biochemical markers, CSF dynamics and tumour pathology were substantial determinants of postoperative thrombotic risk. Qiao et al. [1] discovered that patients with craniopharyngiomas had a higher risk of VTE (OR 1.86, p = 0.036). This risk was likely caused by the perturbation of the hypothalamic-pituitary axis and the resulting metabolic derangements as a result of the tumour. Additionally, patients who experienced high-volume CSF leakage had an even higher risk of thrombosis (OR 4.24, p < 0.001). This suggests that postoperative CSF dynamics, such as excessive drainage and consequence haemodynamic instability, may contribute to the development of DVT.

Risk of bias evaluation
The NOS risk of bias assessment revealed that all three included studies exhibited high methodological quality, with total scores of ≥7 out of 9 as shown in Table 2. Qiao et al. [1] exhibited the highest score (9/9) in the selection, comparability, and outcome domains, suggesting a high level of methodological rigour. Shi et al. [6] and Castelli et al. [5] also received high scores (8/9 and 7/9, respectively), which indicates that the study designs were robust and there was minimal risk of bias. These two studies received slightly lower scores in comparability, indicating minor limitations in adjusting for potential confounders, despite the fact that all studies performed well in selection and outcome assessment.

Discussion
The 10-day postoperative period following major sellar tumor surgeries is a critical window for DVT development, with evidence indicating that peak incidence occurs within postoperative days 7–10, coinciding with significant fluid-electrolyte shifts. The presence of VTE in patients with intracranial tumors post-craniotomy is well-documented, with incidence rates of 10.2% (DVT, 6.8%; PE, 1.7%; and concurrent DVT and PE, 1.7%). Previous studies on VTE within neurosurgical cohorts have predominantly concentrated on general brain tumour cases, documenting inconsistent rates of DVT without identifying a distinct peak risk period [8]. Our findings identify a new, time-sensitive thrombotic risk associated with specific neuroendocrine pathway changes and postoperative biochemical disturbances.
The transition from CDI to SIADH represents a period of high thrombotic risk, as abrupt sodium fluctuations may induce endothelial dysfunction, hypercoagulability, and venous stasis. The main factors contributing to elevated DVT risk in patient post major sellar tumour surgeris are enhanced coagulation activity and extended postoperative immobilisation. Neurosurgical interventions in the sellar region cause considerable tissue damage and vascular alteration, leading to the release of tissue factor, a crucial starter of the extrinsic coagulation pathway [5]. Endothelial damage intensifies this hypercoagulable condition by enhancing the expression of von Willebrand factor and factor VIIa, which further stimulate thrombogenesis. Prolonged immobility after surgery leads to venous stasis, especially in the lower limbs, creating conditions favourable for thrombus development [5,8,9].
In addition to standard coagulation routes, postoperative neuroendocrine disruptions significantly increase the risk of thrombosis. Patients having major sellar tumor surgery often experience dysregulation of the hypothalamic-pituitary axis, resulting in sudden changes in antidiuretic hormone (ADH) output. The shift from CDI to the SIADH significantly impacts intravascular volume, blood viscosity, and endothelial function. The timing of these phases varies among individuals, with CDI typically occurring between postoperative days 1 and 7, followed by a transition to SIADH, which may emerge as early as day 2 and persist up to day 14, though it is most commonly observed between days 7 and 10 in some studies. In the CDI phase, usually occurring within the initial five postoperative days, ADH shortage results in severe diuresis, dehydration, and subsequent hemoconcentration. Hypernatraemia during this phase elevates plasma osmolarity, induces endothelial dysfunction, and enhances platelet aggregation, thereby promotes a prothrombotic condition. As SIADH develops, marked by fluid retention and dilutional hyponatremia, vascular autoregulation becomes impaired, leading to endothelial cell edema and inflammation-driven hypercoagulability. The biphasic fluid and electrolyte imbalances have a synergistic effect on thrombosis risk, elucidating the notable surge in DVT incidence seen during this crucial postoperative period, as shown in Figure 2 [1,7,10].
Sodium dysregulation, observed in CDI-SIADH transitions, affects fluid balance and modifies endothelial integrity and vascular responsiveness. Acute hypernatraemia inhibits nitric oxide (NO) production, diminishing vasodilation ability and exacerbating oxidative stress and microvascular thrombosis. Conversely, hyponatraemia disrupts endothelial cell junctions, intensifies inflammatory cytokine release, and further predisposes the vascular endothelium to thrombotic progression. These dynamic variations exert significant stress on the cerebrovascular and systemic circulation, creating a previously unrecognised connection between neuroendocrine fluctuation and hypercoagulability [1,7,10].
Resections of sellar tumours also often impair normal CSF production and absorption, hence indirectly influencing systemic haemostasis. Modified intracranial pressure dynamics affect baroreceptor-mediated vascular responses, exacerbating fluid dysregulation. Surgical disruption of the hypothalamic-pituitary-adrenal (HPA) axis results in dysregulated release of cortisol and catecholamines, both of which have direct prothrombotic consequences. The immediate postoperative hypercortisolemic condition promotes platelet activation, fibrin deposition, and endothelial dysfunction, while dysregulated catecholamine release induces vasospasm, prevents microcirculatory perfusion, and worsens regional blood flow stasis, increasing the risk of clot formation in patients [1,7,10]. In addition to perioperative risk factors, metabolic dysregulation linked to sellar tumours further increases patients’ susceptibility to thrombotic events. Pituitary adenomas frequently cause hyperprolactinemia or dysregulation of growth hormone, each having distinct impacts on systemic metabolism and coagulation [10–12].
In patients with major sellar region tumours, hormonal disturbances are frequently present even before surgical intervention, often contributing to a baseline prothrombotic state. Functioning pituitary adenomas, especially those producing growth hormone (GH), adrenocorticotropic hormone (ACTH), or prolactin, affect systemic hemostatic and metabolic processes. Excess growth hormone in acromegaly, for example, induces insulin resistance, endothelial dysfunction, and increased concentrations of pro-coagulant substances, including fibrinogen and plasminogen activator inhibitor-1 (PAI-1), all of which contribute to thrombus formation. In ACTH-secreting adenomas, the subsequent hypercortisolemia promotes coagulation through increased platelet activation, inhibition of fibrinolysis, and overexpression of clotting factors such as factor VIII and von Willebrand factor. Prolactinomas, although less commonly stated are associated with endothelial changes and low-grade inflammatory conditions that contribute to a prothrombotic phenotype [13–15].
Both functioning and non-functioning tumors may cause mass effects on the HPA, resulting in various levels of hypopituitarism. In a study of 164 individuals with non-functioning pituitary adenomas (NFPAs), 80.4% had anterior pituitary insufficiency, with hypogonadism and hypothyroidism observed in 48.4% and 36.5% of cases, respectively. GH insufficiency occurred in 27.4%, and tumor size had a high correlation with the probability of hormonal deficit, with a volume threshold of 3105.1 mm³ being indicative of hypopituitarism [16]. Preoperative hormone deficiencies lead to a vascular condition that is both metabolically impaired and fundamentally more thrombogenic, due to diminished fibrinolysis, altered vascular tone, and systemic inflammation [16–18].
This endocrine vulnerability is further heightened in pediatric patients, especially those having surgery for craniopharyngioma (CP). A study involving around 200 children treated at two tertiary centers revealed that central adrenal insufficiency and central hypothyroidism occurred in 56.2% and 70.3% of patients, respectively, during the initial postoperative week. An remarkable 85.5% required desmopressin for the management of diabetes insipidus, and 94.4% continued on a minimum of one hormone replacement during follow-up. Significantly, 74.7% of patients need three or more hormonal replacements—commonly levothyroxine, hydrocortisone, and desmopressin—while 54.4% experienced obesity, thus exacerbating their thrombotic risk [19]. Some studies highlight the substantial neuroendocrine burden experienced by CP patients, which persist beyond post-tumour resection and directly affects haemostatic equilibrium [19,20].
The hemostatic implications of these endocrine disturbances are multifaceted and clinically significant. Cortisol deficiency, particularly in the absence of timely replacement, impairs vascular tone and disrupts anti-inflammatory regulation, increasing the risk of endothelial injury. Hypothyroidism contributes to hypercoagulability by reducing fibrinolytic activity and raising fibrinogen concentrations, while hypogonadism is associated with pro-inflammatory cytokine release and endothelial dysfunction. These hormonal imbalances often converge in the early postoperative period—precisely when patients are immobilised and recovering from extensive surgical manipulation—compounding the risk of venous thromboembolism [14,16,18,19].
Besides direct hormonal influences, systemic inflammation is essential in determining the postoperative thrombotic state. The surgical manipulation of hormonally active or adherent tumours may provoke a widespread inflammatory response, exacerbated by abrupt endocrine transitions. Cytokines, including interleukin-6 and TNF-α, which are increased in this environment, promote tissue factor production and endothelial activation, hence enhancing the coagulation cascade. The inflammatory environment, together with existing hormone deficiencies and metabolic imbalances, creates a strong basis for postoperative thrombotic occurrences [21,22].
Moreover, biochemical markers are essential for predicting and monitoring the risk of DVT after complicated major sellar region tumour surgery. D-dimer serum is acknowledged as a highly sensitive indicator of VTE; nevertheless, its specificity in postoperative neurosurgical patients is limited. Postsurgical increased D-dimer serum levels frequently arise from extensive tissue damage, systemic inflammation, and fibrinolysis, complicating the distinction between normal surgical responses and actual thrombotic occurrences. Qiao et al. [1] found that postoperative D-dimer levels were markedly increased in patients who experienced VTE, with an odds ratio (OR) of 1.11 (p < 0.001), underscoring its utility as a valuable but non-specific biomarker. Activated partial thromboplastin time (aPTT) and CRP also have become important biochemical markers linked to thrombotic problems in sellar tumour procedures. A lowered aPTT (< 26s) postoperatively has been associated with an elevated risk of VTE, as indicated by Castelli et al. [5], where patients experiencing thrombotic events demonstrated markedly lower aPTT, implying a hypercoagulable condition resulting from excessive thrombin production and fibrin synthesis. This discovery corresponds with the hypercoagulability paradigm noted in neurosurgical patients, wherein significant vascular manipulation and endothelial damage trigger the coagulation cascade. Simultaneously, prolonged postoperative CRP rise (> 10 mg/L) has been associated with an increased risk of thrombosis, suggesting a persistent inflammatory response that aggravates endothelial dysfunction and facilitates thrombus formation. A comprehensive assessment of D-dimer, aPTT, and CRP may improve risk classification and facilitate early identification of VTE.
The relationship between electrolyte imbalances and thrombotic risk has been acknowledged as a significant component in the development of postoperative VTE. Variations in sodium levels, especially the shift from hypernatraemia in CDI to hyponatraemia in SIADH, have been significantly associated with a heightened risk of thrombosis. Moreover, although conventional risk assessment instruments like the Padua and Caprini scores are extensively utilised in medical and surgical cohorts, their utilization in neurosurgery patients is constrained. Recent studies indicate that these ratings often overstate VTE risk because they depend on generic risk factors and exclude neurosurgery-specific elements such as electrolyte imbalances, neuroendocrine disruptions, and variations in intracranial pressure [4,7,9,23].
The interaction between postoperative neuroendocrine changes and coagulation mechanisms requires individualized thromboprophylaxis approaches. Pharmacologic prophylaxis, comprising low molecular weight heparin (LMWH) or unfractionated heparin (UFH), is the standard; however, its timing and dosage require careful adjustment to optimise efficacy while mitigating the risk of cerebral haemorrhage. A meta-analysis conducted by Alshehri et al. [24] shown that whereas LMWH substantially decreases the incidence of VTE in comparison to mechanical prophylaxis alone, it concurrently doubles the risk of postoperative haemorrhage. Oberle et al. [25] highlighted the benefits of combining intermittent pneumatic compression (IPC) with anticoagulation, reporting a significant reduction in postoperative thromboembolic events without a significant increase in bleeding risk. This indicates that personalised thromboprophylaxis, initiated according to a patient’s unique risk profile and adjusted during the perioperative phase, may enhance safety and effectiveness. Future developments may concentrate on targeted strategies incorporating real-time monitoring of sodium variations to proactively alleviate endothelial stress, individualised hydration and volume management protocols based on CDI-SIADH transitions, early mobilisation and intermittent pneumatic compression to prevent venous stasis, and supplementary pharmacological agents that enhance endothelial function, to prevent risk of DVT while reducing negative side effects.
The main limitation of this review is the limited number of included studies (n = 3), which constrains the capacity to reach widely generalisable conclusions. Due to the restricted data and the diversity of patient populations, including both pediatric and adult cases, as well as the variability in tumor pathology (e.g. craniopharyngiomas versus pituitary adenomas), no definitive evidence-based recommendations could be established concerning the optimal timing for LMWH prophylaxis, the ideal therapeutic sodium range for risk reduction, or standardized perioperative fluid protocols. Future large-scale, prospective studies focusing on these specific parameters are essential to establish clear clinical guidelines tailored to the neuroendocrine and hemostatic dynamics of sellar region surgeries.

Conclusion
This systematic review highlights the transition from CDI to the SIADH as the crucial postoperative phase associated with the greatest risk of DVT after major sellar region tumor surgeries. The 7–10-day period indicates substantial neuroendocrine and fluid-electrolyte variations that disrupt vascular homeostasis and increase thrombogenic risk.
The review clarifies that perioperative neuroendocrine disturbances, specifically sudden alterations in antidiuretic hormone activity, sodium imbalance, and hormonal dysregulation of cortisol and thyroid axes, directly contribute to endothelial dysfunction, hemoconcentration, and inflammation-induced hypercoagulability. Future research should focus on establishing the ideal timing and dosage of thromboprophylaxis concerning these endocrine fluctuations, incorporating sodium and volume-monitoring protocols, and validating neuroendocrine biomarkers as indicators of thrombotic risk through extensive, prospective, multicenter studies.

Supplementary Material

PRISMA_2020_checklist_DVT.docx