Current understanding of tandem spinal stenosis: epidemiology, diagnosis, and surgical strategy.

Tandem spinal stenosis (TSS) is defined as the concomitant occurrence of stenosis in at least two or more distinct regions (cervical, thoracic, or lumbar) of the spine and may present with a constellation of signs and symptoms. It has four subtypes, including cervico-lumbar, cervico-thoracic, thoraco-lumbar, and cervico-thoraco-lumbar TSS. The prevalence of TSS varies depending on the different subtypes and cohorts. The main aetiologies of TSS are spinal degenerative changes and heterotopic ossification, and patients with developmental spinal stenosis, ligament ossification, and spinal stenosis at any region are at an increased risk of developing TSS. The diagnosis of TSS is challenging. The clinical presentation of TSS could be complex, concealed, or severe, and these features may be confusing to clinicians, resulting in an incomplete or delayed diagnosis. Additionally, a consolidated diagnostic criterion for TSS is urgently required to improve consistency across studies and form a basis for establishing treatment guidelines. The optimal treatment option for TSS is still under debate; areas of controversies include choice of the decompression range, choice between simultaneous or staged surgical patterns, and the order of the surgeries. The present study reviews publications on TSS, consolidates current awareness on prevalence, aetiologies, potential risk factors, diagnostic dilemmas and criteria, and surgical strategies based on TSS subtypes. This is the first review to include thoracic spinal stenosis as a candidate disorder in TSS and aims at providing the readers with a comprehensive overview of TSS.

Introduction

Spinal stenosis is characterised by a reduction in the cross-sectional area of the spinal canal that leads to upper or lower motor neuron deficits and related neurological symptoms depending on the location of the compression. Spinal stenosis can affect more than one level in any segments (1, 2, 3). Tandem spinal stenosis (TSS) refers to concomitant stenosis that affects at least two regions (4).

TSS was first reported in 1957 by Brain et al., who described a patient with cervical spinal stenosis (CSS) and lumbar spinal stenosis (LSS) (5). Subsequently, cases of seven patients with concomitant CSS and LSS were reported by Teng and Papatheodorou (6). The term ‘tandem spinal stenosis’ was first proposed by Dagi et al. to distinguish patients with concurrent cervicolumbar TSS (CLTSS) (4). In 1962, Koizumi reported a case of CSS and thoracic spinal stenosis (ThSS) (7) that broadened the scope of TSS by adding ThSS to stenotic candidate regions (8, 9). To date, TSS is defined as distinct concomitant stenosis in at least two regions of the cervical, thoracic, and lumbar spine and may present with both upper and lower motor neuron symptoms and neurogenic claudication (10, 11, 12, 13).

TSS is classified into four subtypes according to the stenotic region: cervicothracic TSS (CTTSS), thoracolumbar TSS (TLTSS), cervico-thoraco-lumbar TSS (CTLTSS), and CLTSS (14) (Fig. 1). It presents with complicated manifestations that cause difficulties in both diagnosis and treatment. Herein, we review the current understanding of TSS including the subtypes with ThSS, aimed to clarify the epidemiology and diagnosis of TSS and to discuss surgical strategies for the different subtypes of TSS.

Figure 1

The representative MRI of the subtypes of TSS. The whole spine MRI of a CLTSS patient shows CSS at C4-7 (A1, arrows) and LSS at L3-4 (A3, arrows); In the MRI of a CTTSS patient, extensive hypertrophy of posterior longitudinal ligament causes CSS (B1, arrows) and ThSS (B2, arrows), the stenotic change affects C4-T5 (B1, 2); In the MRI of a TLTSS patient, T3-5 ThSS (C2, arrows) and L2-S1 LSS (C3, arrows) occurs concomitantly; A CTLTSS patient has concurrent C2-6 CSS (D1, arrows), T4-5, T10-12 ThSS (D2, arrows), and L2-S1 LSS (D3, arrows), which are caused by heterotopic ossification and degeneration changes (D1-3). CSS, cervical spinal stenosis; CLTSS, cervico-lumbar tandem spinal stenosis; CTTSS, cervico-thoracic tandem spinal stenosis; CTLTSS, cervico-thoraco-lumbar tandem spinal stenosis; LSS, lumbar spinal stenosis; TSS, tandem spinal stenosis; ThSS, thoracic spinal stenosis; TLTSS, thoraco-lumbar tandem spinal stenosis.

Many studies have attempted to determine the prevalence of TSS (Table 1). According to previous reports, the incidence of radiological and symptomatic TSS are 8–60% and 5–28%, respectively (2, 4, 12, 15, 16, 17). However, these data do not reflect the precise status. Based on studies of volunteers and unrelated patients, the incidence of CLTSS is 0.12–11% in the general population (16, 17, 18, 19, 20). However, the incidence of other subtypes was not mentioned. In anatomical studies, two research groups examined the prevalence of TSS by measuring the same collection of cadavers. The incidence of CLTSS in the cadavers ranged from 0.9 to 5.4% (21), and the incidence of CTTSS and TLTSS is 1 and 1.42%, respectively (22, 23). Limited by the cadaveric nature, despite the good consistency, the results can merely reflect TSS caused by developmental or acquired narrowing of bony structures (24). Most prevalence studies on TSS are based on patients with spinal stenosis. The radiological CLTSS incidence among patients with CSS or LSS is 32.2–84.6%, and the symptomatic CLTSS incidence was 0.88–33.3% (25, 26, 27, 28, 29, 30, 31, 32). The incidence of radiological CTTSS and TLTSS ranges from 12.2 to 70% among patients with spinal stenosis, and the incidence of radiological CTLTSS among patients with spinal stenosis is 10–25.7% (14, 33, 34).

Table 1

Prevalence of TSS.

Study design	Sample source	Sample composition			Conclusion	Reference
		n	Sex	Age, years
TSS prevalence in cadavers
Cadaveric study	Cadaver specimens	440	N/A	N/A	LSS in 16.8%; CSS in 21.5%; CLTSS in 5.4%; CSS in LSS 32.4%; LSS in CSS 25.3%	Lee et al. (21)
Cadaveric study	Cadaver specimens	1072	M:882; F: 190	15 – 114*	CLTSS in 2% and CTTSS in 1%; A cadaver with a higher number of LSS segments is more likely to have CSS at the same time, and vice versa.	Bajwa et al. (22), Bajwa et al. (23)
CLTSS prevalence
Retrospective study	Patients receiving surgical decompression for cervical, thoracic, or lumbar stenosis	1603	N/A	N/A	TSS in 2.06%	Bhandutia et al. (11)
Prospective cross-sectional study	Patients with symptomatic LSS	78	M:48; F:30	66 (53–82)	Asymptomatic CSM in 84.6%; Symptomatic CSM in 16.7%; 33.3% CSM patients had symptomatic LSS	Adamova et al. (25)
Retrospective study	Patients who had undergone cervical laminoplasty for CSM	214	M:153; F:61	62.8 (29–85)	LSS in 32% (symptomatic 13%, asymptomatic 19%)	Tsutsumimoto et al. (26)
Retrospective study	Patients who had undergone cervical surgery for CSM	297	M:201; F: 96	65.6 (27–93)	CLTSS in 57.9% (15% underwent a lumbar operation)	Yamada et al. (27)
Retrospective study	Patients who had undergone lumbar surgery for symptomatic LSS	565	M:279; F:286	70.7 (32–92)	CLTSS in 35.8% (5% underwent a cervical operation)	Yamada et al. (28)
Retrospective study	Patients with symptomatic LSS	237	M:117; F: 120	68.8 (45–87)	CSM in 8.86%; Men have a significantly higher incidence of CSM.	Iizuka et al. (29)
Retrospective study	Elderly patients with symptomatic LSS	101	M:39; F: 62	71 (65–86)	CSS in 77 (76.2%); ThSS in 30 (29.7%); CTTSS in 26 (25.7%); There was a correlation between the symptom duration of LSS and the prevalence of both ThSS and CSS.	Lee et al. (30)
Retrospective study	Patients received surgery for symptomatic spinal stenosis	1023	N/A	N/A	TSS in 0.88%	Molinari et al. (32)
Retrospective study	Patients received surgery for symptomatic spinal stenosis	230	N/A	N/A	CLTSS in 3.4%	Aydogan et al. (70)
Retrospective study	Patients received surgery for symptomatic spinal stenosis	158	N/A	N/A	CLTSS in 7.6%	Hsieh et al. (71)
CTLTSS and TSS with ThSS prevalence
Retrospective study	Patients who had undergone thoracic surgery for ThSS	50	M:32; F:18	68.4 (40–86)	Concurrent LSS or CSS in 70%; CTLTSS in 10%	Uehara et al. (14)
Retrospective study	Elderly patients with LSS	460	M:101; F:359	83.3 (80–98)	CSS in 110 (23.9%); ThSS in 112 (24.3%); CLTSS in 26 (11.7%); CTTSS in 56 (12.2%); CTLTSS in 56 (12.2%)	Park et al. (33)
Retrospective study	Elderly patients with symptomatic LSS	101	M:39; F:62	71 (65–86)	CSS in 77 (76.2%); ThSS in 30 (29.7%); CTTSS in 26 (25.7%); There was a correlation between the symptom duration of LSS and the prevalence of both ThSS and CSS.	Kim et al. (34)
TSS prevalence in specific population
Retrospective study	Japanese residents	931	M:627; F:304	68.2 (40–93)	Radiographic CLTSS in 11.0%; Radiographic CLTSS more prevalent in those with developmental canal stenosis; Symptomatic LSS in radiographic CLTSS was 18.6%; CSM in radiographic CLTSS was 9.8%; Symptomatic CLTSS in LSS was 6.1%	Nagata et al. (16)
MRI study	Asymptomatic patients who have undergone cervical and lumbar MRI	94	M:48; F:46	48.0 ± 13.4	CSS in 13.8%; LSS in 12.8%; CLTSS in 4%	Matsumoto et al. (17)
Retrospective study	Patients diagnosed with acoustic tumor by myelography	300	M:159; F:141	51 (18–76)	LSS in 24%; CSS in 21%; CLTSS in 8%	Hitselberger et al. (18)
Retrospective study	Hospital admissions	460964	N/A	N/A	CLTSS in 0.12%	LaBan et al. (19)
MRI study	Patients underwent lumbar spine MRI	2113	N/A	N/A	CLTSS in 1.9%	Seo et al. (20)

*Age of death.

CSM, cervical spondylotic myelopathy; CSS, cervical spinal stenosis; CLTSS, cervico-lumbar tandem spinal stenosis; CTTSS, cervico-thoracic tandem spinal stenosis; CTLTSS, cervico-thoraco-lumbar tandem spinal stenosis; LSS, lumbar spinal stenosis; N/A; not available; TSS, tandem spinal stenosis; ThSS, thoracic spinal stenosis; TLTSS, thoraco-lumbar tandem spinal stenosis.

Aetiology and potential risk factors of TSS

TSS has two main aetiologies that can lead to stenosis in combination or separately. Spondylotic changes caused by degeneration is the main cause of LSS and CSS. CLTSS is caused by degenerative spondylosis due to aging, lifestyle, and other stimuli. The other aetiology is heterotopic ossification, including ossification of the posterior longitudinal ligament (OPLL) and the ligamentum flavum (OLF), which are commonly seen in ThSS and some patients with CSS (7, 35). CTTSS, TLTSS, and CTLTSS are more likely to be caused by heterotopic ossification, which is associated with genetic abnormalities in backgrounds and metabolism (36, 37). Based on the aetiologies, we reviewed the potential risk factors of TSS, which are important to arrive at a timely diagnosis.

Developmental spinal stenosis (DSS) is defined as the reduction of the area of the spinal canal crossing at the pedicle level in multiple or all segments (1). Compared with other patients, the incidence of TSS is higher among patients with DSS (16, 38). Torg-Pavlov rate (TPR) is an index for DSS diagnosis and is the only TSS predictor with statistical significance (39, 40), with a predictive value of less than 0.75–0.78 (16, 29). Although poor reliability of TPR as a cervical anatomical measurement has been reported (41), low TPR in one region is still a predictor of co-existing stenosis in valid cases (42, 43). Thus, TPR assessment should be routinely performed for patient stratification.

Patients with stenosis in one region are likely to develop TSS (29, 44). Several studies have shown that patients who already have CSS or LSS are likely to have stenosis in other regions (21, 30, 38, 45). The presence of CSS positively predicts LSS in 16.7% of cases, and the presence of LSS predicts CSS in 15.3% of cases (21). Additionally, the incidence of coexisting stenosis correlates with the severity and morphology of pre-existing stenosis (30, 43).

OPLL and OLF can cause extensive stenosis, leading to TSS (9). Cervical OPLL frequently extends to the upper thoracic spine and leads to CTTSS, and the incidence of CLTSS is significantly higher among patients with cervical OPLL than among non-OPLL patients (9, 27, 46, 47). OPLL and OLF often occur simultaneously. In the Japanese population, 64.6% of patients with cervical OPLL also have a coexisting OLF (36), over 50% of those with thoracic OPLL have a coexisting cervical OPLL, and 46% of individuals with thoracic OPLL also have thoracic OLF (46). Furthermore, Liang et al. showed that nearly 50% of Chinese patients with OPLL have thoracic OPLL (47). Patients with OPLL or OLF in the lumbar region have a high rate (60%) of coexisting stenosis in other regions, and the rate of concurrent stenosis could increase to 75% when patients have both OPLL and OLF in their lumbar region (47, 48). The extensive and skipping occurrence of OPLL and OLF and their close association with spinal stenosis might play a critical role and have predictive significance in the development of CTLTSS.

Diagnosis dilemmas of TSS

Clinical features and diagnostic difficulties

TSS symptoms are composites of cord compression with or without LSS (Table 2). TSS manifests as features of complexity, concealment, and severity. It usually presents as a mixture of upper and lower motor neuron deficit symptoms and signs. In TSS, upper motor neuron symptoms and long track signs can be blunted or even masked by lower motor neuron symptoms. Hyperreflexia caused by CSS or ThSS may be obscured by compression of the cauda equina, and lower-limb pain induced by conus compression can be covered by radiculopathy (14, 49). Until the aggravation of concomitant compressions, referable findings may gradually emerge, composing the symptom constellation of typical TSS.

Table 2

The subtypes and the clinical manifestation of TSS.

Subtype	Definition	Clinical presentation	Characteristic
CTTSS	TSS involves CSS and ThSS	Neck and back pain; Neurological symptoms: muscle weakness on upper and lower extremities, radiculopathy and hypoesthesia on upper extremities, hypoesthesia on lower extremities, gait disturbance, tendon hyperreflexia or hyporeflexia on upper extremity, tendon hyperreflexia on lower extremity, Hoffmann’s sign and Babinski’s sign positive; girdle feeling and sensory disturbance level on trunk.	CTTSS is mainly caused by heterotopic ossification. Cervical and thoracic lesion is usually close to each other at the cervico-thoracic junction. The symptoms are caused by cord compression, the responsible segment needs to be determined by careful examination. ThSS is usually underdiagnosed due to its low incidence and slower progression.
CLTSS	TSS involves CSS and LSS	Neck and low back pain; Neurological symptoms: muscle weakness on upper and lower extremities, radiculopathy and hypoesthesia on upper and lower extremities, neurogenic claudication, gait disturbance, tendon hyperreflexia or hyporeflexia on upper extremity, tendon hyporeflexia on lower extremity, Hoffmann’s sign and Babinski’s sign positive; sensory disturbance level on trunk.	CLTSS is the most common subtype of TSS. The manifestations caused by upper motor neuron deficit such as tendon hyperreflexia can be masked by those caused by lower motor neuron deficit. Additionally, CSS can also cause lower extremity symptoms, which may confuse and mislead clinicians to the diagnosis of LSS.
TLTSS	TSS involves ThSS and LSS	Back and low back pain; Neurological symptoms: muscle weakness on upper and lower extremities, radiculopathy and hypoesthesia on lower extremities, neurogenic claudication, tendon hyporeflexia on lower extremity, Babinski’s sign positive; girdle feeling and sensory disturbance level on trunk.	OLF at the lower thoracic spine combined with LSS is a common cause of TLTSS. ThSS is usually underdiagnosed due to its low incidence and slower progression.
CTLTSS	Concomitant occurrence of CSS, ThSS and LSS	Neck, back and low back pain; Serious neurological symptoms: muscle weakness on upper and lower extremities, radiculopathy and hypoesthesia on upper and lower extremities, hypoesthesia on lower extremities, neurogenic claudication, gait disturbance, tendon hyperreflexia or hyporeflexia on upper extremity, tendon hyperreflexia on lower extremity, Hoffmann’s sign and Babinski’s sign positive; girdle feeling and sensory disturbance level on trunk.	The most uncommon and severe TSS subtype. It is caused by defused heterotopic ossification and degeneration changes. The clinical presentation is complicated, the neural deficit can affect adjacent effector region and aggravate the symptoms. The surgical strategy should be designed on individual basis.

CSS, cervical spinal stenosis; CLTSS, cervico-lumbar tandem spinal stenosis; CTTSS, cervico-thoracic tandem spinal stenosis; CTLTSS, cervico-thoraco-lumbar tandem spinal stenosis; LSS, lumbar spinal stenosis; OLF, ossification of ligamentum flavum; TSS, tandem spinal stenosis; ThSS, thoracic spinal stenosis; TLTSS, thoraco-lumbar tandem spinal stenosis.

In patients with TSS who mainly present with lower-limb symptoms, cord compression is concealed. When symptoms caused by different lesions occur at adjacent or the same region(s), clinicians might lose sight of more distant stenosis (19, 32). For example, ThSS is often ignored in patients with TLTSS because of its nonspecific symptoms and slow progression (50, 51, 52, 53). It should be noted that ‘asymptomatic’ patients may develop or have changes that merged with the predominant symptoms (30); however, careful physical examination, such as evaluation of pathological gaits and lower-extremity weakness, can provide solid evidence of myelopathy and avoid misdiagnosis of TSS (34, 54). Furthermore, a thorough neurological examination that reflects possible diseases along the extent of the spinal column is recommended to avoid underdiagnosed concomitant stenosis (1, 20, 36, 49). Clinicians are also expected to maintain a high index of suspicion of TSS in the management of patients with spinal stenosis (19, 55, 56).

Patients with TSS may present with more severe preoperative symptoms owing to multiregional compression. The symptoms of one compression can be aggravated by neurological deficits caused by coexisting lesions, and the neurological response to single regional decompression is also lower in patients with TSS (57). Some studies used ‘double crush syndrome’ to explain the severity of TSS, in which proximal lesions, although sometimes asymptomatic, may increase susceptibility to compression at distal lesions (58, 59).

Diagnosis criteria of radiological TSS

In addition to the clinical manifestations, radiological evidence is essential to diagnose TSS. The diagnostic criteria for radiological TSS are based on the criteria for isolated spinal stenosis (60, 61, 62). The diagnostic criteria for TSS are not uniform in the literature, and this has led to variations in the reported prevalence of TSS and heterogeneity of clinical trials. Moreover, surgeons diagnose TSS using varied radiological examinations according to the routine practice in different centres and eras, including digital radiography (DR) (4), discography (63), CT (64), myelography (2, 26), and MRI (16), and this also leads to inconsistencies in the diagnostic criteria used in studies.

Previously, surgeons used DR to diagnose TSS, and most studies within this period used a diameter of <10–12 mm to define radiological TSS (15, 21, 65, 66). In myelography assessment, Tsutsumimoto et al. used dura sac narrowing >50% to diagnose LSS in radiological TSS (26). In CT, a midsagittal diameter of <12 mm was used to diagnose radiological TSS (64). With the emerging applications of MRI, a large variety of diagnostic criteria and grading systems for spinal stenosis have been proposed (Table 3). In recent studies, the spinal stenosis grading systems proposed by Kang and Lee are often combined or applied to diagnose radiological TSS, since they are proposed in similar patterns consistently (61, 62, 67, 68, 69). However, the diagnostic criteria for radiological TSS with ThSS are yet to be clarified, and a comprehensive grading system with a treatment algorithm is urgently required.

Table 3

Useful radiographic diagnosis and grading criteria of partial spinal stenosis in TSS research.

Classification	Radiographic diagnosis criteria	Radiographic grading criteria	Reference
CSS	Based on CT or T2-weighted image of MRI:Dorsal-ventral diameter of the spinal canal :< 12 mm or 10 mm	Based on T2-weighted image of MRI:Grade 0: the absence of central canal stenosis Grade 1: nearly complete obliteration of subarachnoid space, including obliteration of the arbitrary subarachnoid space exceeding 50%, without signs of cord deformity. Grade 2: central canal stenosis with cord deformity but without spinal cord signal change. Grade 3: the presence of spinal cord signal change near the compressed level on T2-weighted images.	4, 61
ThSS	There are no quantitative diagnostic criteria, and the diagnosis is often made through clinical experience.	No radiographic grading system	N/A
LSS	Based on CT or T2-weighted image of MRI:1. Antero-posterior diameter of spinal canal:relative LSS: 10–12 mmabsolute LSS: <10 mm 2. Cross-sectional area of dural tube or sac:relative LSS: < 100 mm2absolute LSS: <75 mm2 or 70 mm23. Ligamentous interfacet distance:< 10 mm (L2 - L3)< 10 mm (L3 - L4)< 12 mm (L4 - L5)< 13 mm (L5 - S1)4. Transverse diameter of spinal canal:< 15 mm or 16 mm	Criteria one (based on T2-weighted image of MRI) :No or minor stenosis: there is clearly CSF visible inside the dural sac, but its distribution is inhomogeneousModerate stenosis: the rootlets occupy the whole of the dural sac, but they can still be individualized. Some CSF is still present giving a grainy appearance to the sac. Severe stenosis: no rootlets can be recognized, the dural sac demonstrating a homogeneous grey signal with no CSF signal visible. There is epidural fat present posteriorly. Extreme stenosis: in addition to no rootlets being recognizable there is no epidural fat posteriorly. Criteria two (Based on T2-weighted image of MRI) :Grade 0: no lumbar stenosis without obliteration of anterior CSF space;Grade 1: mild stenosis with separation of all cauda equina;Grade 2: moderate stenosis with some cauda equina aggregated; Grade 3: severe stenosis with none of the cauda equina separated	59, 61, 67

CSF, cerebrospinal fluid; CSS, cervical spinal stenosis; LSS, lumbar spinal stenosis; TSS, tandem spinal stenosis; ThSS, thoracic spinal stenosis.

Treatment of TSS

Patients with TSS who are not well managed conservatively require surgical intervention. The operative rate of two or more regions is as low as 0.88–7.6% among patients with spinal stenosis (11, 32, 70, 71). Under the basic goal of TSS surgery, which is strategic decompression, maintenance of nerve function, and spinal stability and alignment reconstruction (72), the optimal surgical strategy for TSS is still debatable. Controversies include region selection, surgical patterns, and order.

Does asymptomatic lesions require surgery?

Although some studies have demonstrated that coexisting asymptomatic stenosis does not interfere with postoperative improvement (26, 28, 73), clinicians have observed that latent LSS does not interfere with the recovery of CSM, while asymptomatic CSS can negatively affect patients’ improvement after lumbar surgery (1, 30, 74, 75, 76). After selective surgery, some patients with TSS have incomplete recovery or develop new symptoms that might be masked preoperatively or aggravated postoperatively (12). Another consideration is that asymptomatic lesions may cause intra-operative position-related complications. In the prone position, the area of the stenotic spinal canal decreases by up to 67%, which could aggravate neural compression and decrease blood supply (77). It has been reported that some patients with radiological TSS develop myelopathic symptoms and cauda equina symptoms after long-term surgery in other regions or even after sleeping (11, 78, 79, 80, 81). Thus, the surgical strategy for TSS requires a comprehensive vision for development, as the symptoms cannot completely reflect the neurological vulnerability and deficit (34, 82). Surgeons should pay close attention to latent CSS and ThSS in patients with TSS and consider secondary decompression for those with incomplete resolution or new symptoms after the initial lumbar surgery.

Surgical pattern for CLTSS

Previous studies have demonstrated similar efficacy and risks of staged and simultaneous surgery in patients with CLTSS (27, 83). For nuances, simultaneous surgery could lead to significant short-term improvement, whereas slower recovery or deterioration has been observed during long-term follow-up (57, 69). Staged surgery could lead to better early outcomes than simultaneous surgery; however, the results of simultaneous surgeries in later follow-up were superior (70). For surgical complications, Eskander et al. found that the complication rate was higher among cases in patients with CLTSS who were above 68 years of age, estimated blood loss >400 mL, or combined operation time >150 min (83). Interestingly, Molinari et al. performed simultaneous surgery in a group of patients with TSS whose average age was 67 years, mean operative time was 159 min, and mean intraoperative blood loss was 558 mL; they reported no major complications, and all patients had a high degree of satisfaction (32), indicating a rather non-absolute boundary that influenced the complication rate. Thus, surgeons should consider intrinsic advantages when planning surgeries. Simultaneous surgery can improve operative length, blood loss, and position-related complication rate in patients with severe stenosis (12). On the other hand, staged surgery may be preferred for its ability to lower one-time invasiveness, target more symptomatic regions, and avoid unnecessary surgery by allowing time to recover (72).

These results indicate a comparable outcome between the two strategies when properly selected. In the CLTSS treatment algorithm, patients with CSS and LSS with similar severity, high motivation, high percentage of developing neurological complications in prone positioning, and good general condition are eligible for simultaneous surgery, while patients with poor status, or those who predominantly manifest with CSS and minor presentation of LSS, with corresponding radiological evidence, or vice versa, are suitable for staged surgery (72).

Surgical order of CLTSS

Most patients with CLTSS are elderly with poor general condition and unequal CSS and LSS (31, 34), and these patients are inclined to undergo staged surgery. Thus, the discussion of the surgical order has come alive.

In clinical practice, the priority resolution of cervical compression with or without myelopathy is supported (13). Cervical procedures do not only resolve typical CSS symptoms but can also improve lumbar symptoms, including radiculopathy and low back pain (84, 85, 86, 87), which decreases the necessity of subsequent lumbar surgery (26, 84). Compared with starting from the lumbar region, initial cervical surgery results in a higher complete resolution rate, lower rates of developing new symptoms and secondary surgery, and longer surgical intervals (11, 67, 72). This is because cervical decompression may induce functional recovery of the spinothalamic tract and corticospinal tract, resulting in neurological improvement (85, 88), while initial lumbar surgery may alter the pressure at other lesions, which could worsen the pressure on neural tissues and causes new symptoms (89).

Despite the improvement in lumbar symptoms after the cervical procedure, the effect of cervical surgery cannot be overclaimed. Some lumbar symptoms, such as intermittent claudication, cannot be relieved by cervical decompression (2), and most lumbar symptoms were temporarily relieved; only 14% of patients with CLTSS experience improvements in long-term follow-ups, while the rest would need to undergo secondary surgery (87). Sometimes, lumbar surgery alone provides better outcomes than cervical surgery or both procedures in patients with CLTSS who predominantly present with lumbar symptoms (15). Therefore, instead of starting from the cervical spine, several surgeons and our group have proposed individualised strategies. In the algorithm, after evaluating clinical manifestation and radiological and electrophysiological examinations, patients receive cervical surgery first if they predominantly present with upper motor neuron signs or upper extremity symptoms, and lumbar surgery is performed initially if the patients present with lower extremity symptoms without upper motor neuron signs (1, 2, 50, 71) (Fig. 2). This strategy has resulted in excellent outcomes in several studies (15, 70, 71), and can be used to stratify patients with TSS to avoid unnecessary interventions (12).

Figure 2

The treatment algorithm of the subtypes of TSS. For CLTSS, the main determinants of surgical pattern are predominant symptoms, complication rate, and general condition; while lesion distance alters the surgical pattern greatly in the other TSS subtypes with ThSS. The surgical sequence is determined by the predominant symptoms and subjective examination results. CSS, cervical spinal stenosis; CLTSS, cervico-lumbar tandem spinal stenosis; CTTSS, cervico-thoracic tandem spinal stenosis; CTLTSS, cervico-thoraco-lumbar tandem spinal stenosis; LSS, lumbar spinal stenosis; TSS, tandem spinal stenosis; ThSS, thoracic spinal stenosis; TLTSS, thoraco-lumbar tandem spinal stenosis.

Surgical strategy for TSS with ThSS

Due to extensive spinal cord violation in TSS with ThSS, the treatment does not share the same strategy and prognosis as CLTSS. In CT/TLTSS treatment, simultaneous and staged surgery can lead to comparable clinical outcomes (7, 90), and the strategy is largely determined by patient factors, such as general condition, location and distance of the lesion, predominant compression, and complication rate. When ThSS occurs adjacent to concurrent CSS or LSS, simultaneous surgery is more feasible by performing a single incision under one-time anaesthesia and hospitalisation. In the treatment of CTTSS, simultaneous surgery for adjacent lesions (lesion departing less than three levels) results in good outcomes and high cost-effectiveness (90, 91) (Fig. 2). However, safety concerns associated with simultaneous surgeries, such as deterioration, complications, and secondary surgery rate, impede its application, in addition to the adjacent type (7, 35, 90). Staged surgery is more reasonable for skipped CT/TLTSS. The principle of staged surgery is to resolve the predominant symptom first, which is evaluated via imaging, electrophysiological examinations, neurological status, and medical comorbidities, and subsequently decompress the remaining stenosis (10, 55). The prognosis of CT/TLTSS is poorer than that of CLTSS because of its severe nature, including its prolonged course, extensive cord compression, severe invasion, and dural adhesion (8, 92, 93, 94). Consequently, any intervention may induce unfavourable outcomes, and neither simultaneous nor staged surgery can decrease the complication rate (55).

Recently, treatment with CTLTSS has been reported (95). Some surgeons have applied simultaneous surgery to all lesions and reported improvement without complications (95). However, this strategy is not suitable for all patients because of its invasiveness, long surgical time, high skill dependency, and multiple technical variables, which lead to difficulty in tracing back. Staged surgery is less challenging and more acceptable to both surgeons and patients. By decompressing several adjacent lesions together in separate stages or resolving one region per stage, patients significantly improve the attenuated risk (93, 96). In the algorithm of our group, if the thoracic stenosis is adjacent to the cervical or lumbar lesion, initial cervicothoracic or thoracolumbar combined surgery should be performed to resolve the most severe lesion, and decompression of the other site should be performed after an interval of at least 3 months; if the lesion sites are three levels apart from each other, multi-staged surgery ought to be initialised from the most myelopathic site, if it exists, or from the cervical spine (Fig. 2).

In addition to the predominant symptoms, compression location, and cord priority, non-spinal factors also determine the surgical strategy of TSS. With the exacerbation of the aging society, the elderly account for a large proportion of patients who undergo spinal surgery (97). The high surgical risk due to poor general condition, anaesthesia risk, high revision surgery rate, and complicated underlying diseases make it challenging for elderly patients to undergo surgeries. Although some studies reported similar complication rates in elderly and young patients who underwent lumbar surgeries (98), complication rates, morbidity, and mortality were significantly increased among patients >80 years of age (97, 99). Some common spinal surgical complications, such as cerebrospinal fluid leakage and delayed wound healing, can cause severe consequences in elderly patients owing to their frailty (100). Thus, overall consideration is necessary for elderly patients with TSS. Underlying diseases can also alter treatment strategies. For instance, diabetes mellitus (DM) increases the risk of complications of spinal surgery. Patients with DM are also more likely to develop postoperative infection and deep venous thrombosis, which may lead to prolonged hospitalisation or even death (101). Therefore, the treatment strategy of TSS is also largely determined by non-spinal aspects, and staged surgery or selective surgery might be alternatives for TSS patients with such conditions.

Conclusions

TSS refers to concomitant stenosis in at least two regions of the spine that results in associated symptoms and signs. Clinicians should be aware of the incidence of TSS among patients with relevant aetiologies and risk factors and make a complete and timely diagnosis via comprehensive evaluations. However, the current diagnostic criteria for TSS must be unified and standardised. For treatment, both simultaneous and staged surgeries are effective strategies, and surgeons should design the optimal treatment by analysing the general condition, subtypes, predominant symptoms, cord decompression priority, lesion factors, and prognosis. The concept of TSS has evolved from stenosis that occurs only in the cervical and lumbar spine to stenotic changes that affect any level of the entire spine. In this process, the clinical perception of TSS is largely altered. Stenotic regions are no longer pathologies that accidentally occur concurrently; instead, they are related to certain mechanisms to some extent. Similarly, the treatment of TSS requires a more delicate design with the overall situation. However, among the documented studies, optimal strategies with solid evidence have not been proposed, and more high-level clinical trials are required to determine the suitable surgical pattern for TSS.

ICMJE Conflict of Interest Statement

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding statement

This work was supported by Jilin Provincial Natural Science Foundation (no. 20200201531JC) and National Natural Science Foundation of China (NO. 82072508).

Author contribution statement

All authors listed have made a substantial, direct, and intellectual contribution to the work, and approved it for publication. Bai Q and Wang Y drafted the manuscript; Zhai J and Wu J collected the reference and data; Zhang Y extracted and visualized the figures; Zhao Y designed the study.