Introduction

Cohort studies have demonstrated increased incidence of Hodgkin lymphoma (HL) and non-Hodgkin lymphoma (NHL) among HIV-infected individuals during the first 6 months after antiretroviral therapy (ART) initiation, perhaps due to unmasking immune reconstitution inflammatory syndrome (IRIS). Unmasking IRIS is characterized by the diagnosis of a new HIV-associated condition soon after ART that was not apparent prior to ART, coupled with evidence of ART effectiveness. It has been well described for opportunistic infections and Kaposi sarcoma. However, clinical characteristics and survival for unmasking lymphoma IRIS have not been described.

Methods

We studied lymphoma patients in the Centers for AIDS Research Network of Integrated Clinical Systems (CNICS) from 1996 until 2011. Unmasking lymphoma IRIS was defined as HL or NHL occurring within 6 months after ART initiation accompanied by a ≥0.5 log10copies/mL reduction in HIV RNA between values taken prior to ART and at lymphoma diagnosis. Differences in presentation and survival were examined between lymphoma IRIS and non-IRIS cases.

Results

Of 482 lymphoma patients, 48 (10%) met criteria for unmasking lymphoma IRIS. Of these, 10 (21%) had HL, 19 (40%) diffuse large B-cell lymphoma (DLBCL), 4 (8%) Burkitt lymphoma (BL), 9 (19%) primary central nervous system lymphoma (PCNSL), and 6 (12%) other NHL (Table). Median CD4 cell count at lymphoma diagnosis among IRIS cases was 163 cells/mL (interquartile range 67-302), and 54% had suppressed HIV RNA (< 400 copies/ml). No significant differences were identified between lymphoma IRIS and non-IRIS cases, with the exception of possible earlier stage (47% stage I/II versus 24%, p=0.03), more frequent hepatitis B/C co-infection (31% versus 19%, p=0.05), and more frequent prior AIDS illness (92% versus 79%, p=0.05), as well as expected lower HIV RNA at lymphoma diagnosis likely resulting from the IRIS case definition. Additionally, no differences in cumulative mortality 5 years after lymphoma diagnosis were identified between IRIS and non-IRIS cases, although there was a suggestion of increased early mortality among IRIS cases (Figure).

Table

Characteristics of 482 HIV-infected adults in CNICS with lymphoma between 1996 and 2011, stratified by unmasking IRIS case status

IRISnon-IRISP value
Total, n (%) 48 (10.0) 434 (90.0) ― 
Age (years) at lymphoma diagnosis, mean (SD) 41.5 (6.2) 42.4 (9.0) 0.38 
Male, n (%) 45 (93.8) 381 (87.8) 0.34 
Race/ethnicity*    
White, n (%) 23 (47.9) 216 (50.4) 0.75 
Black, n (%) 20 (41.7) 135 (31.5) 0.15 
Other, n (%) 5 (10.4) 78 (18.2) 0.23 
Lymphoma diagnosis year, median (IQR) 2003 (2000-2007) 2003 (2000-2006) 0.64 
Lymphoma category    
HL, n (%) 10 (20.8) 70 (16.1) 0.41 
DLBCL, n (%) 19 (39.6) 186 (42.9) 0.66 
BL, n (%) 4 (8.3) 53 (12.2) 0.64 
PCNSL, n (%) 9 (18.8) 45 (10.4) 0.08 
Other NHL, n (%) 6 (12.5) 80 (18.4) 0.31 
Lymphoma stage I/II, n (%)* 9 (47.4) 38 (23.6) 0.026 
Hepatitis B/C co-infection 15 (31.3) 83 (19.1) 0.047 
AIDS illness prior to lymphoma diagnosis, n (%) 44 (91.7) 344 (79.3) 0.053 
CD4 count (cells/µL) at lymphoma diagnosis, median (IQR)* 163 (67-302) 123 (37-288) 0.43 
CD4 percentage at lymphoma diagnosis, median (IQR)* 13.5 (6.7.4-21.2) 12.0 (5.5-22.0) 0.49 
CD4 count (cells/µL) nadir, median (IQR)* 69 (24-197) 73 (20-198) 0.97 
HIV RNA (log10copies/mL) at lymphoma diagnosis, median (IQR)* 2.33 (1.40-3.75) 4.50 (2.30-5.30) <0.0001 
HIV RNA<400 copies/mL at lymphoma diagnosis, n (%)* 26 (54.2) 94 (29.8) 0.0008 
IRISnon-IRISP value
Total, n (%) 48 (10.0) 434 (90.0) ― 
Age (years) at lymphoma diagnosis, mean (SD) 41.5 (6.2) 42.4 (9.0) 0.38 
Male, n (%) 45 (93.8) 381 (87.8) 0.34 
Race/ethnicity*    
White, n (%) 23 (47.9) 216 (50.4) 0.75 
Black, n (%) 20 (41.7) 135 (31.5) 0.15 
Other, n (%) 5 (10.4) 78 (18.2) 0.23 
Lymphoma diagnosis year, median (IQR) 2003 (2000-2007) 2003 (2000-2006) 0.64 
Lymphoma category    
HL, n (%) 10 (20.8) 70 (16.1) 0.41 
DLBCL, n (%) 19 (39.6) 186 (42.9) 0.66 
BL, n (%) 4 (8.3) 53 (12.2) 0.64 
PCNSL, n (%) 9 (18.8) 45 (10.4) 0.08 
Other NHL, n (%) 6 (12.5) 80 (18.4) 0.31 
Lymphoma stage I/II, n (%)* 9 (47.4) 38 (23.6) 0.026 
Hepatitis B/C co-infection 15 (31.3) 83 (19.1) 0.047 
AIDS illness prior to lymphoma diagnosis, n (%) 44 (91.7) 344 (79.3) 0.053 
CD4 count (cells/µL) at lymphoma diagnosis, median (IQR)* 163 (67-302) 123 (37-288) 0.43 
CD4 percentage at lymphoma diagnosis, median (IQR)* 13.5 (6.7.4-21.2) 12.0 (5.5-22.0) 0.49 
CD4 count (cells/µL) nadir, median (IQR)* 69 (24-197) 73 (20-198) 0.97 
HIV RNA (log10copies/mL) at lymphoma diagnosis, median (IQR)* 2.33 (1.40-3.75) 4.50 (2.30-5.30) <0.0001 
HIV RNA<400 copies/mL at lymphoma diagnosis, n (%)* 26 (54.2) 94 (29.8) 0.0008 

IRIS=immune reconstitution inflammatory syndrome. HL=Hodgkin lymphoma. DLBCL=diffuse large B-cell lymphoma. BL=Burkitt lymphoma. PCNSL=primary central nervous system lymphoma.

*

N(%) missing observations IRIS/non-IRIS: Race/ethnicity=0(0%)/5(1.2%); Lymphoma stage=29(60.4%)/273(62.9%); CD4 count at lymphoma diagnosis=0(0%)/97(22.4%); CD4 percentage at lymphoma diagnosis=0(0.0%)/123(28.3%); CD4 count nadir=0(0.0%)/48(11.1%); HIV RNA at lymphoma diagnosis=0(0.0%)/119(27.4%).

Conclusions

In a large HIV-associated lymphoma cohort in the United States, 10% of patients met a standardized unmasking lymphoma IRIS case definition. No major differences between lymphoma IRIS and non-IRIS cases were identified. Possible effects of subclinical lymphoma leading to care seeking behavior and subsequent ART initiation prior to lymphoma diagnosis could not be excluded.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution