Antibiotics should not be used for back/leg pain

© 2021 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group, on behalf of the Nordic Orthopedic Federation. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. DOI 10.1080/17453674.2020.1871190 Sir,—It is rather depressing to read the article addressing the topic above in Acta Orthopaedica (Fritzell et al. 2021). By using the definitive title: ”Antibiotics should not be used to treat back pain” supported by a highly selective and extremely limited reference list, Fritzell et al. attempt to shut down the low virulent infection hypothesis leading to Modic changes (MC) and chronic low back pain (Fritzell et al. 2021). Readers have to plow through a long and basically irrelevant introduction (Albert et al. 2008, Fritzell et al. 2021) prior to encountering the article’s point of contention. Are antibiotics effective for the treatment of patients with MCs? 1. The authors’ own publication, in which circa 50% of a child cohort complete a questionnaire 13 years after inclusion is likely just an abstraction for most readers as regards making the case for or against the usage of antibiotic treatment for patients with back pain. 2. Fritzell et al. (2019) also refer to their own biopsy article in which they found no evidence of bacteria in the disc material from several patients and in which they conclude that if bacteria were found in an individual patient that this would be due to a contamination process due to the biopsy itself. Many studies (20+) have disproven the contamination hypothesis during the years (Capoor et al. 2019, Manniche and O’Neill 2019. Pradip et al. 2020). And by using fluorescence in situ hybridization microscopy C. acnes bacteria can be seen in aggregates and biofilms in human disc material which has initiated a local inflammatory response (Capoor et al. 2017, Ohrt-Nissen et al. 2018). Several leading experts in this field have criticized the results of this study and point to several methodological problems as the reason for their lack of bacterial identification (Capoor et al. 2017). The interesting considerations in the article are hidden in the last 20 lines (Fritzell et al. 2021). Fritzell describes how a controversial RCT from Bråten et al. (2019) tested whether it was possible to reproduce the same large effect with antibiotic treatment on patients with MCs as the Danish trial (Albert et al. 2013). The Norwegian trial concluded that they did not find a “significant” effect (Bråten et al. 2019). Bråten et al. chose – despite reviewer objections (BMJ 2019) – to mix results for MC type 1 with MC type 2 in their analyses. This is the equivalent of mixing cold and warm water! The Tables in the Bråten et al. article’s supplementary appendix (2019) told a different story when data for patients with MC1 was presented distinctly from patients with MC2. Data for MC1 patients demonstrated a statistically significant difference and meaningful improvement rarely seen in RCTs involving supervised exercise, manipulation or even spinal surgery for chronic back pain. Patients with MC2 did less well than the placebo group! Already at the start of the reviewer process and since then a large number of back pain experts have written in different forums about the methodological weaknesses of the study (including the mixing up of data sets) in the Norwegian trial and highlighted the misleading conclusions based upon mixed MC1 and 2 patients (BMJ 2019, Albert 2019, Creaney 2019, Fairbanks 2019, Joffe 2019, Lambert 2019). The Norwegian authors have recently published a new sub-group analysis of their RCT (Kristoffersen et al. 2020). The conclusions have been significantly modified. A substantial subgroup of patients with MC1 and oedema seen on STIR sequences demonstrated a large difference between the actively treated group and the placebo group as measured by the Roland Morris Disability Questionnaire (RMDQ), the primary outcome measure; –5.1 RMDQ points; 95% CI –8.2 to –1.9; p = 0.008). The clinical improvements were already seen at 3 months and were consistent at 1-year follow-up which showed that 27% of the actively treated group experienced improvements of more than 75%! The Number Needed to Treat (NNT) in this subgroup was 3.1. The overall conclusion can be that the disc low grade infection hypothesis is a most interesting area of research. This patient group which suffers from longstanding and severe back pain worldwide is deserving of more than simplistic attempts to block further research in this space.