From January 1, 2018, to December 31, 2023, 55 patients were admitted to hospital for acute spinal cord compression, including 17 cases of spinal cord compression of tumorous or degenerative origin and 38 cases of spinal cord compression due to spondylodiscitis, representing a frequency of 69.09%. In this study, 31 patients were selected based on the inclusion criteria and 6 patients were excluded because they give up the treatment after a month. The average hospital admission rate was 5 cases per year.

In our study, the average age of patients was 38.35 ± 15.20 years, ranging from 10 to 72 years. Males accounted for 61.29% of patients, with a sex ratio of 1.58 M/F. The distribution by age, gender and occupation is reported in Table 1 .

The over-60 age group accounted for 25.80%. The 20 - 30 and 40 - 50 age groups were the most active. In our study, 90.32% of patients did not have insurance or health coverage; 54.84% of patients came from rural areas.

In our study, the average consultation time was 5.74 ± 5.92 months, ranging from 1 month to 24 months. Back pain accounted for 41.94% of the reasons for admission. Claudication accounted for 41.94% of other symptoms at admission. This distribution is reported in Table 2 .

In our study, fever was reported in 63.33% of patients. Tuberculosis infection was present in 35.48% of patients. Vaccination status was not specified for patients. Data related to comorbidities and lifestyle. Among comorbidities, diabetes was present as a history in 25.81% of cases. Regarding lifestyle, consumption of unpasteurized cow’s milk cheese was noted as a dietary habit in 83.87% of patients.

In our study, 100% of patients had a lesion syndrome and at least two out of three syndromes as reported on Figure 1 .

The location of the spinal syndrome was lumbar in 26 cases (83.87%), dorsal in 7 cases (22.58%), and cervical in 1 case (3.23%). It consisted of spinal and paravertebral muscle pain in 93.55% of cases, spinal kyphosis deformity in 9.68% of cases, scoliosis in 2.23% of cases, and gibbosity in 35.48% of cases. Spinal involvement affected two spinal segments in 3 patients.

Radicular involvement was the main symptom of the lesion syndrome and was present in 64.52% of cases, with lumbosciatic pain in 75% of radiculopathies. This distribution is reported in Table 3 .

In our study, motor deficit was the main element of the sublesional syndrome, accounting for 90.32% of cases. Paraparesis was the most common motor deficit; 51.61% of patients presented with paraparesis in cases of thoracolumbar involvement.sublesional data are reported in Table 4 .

Using Franckel’s classification, 67.74% of patients were classified as grade C according to Franckel’s classification. Figure 2 summarizes these clinical arguments.

In our study, 70.97% of patients presented at least one of the functional and general signs of tuberculosis. 41% of patients had an altered general condition. Figure 3 showed this distribution.

In our series, 67.74% of patients underwent spinal CT and 38.71% underwent spinal MRI. Compressive epiduritis was present in 54.84% of patients. lesions distribution is showed in Table 5 .

In our series, 54.84% of patients presented with compressive epiduritis. Bone sequestration and spinal instability were present in 11 patients; these factors indicate the need for surgical intervention. Signs and lesions are reported in Table 6 .

Spinal involvement affected two adjacent vertebrae in 83.87% of patients. Spinal involvement affected more than two vertebrae in one patient, and two cases of multi-level involvement were noted.

Following paraclinical examinations, a diagnosis of tuberculous spondylodiscitis was made in 77.42% of patients and non-tuberculous spondylodiscitis in 22.58%.

The average hospital stay was 10.35 ± 7.71 days, ranging from 5 to 40 days. All patients (100%) received drug treatment; 22.58% underwent surgery as showed on Figure 4 .

In our study, 100% of patients received analgesics. For all cases of tuberculous spondylodiscitis, anti-tuberculosis treatment was initiated according to the standardized 12-month regimen recommended by the WHO [17] with an intensive 2-month phase of ERHZ* quadruple therapy and a 10-month continuation phase of RH dual therapy. The average duration of anti-tuberculosis treatment in our study was 11.14 ± 1.93 months, ranging from 9 to 15 months; 52.38% of cases of tuberculous spondylodiscitis underwent anti-tuberculosis treatment for 12 months. 100% of cases of non-tuberculous spondylodiscitis received broad-spectrum antibiotic therapy with bone and parenteral diffusion during their hospital stay. (E: ethambutol; R: rifampicin; H: isoniazid; Z: pyrazinamide).

Surgery was indicated in 11 patients; 7 (63%) actually underwent surgical treatment. The four (4) patients who did not undergo surgery did so because of contraindications to anesthesia and technical problems.

The average time to surgery was 2.71 ± 2.36 days, with extremes ranging from 1 to 7 days. 42.86% of patients were treated within 48 hours. 85% of cases were operated on via the posterior approach. The anterior approach was used in one case of cervical involvement. Instrumentation was performed in 70.42% of patients who underwent surgery ( Figure 5 ).

In our series, 96.77% of patients experienced an improvement in spinal signs; in 90.32% of patients, spinal pain improved completely within two weeks. In 48.39% of patients, there was a one-grade improvement in Franckel’s scale at discharge; Franckel grade D was the most common at 41.94%. No patients had infectious complications; one case (3.22%) of pressure ulcer-type complication was noted. 96.78% of patients had no complications. No patients died in our series ( Table 7 ).

In our study, 29 out of 31 patients were reviewed after 6 months; 21 out of 31 patients were reviewed at 6 months and 12 months; twelve cases were reviewed at 24 months. The assessment was made in relation to the last contact with the patient. Radiological follow-up was performed every 3 months in patients who were reviewed.

Progression to Franckel grade was noted in 65.52% of patients. Functional motor recovery was present in 64.52% of patients. 19.35% had complete neurological recovery. 85.71% of patients who underwent surgery had progressed to Franckel grade as reported in Table 8 .

In our study, 42.86% of patients reviewed at 6 months had progressed to Franckel grade; 47.61% had made a full neurological recovery. All patients who underwent surgery made a functional motor recovery (Franckel grade D); and 57.14% made a full neurological recovery. Detailed results are provided in Table 9 .

In our study, 28.57% of patients reviewed at 12 months had progressed to Franckel grade. 71.43% of patients who underwent surgery had complete neurological recovery. Detailed results are provided in Table 10 .

A return to work was noted in 79.17% of patients in our study after 12 months.

Illustration of a surgical procedure for cervical spinal cord compression due to spondylodiscitis is shown on Figure 6 .

The average age in our study was 38.35 ± 15.20 years, ranging from 10 to 72 years. This finding is similar to the average of several series that found subjects with an average age of 42.25 ± 18 years [2] [36] [37] . In our study, the over-60 age group was the most represented with 8 cases, or 25.80%. This result is consistent with what is reported in the literature, as those over 60 years of age were the most affected because advanced age is a predisposing factor for extrapulmonary tuberculosis [15] [32] [33] .

There was a clear male predominance at 61.29%. This trend is confirmed by several studies [19] [24] [26] [29] [30] . This can be explained by two factors: first, tuberculosis affects men more than women; out of 10 million cases in 2022, 5.8 million were men [38] - [40] , and tuberculosis is the most common cause of spondylodiscitis worldwide, with tuberculous spondylodiscitis being one of the most serious forms of tuberculosis responsible for spinal cord compression [40] [41] . On the other hand, social barriers such as pronounced patriarchy or unequal access to healthcare can limit women’s access to health services, thus creating a recruitment bias.

In our study, traders were the socio-professional group most affected. This trend among occupations in the tertiary sector is also reported in several studies, notably in Cameroon and Senegal [20] [30] . This situation can be explained by the fact that the tertiary sector often includes older people. In Benin, more specifically in Parakou, for example, the highest activity rate by age is in the over-50 age group, and the tertiary sector accounts for around 80% of economic activity.

In our study, the admission delay was 5.74 months, ranging from 1 to 24 months. A similar delay of around 7 months was reported by other authors [3] [6] [11] [14] [32] . Other authors reported delays of up to 11.58 ± 8.17 months [14] [19] . This disparity can be explained by the difference in the types of studies, their duration, and the number of participants recruited, as some authors studied larger groups over a short period of 27 months [14] . On the other hand, the general trend toward long consultation delays can be explained by several factors: the acute progression of symptoms, limited access to appropriate health infrastructure, stigma associated with certain symptoms, particularly spinal deformities, and the cultural tendency to seek traditional alternative therapies before consulting a specialized center. The fact that the majority of patients come from Borgou would therefore be an advantage in terms of shorter admission times in our series [2] [31] [32] .

Back pain was the most common reason for consultation, followed by functional impotence, as in several other studies [29] - [32] . Functional impotence ranked second [2] [19] [31] .

This disparity can be explained by the type of pain that occurs first in spinal cord compression in general. It should be noted that, depending on the pathophysiology, spinal damage precedes spinal cord damage, hence the predominance of back pain in the early stages, especially in the case of disabling inflammatory back pain. This is the main symptom of spondylodiscitis [3] [23] [26] [41] ; however, in compressions of other origins, there may be no spinal involvement, or it may be secondary.

In our study, no medical history was noted; however, one-third of patients had tuberculosis infection; this is lower than the 4.5% tuberculosis infection rate reported in a study in Niamey in 2020 [29] . Tuberculosis infection can be explained by the endemic nature of tuberculosis in sub-Saharan Africa [1] [19] [33] .

The most common comorbidity found in this study was diabetes at 25.81%; this is lower than the 54.5% reported in Saudi Arabia in 2023 [10] but higher than the 15.4% of diabetes reported in Tunisia in 2023 [21] . This confirms diabetes as one of the predisposing factors for spondylodiscitis [35] [37] .

Our study noted a deterioration in general condition in 41% of patients, as did the authors of the literature [21] [37] . This proportion peaks at 81.4% in Cameroon. The deterioration in general health is related to an advanced stage of the disease at the time of consultation. The higher rates of deterioration in general health in other studies can be explained by several factors. First, the deterioration in general health is related to an advanced stage of the disease. In one publication, nearly one-fifth of the population was in the final stage of tuberculosis, unlike in our study. This explains why the deterioration in general health was more significant [19] . Second, this difference can also be explained by the presence and significance of comorbidities, such as diabetes. Finally, this disparity can also be explained by socioeconomic factors such as possible difficulties in accessing care, financial constraints, or delays in seeking medical attention, which would favor the progression of the disease.

This syndrome was present in almost all patients; this was also the case for 93.3% of patients in other series [2] . In our study, the two segments most affected were the lumbar region in more than three-quarters of the sample, followed by the thoracic region. Several studies confirm this location as predominant; in fact, lumbar involvement was found in the majority of studies (62%) [12] .

The predominant lumbar involvement reported in several studies is related to the fact that the lumbar spine, which supports a large part of the body’s weight, is subject to mechanical stresses that make it vulnerable to infection. The constant loads on this region could indeed be responsible for microtrauma, and the cumulative effect of this microtrauma and biomechanical stress could make the lumbar region vulnerable to infection due to localized inflammation. Pain on palpation of the spinous processes was the predominant sign of the syndrome, affecting 9 out of 10 cases, and gibbosity-type deformity was found in one-third of patients. We see the same trend in several studies between 2018 and 2023 [22] [29] [33] .

This lesional syndrome was present in all patients; it was dominated by radicular pain, present in nearly two-thirds of the population. Three-quarters of cases involved lumbosacral radiculopathy. It is thus represented due to constant spinal cord involvement in cases of spinal cord compression [2] [11] [14] ; and the frequency of lumbosacral radiculopathy corresponds to the lumbar location, which is the most common. Some authors have reported a lower frequency of around 50% in their series [19] [29] .

Motor deficit is constant in sublesional syndrome; nearly 9 out of 10 patients have motor deficit. Paraparesis was present in more than half of the patients, making it the most common motor deficit. This result is similar to that observed in the series on spondylodiscitis with spinal cord compression [19] , with 73.41% of cases presenting with predominantly paraparesis.

These are present in 70.97% of patients. Hypoesthesia is predominant, affecting more than a third of the population. This is also reported by several studies [2] [3] [19] [40] .

A decrease in reflexes was predominant among the reflex disorders in our study. One in four patients was affected; one in ten patients had abolished or hyperactive reflexes. Two patients had Babinski’s sign. This result is consistent with the literature; these various signs are also reported in several other studies [1] [2] [12] [19] [29] .

In our study, approximately one in four patients had trophic or vesico-sphincteric disorders (25.81%). This frequency is below the 53% of vesico-sphincteric disorders reported in the Brazzaville studies in 2020 [3] . There is a disparity in the signs of sublesional syndrome in our study, both in terms of their frequency and their types. The predominance of paraparesis and the relatively low frequency of genito-sphincteric disorders reflect a relatively low intensity of sublesional syndrome in general in our study. This difference can be explained by the difference in the stage of presentation of the pathology. Acute spinal cord compression progresses from the spasmodic deficit stage to the deficit stage. Despite a long average consultation delay of nearly 6 months, the patients in our study were seen relatively early compared to comparative studies.

In our study, neurological impairment was assessed using the Frankel classification; just over half of the patients in our study were classified as Frankel grade C, and around one-fifth were classified as grade B. We did not observe any cases classified as grade A. The same trend has been reported in several studies [2] [3] [12] [27] .

Inflammatory syndrome was present in 8 out of 10 patients; this consisted of elevated CRP. This confirms the consistency of biological inflammatory signs in spondylodiscitis [14] [37] .

The predominant sign on CT scan in our series was disc narrowing (32.26%), followed by vertebral compression (19.35%) and blurred erosion of the vertebral endplates, as reported in several studies as a mirror geode appearance and vertebral compression in half of the cases [6] [17] [38] [41] .

Magnetic resonance imaging revealed compressive epiduritis in half of the patients, as well as bone lesions, namely vertebral compression, bone sequestration, and soft tissue lesions such as paravertebral abscesses and intra-canal abscesses. This correlates with clinical signs indicating vertebral involvement on the one hand and secondary spinal cord compression on the other. This can be explained by the stage of the disease at the time of diagnosis.

The results of this study highlight the endemic context of tuberculosis, with more than 7 out of 10 cases of tuberculous spondylodiscitis involving spinal cord compression. The diagnosis was based on a combination of epidemiological, clinical, and paraclinical evidence.

The average duration of anti-tuberculosis treatment in our study was 11.14 ± 1.93 months, with extremes of 9 and 15 months, according to the recommendations for the treatment of extrapulmonary tuberculosis [29] . This difference can be explained by the fact that treatment is sometimes prolonged depending on whether or not the patient responds favorably to treatment.

This physiotherapy treatment is found in several studies as a complement to surgery or conservative treatment [37] [39] . As the extent of nerve damage varies, the proportion of patients who have benefited from rehabilitation is directly related to the frequency and severity of nerve damage.

Among the patients in this series, nearly a quarter underwent surgery (22.58%). This finding is consistent with other studies. For example, 27.84% of patients underwent surgery [19] ; however, in purely surgical series, treatment was performed in 82% [5] [9] [30] and 42.8% [3] of patients. The majority of patients underwent posterior approach surgery on the thoracolumbar spine. One case of cervical compression was treated with anterior surgery involving a corporectomy, iliac graft, and anterior cervical plate. This is consistent with several other studies [2] [3] [12] [24] [30] .

The short-term outcome in our study was favorable, as almost all patients experienced symptom improvement and nearly half had an improvement in Franckel grade, indicating an improvement in sensory-motor deficit; nearly half of the patients had functional motor recovery (Franckel grade D). Furthermore, no deaths or infectious complications were recorded, unlike several other studies that reported deaths [22] [27] [36] .

Regarding long-term outcomes, there were two cases of permanent vision loss, and nearly one-third of patients were seen again at one year. The outcome was favorable at one year; nearly three out of four patients had complete neurological recovery allowing them to walk, and this ratio was higher among patients who underwent surgery. Lower figures are reported for some series, such as 50.63% complete recovery [19] or 77.8% complete recovery [12] [22] . This can be explained by the stage of progression of the disease at the time of treatment and may also depend on professional experience and surgical expertise.