Personally, I think it is case dependent. First, it has to be determined whether its an unmasking or paradoxical IRIS. Starting steroids in under-controlled infections can be detrimental. 

Although the manifestations of IRIS may be florid and severe it usually runs a self-limiting course with rare progression to mortality. ARV therapy is not discontinued but a search should be intensified for the associated condition (e.g. opportunistic infections, malignancies) and treatment of the condition be prioritised. Corticosteroid therapy is indicated for severe cases of IRIS such as CNS involvement with raised intracranial pressures (e.g. cryptococcal meningitis, CNS tuberculosis with tuberculoma or TB meningitis), respiratory failure from airway obstruction or ARDS. However, caution must be exercised with the use of corticosteroid therapy in order to reduce the risk of additional immunosuppression and other serious steroid-related adverse effects. Adjunctive steroid therapy may be harmful for patients with untreated or poorly treated opportunistic infections (OIs) and may lead to dissemination of such OIs with increased risk of mortality. 

Refs:

1. Surendra K. Sharma, Manish Soneja. HIV & immune reconstitution inflammatory syndrome (IRIS). Indian J Med Res 134: 2011; 866-877.

2. Meintjes G, Wilkinson R, Morroni C, et al. Randomized placebo-controlled trial of prednisone for the TB immune reconstitution inflammatory syndrome. [Abstract 34.] 16th Conference on Retroviruses and Opportunistic Infections. February 8-11, 2009; Montreal, Canada.

Systemic corticosteroids are indicated for severe IRIS manifesting in an organ or system mild to moderate are managed conservatively. IRIS is a diagnosis of exclusion, as there is no reliable marker to confirm IRIS.

The fear of further compromising the already immuno-compromised state by my experience is theoretical once the corticosteroids don't exceed three weeks at moderate doses.

Due to the role of many inflamatory boimarkers,systemic steroids couls improve symptoms in patients with IRIS.

Patients with immune reconstitution may develop  severe conditions and multy-systemic compromise  linked to a delayed manifestation of systemic infection, so the priority is treat the underlaying infection than systemic corticoesteroids

IRIS is not the worsening of the previous disease but an exagerated response due to the lysis of the infectious agent and in the other hand, the recovery of the immune system. Steroids are helpful in the management of related symptoms we have  experienced that condition in many HIV/TB patients

corticosteroids are beneficial in iris especially in early time.

They do play a role. Below you find  studies on the topic.

1. Kasang C, Ulmer A, Donhauser N, Schmidt B, Stich A, Klinker H, Kalluvya S, Koutsilieri E, Rethwilm A, Scheller C. HIV patients treated with low-dose prednisolone exhibit lower immune activation than untreated patients. BMC Infect Dis. 2012 Jan 20; 12: 14-2334-12-14. PMCID: PMC3282641.

2. Kremer H, Sonnenberg-Schwan U, Arendt G, Brockmeyer NH, Potthoff A, Ulmer A, Graefe K, Lorenzen T, Starke W, Walker UA, German Competence Network HIV/AIDS. HIV or HIV-therapy? causal attributions of symptoms and their impact on treatment decisions among women and men with HIV.Eur J Med Res. 2009 Apr 16; 14(4): 139-146.

3. Ulmer A, Muller M, Bertisch-Mollenhoff B, Frietsch B. Low-dose prednisolone has a CD4-stabilizing effect in pre-treated HIV-patients during structured therapy interruptions (STI). Eur J Med Res. 2005 Jun 22; 10(6): 227-232.

4. Ulmer A, Muller M, Bertisch-Mollenhoff B, Frietsch B. Low dose prednisolone reduces CD4+ T cell loss in therapy-naive HIV-patients without antiretroviral therapy. Eur J Med Res. 2005 Mar 29; 10(3): 105-109.

agreed with the above answers. case dependent especially used for CNS IRIS conditions