Clinical characteristics and outcomes of newly diagnosed patients with HIV‐associated aggressive B‐cell NHL in China

Abstract Little is known about the incidence, clinical characteristics and prognostic factors in HIV associated lymphoma as these are less common than HIV‐negative lymphoma in China. Currently, there are no standard guidelines for treatment of these patients. Therefore, we performed a study to analyse the clinical characteristics and outcomes of newly diagnosed HIV‐associated aggressive B‐cell non‐Hodgkin's lymphoma (NHL) patients in Chongqing University Cancer Hospital (CUCH). Totally 86 newly diagnosed HIV‐associated aggressive B‐cell NHL patients in CUCH, southwest China, from July 2008 to August 2021, were analysed. In the entire cohort, median age was 48 years (range, 23–87 years), and more patients were male (87.2%). Most patients had elevated lactate dehydrogenase (LDH) (82.6%), advanced ann arbor stage (80.2%) and high IPI score (IPI score, 3–5) (62.7%) at diagnosis. Median CD4+ T‐cell count at diagnosis was 191/μl (range, 4–1022), 84 patients (97.7%) were on combination antiretroviral therapy (cART) at lymphoma diagnosis. In DLBCL patients, cox multivariate analysis showed that age ≥ 60 (HR = 2.251, 95%CI 1.122–4.516; p = 0.012), elevated LDH (HR = 4.452, 95%CI 1.027–19.297; p = 0.041) and received less than two cycles of chemotherapy (HR = 0.629, 95%CI 0.589–1.071; p = 0.012) were independent risk factors for adverse prognosis based on PFS. Age ≥ 60 (HR = 3.162, 95%CI 1.500–6.665; p = 0.002) and received less than two cycles of chemotherapy (HR = 0.524, 95%CI 0.347–0.791; p = 0.002) were also independent risk factor for adverse prognosis based on OS. In BL patients, cox multivariate analysis showed that elevated LDH and received less than two cycles of chemotherapy were independent risk factors for adverse prognosis. In the DLBCL group, median PFS times in the received rituximab and no received rituximab groups were not reached and 12 months, respectively (p = 0.006). Median OS times were not reached and 36 months, respectively (p = 0.021). In the BL group, median PFS times in the received rituximab and no received rituximab groups were not reached and 4.8 months, respectively (p = 0.046). Median OS times were not reached and 10.1 months, respectively (p = 0.035). Overall, these data indicated that standardized anti‐lymphoma therapy and rituximab administration were significantly associated with improved outcomes in patients with HIV‐associated DLBCL and BL.


| BACKG ROU N D
According to the 1993 revised classification system for human im- infection. 1,2 Since 1996, with the wide use of combination antiretroviral therapy (cART), the incidence of malignant cancers in individuals living with HIV has decreased significantly. Since 2017, the incidence of HIV-associated lymphoma has been higher than that of KS, becoming the highest incidence rate among HIV-associated cancers in the United States. 2 Diffuse large B-cell lymphoma (DLBCL) and Burkitt's lymphoma (BL) are the most common subtypes of HIVassociated lymphoma. 3 Treatment of HIV-associated DLBCL and HIV-associated BL prior to the advent of cART was frequently complicated by opportunistic infections. Complete remission (CR) rates in the range of 16%-56% and median survival times of 5-8 months were reported regardless of chemotherapy regimen or dose intensity. 4,5 Since the introduction of cART, studies have suggested better tolerance of chemotherapy and significantly better survival. 6,7 A phase 2 study assessed CHOP plus rituximab (R-CHOP) in HIV-associated B-cell NHL, revealing a CR rate of 77% and a 2 year-survival rate of 75%, which suggests that rituximab is beneficial and safe. 8 In another phase 2 clinical trial of concurrent versus sequential rituximab with the EPOCH regimen in HIV-associated B-cell NHL, CR occurred in 21 of 31 evaluable patients (68%, 95% confidence interval [CI] 49%-83%) in the concurrent arm versus 18 of 37 patients (49%, 95%CI 32%-66%) in the sequential arm. The two-year overall survival rates were 70% (95%CI 57%-83%) and 67% (95%CI 54%-80%) in the concurrent and sequential arms, respectively. 9 With 5 years of follow-up, the PFS and OS rates of short-course EPOCH and dose-dense rituximab (SC-EPOCH-RR) regimen were 84% and 68%, respectively, in HIVassociated B-cell NHL. 10 To date, only a few studies have assessed the clinical features of Chinese HIV-associated lymphoma patients. 11,12 Meanwhile, CUCH is the only cancer centre in Western China that treats HIV-associated lymphoma patients. Herein, we analysed the clinical characteristics and prognosis of 86 newly diagnosed HIV-associated aggressive Bcell NHL patients, as the largest cohort reported in China to date.

| Patients
All newly diagnosed patients with HIV-associated lymphoma be-  independent risk factors for adverse prognosis based on PFS. Age ≥ 60 (HR = 3.162, 95%CI 1.500-6.665; p = 0.002) and received less than two cycles of chemotherapy (HR = 0.524, 95%CI 0.347-0.791; p = 0.002) were also independent risk factor for adverse prognosis based on OS. In BL patients, cox multivariate analysis showed that elevated LDH and received less than two cycles of chemotherapy were independent risk factors for adverse prognosis. In the DLBCL group, median PFS times in the received rituximab and no received rituximab groups were not reached and 12 months, respectively (p = 0.006). Median OS times were not reached and 36 months, respectively (p = 0.021). In the BL group, median PFS times in the received rituximab and no received rituximab groups were not reached and 4.8 months, respectively (p = 0.046).
Median OS times were not reached and 10.1 months, respectively (p = 0.035). Overall, these data indicated that standardized anti-lymphoma therapy and rituximab administration were significantly associated with improved outcomes in patients with HIVassociated DLBCL and BL.

K E Y W O R D S
BL, Clinical characteristics, DLBCL, HIV, Prognosis was approved by the institutional review board of CUCH and conducted according to the Declaration of Helsinki (registered at www. chictr.org.cn as #ChiCTR2100054581).

| Anti-HIV treatment
In this study, 84 patients (97.7%) were administered cART. Two of them gave up cART for fear of discrimination. cART included two nucleoside reverse transcriptase inhibitors and one nonnucleoside reverse transcriptase inhibitor.

| Statistical analysis
PFS was defined as the time from lymphoma diagnosis to disease progression, relapse or death from any cause. OS was defined as F I G U R E 1 Flow diagram of the study design. Datasets were queried for patients with HIV-associated lymphoma from July 2008 to August 2021 in Chongqing University Cancer Hospital the time from lymphoma diagnosis to last follow-up or death from any cause. All statistical data were analysed with SPSS version 26 or GraphPad Prism 9. Survival was estimated using Kaplan-Meier curves and compared by the log-rank test. The Cox proportional hazards regression model was used in multivariate analysis to determine prognostic factors. p < 0.05 was considered statistically significant.

| Patient characteristics
Eighty-six newly diagnosed HIV-associated aggressive B-cell NHL patients were included in this analysis. The baseline clinical features of these patients are summarized in Table 1   Median OS times in the DLBCL and BL groups were not reached and 11.5 months, respectively. The overall 2 years OS rates were 58.5% and 27.2% respectively (p = 0.047) ( Figure 2B). These data showed that the survival of HIV-associated DLBCL was significantly better than that of patients with HIV-associated BL.

| Treatment outcomes
In the DLBCL group, median PFS times in the age < 60 and age ≥ 60 groups were not reached and 2 months, respectively (p = 0.003) ( Figure 3A). Median OS times in the age < 60 and age ≥ 60 groups were not reached and 4 months, respectively (p < 0.001) ( Figure 3B). Median PFS times in the no received anti-lymphoma chemotherapy, only received one or two cycles and more than two cycles groups were 2.5 months, 20 months, and 37 months, respectively (p = 0.009) ( Figure 3C). Median OS times in the no received anti-lymphoma chemotherapy, only received one or two cycles and more than two cycles groups were 3.5 months, 36 months, and not reached, respectively (p < 0.001) ( Figure 3D). These results demonstrated that age < 60 and standardized anti-lymphoma therapy were significantly associated with improved outcomes in HIV-associated DLBCL patients. In the BL group, standardized anti-lymphoma therapy was significantly associated with improved outcomes.
In this study, there was no patient with primary HIV-associated
In BL patients, Cox multivariate analysis showed that elevated LDH and received less than two cycles of chemotherapy were independent risk factors for adverse prognosis ( Table 4).

| Rituximab administration and patient outcome
In the DLBCL group, median PFS times in the received rituximab and no received rituximab groups were not reached and 12 months.
The overall 2 years PFS rates were 63.2% and 30.3%, respectively

| DISCUSS ION
In the 40 years since AIDS was first reported, the introduction of cART has greatly improved the prognosis of individuals living with HIV, resulting in significantly increased life expectancy, which is currently close to that of the general population. 13 After 2017, the incidence of HIV-associated NHL surpassed KS at 100/100,000 to 300/100,000, making it the most common Because early initiation of cART reduces the risk of developing HIV-associated NHL, the current treatment recommendation is that antiretroviral therapy should be started as early as possible in patients diagnosed with AIDS. 26,27 In general, the drugs to be avoided in anti-lymphoma therapy include those that highly inhibit the bone marrow (e.g. zidovudine) as well as strong cytochrome P450 inhibitors and inducers, which have strong interactions with chemotherapy. 28 Therefore, we recommend strengthening communication with AIDS specialists and developing personalized cART regimens for the treatment of HIV-associated NHL.
Rituximab significantly improves the prognosis of HIV-negative B-cell NHL patients. 29 Barta et al. showed that the combination of rituximab and CHOP (R-CHOP) does not increase mortality due to infectious complications. 30 Noy and Roschewski also confirmed that rituximab is safe and effective for HIV-associated BL patients. 31 38,39 Similarly, Noy and colleagues prospectively treated 34 HIV-associated BL patients with CODOX-M-IVAC-R in AMC048.
Although most of the patients had advanced disease and high risk, 1 year PFS and OS were 69% and 72%, respectively. These results were similar to those of HIV-negative BL patients administered the same regimen. 31 More prospective clinical studies are needed to establish a prognostic risk assessment system for HIV-associated lymphoma.
HIV-associated lymphoma is a highly heterogeneous disease In summary, our study suggested that elevated LDH, received less than two cycles of chemotherapy and no received rituximab predicted poor OS and PFS in HIV-associated DLBCL and BL patients.
More high-quality randomized controlled studies were needed to test our findings.