Document resource
Introduction and Objectives Lymph node tuberculosis represents a common manifestation of tuberculosis (TB). Diagnosis can be challenging, particularly if the condition is not suspected.Multiple complications are associated with TB treatment, including development of paradoxical reactions. Corticosteroids are not routinely recommended for treatment of lymph node TB, however may be used in clinical practice. We aimed to conduct a multi-centre audit to review diagnostic procedures and treatment complications during management of lymph node TB, referring to NICE guidance.1 Methods Three UK NHS Trusts identified patients diagnosed with confirmed (culture-confirmed, acid-fast bacilli or molecular diagnostic positive, or histological evidence) or presumed lymph node tuberculosis between 1st January 2019 to 31st December 2023. A retrospective case note review was performed.Results Data was available on 406 individuals who were diagnosed with lymph node tuberculosis. Results are summarised by site in table 1. 174 (43%) were female. 269 (66%) were microbiologically confirmed. In 213 (52%), presentation was lymph node involvement only, while in 191 (47%), presentation was lymph node TB with other sites involved.The median number of procedures for diagnosis was 1, (min=0, max =7). 269 (66%) cases were microbiologically confirmed. Overall, in 21 (5%) cases a paradoxical reaction on anti-tuberculous therapy (ATT) was diagnosed. However, this differed by site (Site 1 = 13%, Site 2 = 8%, Site 3 = 1%, p≤0.01), whilst 25 (6%) patients were managed with corticosteroids during ATT. There were high rates of treatment success.Abstract M42 Table 1 Trust 1 Trust 2 Trust 3 Total Number of patients identified, n 61* 144 201 406 Age, median years (IQR) 33 (27 – 47) 37 (28- 49) 43 (33 - 55) Female, n (%) 18 (30) 56 (39) 100 (50) 174 LN TB only, n (%) 28 (46) 96 (67) 89 (44) 213 (52) LN with other site involvement, n (%) 31 (51)** 48 (33) 112 (56) 191 (47) Number of procedures, median (IQR) 2 (1–2) 1 (1 – 1) 1 (1–1) 1 Microbiologically confirmed, n (%) *Hospital: 19 (90)*Lab: 40 (100) 79 (55) 131 (65) 269 (66) Diagnosed/Treated as Paradoxical reaction, n (%) 8 (13) 11 (8) 2 (1) 21 (5) Corticosteroids during treatment, n (%) 12 (20) 4 (3) 9 (4) 25 (5) Treatment outcome = completed treatment, n (%) 55 (90) 139 (97) 197 (98) 391 (96) Treatment outcome = care transferred or lost to follow-up, n (%) 4 (7) 1 (0.7) 1 (0.5) 6 (1) Treatment outcome = treatment failure or died, n (%) 0 (0) 4 (2.7) 0 (0) 4 (1) *In Trust 1, records were identified by hospital admission coded as lymph node tuberculosis or if microbiology laboratory records identified a positive Mtb culture from a non-sputum sample. Lymph node TB managed entirely in the outpatient setting may not have been included. **Included if significant lymph node diseaseConclusions Lymph node TB is responsible for considerable morbidity. Specific standards for diagnosis and management of lymph node TB and guidance for use of corticosteroids in this condition are lacking. 1 However, our data suggests achieving microbiological diagnosis is challenging. Patients usually have at least one procedure, but may undergo multiple diagnostic procedures. In a third of our cases, diagnosis was not microbiologically confirmed. Paradoxical reactions occur during treatment, although due to lack of specific diagnostic tests or consensus definitions, these may be difficult to recognise. Diagnosis of paradoxical reaction and treatment with corticosteroids was relatively uncommon but may vary by setting.Reference Tuberculosis (2016) NICE guideline NG33. Last updated 16/02/2024.