1 Introduction

Developmental Cervical Spinal Stenosis (DCSS) is a congenital condition characterized by abnormally narrow bony cervical canal dimensions, predisposing individuals to neural compression and functional deficits.¹ DCSS can impair spinal cord function, leading to symptoms such as sensory disturbances, motor dysfunction, and muscle weakness.² Patients with DCSS face an increased risk of spinal cord injury during daily activities due to degenerative changes like disc herniation or osteophyte formation causing neural compression and degeneration. Under certain circumstances, such as trauma or progressive degeneration, this stenosis can worsen, resulting in more severe neurological impairment.³

The modified Japanese Orthopaedic Association (mJOA) score is an investigator-administered tool used to evaluate neurological function in patients with DCM.⁴ The mJOA is the second most commonly used outcome measure to quantify severity in patients with DCM.⁵ The mJOA consists of upper (5 points) and lower extremity (7 points) motor function, sensation (3 points), and micturition (3 points) for a total of 18 points, with a perfect score indicating no neurological deficits. Although the mJOA score is widely used in clinical practice, it may not adequately reflect whether an intervention provides meaningful benefits for individual patients. The Minimum Clinical Important Difference (MCID) addresses this gap by defining the smallest change in patient-reported or functional outcome measures that can be perceived as beneficial by patients or clinicians.⁶ The MCID has increasingly become an important indicator in assessing postoperative recovery in patients with cervical spondylosis, as even minor improvements in neurological function or pain relief can significantly impact a patient's quality of life. In cervical spine research, MCID thresholds for key outcomes have been validated through consensus and longitudinal studies: for the mJOA score, an absolute increase of ≥2 points is widely accepted as clinically meaningful,⁷ corresponding to a noticeable reduction in neurological deficits (e.g., improved gait, hand dexterity).

DCSS is recognized as a known predisposing factor for cervical spondylosis, particularly Degenerative Cervical Myelopathy (DCM). Its potential multilevel nature influences the clinical presentation and treatment strategies for cervical spondylosis.⁸ Concurrently, selecting the optimal surgical strategy for multilevel cervical spondylosis remains complex. Options typically include anterior, posterior, or combined approaches, each with specific advantages and disadvantages. The choice involves multiple factors, including the patient's specific condition, symptom severity, neurological status, and imaging characteristics. Controversy exists regarding the optimal surgical management for multilevel cervical spondylosis with varying degrees of stenosis. This study aims to investigate appropriate surgical strategies by integrating imaging findings and clinical functional assessments.

While existing studies broadly support anterior approaches for mild stenosis and posterior for severe cases, there remains a lack of consensus on the optimal approach for moderate stenosis and quantitative thresholds for decision-making. This study uniquely combines Pavlovian ratio stratification with clinical indicators as well as imaging parameters to establish a severity-based strategy for surgical selection.

2 Methods

2.4

2.4 Radiological and imaging evaluations

Group stratification was based on the preoperative Pavlov ratio measured on lateral cervical radiographs. The Pavlov ratio was defined as the sagittal diameter of the cervical canal divided by the sagittal diameter of the corresponding vertebral body (Fig. 1). The mean Pavlov ratio was calculated as the average of the ratios at C3-C7 levels. Patients were classified as: Mild stenosis group (mean Pavlov ratio 0.71–0.75), Moderate stenosis group (mean Pavlov ratio 0.66–0.70), and Severe stenosis group (mean Pavlov ratio ≤0.65). Cervical radiographs were obtained preoperatively and at 6 months postoperatively to measure the C2-7 Cobb angle and Sagittal vertical axis (SVA). Preoperative magnetic resonance imaging (MRI) was performed to measure maximum spinal cord compression (MSCC), spinal cord occupation ratio (SCOR), compression ratio (CR).

Close

Fig. 1 Schematic diagram for measuring Pavlov ratio.

Export to PPT

2.4.1

2.4.1 C2–7 Cobb angle

The C2–7 Cobb angle was defined as the angle between lines drawn tangential to inferior endplates of C2 and C7(Fig. 2).

Close

Fig. 2 (a–e) Schematic diagram of the measurement of C2-7 Cobb angle, SVA, MSCC, SCOR and CR.

Export to PPT

2.4.2

2.4.2 Sagittal vertical axis (SVA)

The C2–7 sagittal vertical axis was defined as the horizontal distance between the vertical line from the center of the C2 vertebral body and the posterosuperior corner of the C7 vertebral body(Fig. 2).

2.4.3

2.4.3 Maximum spinal cord compression (MSCC)

The maximum spinal cord compression was defined as the ratio of the midsagittal diameter of the spinal cord at the compression site divided by the average diameter of the spinal cord at the closest non-compressed regions above& below(Fig. 2).

2.4.4

2.4.4 Spinal cord occupation ratio (SCOR)

The spinal cord occupation ratio was defined as the ratio of the sum of the cord width above and below, and the sum of the canal width above and below the point of compression(Fig. 2).

2.4.5

2.4.5 Compression ratio (CR)

The compression ratio was defined as the ratio of the sagittal diameter divided by the transverse diameter of the spinal cord observed on axial T1WI(Fig. 2).

3 Results

3.2

3.2 Clinical data

Table 2 shows the differences in clinical data among the three stenosis groups stratified by surgical approach. In the severe stenosis group, there were significant differences between the anterior and posterior approaches in terms of intraoperative blood loss, postoperative mJOAs, RR, and postoperative VAS scores. The differences in postoperative mJOAs and RR remained statistically significant after Bonferroni correction. The posterior approach group showed a trend towards lower postoperative VAS scores compared to the anterior group (P = 0.049), though this difference did not reach statistical significance after Bonferroni correction, suggesting limited clinical relevance. In the moderate stenosis group, intraoperative blood loss was significantly higher in the posterior group than in the anterior group, but no significant differences were found in other parameters. In the mild stenosis group, intraoperative blood loss was higher in the posterior group, whereas postoperative mJOAs and RR were higher in the anterior group, and all three differences were statistically significant after Bonferroni correction.

Table 2 Differences in clinical indicators in patients with cervical spinal stenosis undergoing different surgical approaches.

	Severe stenosis group				Moderate stenosis group				Mild stenosis group
	AA(n = 12)	PA(n = 32)	P value	95 %CI	AA(n = 24)	PA(n = 28)	P value	95 %CI	AA(n = 50)	PA(n = 22)	P value	95 %CI
Surgery time	118.33 ± 23.17	152.00 ± 63.01	0.310	(-83.45, 16.11)	111.25 ± 36.03	122.00 ± 54.04	0.673	(-43.18, 21.68)	103.93 ± 48.76	110.00 ± 25.82	0.816	(-37.59, 25.45)
Intraoperative hemorrhage	35.00 ± 20.74	120.00 ± 44.72	0.004∗∗	(-132.64, −37.36)	56.25 ± 17.68	192.00 ± 148.73	0.006∗∗	(-240.52, −30.98)	49.29 ± 12.07	220.00 ± 155.78	0.001∗∗∗#	(-246.59, −94.83)
Preoperative mJOAs	12.00 ± 0.63	12.25 ± 0.96	0.629	(-0.81, 0.31)	14.38 ± 0.52	14.40 ± 1.52	0.966	(-0.69, 0.65)	15.21 ± 0.97	14.50 ± 0.53	0.072	(-0.04, 1.46)
Postoperative mJOAs	14.17 ± 0.75	16.25 ± 0.50	0.001∗∗∗#	(-2.81, −1.35)	17.25 ± 0.89	16.60 ± 0.89	0.226	(-0.22, 1.52)	17.07 ± 0.92	15.63 ± 0.74	0.001∗∗∗#	(0.79, 2.09)
Recovery rate	0.36 ± 0.11	0.70 ± 0.08	0.001∗∗∗#	(-0.43, −0.25)	0.80 ± 0.23	0.62 ± 0.11	0.078	(-0.01, 0.37)	0.72 ± 0.25	0.33 ± 0.14	0.000∗∗∗##	(0.27, 0.51)
MCID achievement rate	58.3 % (7/12)	90.6 % (29/32)	0.002∗∗	(-57.6, −7.0)	79.2 % (19/24)	82.1 % (23/28)	0.821	(-22.1, 16.3)	56.0 % (28/50)	31.8 % (7/22)	0.008	(2.6, 45.8)
Preoperative VAS	7.66 ± 1.97	7.02 ± 0.82	0.194	(-0.22, 1.50)	3.89 ± 1.27	5.00 ± 1.67	0.166	(-2.44, 0.22)	4.36 ± 1.55	4.77 ± 1.42	0.479	(-1.25, 0.43)
Postoperative VAS	3.17 ± 0.75	2.05 ± 0.82	0.049∗	(0.01, 2.23)	1.67 ± 0.71	2.17 ± 1.17	0.318	(-1.23, 0.23)	1.64 ± 1.55	2.31 ± 1.11	0.215	(-1.54, 0.20)

Table 3 shows the differences in clinical data for the three stenosis groups stratified by surgical approach. In the anterior surgical group, the preoperative mJOAs, postoperative mJOAs, RR, and postoperative VAS were statistically different among the three stenosis groups, and the differences in preoperative mJOAs and postoperative mJOAs remained statistically significant after Bonferroni correction. The mild stenosis group had the highest preoperative mJOAs among the three groups, whereas the moderate stenosis group had the highest postoperative mJOAs among the three groups. In the posterior surgery group, the preoperative mJOAs and RR values were statistically different among the three groups.

Table 3 Differences in clinical indicators in patients with different levels of cervical stenosis under the same surgical approach.

	AA					PA
	Severe stenosis	Moderate stenosis	Mild stenosis	P value	95 %CI	Severe stenosis	Moderate stenosis	Mild stenosis	P value	95 %CI
Surgery time	118.33 ± 23.17	108.89 ± 34.44	103.93 ± 48.76	0.770	(105.62, 116.58)	160.35 ± 69.76	116.67 ± 50.07	110.00 ± 25.82	0.349	(118.73, 158.27)
Intraoperative hemorrhage	35.32 ± 20.74	51.11 ± 22.60	49.29 ± 12.07	0.195	(39.85, 50.67)	125.47 ± 50.09	176.67 ± 138.23	220.28 ± 155.79	0.583	(146.32, 209.43)
Preoperative mJOAs	12.01 ± 0.63	14.56 ± 0.73	15.21 ± 0.97	0.000∗∗∗##	(13.26, 14.84)	12.25 ± 0.96	14.33 ± 1.37	14.75 ± 0.50	0.014∗	(13.01, 14.49)
Postoperative mJOAs	14.17 ± 0.75	17.22 ± 0.83	17.07 ± 0.92	0.000∗∗∗##	(15.31, 16.93)	16.25 ± 0.50	16.67 ± 0.82	16.00 ± 0.82	0.391	(16.09, 16.58)
Recovery rate	0.36 ± 0.11	0.80 ± 0.21	0.71 ± 0.25	0.002∗∗	(0.52, 0.70)	0.70 ± 0.08	0.64 ± 0.11	0.40 ± 0.18	0.016∗	(0.51, 0.65)
Preoperative VAS	4.50 ± 1.97	3.89 ± 1.27	4.36 ± 1.55	0.709	(3.98, 4.52)	6.00 ± 0.82	5.00 ± 1.67	4.75 ± 1.50	0.445	(4.96, 5.84)
Postoperative VAS	3.17 ± 0.75	1.67 ± 0.71	1.64 ± 1.55	0.038∗	(1.86, 2.68)	2.00 ± 0.82	2.17 ± 1.17	2.25 ± 1.50	0.955	(2.03, 2.27)

Fig. 3 illustrates the trends in preoperative and postoperative mJOAs for each group. All patients showed improvement in mJOAs postoperatively, with scores gradually increasing over time. In the severe stenosis group at 2 months postoperatively, patients undergoing posterior surgery showed a substantial increase in mJOAs compared to anterior surgery, remaining significantly higher at 1 year. In the moderate stenosis group at 2 months, improvement was similar between anterior and posterior approaches. At 6 and 12 months, mJOAs were slightly higher in the anterior group. In the mild stenosis group, mJOAs improvement was slightly greater after anterior surgery compared to posterior surgery.

Close

Fig. 3 The change trends of mJOAs before and after surgery in patients with cervical spinal stenosis in different subgroups.

Export to PPT

MCID achievement rates: Severe stenosis-Anterior 58.3 % (7/12) vs Posterior 90.6 % (29/32); Moderate stenosis-Anterior 79.2 % (19/24) vs Posterior 82.1 % (23/28); Mild stenosis-Anterior 56.0 % (28/50) vs Posterior 31.8 % (7/22).

4 Recovery Rate

Patients were divided into two groups based on an RR threshold of 50 %. Table 6 results show that the group with RR <50 % had significantly lower postoperative mJOAs, higher postoperative VAS scores, lower mean Pavlov ratio, and smaller CR.

A binomial logistic regression analysis was performed to identify the risk factors associated with poor postoperative functional recovery (mJOA improvement rate <50 %) in patients. The results revealed that preoperative mJOAs, postoperative mJOAs, postoperative VAS, average Pavlov ratio, and CR were independent risk factors (P < 0.05). Among these, a decrease in the CR value exhibited the strongest predictive efficacy (OR = 0.012, Table 7).

5 Discussion

Multilevel developmental cervical spinal stenosis (DCSS) is a congenital bony spinal canal narrowing disorder, commonly complicated by degenerative cervical pathologies. This results in multilevel compression of the spinal cord and nerve roots, leading to myelopathy (e.g., paresthesia, motor deficits) and radiculopathy.⁹ For patients with developmental spinal stenosis, the reduced spinal cord buffer space means even minor degeneration or trauma can lead to significant cord compression and related clinical symptoms, posing considerable treatment challenges. Significant controversy exists regarding the optimal surgical strategy for multilevel cervical stenosis with cord compression: Anterior approaches (e.g., ACDF) directly address ventral compression, restore cervical lordosis, and restore disc height through structural interbody grafting.¹⁰ However, they carry a risk of incomplete decompression in severe stenosis and have been reported¹¹ to have higher failure rates when treating three or more levels. Posterior approaches (e.g., open-door laminoplasty) are considered ideal for multilevel degenerative cervical disease,¹² providing extensive decompression by enlarging the spinal canal volume, allowing the cord to float away from ventral compression.^11,13 However, they disrupt posterior elements, potentially affecting sagittal balance, and are generally more invasive than anterior approaches. Dai et al.¹⁴ compared ACDF and laminoplasty for multilevel cervical spondylotic myelopathy with developmental canal stenosis, suggesting ACDF is preferable for patients with poor cervical lordosis or significant neck pain, while laminoplasty suits those with good lordosis. While their study compared clinical and radiological outcomes of anterior and posterior surgery for multilevel developmental stenosis, it did not specifically analyze outcomes based on the degree of stenosis. Therefore, selecting the optimal surgical approach based on stenosis severity requires further investigation.

This retrospective analysis of 168 patients with multilevel cervical stenosis and varying DCSS severity (Severe: Pavlov ratio ≤0.65; Moderate: 0.66–0.70; Mild: 0.71–0.75) evaluated differences in outcomes between anterior and posterior approaches. The mJOA score, a key indicator of cervical myelopathy severity, was primarily applied for the assessment of changes in spinal neural function. The RR and VAS scores provided further assessment of symptomatic improvement. The results demonstrated: For patients with severe stenosis, posterior surgery significantly outperformed anterior surgery in improving postoperative mJOAs and RR (P < 0.05). Patients with severe DCSS exhibit significant stenosis across multiple levels; posterior surgery achieves adequate canal volume expansion, more thoroughly decompressing the multilevel cord compression. Anterior decompression primarily targets the disc space level; even with corpectomy (ACCF), decompression is mainly effective at the resected vertebral body level, leaving stenosis at other levels unaddressed, limiting overall decompression efficacy. For moderate stenosis patients, outcomes between the two approaches showed no significant difference, suggesting surgical choice can be flexible, based on the primary source of compression (anterior vs. posterior) and cervical sagittal parameters. For mild stenosis patients, anterior surgery yielded significantly higher postoperative mJOAs and RR (P < 0.05). This is likely because the primary compressive pathology in mild DCSS stems from anterior structures (e.g., discs, osteophytes), with the developmental stenosis itself contributing minimally to the neural compromise. Anterior surgery directly removes anterior compressive elements, is less invasive, and better preserves physiological cervical lordosis.

In terms of imaging, the lack of statistically significant differences between groups does not mean that surgery has no effect on imaging parameters. On the contrary, these data indicate that although surgery may alter certain radiological parameters, the extent or pattern of these changes does not differ significantly between anterior and posterior surgical approaches, or between groups with different degrees of stenosis. In Table 4, within each stenosis subgroup, both approaches showed measurable changes in lordosis: Severe stenosis: anterior approach resulted in a net increase (ΔC2-7 Cobb angle = 5.78 ± 4.03), and posterior approach in a smaller increase (ΔC2-7 Cobb angle = 2.75 ± 1.71). While the difference between approaches was not significant (P = 0.342), the positive ΔC2-7 Cobb angle values confirm that both surgeries tended to improve lordosis in this subgroup. Moderate stenosis: anterior approach showed a mild reduction (ΔC2-7 Cobb angle = −4.37 ± 1.71), and posterior approach a mild increase (ΔC2-7 Cobb angle = 1.58 ± 0.67). Again, no significant intergroup difference (P = 0.015, uncorrected) was observed, but the directional changes indicate surgical effects on spinal alignment. Mild stenosis: anterior approach yielded a small increase (ΔC2-7 Cobb angle = 3.80 ± 0.11), and posterior approach a reduction (ΔC2-7 Cobb angle = −11.31 ± 17.43). The lack of significance (P = 0.103) reflects variability in mild cases but does not negate the fact that surgery altered lordosis in individual patients. Changes in the SVA also indicate the impact of surgery on sagittal balance, even though there is no statistical significance between groups. The Pavlov ratio, a static indicator of baseline stenosis severity, showed no significant postoperative changes between groups, which aligns with its role as an anatomical descriptor rather than a dynamic outcome. However, this stability does not negate surgical impact on other parameters; instead, it confirms that the Pavlov ratio serves primarily to stratify preoperative stenosis, not to reflect surgical effects. Our radiological data (Tables 4 and 5) indicate that while surgery does alter the curvature of the cervical spine to some extent, the lack of significant differences between groups means we cannot conclude that one method is “superior” in radiological terms. We believe that one should not claim the “superiority” of a particular surgical approach based solely on radiological results; rather, these results reflect individual patient responses to surgery, and their clinical significance should be interpreted in conjunction with functional improvements.

The minimal clinically important difference (MCID) is defined as the smallest change recognized as clinically meaningful by a patient.¹⁵ When evaluating surgical efficacy for cervical myelopathy, MCID provides more clinically relevant information than statistical significance alone. Tetreault et al.⁷ established an MCID threshold of 2 points for the mJOAs in DCM. Applying this standard, our study assessed the proportion of patients achieving clinically meaningful improvement across stenosis severities and surgical approaches. The results showed that for patients in both the severe and moderate stenosis groups, regardless of whether they underwent anterior or posterior surgery, the average improvement in mJOA scores exceeded the 2-point MCID threshold. This strongly supports the clinical effectiveness of surgical intervention for patients with moderate-to-severe DCSS and multilevel cord compression.

Notably, within the mild stenosis group, although anterior surgery yielded significantly better postoperative mJOA scores and RR compared to posterior surgery (P < 0.05), the average mJOA score improvement for the entire mild stenosis group did not reach the 2-point MCID threshold (only 58 % of these patients achieved the MCID threshold). This is primarily attributed to the relatively higher preoperative mJOA baseline scores in this group (indicating milder initial neurological deficits), limiting the potential for substantial postoperative gains. However, this does not mean that surgery is ineffective for patients with mild stenosis, as statistical significance and MCID are two different concepts. The actual surgical effect should be considered in combination with the patient's symptoms, clinical scores, and imaging.

The MCID analysis provides an additional dimension of support for the core conclusions of this study: For patients with moderate-to-severe stenosis, surgery via the appropriate approach delivers a patient-perceivable, clinically meaningful improvement in neurological function. For mild stenosis patients, while the average improvement fell short of the MCID threshold, surgical intervention, particularly anterior surgery, still show a more positive trend in efficacy compared to posterior surgery, requiring more personalized evaluation of surgical indications and expected benefits.

A binary logistic regression analysis indicated that preoperative mJOAs, postoperative mJOAs, postoperative VAS, average Pavlov ratio, and CR were independent risk factors for poor postoperative functional recovery. Among these, a decreased CR value exhibited the strongest predictive efficacy (OR = 0.012). CR reflects the flattening degree of the spinal cord cross-section, and a reduced ratio suggests sagittal diameter narrowing due to anterior-posterior compression of the spinal cord. When CR < 0.35, the risks of spinal cord microcirculatory disturbance and axonal injury significantly increase, which is consistent with the "spinal cord-canal mismatch" theory proposed by Nouri et al..²

This study is the first to systematically investigate the relationship between stenosis severity and surgical approach, providing a basis for individualized treatment: For patients with severe stenosis, posterior surgery may show a certain tendency in maintaining cervical spine curvature stability. For patients with moderate stenosis, there are no significant differences in imaging parameters between anterior and posterior surgeries. The choice of surgical approach should focus more on the patient's specific source of compression, baseline cervical spine curvature status, and the surgeon's experience preferences. The weight of individualized assessment should be higher than the simple grading of stenosis severity. For patients with mild stenosis, anterior surgery may be more suitable as an initial consideration; however, this preference should be adjusted based on individual factors such as the severity of clinical symptoms (e.g., presence of persistent neurological deficits) and cervical spine mobility, to avoid making decisions solely based on radiographic stenosis severity. In addition, higher intraoperative blood loss during posterior surgery suggests that patient-specific factors (such as anemia or coagulopathy) should be considered. Therefore, intraoperative blood loss is also an important factor when considering the specific surgical approach for a patient.

Key limitations include: (1) Single-center retrospective analysis, operative selection may be influenced by surgeon experience, leading to selection bias; (2) Short follow-up (≤1 year) insufficient to assess long-term sagittal balance deterioration, adjacent segment degeneration or reoperation rates. While long-term outcomes are essential, our 1-year data provide meaningful insights into early safety (e.g., no major complications) and preliminary efficacy (sustained neurological improvement), which are clinically relevant for guiding initial surgical planning. In the study, the Bonferroni method was used to correct for multiple comparisons, and some of the original significant outcomes were no longer significant after correction, suggesting that the between-group differences in these indicators may be affected by random errors, and that clinical decision-making should prioritize outcomes that are still significant after correction (e.g., mJOAs and RR). In addition, the limited sample size of the subgroups may have resulted in insufficient efficacy of some analyses, and future studies may adopt a multicenter design to expand the sample size to enhance the reliability of the results.

6 Conclusion

The selection of a surgical approach for multilevel cervical canal stenosis should be primarily guided by the degree of stenosis. This study confirms that the efficacy difference between anterior and posterior surgery is closely related to stenosis severity. Precise preoperative assessment (e.g., Pavlov ratio, source of cord compression) is crucial for optimizing surgical strategy and improving prognosis.

Patient's consent

This study was conducted in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Written informed consent was obtained from all individual participants included in the study.

Ethics approval and consent to participate

Approval for the current study protocol was obtained from the ethics committees of Shanghai Changzheng Hospital (2021SL004). The norms on which the study is based are in accordance with the "Declaration of Helsinki".

Consent for publication

Not Applicable.

Availability of data and materials

All the data of the manuscript are presented in the paper.

Author contributions

Conception and design: Yang Liu and Xuhong Zhang; Acquisition of data: Hanlin Song, Zichuan Wu, Junbin Liu and Junzhe Sheng; Data analysis and interpretation: Baifeng Sun, Chen Xu, and Min Qi; Statistical analysis: Hanlin Song, Zichuan Wu, and Xuhong Zhang; Manuscript Preparation: Xuhong Zhang, Zichuan Wu; Manuscript revision and modification: Min Qi and Yang Liu.

Funding

This work is funded by the Shanghai Municipal Education Commission (2023 Science and Technology 05–60).