The advent of antiretroviral therapy (ART) revolutionised the care of patients with human immunodeficiency virus (HIV) but even in the late 1980s, clinicians began to recognise that despite improvement in their patients’ CD4 T-lymphocyte (CD4) cell count with ART, some patients paradoxically worsened with a new, or exacerbation of a previously treated, opportunistic infection. These disease episodes were characterised by exaggerated and/or atypical inflammation and involved a wide variety of co-infections, including atypical mycobacteria, Mycobacterium tuberculosis, Pneumocystis, Cryptococcus, varicella zoster virus (VZV), John Cunningham (JC) virus and cytomegalovirus. Collectively, these conditions were referred to as immune restoration disease, though now the term immune reconstitution inflammatory syndrome (IRIS) is more commonly used. IRIS events occur mostly in the first 6 months of ART commencement, but may also occur in the setting of ART intensification. Importantly, IRIS events are not limited to individuals with HIV but also occur in non-HIV settings, particularly solid organ transplant and haematopoietic stem cell transplant recipients in the setting of immune recovery and withdrawal of immunosuppressant drug therapy.
Terminology and classification
An IRIS is best described in relation to the type of provoking co-infection e.g. tuberculosis-associated IRIS (TB-IRIS) or cryptococcosis-associated IRIS (C-IRIS) because the immunopathogenesis may differ for different types of pathogen. Further classification is based on locality or organ of involvement, e.g. central nervous system (CNS) C-IRIS, or Mycobacterium avium complex (MAC)-IRIS lymphadenitis. This classification reflects the triggering co-infection and the target organ involved and helps the clinician predict outcomes and assists with a management algorithm. For example, a patient with CNS C-IRIS, should be assessed for meningoencephalitis symptoms, have a lumbar puncture performed for opening pressure and CNS fungal culture and be managed with anti-inflammatories targeting symptoms such as headache and anti-emetics for nausea. Further, this detailed classification assists in research on the immunopathogenesis of the various forms of IRIS.
The clinical manifestations of an IRIS are largely determined by the provoking pathogen. Indeed different pathogens are likely to result in different IRIS disease mechanisms, with mycobacterial and fungal IRIS being associated with granulomatous and suppurative inflammation, while IRIS associated with viruses such as herpes simplex virus (HSV) or JC virus is usually characterised by a CD8+ T-cell response. Thus, when exploring the literature on IRIS pathogenesis and management, studies targeted toward specific forms of IRIS are likely to be more helpful than studies on all-comer IRIS.
The temporal association between the recognition of a co-infection, IRIS and ART commencement is also important - unmasking IRIS when ART unmasks a previously unrecognised infection or paradoxical IRIS when a previously well-treated co-infection flares post-ART. For example, unmasking TB-IRIS is diagnosed when an ART-naive patient with HIV and no recognisable features of co-infection, commences on ART and soon after develops symptoms suggestive of tuberculosis (TB). The term ‘ART-associated TB’ is also sometimes used if the TB is not characterised by unusual inflammation. In contrast, paradoxical CNS C-IRIS is said to occur when an ART-naive patient with HIV who presents with cryptococcal meningitis (CM), undergoes and displays improvement on CM treatment, then starts ART, after which a flare of CM symptoms develops.
An IRIS is thought to be due to an immune response against an active (often quiescent) infection by an opportunistic pathogen or antigens of non-viable pathogens. The strongest risk factors for an IRIS include a low CD4cell count and an active or subclinical opportunistic infection with a high pathogen load, but other host, pathogen and immune factors have been recognised. The inflammation that arises from an IRIS is not uncommonly misinterpreted as an opportunistic infection, but the process is one of restoration, rather than loss of pathogen-specific immune response. Discerning between these conditions is essential as clinical management differs.
Diagnosis of an IRIS is made clinically on the basis of a temporal association between starting ART and the clinical phenomena; exclusion of alternative diagnoses; and is usually accompanied by a fall in HIV viral load but may not be associated with alterations in blood CD4 cell count. The major differential diagnosis is relapse of the infection that triggered the IRIS or occurrence of another infection. Clinicians should think broadly for other causes of presenting symptoms.
Research groups have attempted to describe consensus definitions specific for TB-IRIS7 and C-IRIS. However, there are no accepted diagnostic tests for an IRIS and the validity of an IRIS diagnosis may vary both in clinical practice and also in research. In research studies, an IRIS is often classified as definite, probable and possible – this approach may also be helpful in routine clinical management.
Immunopathogenesis of an IRIS
Broadly, an IRIS is driven by both host immune responses and pathogen load. Patients with lower CD4 cell counts, poor CD4 cell recovery to ART, higher antigen load and higher frequencies of antigen-specific T-cells ready to be mobilised once immunosuppression reverses are predisposed to an IRIS. The immunopathogenesis of an IRIS varies for different pathogens, organ involvement and also between unmasking and paradoxical IRIS. In paradoxical TB-IRIS, there is now good evidence that activation of monocytes contributes to the immunopathology. Unfortunately, assays used in IRIS research such as interferon gamma (IFN-) release assays, plasma cytokine levels, antigen-specific T-cells and activated T-cells, and host genetic determinants are yet to be translated into clinical practice and are not readily available in routine practice. Readers are referred to recent reviews for a summary of IRIS immunopathogenesis.
Clinical diagnosis and recognition of an IRIS
An IRIS mostly occurs between 2 weeks and 3 months after ART commencement or ART intensification, but may occur earlier (within days) and as late as a year after treatment. Temporal onset of IRIS is particularly difficult to attribute in patients with erratic ART adherence. Typically, patients with a previously low nadir CD4 cell count, present with fever and symptoms suggestive of an opportunistic infection, albeit usually manifested in an atypical manner. Practically any chronic viral, mycobacterial or fungal infection may be associated with an IRIS.
- Recurrence of fever, cough and worsening radiological features in TB-IRIS
- Recurrence of seizures, headaches and worsening symptoms of raised intracranial pressure in C-IRIS
- Extra-pulmonary manifestations in MAC-IRIS with painful lymph nodes and severe abdominal pain
- Transaminitis, fever and jaundice in hepatitis C virus (HCV) or hepatitis B virus (HBV) flares
- Progressive brain lesions in progressive multifocal leukoencephalopathy (PML)-IRIS
- Fevers and dyspnoea in Pneumocystis jiroveci pneumonia (PJP)-IRIS
- Worsening vision with cytomegalovirus (CMV)-retinitis-IRIS
- New headaches and focal lesions in toxoplasmosis-IRIS
- Meningomyelitis with VZV-IRIS
General approach and management of an IRIS
A suite of investigations to exclude recurrence of the co-infection and other differential diagnoses is necessary. For example, when suspecting paradoxical TB-IRIS in a patient presenting with fever and cough, considerations should include drug non-adherence, drug resistance, new community-acquired pneumonia, PJP and viral infections. Investigations would include sputum microscopy and culture for bacterial infection and TB, chest X-ray or CT scan of chest or both. Similarly, a patient being considered for paradoxical CNS C-IRIS, should have a CT scan of brain, lumbar puncture for opening pressure, cell counts and bacterial, TB and fungal culture and possibly, detection of toxoplasma and JCV DNA by polymerase chain reaction (PCR) or MRI brain or both.
When evaluating cultures, it is important to note that while cultures are often negative in an IRIS (in keeping with the concept that there are residual provoking antigens despite non-viable organisms), the presence of a positive culture does not exclude an IRIS, as viable pathogens are clearly also antigenic. After the exclusion of alternative diagnoses, symptomatic management of an IRIS is usually sufficient. For example, therapeutic lumbar punctures and non-steroidal anti-inflammatory drugs often relieve paradoxical CNS C-IRIS. Escalation in antimicrobial therapy is only necessary in microbiological relapse. Importantly, ARTs should not be ceased.
While many clinicians use steroids in an IRIS event, the evidence for their use is not strong and use should only be considered in cases of severe symptomatic IRIS refractory to simple measures. A randomised placebo-controlled trial in South Africa, used 4 weeks of prednisolone (1.5 mg then 0.75 mg/kg/day) in patients with paradoxical pulmonary TB-IRIS, and showed reduction in a combined endpoint of duration of hospitalisation and number of procedures, and improvement in symptoms and wellbeing but increased concurrent infections and glucocorticoid reactions.
Of note, as IRIS symptoms may be protracted for weeks and may also recur, patience is necessary. IRIS symptoms notoriously recur in the setting of alterations in steroid dosing, thus often resulting in extremely long durations of steroid therapy and its accompanying risks, which lead to further immunosuppression. Tumour necrosis factor blockade in MAC-, TB- and C-IRIS have been reported in a few case reports recently. Surgical measures have occasionally been necessary (e.g. drainage or excision of large mycobacterial-IRIS lymph nodes or abscesses).
Prevention of an IRIS
Early diagnosis and treatment of HIV infection regardless of CD4 cell counts as advocated by the findings of the START study will hopefully minimise the number of patients commencing ART with severe CD4 cell deficiency, thus reducing their susceptibility to co-infections and also their risks for a future IRIS event. Patients with known co-infections, should be treated sufficiently to reduce the pathogen load before ART commencement. Needless to say, thorough assessment and screening for co-infections before ART commencement is vital.
Timing of ART commencement in the setting of opportunistic infections, specifically in TB and in CM are discussed in detail elsewhere. Broadly, early ART commencement is recommended particularly in the patients who have advanced immunodeficiency. To date, the studies examining the timing of ART commencement in the setting of pulmonary TB have been in favour of early commencement, but this is less urgent in persons with relatively preserved CD4 counts (> 220 cells/μL) and should be delayed in TB meningitis. The survival benefit seen in early ART commencement is best seen in patients with CD4 counts < 50 cells/μL. In the setting of CM, randomised studies suggest ART should be deferred until > 10 days post CM diagnosis. Timing of ART commencement has not been systematically assessed in other co-infections.
Currently recommended first-line ART regimens are largely equipotent, thus there is no benefit of one over another in prevention of an IRIS. The role of maraviroc (a CCR5 inhibitor) in prevention of IRIS was explored in the CADIRIS study performed in ART-naïve patients with CD4 counts > 100 cells/μL. While the addition of maraviroc did not show any benefit in this study, there is a possibility that CCR5 inhibitors in the setting CM and paradoxical CNS C-IRIS, in more immune-deplete patients or as adjunctive treatment for severe PML-IRIS may be beneficial.
Autoimmune disease and immune-mediated inflammatory disease associated with immune reconstitution
Apart from an IRIS, other immune reconstitution disorders may occur following treatment of severely immunodeficient HIV patients with ART, though are uncommon. The most common are autoimmune disease, mainly Graves’ disease and sarcoidosis.
Future research in IRIS
There is an urgent need to translate research findings on IRIS into clinical practice. Standardisation of assay stimuli and performance, preferably developed into simple field-based or point-of-care tests for biomarkers of IRIS are necessary. Prevention and treatment strategies including steroids and novel immunomodulators should be trialled.