Indication for surgical interventions

While asymptomatic subdural hematomas (SDHs) are treated conservatively, SDHs that result in worsening clinical symptoms or exhibit radiographic evidence of significant mass effect are managed surgically (Fig. 1). SDH recurrence is defined as a re-accumulation of SDH that becomes symptomatic, requiring intervention. SDHs requiring immediate attention are associated with acute clinical deterioration, e.g., mental status changes, unilateral or bilateral dilated pupils, or extensor posturing consistent with brainstem herniation. Although not imminently surgical, SDHs causing more than 5-mm midline shift or over 10 mm in size are considered operative, in principle. Surgical approaches to SDH treatment include twist drill craniostomy, including subdural evacuating port system (SEPS), burr hole(s) allowing for active or passive drainage with or without irrigation and drain placement, craniotomy to span the extent of SDH and allow for membrane removal with bone replaced at the end of surgery, or decompressive hemicraniectomy for acute SDH with significant mass effect and cerebral edema, while keeping the bone off to allow for cerebral edema resolution.

Fig. 1
figure 1

Surgical management of SDH.

Full size image

Twist drill craniostomy

Twist drill craniostomy (TDC) for chronic SDH was described in 1960; the use of closed drainage system was trialed in late 1970s (Table 1). Analogous pediatric procedures (not discussed) include percutaneous tapping and subdural shunt placement [1,2,3,4]. TDC variations introduced over time include a variety of drilling angles, craniostomy entrance points, mechanical vs automated drilling, insertion of hollow screws, cannulas or trocars with closed suction systems, and attachment of irrigation ports. TDC complication rates are low and include infection, seizures, and SDH re-accumulation, less commonly serious conditions, e.g., tension pneumocephalus [5, 6]. Other studies report 8% complications [7], including development of new acute SDH/EDH requiring surgical evacuation due to damaged dural/cortical artery or bridging veins, or violation of brain parenchyma by improper drain placement [8, 9]. SEPS introduced in 1999 resulted in clinical outcome improvement and reduced side effects associated with drain placement; a hollow port with self-tapping threads is introduced over the thickest SDH portion without entering the subdural space, allowing for slow drainage via mild negative pressure [10,11,12]. Neal et al. reported a 77.8% success rate with SEPS, less likely with mixed density or septated SDH [13]. Miele et al. concluded that the degree of head elevation does not affect drainage or recurrence [14]. An RCT by Sindou et al. showed similar success rates with 48- vs 96-h drainage, with twice complications (26.9%) and mortality (11.4%) with longer drainage [15]. A study of older SDH patients by Miranda et al. (> 65 years, mean 80) demonstrated a ~ 10% recurrence with TDC and 26.5% 6-month mortality [16]. Escosa et al. reported closed drain system to be beneficial, while presence of post-operative midline shift > 5 mm, SDH > 10 mm and neurologic deficits were associated with recurrence, regardless of dexamethasone, aspirin, or AC use [17]. Krieg et al. reported a 63.3% success with hollow screws; 16.2% patients required repeat screw placement and 20.5% required burr holes with membranectomy [18].

Table 1 Twist drill craniostomy studies
Full size table

TDC under local or monitored anesthesia care (MAC) may be optimal in older patients with comorbidities; bilateral SDH drainage is accomplished without position change. Yadav et al. proposed modified angle TDC, using an infant feeding tube over guidewire for drain placement to avoid violating parenchyma [19]. Balser et al. compared the time to intervention and length of stay with SEPS vs burr holes, confirming that TDC is rapid and safe [20]; other studies suggest shorter procedure times but longer drainage with TDC [21]. SDH entry site and continued drainage may play a role. Jablawi et al. reported 67% effectiveness with TDC, 43% patients requiring open evacuation; entry points were over thickest SDH width rather than along the superior temporal line 1 cm anterior to the coronal suture, and scalp incision was closed without drain placement [22]. Further modifications include irrigation ports allowing instillation of adjunct therapies into SDH cavity. Wang et al. reported SEPS-like system with two ports, one for drainage and one for irrigation [23]. Neils et al. evaluated intra-cavity tPA with TDC, reporting 0% recurrence with tPA vs 30% without tPA; the mean drainage volume increased from 90 to 427.33 cm3 [24]. Urokinase likewise decreased duration of indwelling drain and hospital stay, with 0.43% recurrences [25]. Traumatic brain injury (TBI) causes tissue plasminogen activator release, increased plasmin levels, activation of fibrinolysis, inflammatory response, and increased vascular permeability via kallikrein system; tranexamic acid (TXA) counteracts this by binding to lysine sites on plasminogen with a decrease in plasmin levels [26]. The use of oral TXA 650 mg daily as SEPS adjunct resulted in a ~ 40.74% SDH reduction upon SEPS placement, with an additional 91.31% with continued TXA [27].

Burr hole evacuation

Over 700 articles and 25 clinical trials are published on burr hole SDH evacuation (BHE) (Table 2). OR environment is required due to need for perforator drills, electrocautery, and other surgical instruments; the remainder is flexible to tailor to patient needs. The standard burr hole diameter (14 mm) allows for larger dural opening, better visualization of clot and membranes, easier access to control bleeding, and possible introduction of additional instruments, e.g., endoscopes. While TDC introduces minimal air due to small cranial opening and closed drainage, BHE allows significant air into SDH space, requiring irrigation and drain placement for evacuation.

Table 2 Burr hole craniotomy studies
Full size table

Anesthesia

Similar to TDC, BHE can be performed under local or MAC. Khadka et al. described BHE under local anesthesia with 98.6% evacuation and 4.7% recurrence [28]. Guzel et al. described successful BHE via single burr hole under MAC with no complications, beneficial for patients who cannot tolerate intubation or general anesthesia [29].

Single vs multiple BHE

Studies from 1970s describe chronic SDH patients treated with multiple BHE with 41% improvement, 25% deterioration, and 22.7% death [4]. Taussky et al. compared single vs multiple BHE in chronic SDH patients, reporting 29.4 vs 4.8% recurrence, respectively [30]. Zumofen et al. reported 86.9% success with 13.1% recurrence with BHE and overlying subgaleal drain; patients received antibiotics and phenytoin intra-operatively after reversal of asa/AC and remained flat for 48 h with drains below the head [31]. This study reported 6.6% seizures and 1.6% infections [31]. An RCT by Nayil et al. compared one vs two burr holes and reported 4–6% recurrences; patients remained flat for 48 hours [32]. Prospective study of BHE with irrigation and drain placement reported 100% SDH resolution with 3.2% recurrence [33].

Continued drainage

Drain placement is key, associated with lower recurrence of 3.1–10.5% with BHE and drain placement vs 17–33% without [34,35,36,37]. An RCT by Santarius et al. reported less recurrence with subdural drain placement (24 vs 9.3%); the trial was stopped early due to clear difference [38]. While Javadi et al. showed similar recurrence [39], Ramcharan et al. reported 4% recurrence with drain placement and 30% without [40]. An RCT by Singh et al. reported less recurrence with drain placement (26 vs 9%), with similar complications and mortality [41•]. An RCT by Kutty and Johny reported lower recurrence with single BHE with drain placement vs two BHE (1.4 vs 15.7%) [42]. Guilfoyle et al. published long-term follow-up from the Cambridge Chronic SDH Trial (CCSHT), indicating improved evacuation, less recurrence, and survival benefit 5 years post-operatively [43]. The timing of drainage may matter; Yu et al. reported 6.6% recurrence with drain placement, ranging from 16.3% with drain removal prior to 3 days to 1.3% if removed thereafter [44]. Drain location may likewise matter; subperiosteal drainage reduced seizure and infection risks avoiding SDH capsule, but was associated with higher recurrence [31]. Bellut et al. reported 1.8% recurrence with subperiosteal drain and 3.1% with subdural drains [45]. Subgaleal drains with tip [37] or holes overlying burr hole [46] reduced recurrence as compared to no drain, avoiding injury to SDH membranes/parenchyma and pneumocephalus [46]. An RCT by Kaliaperumal et al. compared subdural vs subperiosteal drain placement with no recurrences with either, but better functional recovery with subperiosteal drains [47•].

Open vs closed system

Weir compared BHE with closed drainage vs catheter placement with passive drainage, former being superior [48]. An RCT by Laumer et al. compared closed drainage BHE vs implantation of silicone catheter with Rickham reservoir within burr hole, reporting similar recurrence; patients with Rickham required fourfold less repeat surgical interventions due to bedside access to reservoir [49]. Kwon et al. reported > 200-cm3 drainage with closed system BHE associated with 0% recurrence, < 200-cm3 drainage with open system BHE with 4.1% recurrence [50]. Kuroki et al. compared BHE with rapid drain placement vs BHE with irrigation and drain placement and found 1.8 vs 11.1% recurrence, respectively [51]. Kwon et al. further supported the reverse correlation between the amount of drainage and recurrence: re-accumulation was common with mixed density SDH and drainage volume < 200 cm3 [50].

Active vs passive irrigation

Post-operative pneumocephalus is reduced by irrigation, supine positioning while filling SDH cavity, drain placement, and avoidance of intraoperative nitrous oxide [52]. Kitakami et al. reported BHE with saline irrigation followed by CO2 resulting in rapid SDH cavity disappearance [53]. Hennig and Kloster reported lower (2.6%) recurrence with irrigation; BHE without irrigation showed high recurrence (32.6 vs 23.8%) with or without drain; 44.4% patients with recurrence required craniotomies [54]. Ishibashi et al. also reported 2.9% recurrence post-BHE with irrigation, as compared to 10.3% without irrigation [55]. A study of BHE by Zakaraia et al. with closed drainage with or without irrigation showed comparable recurrence [56].

Mobilization

Abouzari et al. evaluated BHE with irrigation and closed drainage; patients were randomized to remain flat for 3 days vs mobilized to 30–40° [57]. Patients remaining flat had 2.3% SDH recurrence, while those mobilized had 19% recurrence; pulmonary/thrombotic complications were similar [57]. Adeolu et al. compared mobilization day 2 vs 7, reporting no recurrence in either group [58]. Tsutsumi et al. reported higher recurrence with T1 hypointense SDH (11.6 vs 3.4%) [35]. Kaplan et al. used transcranial Dopplers to assess ipsilateral middle cerebral artery (MCA) flow upon evacuation of > 2 cm SDH; MCA flow normalized with improvement in mental status and contralateral weakness [59].

Schwarz et al. reported midline shift, hypertension, bilateral SDH, and vitamin K antagonists as recurrence predictors [60]. Nakaguchi et al. reported 12% recurrence with < 10-mm SDH vs 45% with > 10-mm SDH, 26% recurrence with pneumocephalus vs 8% without, and 48% recurrence if > 30% SDH cavity was air by volume [61]. Patients with frontal drains had 5% recurrence vs 38% with parietal, 36% with occipital, and 33% with temporal drains; patients with frontal drains had the least subdural air [61]. Unterhofer et al. reported comparable recurrence with or without membranectomy (21 vs 28%) [62•]. Mobbs and Khong suggested endoscopic visualization and subdural catheter placement, with a sharp division or electrocautery of membranes [63].

Adjunct therapies

Platelet-activating factor (PAF) levels are increased within acute SDH, localizing to perisinusoidal vessels in SDH membranes [64]. Hirashima et al. reported 0% recurrence with 1.5–3mg daily etizolam (PAF-dependent platelet aggregation inhibitor) vs 29.2% in controls [65]. Sun et al. reported less recurrence with dexamethasone [66]. Neils et al. evaluated intra-cavity tPA, reporting 30% recurrence with TDC, 11.8% with BHE, and 0% with BHE and tPA; the mean evacuated volume increased from 26 to 100 cm3 [24]. Shimamura et al. RCT reported a fivefold reduction in recurrence (25.6 vs 5.5%) with BHE, drain, and SDH cavity irrigation with 100 u/ml thrombin regardless of systemic anti-platelets; TXA was administered for 24 h [67]. Thrombin may aid hemostasis by inducing vessel constriction and fibrin deposition by binding its receptor on sinusoidal endothelium that undergoes microhemorrhaging in SDH hyperfibrinolytic state [68]. Yeon et al. reported no change in SDH recurrence if warfarin was restarted 3 days post-operatively [69].

Twist drill vs burr hole evacuation

Smely et al. reported 18.2% recurrence and 6% death with TDC and 33% recurrence and 10% death with BHE [70]. Williams et al. reported 63.6% deterioration with TDC, compared to 16% deterioration with BHE and 7% with BHE and drain placement [71]. An RCT by Muzii et al. compared SEPS vs BHE with irrigation and drain placement; recurrence, mortality, and recovery were similar (18.2 vs 33.3%) [72]. Gokmen et al. reported similar SDH recurrences, 2.6% in TDC and 6.3% in BHE group [73]. Lin reported improved cure rate with TDC vs BHE (88.8 vs 75.5%) with less recurrence requiring repeat interventions (7.9 vs 11.9%) and less complications (7.9 vs 20.7%) [74]. Kim et al. reported similar outcomes [75]. Wang et al. treated patients with twist drill YL-1 needle vs BHE with irrigation, reporting 100 vs 94.3% improvement and 13.2 vs 17% recurrence; another prospective study reported 18.4 vs 11.1% recurrence with TDC vs BHE [23, 76]. Finally, Brennan et al. reported 2% mortality, 14% morbidity, 9% recurrences, and 22% unfavorable outcomes associated with BHE, with patient age, bed rest, and single burr hole being independent predictors of poor functional outcome, and failure to insert the drain as a predictor of recurrence and unfavorable functional outcome [77].

Craniotomy for acute SDH

Craniotomy is useful for treating acute SDH with thick clot and chronic or subacute SDH that failed other treatments with significant radiographic progression, or SDH with septations (Table 3). Management of acute SDH depends on patient prognosis, goals of care, and hospital course; it is life-threatening due to the underlying structural TBI associated with up to 92% mortality. Surgical evacuation via open craniotomy is indicated for acute SDH with > 1-cm thickness or > 5-mm midline shift [78], deteriorating patient with the Glasgow Coma Scale (GCS) < 8, unilateral or bilateral fixed dilated pupils, or evidence of elevated increased intracranial pressure (ICP) > 20 [79]. Modifications include craniotomy or craniectomy, with or without partial or complete resection of SDH membranes. SDH etiology, patient age, time to surgical treatment, pre- and post-operative ICPs, and extent of craniotomy or decompressive craniectomy affect outcomes. Subdural hygromas require surgical interventions if behaving aggressively, including evacuation, placement of ventriculoperitoneal or subduroperitoneal shunt, or cranioplasty.

Table 3 Craniotomy studies
Full size table

Morbidity and mortality in acute SDH patients remain high due to underlying parenchymal injury, loss of autoregulation, poor preoperative neurologic state, and uncontrolled post-operative ICPs. Britt and Hamilton proposed large decompressive craniectomy for acute SDH in 1978 [80]. Koc et al. reported 38% functional recovery in 91% acute SDH patients with GCS 9–15, with a mortality over 90% in patients with GCS 3–4, bilaterally nonreactive pupils, or intracerebral hematomas; 80% mortality was reported in patients with unilaterally nonreactive pupil or subarachnoid hemorrhage [81]. Guilburd and Sviri described acute SDH evacuation through multiple dural fenestrations with > 80% success; however, 51.6% expired and over third sustained persistently elevated ICPs > 25 refractory to medical management [82]. An RCT by Jiang et al. investigated large frontotemporoparietal decompression vs temporoparietal decompression in patients with severe TBI and refractory ICPs [83]. More patients with wide decompression showed improvement (39.8 vs 28.6%) with better survival and less neurodeficits [83]. Smaller studies showed no difference in outcomes [84].

A study of severe TBI patients by Ucar et al., conducted to identify GCS cutoff for patients that benefit from decompression and address its timing, indicated that decompression should occur ≤ 4 h from injury and is futile with GCS ≤ 5 [85]. Woertgen et al. reported no difference in outcomes in acute SDH patients undergoing craniotomy or decompressive hemicraniectomy; however, mortality due to herniation was higher with decompression (53 vs 32.3%) [86]. Missori et al. used double dural sheets with decompression to aid with temporalis dissection during cranioplasty [87]. Nguyen et al. performed fenestration on bone prior to replacement, with improved post-operative SDH volumes, better drainage, and reduced recurrences [88]. Other studies described floating or hinge craniotomies (bone attached to cranium with sutures [89], Y-shaped titanium plates [90], or temporalis muscle [90, 91]), or using divided bone flaps with multiple hinges [92].

Craniotomy for chronic SDH

Chronic SDH evacuation via craniotomy is associated with better outcomes. Tabaddor and Shulmon reported 39.3% clinical improvement, 21.4% deterioration, and 28.6% mortality; a later study by Hamilton et al. reported 71.7% improvement, 10.6% recurrence, and 14.9% complications [4, 93]. Ernestus et al., Callovini et al., and Van der Veken reported similar numbers [94,95,96]. Godlewski et al. compared patients with acute, subacute, and chronic SDH; 83.5% chronic, 75% subacute, and 10.5% acute SDH patients underwent BHE, with 37.1, 26.7, and 25% recurrences, respectively; the remaining patients were treated with craniotomy/craniectomy [97]. In craniotomy/craniectomy group, 5.9% of acute SDH, 60% of subacute, and 42.9% of chronic SDH patients had recurrences [97]. Craniotomy/craniectomy patients with acute SDH did well, while subacute/chronic SDH patients best responded to BHE.

Extensive craniectomy may not be important in treating chronic SDH patients. Beatty et al. reported 91% improvement with 8.7% mortality post-minicraniectomy [98]. A study by Lee et al. comparing BHE vs 3-cm craniotomy with membranectomy vs extended craniotomy with membranectomy reported 16, 18 and 23% recurrences, respectively [99]. Lindval and Koskinen likewise reported 17% recurrences irrespective of anti-platelet use [100]. Mahmoud et al. reported 2.86% recurrence with minicraniectomies [101]. Horn et al. reported 8% recurrence in BHE vs craniotomy groups [102], while White et al. reported 17.7 vs 19.8% recurrence, with 2.5-fold higher (17.2%) mortality in the latter [103]. Regan et al. showed 6.6% recurrence and 3.3% mortality in patients undergoing BHE; recurrence post-minicraniotomy was 24.1% with mortality of 6.9% [104•]. Mondorf et al. reported 27.8 vs 14.3% recurrence with craniotomy vs BHE [105]. Kim et al. reported ~ 90% improvement regardless of technique; patients undergoing BHE had 8.9% recurrences with 8.1% mortality, vs 50% recurrence post minicraniotomy, and 9.5% recurrence with 9.5% mortality after extensive craniotomy [106].

Membranectomy may be beneficial. Tanikawa et al. reported 0% recurrence with patients undergoing craniotomy, including chronic SDH patients with multiple membranes [107]. Mohamed et al. reported no recurrence with craniotomy, durectomy, outer membranectomy, and drain placement, with two patients requiring percutaneous tapping [108]. Rocchi et al. reported zero recurrence with membranectomy, with 21.4% complications including seizures and a hemorrhagic infarct [109]. Balevi et al. reported 11.9% recurrence following large craniectomy with membranectomy within 24 h; 22.8% patients developed tension pneumothorax, 25% had seizures, and 14.28% expired [110]. An RCT by Unterhofer et al. reported a comparable 21.4% recurrence with minicraniotomy vs 28.6% with additional membranectomy, possibly due to inability of atrophied brain to re-expand and formation of neomembranes [62•].

Bilateral chronic SDH

Bilateral chronic SDHs have an incidence of 15–25% and are associated with higher recurrence and worse outcomes, with advanced age, diabetes, coagulation abnormalities, altered mental status, gait instability, and less midline shift associated with bilateral SDH [111,112,113,114]. Hsieh et al. reported age as a risk factor, with similar complication and recurrence in patients with unilateral vs bilateral BHE [115]. Lee and Park reported similar recurrence post-BHE with bilateral vs unilateral SDH (21.4 vs 16.3%), with favorable outcome and improvement less likely with bilateral SDH [113]. Fujitani et al. reported that SDH’s iso- or hypointensity on MRI were a single predictor of contralateral SDH enlargement in patients with bilateral chronic SDH undergoing unilateral BHE [116].

Few reports describe SAH, IPH, and acute SDH after bilateral BHE; neurologic deterioration may also occur after bilateral craniotomies due to brain sag [117, 118]. Decision-making is complicated if significant differences in SDH size/thickness or lateralization of symptoms are present, suggesting that one SDH is asymptomatic. Lee et al. described TDC for bilateral SDH, 20% patients requiring reoperation [119]. Recent review reported 21.6% recurrences in 136 unilaterally treated patients with bilateral SDH, 28.7% requiring repeat treatment; the absence of post-operative drainage and mixed density SDH were independent predictors for retreatment [120]. A review of 500 chronic SDH patients by Mori et al. showed 49 recurrences, 18 being contralateral to the operated side, and 14 of these having a contralateral thin SDH/effusion on preoperative CT [121]. SDH recurrences may result from ICP alterations and decreased tamponade effect due to brain displacement toward the surgical side, allowing re-expansion of contralateral SDH [121].

Conclusions

Optimizing surgical management for various types of subdural hemorrhages is needed. Recurrence rates are highest with twist drill craniostomies, although they have significant advantages of procedure completion at bedside or under MAC. Closed system, active irrigation, continued drainage, and mobilization have been associated with better outcomes with twist drill and burr hole SDH evacuation. Craniotomy carries the highest morbidity and mortality, especially in older patients who are more likely to be on anti-platelets or anticoagulation, but should be the method of choice in treating recurrent chronic SDH with multiple membranes, or acute SDH in deteriorating patients. Management of bilateral chronic SDH remains under investigation. Patients with SDH are served best by a specialized multidisciplinary team, individualized approach, and rapid surgical interventions, if needed.