Introduction
Total laminectomy (TL) has conventionally been the principal approach used for posterior exposure in intradural spinal tumor resection. Although TL provides liberal access, key stabilizing structures within the spine, including the laminae, facet joints, ligamentous complex, and paraspinal musculature, are often disrupted. Consequently, it is associated with substantial postoperative deformity, particularly kyphosis and scoliosis, after multilevel exposure [1,2]. When the deformity becomes clinically significant, instrumented fusion may be required, adding to the morbidity, complexity, and cost of the procedure [3–5].
Recent advancements have led to a widespread interest in tissue-sparing techniques in spinal oncology, especially regarding minimally invasive strategies aimed at limiting muscle dissection, reducing perioperative morbidity, and preserving spinal stability. Surgical techniques such as endoscopic resection, laminoplasty, and prophylactic instrumentation have been reported in selected cases [3,4]. Among these techniques, unilateral hemilaminectomy (UH) creates a unilateral posterior corridor with limited bone removal, preserving the contralateral lamina-facet-tension-band complex, typically obviating the need for implants. The existing literature on UH suggests that it can achieve gross total resection (GTR) with favorable neurological and perioperative outcomes that are comparable or superior to traditional approaches [1,5]; however, these reports had small samples, were retrospective in nature and methodologically heterogeneous, or had limited long-term data on neurological recovery, mechanical stability without fusion, and resource utilization.
Over the past decade, our spinal oncology team has implemented a standardized UH workflow for adult patients with intradural-extramedullary and intramedullary tumors. The present study aimed to evaluate the oncologic efficacy (tumor control and extent of resection), neurological recovery, perioperative morbidity, and long-term biomechanical stability after UH in patients with intradural spinal tumors. As a secondary aim, we conducted a focused literature review to compare the outcomes of UH and related minimally invasive unilateral approaches with TL. Since previous studies used heterogeneous outcome definitions and incomplete biomechanical follow-up, a formal quantitative meta-analysis was not feasible; hence, we performed a descriptive analysis of key comparative and economic data.
Materials and Methods
Ethics statement
This study was conducted following the principles of the Declaration of Helsinki. Ethical approval for the study was obtained from the İstinye University Human Research Ethics Committee (approval no., 2025-263). The requirement for informed consent was waived owing to the retrospective design and use of de-identified data.
Study design and patients
This single-center retrospective observational cohort study was conducted at a tertiary spine oncology practice between 2015 and 2025. Consecutive adults (age ≥18 years) undergoing UH for intradural spinal tumors were identified from a prospectively maintained operative database. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) recommendations for observational studies were followed throughout the study.
Patient selection
The following inclusion criteria were used: (1) patients showing the presence of an intradural spinal tumor (intradural-extramedullary or intramedullary) at any spinal level confirmed by preoperative magnetic resonance imaging (MRI) and intraoperative findings; (2) UH as the index posterior exposure procedure; and (3) availability of postoperative clinical and radiologic follow-up data. Pediatric patients (age <18 years) and those with extradural tumors were excluded. At our center, UH was not attempted in: (1) patients with severe fixed kyphosis or scoliosis at the index level; (2) those with a previous instrumented fusion at the target motion segment; (3) patients with circumferential or predominantly ventral tumors that may require bilateral or anterior exposure; or (4) multilevel lesions in which more extensive bone removal was anticipated. Such anatomically unfavorable or high-risk cases were treated with alternative approaches, such as multilevel laminectomy with or without instrumentation, and were therefore not included in this UH cohort. The patient screening and selection process is summarized in a flow diagram.
Surgical technique
All resection procedures were performed through a standardized UH approach under an operating microscope. The side of the UH was chosen according to the symptomatic side or the side of maximal tumor bulk on preoperative MRI. Bone removal typically extended from the medial border of the ipsilateral facet joint across the hemilamina to the base of the spinous process, while preserving the contralateral lamina, facet capsule, and interspinous/supraspinous ligaments. Limited undercutting of the ipsilateral facet and spinous base was performed when necessary to visualize the ventral dura or contralateral tumor margin, but the contralateral facet surface was not drilled. Contralateral decompression was performed in selective cases, as required. Likewise, resection of the ligamentum flavum was limited to the ipsilateral side, with contralateral extension only when deemed necessary. A representative UH exposure and working corridor, with corresponding pre- and postoperative imaging and the intraoperative microscopic view, are shown in Fig. 1.
After hemilaminectomy and linear durotomy, microsurgical tumor resection was performed with the aim of GTR, whenever safely achievable. In the case of infiltrative intramedullary astrocytomas, maximal safe debulking (subtotal resection [STR]) was preferred over aggressive dissection at the expense of neurological function. The dura was closed in a watertight fashion using primary suturing; postoperative stereotactic radiotherapy (SRT) was performed after STR.
Outcomes
Primary oncologic efficacy
The primary efficacy endpoint was the extent of resection (GTR versus STR) based on the operative report and early postoperative MRI.
Primary biomechanical safety
The primary biomechanical endpoint was segmental instability, defined as symptomatic mechanical pain plus dynamic radiographic translation of >3 mm or focal kyphosis >10°, or any indication for index-level instrumented fusion during follow-up.
Oncologic durability
Defined as the absence of radiographic recurrence, regrowth, or progression, and absence of need for reoperation for tumor progression at the index level.
Follow-up
Routine postoperative clinical review was performed at approximately 4–6 weeks and 3–6 months, with early MRI at the 3–6 months visit, followed by review at around 12 months, and annually thereafter. Routine screening of asymptomatic patients was not performed; dynamic flexion-extension radiographs were obtained selectively in patients who reported mechanical back or neck pain or in whom clinical examination raised concern for instability. Patients who did not attend scheduled in-person visits were contacted, and those who could not be reached were censored at the time of the last documented assessment.
Statistical analysis
All analyses were performed using Jamovi (The Jamovi Project; https://www.jamovi.org/) based on R (R Foundation for Statistical Computing, Vienna, Austria). Analyses were primarily descriptive; continuous variables are reported as mean±standard deviation, with 95% confidence intervals (CIs), or median with interquartile range (IQR). Distributional assumptions were checked using the Shapiro-Wilk test, visual inspection of histograms, and Q–Q plots. Categorical variables are presented as frequency (n/N) and percentages with 95% CIs (Wilson method; Clopper-Pearson for zero-event outcomes). Between-group differences in proportions (e.g., GTR rates) were assessed using Fisher’s exact test. Paired change in neurological status in the extramedullary subgroup was evaluated using McNemar’s exact test; changes in the smaller intramedullary subgroup were summarized descriptively. Time-to-event analyses (Kaplan-Meier survival analyses) were prespecified only if oncologic or mechanical failures occurred; otherwise, durability was reported as observed event rates with exact CIs. A two-sided α of 0.05 was used.
Literature search
To contextualize cohort findings, we performed a focused literature search of the PubMed/MEDLINE database from its inception to November 1, 2025, without date limits. Search terms combined concepts for intradural (intradural-extramedullary and intramedullary) spinal tumors; unilateral hemilaminectomy and related minimally invasive unilateral approaches; and open or total laminectomy comparators. To mitigate indexing and terminology heterogeneity, the PubMed search was supplemented by backward and forward citation tracking and by examining the “Similar articles” option in PubMed for each eligible record. Record eligibility was restricted to adult, human patients, and English-language studies. We prioritized comparative clinical studies and higher-level evidence; cost/resource-utilization analyses and biomechanical stability studies were retained when directly informative.
Overall, 24 studies met the inclusion criteria: 19 comparative clinical series (including two with cost endpoints and one biomechanical study with finite-element analysis) and five systematic reviews/meta-analyses. Given the heterogeneity in outcome definitions and follow-up, no quantitative meta-analysis was attempted. Instead, key design features and endpoints (oncologic control, perioperative outcomes, biomechanical stability, and cost/resource use) were extracted and summarized descriptively (Supplements 1–4), which were then used to interpret and contextualize the cohort results.
Use of artificial intelligence
We used OpenAI only for language editing and style refinement of the manuscript; no data collection, statistical analysis, or interpretation of results was performed by artificial intelligence. The authors take full responsibility for the accuracy and integrity of the scientific content.
Results
Cohort assembly
The Results section reports outcomes from the institutional cohort; findings from the focused literature review are incorporated qualitatively in the Discussion and summarized in Supplements 1–4. Between 2015 and 2025, 405 patients underwent UH for spinal tumors. After excluding adult extradural tumors (n=250) and pediatric cases (n=19), the final cohort comprised 136 adult intradural tumors (Fig. 2). None of these patients with intradural tumors required revision surgery for known recurrence, planned bilateral laminectomy, posterior approaches other than UH, or prophylactic index-level instrumentation, confirming cohort homogeneity.
Baseline characteristics
The mean age of the study cohort was 44.9±12.7 years (95% confidence interval [CI], 42.8–47.1), and the median age was 43.5 years (IQR, 37.0–53.0). Women accounted for 57.4% of patients (n=78/136). Most lesions were intradural-extramedullary (n=108/136; 79.4%), whereas 28 cases (20.6%) were intramedullary tumors. In the majority of cases, the tumor was located in the thoracic spine (n=78/136; 57.4%); for the remaining, there were 27 cases (19.9%) of cervical spine tumors and 25 of lumbar (18.4%). Histologically, schwannomas (n=66/136; 48.5%) and meningiomas (n=39/136; 28.7%) predominated in the cohort; intramedullary tumors included ependymomas (n=18/136; 13.2%) and astrocytomas (n=7/136; 5.1%). Mean clinical and radiographic follow-up duration was 74.5±39.4 months (95% CI, 67.8–81.2 months). Four patients (2.9%) were lost to follow-up, all with intramedullary tumors (Table 1).
Operative and early postoperative course
UH exposure length was deliberately limited (mean, 1.67±0.58 levels; median [IQR], 2 [1–2]); no additional formal facetectomy was performed. Among the operative factors, the mean operative time was 91.0±13.5 minutes (median [IQR], 90 [80–100] minutes), the mean estimated blood loss was 58.5±16.6 mL (median [IQR], 60 [40–70] mL), and the mean length of hospital stay was 3.97±1.05 days (median [IQR], 4 [3–4] days). No perioperative transfusion was required. Only one patient developed a postoperative complication (0.7%), i.e., a cerebrospinal fluid leak requiring duraplasty after air travel. No patient required instrumented fusion during index admission (Table 2).
Extent of resection
Overall, GTR was achieved in 130 cases (95.6%). On comparing compartments, GTR was achieved in all 108 cases (100%) of extramedullary tumors compared to 78.6% for intramedullary tumors (n=22/28). Six (21.4%) of the remaining cases of intramedullary tumors underwent STR followed by postoperative SRT (STR+SRT). This difference in GTR rates for extramedullary versus intramedullary tumors was statistically significant (p<0.001) (Table 3).
Neurological outcomes
Preoperatively, the majority of patients (n=96/136; 70.6%) had AIS grade E neurology, while 40 (29.4%) were graded as AIS D/C. Postoperatively, 127 patients were able to achieve AIS-E status (93.4%); only nine patients (6.6%) retained the AIS-D/C status. In the extramedullary subgroup, the paired change toward ASIA E was statistically significant (p<0.001). Neurological evolution in the intramedullary subgroup is summarized descriptively in Table 4 due to small cell counts.
Oncologic and mechanical durability
Across mid- to long-term follow-up, no radiographic tumor recurrence, regrowth, or progression was observed at the index level, and no patient developed radiographic or symptomatic segmental instability. Forty-three patients (31.6%) underwent at least one dynamic radiographic assessment during follow-up. Among these, none demonstrated pathologic translation or focal kyphosis meeting our predefined instability criteria. Additionally, no patient required delayed instrumented fusion during follow-up (Table 2). Since no oncologic or mechanical failures were observed, Kaplan-Meier survival curves were not informative and are not presented.
In the six cases of intramedullary astrocytomas treated with STR followed by SRT, serial clinical examinations and follow-up MRI of the spinal cord did not reveal any incidence of radiation-related myelopathy, necrosis, or secondary neoplasia during the observation period. However, the small sample size and limited numbers preclude definitive conclusions regarding late radiation toxicity.
Discussion
In this consecutive 10-year cohort of adult patients with intradural spinal tumors, we were able to achieve high GTR rates with UH, resulting in clinically meaningful neurological recovery, exceptionally low perioperative morbidity, and long-term mechanical stability. None of our patients required perioperative or delayed instrumented fusion, including those with intramedullary lesions. These outcomes support UH as an anatomically conservative yet oncologically effective exposure strategy in appropriately selected cases of intradural tumors.
Although historically, TL is the standard strategy for intradural access, it disrupts posterior stabilizing elements and increases the risk of postoperative deformity and potential need for instrumentation [1,2,6,7]. Such stabilization carries the risk of additional morbidity as well as increases operative complexity and cost [6,8]. In contrast, UH preserves the contralateral lamina-facet-tension band complex, limits muscle dissection, and minimizes collateral tissue trauma [1,9,10]. Even with limited unilateral bone removal, UH provides sufficient exposure for safe microsurgical resection while reducing operative trauma and perioperative morbidity [7,11]. This also eliminates metal artifacts on postoperative MRI, facilitating close monitoring [12].
Consistent with our findings, previous comparative studies have shown that UH offers shorter hospitalization, fewer complications, and GTR rates similar to TL, particularly for laterally positioned intradural-extramedullary tumors [6–8,10,11,13–26]. Numerous comparative reports have shown high GTR irrespective of the unilateral corridor [3,5,13,24,26–28]. UH is also feasible in carefully selected intramedullary lesions when aided by contralateral undercutting, spinous base drilling, or tailored microsurgical working angles [1,2,10,11,16,17,23]. In our study, STR was confined to astrocytoma cases, all of which achieved local control with adjuvant SRT. This aligns with the established literature showing low recurrence rates after GTR for benign intradural tumors and the common need for STR+SRT in astrocytoma [8,11,20]. Within this small SRT subgroup, no clinically or radiographically evident late radiation-related complications were observed, but the numbers were too small to draw firm conclusions about the long-term safety of SRT in this context. Because UH is tissue-sparing, it further improves comfort, reduces postoperative pain, and permits earlier mobilization [9,10,24].
To contextualize this cohort, Supplements 1–4 synthesize the existing literature on UH versus TL into four domains: clinical outcomes (SC1), minimally invasive surgery-related perioperative advantages (SC2), economic considerations (SC3), and aggregated perioperative/clinical-economic indicators (SC4). Collectively, these syntheses reinforce the external validity of our findings.
Perioperative efficiency is a consistent advantage of UH, with multiple studies reporting reduced operative time, blood loss, and length of stay compared with TL [6,7,11,23,26]. Previous meta-analyses of the efficacy of UH also demonstrate preserved oncologic adequacy [3,5,28]. These improvements may be attributed to limited unilateral bone removal and reduced muscle trauma with UH [11,23]. In our cohort, operative time, blood loss, and hospitalization mirrored these established patterns. In terms of financial burden, economic benefits in UH arise from the reduced need for instrumentation [6,7], lower complication rates [6,8,10,13–17,19,20,23], possibility of early mobilization [10,14,15,17,18], and shorter hospital stays [6–8,10,11,18–21,23,26]. Some meta-analyses also report cost reductions of up to 30% compared with TL [6,8]. Although formal economic modeling was beyond the scope of our study, the perioperative and short-stay profiles observed in our study support a resource-efficient trend with UH.
Biomechanically, UH preserves key posterior elements and reduces the risk of post-laminectomy deformity or delayed instability [3,5,27–29]. Although Zander et al. [30] noted increased axial rotation after unilateral facetectomy, such changes were not translated into clinically significant instability when careful microsurgical technique was applied [28,29]. In our cohort, no patient demonstrated radiographic or symptomatic instability or required delayed instrumentation.
Fig. 3 summarizes the degree of cross-study agreement on the advantages of UH—prevention of postoperative instability (84%), reduced blood loss (63%), and shorter hospitalization (58%)—demonstrating particularly strong consensus. Parameters with less consistent reporting highlight the need for standardized outcome frameworks for future comparisons on UH versus TL.
These findings affirm UH as an effective, structurally conservative, and resource-efficient exposure strategy for most intradural-extramedullary tumors. For intramedullary lesions, careful anatomical selection is crucial as visualization challenges can limit resection and necessitate adjunctive radiotherapy. On the other hand, more extensive exposures or staged stabilization may be indicated for ventral, midline-crossing, highly vascular, or multilevel lesions. Prospective multicenter studies incorporating predefined biomechanical and economic endpoints are warranted to enable a clearer delineation of the merits of UH versus TL and other minimally invasive approaches.
The current study has several limitations. First the single-center, retrospective design without a control group containing patients operated on with TL or instrumented-fusion procedures limits our ability to draw causal inferences and the external validity of our findings. Our institutional workflow likely favored UH in anatomically suitable lesions while directing ventral, circumferential, multilevel, or severely deformed cases to alternative approaches; hence, selection bias toward favorable anatomy is probable. Second, heterogeneity of histologies and compartments, the small intramedullary/astrocytoma subset, and four losses to follow-up (all astrocytoma) reduced our precision for subgroup analysis and presumably introduced bias in durability estimates despite censoring. Third, biomechanical assessment relied on clinical evaluation combined with dynamic radiographs obtained only when patients reported mechanical pain or instability symptoms, so asymptomatic or purely radiographic deformity could not be excluded. Our findings primarily address clinically relevant instability requiring intervention. Fourth, the extent of resection was determined from early postoperative MRI and operative notes; microscopic residual disease could not be excluded. Furthermore, the use of STR+SRT in all astrocytoma cases probably introduced treatment-selection bias, limiting the ability to extrapolate the results regarding radiation-related toxicity. Fifth, patient-reported outcomes and prospective cost data were not collected; resource use was inferred from perioperative surrogates. Additionally, the absence of observed oncologic or mechanical failures rendered Kaplan-Meier analyses uninformative. Finally, these results reflect a high-volume team and a standardized UH workflow, which may not be universally generalizable to all centers.
Conclusions
A standardized UH yields high GTR rates, meaningful neurological improvement, very low perioperative morbidity, and durable mechanical stability in adult patients with intradural-extramedullary spinal tumors, typically obviating the need for prophylactic or delayed fusion. For intramedullary tumors, particularly astrocytomas, a UH remains feasible but more often necessitates STR with adjuvant SRT and should therefore be reserved for carefully selected cases within a multidisciplinary treatment framework.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement
