Question | Discussion | References


Introduction and Overview of TB-IRIS

Immune Reconstitution Inflammatory Syndrome (IRIS) refers to a phenomenon experienced by people living with HIV who have recently initiated antiretroviral therapy.  The partial reconstitution of the immune system following initiation of antiretroviral therapy in these patients can result in an exaggerated inflammatory response against any concurrent opportunistic infection.  Sometimes the opportunistic infection pathogen against which the inflammatory response is directed remains clinically 'silent' prior to initiation of antiretroviral therapy, such that antiretroviral therapy 'unmasks' a previously undiagnosed disease.  Tuberculosis-Immune Reconstitution Inflammatory Syndrome (TB-IRIS) refers specifically to IRIS that occurs in the context of a patient with active Mycobacterium tuberculosis infection, and is a relatively common complication for HIV-infected persons who initiate antiretroviral therapy in resource-limited settings, particularly in regions that have a high TB prevalence.  Management of TB-IRIS is complicated by the difficulty in establishing the diagnosis and the absence of randomized controlled trials that clearly define optimal treatment strategies.


Tuberculosis IRIS is estimated to occur in 11% to 45% of patients co-infected with TB and HIV following initiation of antiretroviral therapy[1].  Difficulties associated with the diagnosis of TB-IRIS, and the lack (until recently) of a clear case definition for TB-IRIS, have resulted in the wide range of disease estimates.  Rates of TB-IRIS within a particular geographic area primarily depend on the prevalence of TB in the region and typical CD4 counts when patients commence antiretroviral therapy. Although TB-IRIS often causes significant clinical symptoms, it is rarely fatal: a recent meta-analysis involving more than 13,000 patients with TB-IRIS reported a case-fatality rate of 3.2%[2]. A case series from South Africa reported a significantly higher mortality rate associated with TB-IRIS among patients in whom TB-IRIS involved the central nervous system[3].

Pathophysiology, Risk Factors, and Clinical Presentation, for TB-IRIS

The exact pathophysiology of TB-IRIS remains unclear, but it appears to result from an invigorated inflammatory reaction against Mycobacterium tuberculosis antigens driven by antiretroviral therapy-induced reconstitution of the immune system[4,5].  Specific risk factors identified for TB-IRIS include multiple prior opportunistic infections, low CD4 count at initiation of antiretroviral therapy, rapid rise in CD4 cell count following initiation of antiretroviral therapy, and a rapid drop in HIV RNA levels following initiation of antiretroviral therapy[4,6,7,8,9]. In addition, persons with more advanced HIV disease at the time of initiation of antiretroviral therapy appear to have increased risk of developing IRIS [10].  Patients with pulmonary TB-IRIS typically present with fever, lymphadenopathy, and respiratory complaints[7].  Given that disseminated disease frequently develops in HIV-infected persons with active TB, especially in those with advanced HIV infection, TB-IRIS can present in diverse ways.  The range of manifestations include abdominal lymphadenopathy, cervical lymphadenitis (Figure 1), abscesses, pulmonary effusions and/or pulmonary infiltrates (Figure 2), pericardial effusions (Figure 3), meningitis, brain tuberculomas, and skeletal lesions (usually vertebral) . In most cases, TB-IRIS typically manifests within 2 months of initiation of antiretroviral therapy [12], though late presentations of TB-IRIS have been reported [13].

Diagnosis and Case Definition of TB-IRIS

The diagnosis of TB-IRIS is rarely straightforward, as no single laboratory test can establish or rule out TB-IRIS.  To establish a provisional diagnosis of TB-IRIS, clinicians must rely upon a combination of clinical signs and symptoms, laboratory findings, the likelihood of other conditions, and clinical judgment.  In 2008, a panel of expert clinicians convened in Kampala, Uganda to formulate a case definition of TB-IRIS that could be practically applied in resource-limited settings [14].  The International Network for the Study of HIV-associated IRIS panel distinguished between “paradoxical TB-IRIS” (worsening of previously diagnosed TB disease in a person with HIV infection who initiates antiretroviral therapy) and "unmasking TB-IRIS" (clinical manifestations of previously undiagnosed TB in a person with HIV infection who initiates antiretroviral therapy).  This panel also agreed by consensus on a case definition and a provisional case definition for paradoxical TB-IRIS (Figure 4) and unmasking TB-IRIS (Figure 5). The case definitions deliberately exclude the use of laboratory tests, such as CD4 and viral load testing, recognizing that clinicians managing HIV-infected persons in resource-limited settings do not routinely have access to such tests.  Investigators subsequently validated these case definitions in cohort studies in Thailand [15] and South Africa [16,17], and these definitions remain the most useful case definitions for application in resource-limited settings. 

Timing of Antiretroviral Therapy with Known TB

Recently released data from three randomized trials suggest that although delaying the initiation of antiretroviral therapy in patients with HIV and TB coinfection reduces the risk of TB-IRIS, this approach correlates with higher mortality rates when compared with earlier initiation of antiretroviral therapy, especially in patients with low CD4 counts[18,19,20].  The WHO Guidelines therefore recommend starting antiretroviral therapy as soon as tuberculosis therapy is tolerated —specifically, as soon as possible (and within 8 weeks) of initiation of anti-TB therapy [21,22]

Use of Corticosteroids in TB-IRIS

Numerous anecdotal reports and small case series have suggested corticosteroids help reduce inflammation in the setting of TB-IRIS.  Specific formulations and doses used in reports have included intravenous methylprednisolone 40 mg every 12 hours and prednisone 20 to 70 mg per day. The length of steroid administration and the use of a steroid taper have also varied considerably [12]. In the only randomized controlled trial that has evaluated the safety and efficacy of corticosteroids for TB-IRIS, investigators randomized 110 South African HIV-infected persons with evidence of paradoxical TB-IRIS to receive either placebo or a 4-week course of prednisone (1.5 mg/kg/day, for 2 weeks, followed by oral prednisone dosed at 0.75 mg/kg/day for 2 weeks) . The study excluded patients with life-threatening manifestations of TB-IRIS. The primary combined endpoint (days hospitalized plus therapeutic outpatient procedures) was significantly lower in patients randomized to the corticosteroid arm.  At weeks two and four of the study, patients who received corticosteroids had higher Karnofsky scores, higher quality of life scores, more frequent improvement in chest radiograph findings, and fewer symptoms, when compared with patients who received placebo (Figure 6 series Study Title and SourceStudy DesignWeek 2 Symptom ScoreWeek 4 Symptom ScoreWeek 2 Chest Radiograph ScoreWeek 4 Chest Radiograph ScoreSource & Permission).  Severe infections occurred with equal frequency in the two groups. Corticosteroid therapy was re-initiated in 10 patients after 4 weeks, suggesting that some patients may require an extended course of corticosteroids.

Risk of Drug-Resistant TB in Patients with Suspected TB-IRIS

In a recent cohort study of South African patients, 10% of the patients with suspected TB-IRIS were found to have previously undiagnosed rifampicin-resistant TB[24].  Hence, drug resistance should ideally be ruled out in patients with active TB and a presentation consistent with TB-IRIS, especially if administration of corticosteroids is being considered.

Recommendations for the Management of TB-IRIS

Evidence-based guidelines for the management of suspected or confirmed TB-IRIS in resource-limited setting do not yet exist, owing to the scarcity of randomized controlled trials evaluating management approaches.  Nevertheless, we recommend the following general principles of treatment based on existing reports, trials, and expert opinion.

  • Clinicians should become familiar with the case definitions for TB-IRIS and use these definitions in resource-limited settings.  Because no definitive test for TB-IRIS exists, clinicians should consider alternative diagnoses in patients with suspected TB-IRIS, especially if the severity of the condition warrants administering corticosteroids.  Clinicians should ideally evaluate the patient for drug-resistant TB, as cases of multidrug-resistant-TB have been identified in patients with TB-IRIS.
  • Most cases of TB-IRIS have a self-limited course and will resolve with continuing treatment with little or no change in overall management. Use of non-steroidal anti-inflammatory agents may provide adequate relief of symptoms in mild cases, though the effectiveness of this approach remains unproven.
  • Patients should continue anti-tuberculous therapy without change unless there is a reason to suspect the current regimen is inadequate (for example if MDR-TB is suspected, or drug-drug interactions could result in suboptimal serum anti-TB drug levels).
  • In nearly all cases, the patient with TB-IRIS should remain on antiretroviral therapy.  Certain circumstances, however, may require temporary interruption of antiretroviral therapy, such as when life-threatening complications of IRIS develop (e.g. increasing intracranial pressure in the setting of TB meningitis, or an expanding abscess that compromises the patient’s airway).
  • We recommend performing incision and drainage (with stains and cultures) for localized abscesses, though evidence-based guidelines are lacking.
  • Although criteria for initiating corticosteroids remain poorly defined, we recommend administering prednisone for TB-IRIS cases when patients have persistent fever, abscesses, meningitis, or dyspnea.
  • The dosing and duration of corticosteroids should be tailored to individual patient circumstances.  We recommend patients with moderate to severe TB-IRIS receive prednisone dosed at 1.5 mg/kg/day for 2 weeks, followed by a taper of 0.75 mg/kg/day over 2 additional weeks. Some patients, particularly those with very high mycobacterial loads or severe clinical deterioration, may require prolonged courses of corticosteroids.
  • The patient receiving corticosteroids for TB-IRIS should undergo careful monitoring for response to corticosteroid therapy.  If the patient’s clinical condition does not improve, the clinician should consider an alternative diagnosis and possible termination of corticosteroids, since administering corticosteroids to a patient with a non-tuberculous opportunistic infection or disease could result in further clinical deterioration.  Ideally, the clinician should rule out drug-resistant TB before administering corticosteroids.

[Back to Case Question | See Case References]

Back to Top

  • The following link will open in a new window.
    Figure 1. Cervical Lymphadenitis in Patient with TB-IRIS

    This image shows a patient with a large cervical lymphatic abscess that developed as a result of TB-IRIS.
    Source: Professor OC Abraham, Department of Medicine Unit-1 and Infectious Disease unit, Christian Medical College, Vellore, India

    Figure 1
  • The following link will open in a new window.
    Figure 2. Chest Radiograph in Patient with Pulmonary Manifestations of TB-IRIS

    This chest radiograph shows a large right-sided pleural effusion and scattered patchy infiltrates that developed as a result of TB-IRIS.
    Source: Government Hospital of Thoracic Medicine, Tambaram Sanatorium, Chennai, India

    Figure 2
  • The following link will open in a new window.
    Figure 3. Pericardial Effusion in Patient with and TB-IRIS

    This chest radiograph shows a patient with TB prior to receiving antiretroviral therapy (Panel A) and the same patient with a large pericardial effusion (Panel B) 2 months after receiving antiretroviral therapy.
    Source: Government Hospital of Thoracic Medicine, Tambaram Sanatorium, Chennai, India

    Figure 3
  • The following link will open in a new window.
    Figure 4. INSHI Case Definition for Paradoxical TB-associated IRIS

    Source: Meintjes G, Lawn SD, Scano F, et al. Tuberculosis-associated immune reconstitution inflammatory syndrome: case definitions for use in resource-limited settings. Lancet Infect Dis. 2008;8:516–23.

    Figure 4
  • The following link will open in a new window.
    Figure 5. INSHI Provisional Case Definition for Unmasking TB-associated IRIS

    Source: : Meintjes G, Lawn SD, Scano F, et al. Tuberculosis-associated immune reconstitution inflammatory syndrome: case definitions for use in resource-limited settings. Lancet Infect Dis. 2008;8:516–23.

    Figure 5
  • The following link will open in a new window.
    Figure 6. Efficacy and Safety of a Four-week Course of Corticosteroids for HIV-Infected Persons with Paradoxical TB-IRIS in South Africa. Image 1. Study Title and Source
    Figure 6 (series)