Abstract
Objectives
The objective of the current study was to compare short- and long-term effect on chronic low back pain of intradiscal injection of methylprednisolone with or without presence of Modic type 1 MRI changes.
Patients and methods
Medical charts of patients receiving intradiscal injection of methylprednisolone from January 1, 1995 to December 31, 1998 were retrospectively reviewed. Clinical parameters were recorded at baseline, 24 h after injection and at follow-up (12–14 months). Patients were studied in three groups: Modic I-a, if patients had Modic type 1 changes with no previous surgery or nucleolysis ( n = 30); Modic I-b, if patients had Modic type 1 changes at the level of previous surgery or nucleolysis ( n = 37); Control, if patients had no Modic type 1 changes ( n = 30).
Results
Twenty-four hours after methylprednisolone injection, higher proportion of patients with self-assessed improvement was observed in Modic I-a (90%) and Modic I-b (71%) than in Control (30%). Low back pain decreased in both Modic groups. Low back pain did not vary from baseline in controls. No effect was detected in three groups, neither for radiating pain 24 h after injection, nor for any outcome parameters at the latest follow-up.
Conclusions
We suggest that patients with disabling chronic low back pain and Modic type 1 MRI changes have specific acute response to intradiscal injection of methylprednisolone. Clinical studies are however necessary to further investigate the effectiveness and safety of such injections.
Résumé
Objectif
L’objectif de cette étude était de comparer les effets à court et à long termes d’infiltrations intradiscales de méthylprednisolone au cours de lombalgies chroniques avec ou sans anomalies de signal IRM de type Modic 1.
Patients et méthode
Nous avons conduit une étude rétrospective sur les dossiers médicaux de patients ayant reçu une infiltration intradiscale de méthylprednisolone du 1 er janvier 1995 au 31 décembre 1998. Les paramètres cliniques étaient recueillis lors de la consultation initiale, 24 heures après l’infiltration et le jour de la visite de suivi (12–14 mois). Les patients étaient divisés en trois groupes : groupe Modic I-a, si les patients présentaient des anomalies de signal de type Modic 1 sans antécédent de chirurgie ou nucléolyse ( n = 30) ; groupe Modic I-b, si les anomalies de type Modic 1 étaient apparues à la suite d’une chirurgie ou d’une nucléolyse et au même endroit ( n = 37) ; groupe Témoin, si les patients n’avaient pas de changement de signal IRM de type Modic 1 ( n = 30).
Résultats
Vingt-quatre heures après l’infiltration de méthylprednisolone, une plus grande proportion de patients rapporte une amélioration de la douleur dans le groupe Modic I-a (90 %) et Modic I-b (71 %) que dans le groupe Témoin (30 %). Les douleurs lombaires ont diminué dans les deux groupes Modic, alors qu’aucune différence sur l’intensité de la lombalgie entre la visite initiale et l’évaluation post-infiltration n’a été observée pour le groupe témoin. Aucun effet n’a été détecté dans les trois groupes, pour la douleur irradiant au membre inférieur 24 heures après l’infiltration, ou pour l’ensemble des paramètres lors de la dernière visite de suivi.
Conclusions
Nous suggérons que les patients souffrant de lombalgie chronique invalidante avec anomalies de signal IRM de type Modic 1 ont une réponse spécifique à court terme aux infiltrations intradiscales de méthylprednisolone. Des études complémentaires sont cependant nécessaires pour documenter l’efficacité et la sécurité de ces infiltrations.
1
English version
1.1
Introduction
Morphologic abnormalities detected by standard imaging and corresponding to degenerative discopathy are poorly specific of chronic low back pain . They are found in 40% of asymptomatic subjects. Vertebral degenerative changes at the lumbar spine level are classified according to a 3-stage system (Modic) based on MRI signals of bone marrow adjacent to vertebral end plates . Stage 1 is hyposignal in T1-weighted sequences and hypersignal in T2-weighted sequences (Modic type 1); stage 2 is hypersignal in T1- and T2-weighted sequences (Modic type 2); and stage 3 is hyposignal in T1- and T2-weighted sequences (Modic type 3). Moderate to excellent inter- or intraobserver reliabilities have been reported for Modic classification . In Modic type 1 changes, end plates of vertebral bodies are characterized by subchondral bone fissuring, high concentration of pro-inflammatory cytokines, and by increased vascularity and mononuclear cell content . Modic type 2 corresponds to a fatty infiltration of peritrabecular bone marrow, and Modic type 3 consists in osteosclerosis of vertebral end plates . A prospective study indicates that these three MRI types may be successive stages of a single process . In contrast to other degenerative modifications of the spine, Modic type 1 changes are closely associated with chronic low back pain as suggested by observational studies . Such changes are uncommon in asymptomatic controls . Modic type 1 changes are also associated with specific features in patients with chronic low back pain: patients with Modic type 1 changes have more pain in late night and morning, and higher serum high-sensitivity C-reactive protein level than patients without Modic 1 changes, independent of evidence of spondylarthropathy .
Modic type 1 changes may therefore reflect inflammatory reaction in bone marrow adjacent to vertebral end plates and to intervertebral disc undergoing active degenerative process. This provides a rationale for intradiscal injection of corticosteroids . Furthermore, one level IV study suggests specific response to intradiscal injection of corticosteroids in patients with severe disabling chronic low back pain and Modic type 1 changes . These patients appeared to have lower pain than patients with Modic type 2 changes, 1 month after the therapy. However, such a specific response on shorter and longer follow-up, has not been demonstrated.
We aimed to perform a retrospective chart review of a case series of patients with chronic low back pain who had been treated by intradiscal injection of corticosteroids to investigate short- and long-term clinical response of such injections in patients with or without Modic type 1 changes seen on MRI. Modic type 1 changes have been reported after surgery , but they have not been specifically evaluated . We therefore considered separately Modic type 1 changes in patients with and without previous surgery or nucleolysis treatment.
1.2
Patients and methods
This was a retrospective study of data from medical records for 3 years, 1995 to 1998, which corresponds to a period when intradiscal injections of corticosteroids were not yet restricted to patients with chronic low back pain and Modic type 1 changes in our unit.
1.2.1
Patients
The selection of the population was consecutive. Data were included for patients with severe, disabling chronic low back pain whose disease had not responded to usual conservative treatments and who lacked evidence of systemic inflammatory disorder, metabolic bone disease, local infection or malignancy. All patients had undergone lumbar spine MRI with T1 and T2-weighted sequences. MRI results were assessed by a panel of at least two spine specialists (JB, PD, SP, MR). Data for patients with previous spinal surgery or nucleolysis treatment within the previous 6 months were excluded. In our experience, edema of the bone marrow adjacent to vertebral endplates is common on MRI just after disc surgery or nucleolysis, without any pathologic significance. Conversely to persisting Modic type 1 MRI changes, it disappears within 6 months after the intervention and is not related with clinical features. Patients were investigated in three groups on the basis of Modic type 1 MRI changes: Modic type 1 changes and no previous surgery or nucleolysis treatment (Modic I-a), Modic type 1 changes with previous surgery or nucleolysis treatment (Modic I-b), and degenerative disc disease without Modic type 1 changes (Control).
1.2.2
Intradiscal methylprednisolone injection
During 3-day hospitalization, patients underwent discography under fluoroscopic control as previously described , then received one injection of methylprednisolone (2 ml; Solu-Medrol ® , Pharmacia, Saint-Quentin-en-Yvelines, France) into the center of the disc. Patients were asked to remain in bed for 12 to 24 h and to wear lumbar support for 2 weeks after rest.
1.2.3
Clinical assessment
Clinical parameters were recorded at baseline that was the beginning of the 3-day hospitalization, 24 h after injection and at latest follow-up. Parameters recorded at baseline were age, sex, duration of symptoms, previous surgery or nucleolysis at the level of the injection site of the spine, level of low back pain (on a visual analog scale [VAS], 0–100 mm) and radiating pain in the lower limb (VAS). Outcome parameters were level of low back pain and radiating pain on a VAS and patient self-assessed improvement in pain condition (“yes” if the patient was improved and “no” if the patient was not) at 24 hr after injection and at latest follow-up.
1.2.4
Statistical analysis
Statistical analysis involved use of Statview v4.5 (Abacus Concepts, Berkeley, CA). ANOVA and Fisher’s exact test were used for analysis of quantitative variables, described with means and standard errors (SE). Chi 2 test was used for analysis of qualitative variables, described with numbers and percentages. A P < 0.05 was considered statistically significant.
1.3
Results
We included 97 patients: 30 with Modic I-a changes, 37 with Modic I-b changes and 30 controls ( Table 1 ). Patients from Modic I-a and Modic I-b groups differed in characteristics at baseline for duration of symptoms and level of radiating pain: Modic I-b patients had higher values than Modic I-a patients for duration of symptoms and level of radiating pain. Of Modic I-b patients, 33 had previously undergone spinal surgery, namely 31, discectomy at the level of the injection site, two arthrodesis and four had previously undergone nucleolysis with papain. The previously operated levels were L4-L5 in 16 cases and L5-S1 in 21. In controls, the values were 7 for discectomy 2 for arthrodesis, 7 for L4-L5 and 2 for L5-S1. The mean time between previous intervention and intradiscal injection was 2.6 ± 0.7 years in Modic I-b group and 1.4 ± 0.8 in controls. In the total population, the injected levels were L2-L3 in 4 cases, L3-L4 in 2, L4–L5 in 44, and L5-S1 in 47 and did not differ among groups.
| Control ( n = 30) | Modic I-a ( n = 30) | Modic I-b ( n = 37) | |
|---|---|---|---|
| Age, years | 46 ± 3 | 48 ± 2 | 43 ± 2 |
| Female, no. (%) | 12 (40) | 18 (60) | 16 (43) |
| Previous surgery or nucleolysis, no. (%) | 7 (23) / 2 (7) | 0 / 0 | 33 (89) / 4 (11) |
| Duration of symptoms, years | 4.3 ± 1.3 | 3 ± 0.5 | 7.2 ± 1.4 * |
| Back pain (VAS, 0–100 mm) | 55 ± 6 | 52 ± 5 | 62 ± 4 |
| Radiating pain (VAS, 0–100 mm) | 41 ± 5 | 28 ± 5 | 50 ± 5 * |
Data were available for short-term (24-h) assessment for 95 patients (30 Modic I-a, 35 Modic I-b, 30 controls) and for long-term assessment for 77 (26 Modic I-a, 36 Modic I-b 15 controls). The drop out rate was therefore 2% at short-term assessment (Modic I-a 0%, Modic I-b 5%, controls 0%) and 21% at long-term assessment (Modic I-a 13%, Modic I-b 3%, controls 50%). The mean long-term follow-up was 14 ± 2 months in Modic I-a, 14 ± 1 in Modic I-b and 12 ± 2 in Control. It did not differ among groups.
Twenty-four hours after injection, the proportion of patients with self-assessed improvement was significantly higher for Modic I-a and Modic I-b groups than for controls: Modic I-a, 27 patients (90%); Modic I-b, 25 patients (71%); Control, nine patients (30%) ( Fig. 1 ), with no difference between Modic I-a and Modic I-b groups. Both groups experienced a significant decrease from baseline in level of low back pain (Modic I-a, 52 ± 5 vs. 28 ± 5 mm; Modic I-b, 62 ± 4 vs. 37 ± 5 mm; P < 0.05; Fig. 2 ). Controls showed no significant change from baseline in level of low back pain. Level of radiating pain in the lower limb at 24 h did not significantly differ from that at baseline in the three groups ( Fig. 3 ). Radiating pain level was lower, although not significantly, for Modic I-a and Modic I-b groups than for controls at 24 h.
At the last follow-up, patients did not differ in self-assessed improvement ( Fig. 1 ) or in changes from baseline in level of low back pain ( Fig. 2 ) or radiating pain ( Fig. 3 ).
1.4
Discussion
Our results indicate an acute short-term (24-h) reduction in pain with intradiscal injection of corticosteroids in severe, disabling chronic low back pain associated with Modic type 1 MRI signals. Interestingly, this effect was not observed in patients with chronic low back pain without evidence of Modic type 1 MRI signals. Our results add to the evidence of low back pain associated with Modic 1 changes. Revel et al. first suggested a specific clinical profile of such patients . Patients with Modic type 1 changes seen on MRI were described as experiencing severe worsening of chronic low back pain, and a high proportion reported increased pain during late night and morning . Biological findings agree with the clinical inflammatory profile. Patients with chronic low back pain and Modic type 1 changes show increased serum level of high-sensitivity C-reactive protein. As well, increased expression of tumor necrosis factor alpha in vertebral end plates has been reported . Despite this profile, patients show no further evidence of spondylarthropathy . There was no more evidence for spondylarthropathy or infectious spondylodiscitis in our patients.
No evidence exists for the effectiveness of intradiscal corticosteroid injection in unselected patients with low back pain . By contrast, good clinical response has been suggested for such injections in chronic low back pain with Modic 1 changes . A retrospective study found that patients with Modic type 1 changes had lower pain level than those with Modic type 2 changes 1 month after intradiscal injection of acetate of prednisolone, but not after 6 months. A study with longer follow-up was impaired by low quality of the data preventing any conclusion . Our results therefore agree with some of those previously reported. We found the clinical response in pain level to intradiscal injection of methylprednisolone much better in patients with Modic type 1 changes than in those without Modic type 1 changes. Interestingly, this effect was observed in patients with Modic type 1 changes, with or without previous surgery.
Our study contains some limitations. First of all, it was not a prospective randomized trial. Confounding factors were not completely excluded too. Modic type 1 groups partly differed. We did not detect any significant difference between controls and patients with Modic type 1 MRI changes considering age, sex, duration of symptoms and pain intensity. However disability, psychological distress, fears and believes and work conditions were not controlled in our patients. The acute short-term effect we observed may have persisted for weeks, as was previously described . Unfortunately, we could not verify this hypothesis with a mid-term assessment in our study. We may have also underestimated the long-term effect. Indeed, our study is limited by a high rate of control patients lost to the last follow-up. This situation may be due to lower or higher unsatisfactory results in this group than in others. Taking into account the distribution of the baseline values in our study, a difference of 20% in low back pain and a power of 80%, the number of patients needed to include should have been 220 for a 2-group comparison at a P -value of 0.05. However, our data suggesting the lack of long-lasting effect after intradiscal injection of corticosteroids with Modic type 1 changes agree with those of others.
Side effects from intradiscal injections of corticosteroids have been reported . The main side effects are destruction of the disc after it has been punctured and intradiscal calcification, which is associated with triamcinolone and cortivazol intradiscal injections . Iatrogenic intradiscal calcification has been associated with relapsed, unbearable low back pain and sciatica. To our knowledge, such a complication has not been reported after intradiscal injection of acetate of prednisolone or methylprednisolone .
Multidisciplinary programs of rehabilitation are recommended in persisting disabling low back pain . Treatment has to be adapted to physical, psychological, social and occupational conditions of each patient. Taken together, our results and those of others strongly support the concept of a specific syndrome associated with Modic type 1 changes seen on MRI: worsening symptoms during chronic low back pain, inflammatory clinical and biological profile, and acute short-term response in pain level with intradiscal corticosteroid injection. This syndrome appears to be due to active degenerative disc disease. The involvement of intervertebral instability in Modic type 1 changes has been suggested ; however, we did not observe spondylolisthesis in our patients and did not find evidence of intervertebral instability. Intradiscal injections of corticosteroids remain to be further investigated. Few results exist of the benefit of intradiscal injections of corticisteroids for patients with severe disabling chronic low back pain and Modic type 1 changes. Their effectiveness has not been demonstrated. Cohort studies are therefore necessary to evaluate the benefit – especially considering the potential damage into the disc – of such injections.
Disclosure of interest
The authors declare that they have no conflicts of interest concerning this article.
2
Version française
2.1
Introduction
Les anomalies morphologiques détectées sur la radiographie standard et correspondant aux discopathies dégénératives sont très peu spécifiques de la lombalgie chronique . On retrouve ces anomalies chez 40 % des personnes asymptomatiques. Les discopathies dégénératives lombaires peuvent être classées en trois stades en fonction du signal IRM de la moelle osseuse adjacente aux plateaux vertébraux (classification de Modic) . Le stade 1 se traduit par un hyposignal sur les séquences pondérées T1 associé à un hypersignal sur les séquences pondérées T2 (Modic de type 1); le stade 2 se traduit par un hypersignal sur les séquences pondérées T1 et T2 (Modic de type 2) et le stade 3 se traduit par un hyposignal sur les séquences pondérées T1 et T2 (Modic de type 3). Plusieurs études montrent une fiabilité inter- et intra-observateur de modérée à excellente pour la classification de Modic . Les modifications des plateaux vertébraux de type Modic 1 correspondent à des fissures de l’os sous-chondral, une concentration élevée de cytokines pro-inflammatoires et à une augmentation de la vascularisation et du contenu en cellules mononuclées . Le type Modic 2 correspond à une involution graisseuse de la moelle et le type Modic 3 à une ostéosclérose des plateaux vertébraux. . Une étude prospective indique que ces trois stades sont peut-être des étapes successives d’un processus dégénératif unique . Plusieurs études suggèrent que contrairement aux autres modifications dégénératives du rachis, les changements de type Modic 1 seraient étroitement liés à la lombalgie chronique . Ces modifications sont rares chez les personnes asymptomatiques . Les changements de type Modic 1 sont également associés à des caractéristiques spécifiques chez les patients souffrant de lombalgie chronique : les patients de type Modic 1 ont des douleurs plus marquées avec une recrudescence nocturne tardive et matinale ainsi qu’une élévation spécifique de la CRP ultrasensible à l’inverse des patients sans modifications de type Modic 1, et ce indépendamment de toute spondylarthropathie .
Les modifications de type Modic 1 pourraient ainsi refléter une réaction inflammatoire de la moelle osseuse adjacente aux plateaux vertébraux et aux disques intervertébraux connaissant un processus dégénératif actif. Ces éléments constituent l’argumentaire des infiltrations intradiscales de corticostéroïdes . Une étude de niveau IV suggère une réponse spécifique aux infiltrations de corticostéroïdes chez les patients lombalgiques chroniques avec discopathie de type Modic 1 . Ces patients ont moins de douleur que les patients présentant des discopathies de type Modic 2, un mois après le traitement par infiltrations intradiscales. Cependant, cette réponse n’a pas été documentée lors de suivis plus court ou à plus long terme.
Notre objectif était l’analyse rétrospective des dossiers médicaux de patients souffrant de lombalgie chronique traités par infiltration intradiscale de corticostéroïdes afin d’évaluer les effets cliniques à court et long termes de ces infiltrations chez des patients présentant ou non des changements IRM de type Modic 1. Ces changements de type Modic 1 ont été décrits après chirurgie , sans être spécifiquement étudiés . Nous avons donc différencié les changements de type Modic 1 chez les patients avec ou sans chirurgie ou nucléolyse préalable.
2.2
Patients et méthode
Il s’agit d’une étude rétrospective. Les données des dossiers médicaux des patients ont été collectées sur trois ans de 1995 à 1998, correspondant, dans notre département, à la période où les infiltrations intradiscales de corticostéroïdes n’étaient pas encore limitées aux patients souffrant de lombalgie chronique avec modifications IRM de type Modic 1.
2.2.1
Patients
Les patients ont été inclus de façon consécutive. Ils souffraient de lombalgie sévère, chronique et invalidante ne répondant pas aux traitements habituels. Il n’y avait pas de syndrome inflammatoire biologique, ni de maladie osseuse métabolique, infection locale ou tumeur. Tous les patients ont eu une IRM avec séquences pondérées T1 et T2. Les résultats d’imagerie furent analysés par au moins deux spécialistes du rachis (JB, PD, SP, MR). Les patients opérés du rachis ou traités par nucléolyse au cours des six mois précédant le traitement par infiltrations n’ont pas été inclus dans l’étude. D’après notre expérience, un œdème médullaire et des plateaux vertébraux adjacents est communément observé en IRM juste après une chirurgie discale ou un traitement par nucléolyse sans valeur pathologique particulière et disparaît au-delà de six mois.
Les patients étaient séparés en trois groupes sur la base des changements IRM de type Modic 1 : groupe Modic I-a ou changements de type Modic 1 sans antécédent de chirurgie discale ni traitement par nucléolyse ; groupe Modic I-b ou changements de type Modic 1 avec antécédent de chirurgie discale ou traitement par nucléolyse ; groupe Témoin ou discopathie dégénérative sans changement IRM de type Modic 1.
2.2.2
Infiltration intradiscale de méthylprednisolone
Durant une hospitalisation de trois jours, les patients ont eu une discographie sous contrôle fluoroscopique comme décrit précédemment , couplée à une infiltration de méthylprednisolone (2 mL ; SoluMedrol ® , Pharmacia, Saint-Quentin-en-Yvelines, France) directement au centre du disque. Après l’infiltration, les patients devaient rester allongés pendant une période allant 12 à 24 heures puis porter une ceinture de maintien lombaire durant les deux semaines.
2.2.3
Évaluation clinique
Les paramètres cliniques étaient collectés à plusieurs moments, tout d’abord le premier jour d’hospitalisation avant traitement, puis 24 heures après l’infiltration et lors de la dernière consultation de suivi. Durant la visite initiale, les paramètres cliniques recueillis étaient les suivants : âge, sexe, durée des symptômes, antécédents de chirurgie ou nucléolyse à l’étage de l’infiltration, intensité de la douleur lombaire (mesurée sur une échelle visuelle analogique [EVA] de 0 à 100 mm) et intensité de la douleur irradiant dans les membres inférieurs (EVA). Les critères de jugement retenus pour cette étude étaient la douleur lombaire et la douleur irradiant dans les membres inférieurs sur l’EVA ainsi que l’auto-évaluation de l’amélioration des douleurs (« oui » si le patient ressent une amélioration et « non » si le patient ne ressent aucune amélioration) 24 heures après l’infiltration et lors de la dernière visite de suivi.
2.2.4
Analyse statistique
Les calculs ont été faits avec le logiciel Statview v4.5 (Abacus Concepts, Berkeley, CA). Les variables quantitatives ont été exprimées en moyennes et erreurs standard (SE) et traitées par analyse de variance Anova et par le test de Fisher. Les variables qualitatives ont été exprimées en nombres et pourcentages et comparées avec le test du Chi 2 . La valeur p < 0,05 était considérée comme statistiquement significative.
2.3
Résultats
Nous avons inclus 97 patients : 30 dans le groupe Modic I-a, 37 dans le groupe Modic I-b et 30 dans le groupe témoin ( Tableau 1 ). Les patients des groupes Modic I-a et Modic I-b avaient des caractéristiques différentes à la visite initiale concernant la durée des symptômes et l’intensité de la douleur irradiée. Dans le groupe Modic I-b, 33 patients avaient des antécédents de chirurgie du rachis, 31 de discectomie lombaire au niveau de l’infiltration, deux d’arthrodèse et quatre de nucléolyse à la papaïne. Les niveaux opératoires étaient L4-L5 dans 16 cas et L5-S1 dans 21 cas. Le groupe témoin, comprenait sept cas de discectomie, deux cas d’arthrodèse, sept de niveau L4-L5 et deux de niveau L5-S1.
Related posts:
Back pain during pregnancy and living conditions – a comparison between Beninese and Canadian women
Therapeutic education and intermittent self-catheterization: Recommendations for an educational program and a literature review
Pain assessment in Charcot-Marie-Tooth (CMT) disease
Dystonic pattern in a fibromyalgia patient: Relevance of a biopsychosocial rehabilitation approach
Patients with complicated Pott’s disease: Management in a rehabilitation department and functional prognosis
Returning to work after a stroke: A retrospective study at the Physical and Rehabilitation Medicine Center “La Tour de Gassies”
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
