Introduction
Hypertensive intracerebral hemorrhage (ICH) accounts for 50%–70% of all spontaneous ICH; Reference Ariesen, Claus, Rinkel and Algra1 its morbidity and mortality rates are higher than any other stroke type. Reference Krishnamurthi, Feigin and Forouzanfar2 Regardless of aggressive treatment or newer management strategies, the outcome of patients remains very poor; mortality at 1 month is over 40% and has not improved in the last decades, Reference Gross, Jankowitz and Friedlander3–Reference Dennis, Burn, Sandercock, Bamford, Wade and Warlow6 and patients achieving a favorable functional outcome or an independent life after 1 year vary between 12% and 39%. Reference Sacco, Marini, Toni, Olivieri and Carolei4–Reference Moulin and Cordonnier8 These numbers reveal that spontaneous ICH is not only a very lethal disease but that survivors constitute a significant burden on health care resources, even higher than that of ischemic stroke. Reference Krishnamurthi, Feigin and Forouzanfar2
In patients with a hypertensive ICH, primary injury occurs during the first hours due to hematoma expansion and brain parenchyma compression. Secondary injury results from perihematomal edema and hemotoxicity of blood degradation products. Reference Wilkinson, Pandey, Thompson, Keep, Hua and Xi9 Surgical hematoma evacuation and conservative therapy are the main treatments for ICH. Reference Hemphill, Greenberg and Anderson10,Reference Steiner, Salman and Beer11 Surgery is conducted because an early volume reduction of the hemorrhagic lesion improves cerebral perfusion and avoids damage to neural tissue. Reference Nehls, Mendelow, Graham and Teasdale12 The most common hypertensive ICH sites are deep brain structures, such as the basal ganglia and the thalamus; a considerable layer of brain tissue must be crossed during surgery, which may cause iatrogenic damage to healthy cerebral tissue.
The International Surgical Trial in Intracerebral Hemorrhage (STICH) showed no overall benefit of open craniotomy over medical treatment alone. Reference Mendelow, Gregson and Fernandes13,Reference Mendelow, Gregson and Rowan14 Nevertheless, the high crossover rates and methodological issues left unclarified whether surgery may benefit specific groups of patients with supratentorial ICH, showing a slight but clinically relevant survival advantage for patients with spontaneous superficial ICH without intraventricular hemorrhage. Moreover, supratentorial hematoma evacuation might be considered a life-saving measure in deteriorating patients. Reference Hemphill, Greenberg and Anderson10 Surgery with less invasive approaches, such as endoscopic aspiration, or endoport-assisted evacuation, has shown potential benefits in the surgical management of spontaneous ICH by decreasing brain injury caused by surgical manipulation. Reference Marenco-Hillembrand, Suarez-Meade and Ruiz Garcia15–Reference Awad, Polster and Carrión-Penagos19 Therefore, we propose a low-cost 3D-printed endoport for a less invasive approach in the surgical management of hypertensive ICH patients. In this study, we compared the functional outcome and mortality rate of patients with deep supratentorial hypertensive ICH treated by early evacuation (less than 24 hours post-stroke) using two different techniques: endoport-assisted vs. standard open surgery (OS).
Material and Methods
Patient Selection
A retrospective analysis was conducted from medical records Reference Garcia-Estrada, Morales-Gómez, Delgado-Brito, Martínez-López, Flores-Huerta and Martínez-Ponce de León20 of all adult patients who underwent early surgical evacuation of deep hypertensive ICH by open craniotomy or decompressive craniectomy at a single academic training center between August 2017 and March 2019. Cohorts were identified upon the surgical approach employed: endoport-assisted and OS. Secondary ICH was not included in the protocol. The excluded patients had incomplete radiologic or clinical data, more than a 24-hour evolution before surgery, lobar or infratentorial hemorrhages.
Fabrication of the 3D-Printed Endoport
The endoport was developed by some of the authors with material already in use for medical devices and produced at minimal cost using 3D printing technology. It was designed using an open-source 3D computer graphics software (Blender software, version 2.78, Blender Foundation) and made through additive manufacturing using a desktop-sized and affordable stereolithography 3D printer (Form 2, Formlabs Inc., Somerville, MA, USA) using a biocompatible resin (Dental SG Clear Resin, Formlabs Inc.). Finite element analysis was performed in Ansys Workbench (Ansys Inc., Canonsburg, PA, USA) to examine the behavior of the endoport under boundary loads and ensure the integrity of the device during surgery. The illumination circuit was designed in Proteus Design Suite (Labcenter Electronics Ltd, Grassington, North Yorkshire, England) and enclosed and insulated in the device. A sterilized cable is attached during surgery for connecting the device to the power source. A 14 mm working channel was chosen because, in lab tests, it was the smaller diameter at which the neurosurgeons were comfortable using suction and bipolar cautery. Although different diameters and lengths could be manufactured, this study utilized the 14 mm × 6 cm work channel endoport. The obturator tip geometry was designed using finite element analysis in Ansys Workbench to provide blunt dissection and minimal resistance to brain tissue.
The device consists of two cylindrical components (obturator and sheath) that, when assembled, could be inserted to provide access to the hematoma cavity through blunt dissection of subcortical areas (Figure 1). The obturator includes a central canal for the optional use of a neuronavigational pointer. An enclosed LED circuit board is mounted on the sheath to provide adequate surgical site illumination. Once the endoport is inserted and the obturator disassembled, the sheath includes corridor access and direct visualization of the subcortical surgical site. The fabrication cost was estimated to be about 40 USD, the same device, adequately sterilized, was employed for multiple patients. Each device was utilized only three times unless mechanical/electrical defects were found; before each use, the endoport was carefully inspected by a qualified bioengineer involved in its manufacturing process, and gas plasma sterilization was performed.
Low-cost 3D-printed endoport for a less invasive approach in the surgical management of hypertensive ICH patients.
Surgical Approach
The endoport-assisted ICH evacuation consisted of the device placement through a 6 mm corticotomy in the middle temporal gyrus directed toward the clot’s location (Figure 2). As the endoport entered the hematoma, the obturator (inner part) was pushed out with a variable amount of clot. Once the endoport was placed, the obturator was removed entirely, and the sheath (outer part) provided a 14 mm diameter and 6 cm length work channel with direct visualization of the cavity through which the hematoma was evacuated and hemostasis achieved. The surgery was performed without additional optic devices (i.e., microscope, surgical loupes).
An illustrative case of a large putaminal ICH in a hypertense 47-year-old man. (A) preoperative coronal CT scan without contrast. (B) An illustration of the endoport-assisted ICH evacuation. (C) Intraoperative images of hematoma evacuation, the 3D-printed endoport allows visualization without additional optics. (D) Postoperative CT scan showing complete hematoma evacuation.
In the OS group, a 2 cm corticotomy was performed in the middle temporal gyrus; then, by blunt dissection of the white matter with a bipolar cautery, the hematoma was located; to create a visibility corridor, malleable brain retractors were employed. In both cases, hematoma evacuation was performed using suction and bipolar cautery. Hemostasis was achieved by continuous saline irrigation and pressure packing. The identified bleeding artery was electrically coagulated.
Clinical Outcomes
Our primary outcomes of interest were the functional outcome and the mortality measured at 90-day post-stroke. To assess the functional outcome, the authors used the Modified Rankin Scale for Neurologic Disability (mRS) measured at the moment of hospital discharge and 7-, 30-, and 90-day post-stroke. As a measure of long-term outcome in hypertensive ICH patients, the 90-day post-stroke functional outcome was further dichotomized in favorable and unfavorable outcomes, mRS 0–3 and 4–6, respectively. Reference Uyttenboogaart, Stewart, Vroomen, De Keyser and Luijckx21 These outcomes were identified from hospital records and death certificates.
Our secondary outcomes of interest were the postoperative midline shift and residual hematoma volume, clot evacuation rate, reoperations, hospital length of stay (LOS), and in-hospital mortality. The postoperative midline shift and residual hematoma volume were determined in a brain computed tomography (CT) performed 12–24 hours after surgery.
Covariate Data Collection
Baseline demographic, clinical, and radiologic variables were collected retrospectively from the hospital archives. According to the institutional protocol, surgical intervention was considered for patients with supratentorial hypertensive ICH volume ≥ 30 ml, midline shift ≥ 5 mm, or cerebral herniation signs. Depending on the brain edema observed, the operating surgeon decided, in a nonrandomized fashion, whether to perform a craniotomy or decompressive craniectomy.
Hemorrhage volumes were estimated in brain CT by the ABC/2 method; intraventricular blood was not included. An independent, board-certified neuroradiologist performed the radiological image interpretation without knowledge of the study rationale or hypotheses. All other variables were identified from hospital records and death certificates.
Statistical Analysis
Data are reported using standard statistical methods. Continuous variables with normal distributions are presented as means with standard deviation (SD), whereas those with non-normal distributions (based on the Shapiro–Wilk test) are presented as medians with interquartile range. Categorical variables are presented as percentages. Comparisons of numerical data were performed using an unpaired t-test (parametric data) or Mann–Whitney U-test (nonparametric data). Comparisons of categorical data were made by chi-square (χ 2) and Fisher’s exact tests, when appropriate. Survival analysis was performed by the Kaplan–Meier method censored 90 days after the hypertensive ICH; comparisons between groups were achieved through the log-rank test. Binomial logistic regression analyses were performed to adjust for the effect of potential confounders on the 90-day post-stroke favorable functional outcome and survival; well-known predictors of hypertensive ICH were used as covariates (i.e., age, hemorrhage volume, admission Glasgow Coma Scale (GCS), surgical procedure, ICH score, and the time from ICH to surgery). A p-value <0.05 was considered statistically significant; all p values were two-sided. Analyses were generated with Jamovi software, version 1.2 (The jamovi project, 2020). The local ethics committee approved this study, and the requirement for informed consent was waived due to the retrospective nature of the study.
Results
Patient Numbers
A total of 54 consecutive adult patients who underwent surgical evacuation of primary ICH by open craniotomy or decompressive craniectomy at our hospital between August 2017 and March 2019 were reviewed retrospectively. Twelve patients did not meet our selection criteria, including five patients with more than a 24-hour evolution before surgery, five lobar hemorrhages, and two hemorrhages originating from the mesencephalon and cerebellum; another six patients were excluded because they were transferred to other hospitals, and clinical data could not be obtained. Therefore, analyses were conducted on a final dataset of 36 adult patients who underwent early surgical evacuation of supratentorial hypertensive ICH classified into two groups: endoport and OS; both groups progressed simultaneously. Depending on the on-call schedule, eight neurosurgeons with 3–5 years of clinical experience performed the surgeries: 3 of them had previous experience utilizing endoport systems and used the proposed device in all their cases; the remaining five employed an OS.
Patient Characteristics
As detailed in Table 1, the two cohorts (18 endoport; 18 OS) showed no statistically significant differences in demographic, clinical, and radiologic characteristics, including median admission hemorrhage volume, GCS, and ICH scores. The average time interval between onset of symptoms related to the hypertensive ICH and surgery was similar in both cohorts (14 vs. 12 hours).
Demographic and baseline characteristics of patients
BP = blood pressure; GCS = Glasgow Coma Scale; ICH = intracerebral hemorrhage; CT = computed tomography.
Data are median (IQR) or n (%).
* Data are mean ± SD.
** ICH volumes were estimated in brain CT by the ABC/2 method.
Functional Outcome
The endoport group was associated with a better functional outcome at hospital discharge and 7-, 30-, and 90-day post-stroke (Figure 3). At the 90-day follow-up, 44% of patients in the endoport group had a favorable functional outcome, compared to 17% in the OS group (p = 0.039).
Functional outcome. At the 90-day post-stroke follow-up, 44% of patients in the endoport group had a favorable functional outcome, compared to 17% in the STOS group (p = 0.039). mRS scores range from 0 (no disability) to 6 (death). mRS = modified Rankin Scale.
Binomial logistic regression was performed to determine the effects of admission ICH score, the time interval from symptom onset to surgery, surgical procedure, and surgical approach on the likelihood that participants had a favorable functional outcome at the 90-day follow-up. The logistic regression model was statistically significant, χ 2(4) = 21.3, p < 0.001. The model explained 63% (Nagelkerke R 2) of the variance in favorable functional outcome at the 90-day follow-up and correctly classified 86% of cases. Patients in the endoport group were 26 times more likely (p = 0.049) to have a favorable functional outcome at the 90-day follow-up than the OS group. An increase in the ICH score was associated with a reduction in the likelihood (p = 0.026) of a 90-day favorable functional outcome. The average time from the onset of symptoms to the time of surgery (p = 0.053) and the surgical procedure employed (i.e., craniotomy or decompressive craniectomy) (p = 0.134) did not add significantly to the model.
Mortality and Survival
The endoport group was associated with lower 90-day mortality (33% vs. 72%, p = 0.019). Kaplan–Meier analysis revealed a higher probability of survival at 90 days after stroke in the endoport group (67% vs. 28%; HR 0.32, p = 0.015) (Figure 4).
Overall survival. Kaplan–Meier analysis revealed a higher probability of survival at 90 days after stroke in the endoport group (67% vs. 28%; HR 0.32 (95% CI 0.12, 0.80), p = 0.015). Data were censored at day 90. Shaded areas show 95% CIs.
Binomial logistic regression was performed to determine the effects of admission GCS score, hemorrhage volume, surgical procedure, and approach on the likelihood that participants were alive at the 90-day follow-up. The logistic regression model was statistically significant, χ 2(4) = 25.8, p < 0.001. The model explained 68% (Nagelkerke R 2) of the variance in 90-day survival and correctly classified 83% of cases. Patients in the endoport group were 17 times more likely (p = 0.019) to be alive at the 90-day follow-up than the OS group. An increase of one point of the GCS was associated with 2.7 times increased likelihood (p = 0.012) of 90-day survival. Hemorrhage volume (p = 0.474) and surgical procedure (p = 0.578) did not add significantly to the model.
Secondary Outcomes Analyses
Detailed information on the treatment variables and outcomes is shown in Table 2. Postoperative residual hematoma volume, clot evacuation rate, and postoperative midline shift showed no statistically significant differences between both groups. Seven patients were reoperated; the leading was a rebleed, followed by a surgical site abscess; there were no notable differences between groups.
Treatment variables and outcomes
CT = computed tomography; mRS = modified Rankin Scale.
Data are median (IQR) or n (%).
Median LOS was similar in both groups; however, patients treated with an endoport were associated with lower in-hospital mortality. Fifty-three percent of patients died during the 90-day follow-up. Pneumonia was the leading cause of death (25% of patients), followed by renal failure (11%), pulmonary embolism (8%), heart failure (6%), and central nervous system infection (3%).
Discussion
This study compared the functional outcome and mortality of two different ICH evacuation techniques: endoport-assisted vs. OS. To the best of our knowledge, a low-cost 3D-printed endoport has not been proposed for a less invasive ICH evacuation.
Large-sized clinical trials have failed to show an overall benefit of open or minimally invasive surgery over medical treatment in ICH management. Reference Mendelow, Gregson and Fernandes13,Reference Mendelow, Gregson and Rowan14,Reference Mould, Carhuapoma and Muschelli22–Reference Hanley, Thompson and Rosenblum24 Nevertheless, several small-to-moderate-sized trials of minimally invasive surgery have yielded encouraging results. A meta-analysis of 14 surgical trials, including 2186 patients, concluded that there is evidence that surgery is of benefit if undertaken early before the patient deteriorates. Reference Gregson, Broderick and Auer25 This study was highly influenced by a randomized, controlled minimally invasive surgery trial conducted by Wang et al., where they compared CT-guided aspiration craniopuncture plus lytic infusion to medical management, showing a substantial reduction in 90-day dependence rates. Reference Wang, Jiang and Liu26 Moreover, Zhou et al., in a meta-analysis of randomized controlled trials, concluded that patients with supratentorial ICH might benefit more from minimally invasive surgery than other treatment options. Reference Zhou, Chen and Li16
The Minimally Invasive Surgery Plus Recombinant Tissue Plasminogen Activator for Intracerebral Hemorrhage Evacuation Phase III Clinical Trial (MISTIE III), which involves the placement of a catheter in the hematoma cavity and the administration of a thrombolytic agent, failed to improve functional outcome compared to the standard medical care. Reference Hanley, Thompson and Rosenblum24 It is important to note that the MISTIE technique involves passive drainage through a catheter for several days, allowing secondary injury progression. Nevertheless, research suggests other less invasive approaches, such as endoscopic aspiration (Apollo or Artemis, Penumbra, Alameda CA, USA), Reference Spiotta, Fiorella and Vargas27–Reference Rothrock, Chartrain and Scaggiante31 exoscope-assisted endoport system (BrainPath, Nico Corporation, Indianapolis, Indiana, USA), Reference Labib, Shah and Kassam32,Reference Bauer, Rasmussen and Bain33 and the herein proposed method, may improve clinical outcomes. This improvement is probably obtained by reducing brain injury compared to open surgical manipulation. Reference Kaya, Türkmenoğlu, Ziyal, Dalkiliç, Sahin and Aydin34 Additionally, in contrast to the MISTIE technique, these methods may achieve an immediate clot evacuation, potentially reducing injury progression. Reference Wilkinson, Pandey, Thompson, Keep, Hua and Xi9
The Penumbra Apollo and Artemis endoscopic aspiration system (first and second generation, respectively) consists of an aspiration and irrigation cannula with a vibrational element to eliminate clot buildup. This system fits through the channels of commercially available endoscopes, allowing drainage under continuous endoscopic visualization. Reference Spiotta, Fiorella and Vargas27–Reference Kellner, Song and Pan29,Reference Kellner, Chartrain and Nistal35 As an alternative, endoports or tubular retractors, such as the BrainPath system and the herein 3D-printed device, provide radial retraction that symmetrically distributes the forces to the surrounding brain parenchyma and provide a protected corridor for access, visibility, and evacuation of the hematoma. Reference Kelly, Goerss and Kall36,Reference Mansour, Echeverry, Shapiro and Snelling37 The technique using the BrainPath endoport system for ICH evacuation is described utilizing neuronavigation and high-definition extracorporeal optics. Reference Labib, Shah and Kassam32,Reference Bauer, Rasmussen and Bain33,Reference Sujijantarat, El and Pierson38,Reference Przybylowski, Ding, Starke, Webster Crowley and Liu39 Otherwise, the proposed endoport may provide adequate access and visibility without additional devices and its associated cost. Furthermore, the BrainPath system price is around 4,000 USD Reference Norton, Dickerson, Kulwin and Shah40 compared to the 40 USD of the device described in this study. Thus, this may be of particular benefit for health care systems with limited resources.
The endoport group demonstrated a 27% absolute increase in favorable functional outcomes and a 39% absolute reduction in all-cause mortality contrasted to the OS at the 90-day follow-up. Although no statistically significant differences were noted between both groups, the endoport group showed an 8 ml lower median ICH volume, which may have affected the outcomes. Nevertheless, the statistical analysis showed that the endoport-assisted approach had an independent association with a better functional outcome and lower mortality.
Favorable functional outcomes obtained by the endoport group (44%) are comparable to those obtained by other retrospective studies of endoscopic aspiration (46%) Reference Kellner, Song and Pan29 and endoport-assisted evacuation (44%–63%). Reference Labib, Shah and Kassam32,Reference Sujijantarat, El and Pierson38,Reference Przybylowski, Ding, Starke, Webster Crowley and Liu39 Likewise, the observed 30-day mortality in the endoport group (22%) was similar to endoscopic (9%–28%) Reference Goyal, Tsivgoulis and Malhotra28,Reference Kellner, Song and Pan29 and other endoport systems (0%–36%). Reference Labib, Shah and Kassam32,Reference Bauer, Rasmussen and Bain33,Reference Przybylowski, Ding, Starke, Webster Crowley and Liu39 It is important to note that a wide variation in mortality rates could be associated with data heterogeneity. Given that both endoport and OS cohorts had a median ICH score of 3, the OS group showed comparable 30-day mortality than its associated risk (67% vs. 72%), Reference Hemphill, Bonovich, Besmertis, Manley and Johnston41 supporting previous evidence discouraging the OS over medical treatment. Reference Mendelow, Gregson and Fernandes13,Reference Mendelow, Gregson and Rowan14
In the present study, the endoport group tended to have a non-significant lower clot evacuation rate than the OS group (92% vs. 97%). It is possibly associated with errors in the trajectory that may have led to inadequate ICH removal or because of the device’s narrow corridor, limiting the surgical field’s visibility. Despite this, the herein proposed endoport showed promising efficacy, comparable to evacuation rates described with the exoscope-assisted endoport system (95%), Reference Bauer, Rasmussen and Bain33 and higher than the one reported by the endoscopic aspiration (54%–88%) Reference Spiotta, Fiorella and Vargas27,Reference Kellner, Chartrain and Nistal35 and MISTIE procedure (69%); Reference Hanley, Thompson and Rosenblum24 it is essential to note that the endoport-assisted ICH evacuation was performed without the need for navigation and extracorporeal optics, which may increase surgical costs. The endoport-assisted evacuation technique was safe concerning bleeding, infection, and reoperation rates; moreover, there were no notable differences between both groups in the LOS. These results were similar to those reported with minimally invasive techniques. Reference Marenco-Hillembrand, Suarez-Meade and Ruiz Garcia15,Reference Hanley, Thompson and Rosenblum24,Reference Spiotta, Fiorella and Vargas27,Reference Labib, Shah and Kassam32,Reference Kellner, Chartrain and Nistal35,Reference Przybylowski, Ding, Starke, Webster Crowley and Liu39
This study has some limitations. First, it was a nonrandomized, retrospective study. Therefore, the authors could not deny the possibility of selection bias; the surgical technique’s choice depended upon the surgeons’ preference. Second, the number of patients is small to generalize our results and might not give sufficient power to detect slight differences between groups. Third, our cohorts’ mortality rates are high, explained by the inclusion of patients with high ICH scores and large hematoma volumes. Our study population’s mortality is comparable to prior published literature on patients with severe ICH score. Reference McCracken, Lovasik and McCracken42 Fourth, there may have been differences in individual patient management that may have influenced our results. Fifth, these results may not apply to lobar ICH, more frequently associated with cerebral amyloid angiopathy; adults with lobar ICH are more likely to survive than adults with non-lobar ICH. Reference Samarasekera, Fonville and Lerpiniere43 Sixth, because of the lack of a medical management group, the advantages of endoport surgical evacuation over medical treatment must not be inferred until adequate research is conducted. Finally, because patients were only followed for 3 months, the possibility of continuous improvement after this point was not evaluated.
Conclusions
Overall, this study suggests that endoport-assisted ICH evacuation may have better functional outcomes and lower mortality than standard OS. The proposed device could provide a safe, low-cost alternative for ICH’s surgical treatment and help develop more cost-effective techniques that can be employed in a limited-resource setting. More rigorous research is hence needed and ongoing to assess the potential benefits of this technique.
Supplementary Material
To view supplementary material for this article, please visit https://doi.org/10.1017/cjn.2021.185.
Conflict of Interest
All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript.
Statement of Authorship
All persons who meet authorship criteria are listed as authors, and all authors certify that they have participated sufficiently in the work to take public responsibility for the content, including participation in the concept, design, analysis, writing, or revision of the manuscript.
