Previous studies have shown improvement in the remission rate, prognosis, and survival of patients with AIDS-related lymphomas (ARLs) with highly active antiretroviral therapy (HAART).1-5 In the pre-HAART era, low-dose chemotherapy was the preferred regimen because of a higher risk of infection and hematologic toxicity.6 With the availability of HAART, administration of standard chemotherapy has become feasible as infectious complications have been reduced.1-5
The study on AIDS-related lymphoma by Mounier et al7 is certainly a step forward in understanding the impact of HAART in the management of ARL. The patients in the study were stratified into 3 risk groups based on their HIV score, and low-dose versus high-dose chemotherapy was administered in each group. The authors concluded that the intensity of the chemotherapy regimen did not affect the overall survival in each group. However, we think it would have been more interesting had the authors specified the number of patients that were enrolled into the low-dose chemotherapy and the high-dose chemotherapy arms based on their International Prognostic Index (IPI) score. This information would have potentially impacted the results of the study. For example, it is possible that there was a greater number of patients with a lower IPI score in the less-intensive chemotherapy group compared with those in the high-dose chemotherapy group, thereby giving the false perception of “no difference” in overall survival irrespective of the type of chemotherapy regimen used. As in the case of non-HIV patients with high-grade non-Hodgkin lymphoma, we believe that the IPI score rather than the type of chemotherapy regimen used may be affecting the survival rate of ARL patients. The same argument would hold good for the type/grade of lymphomas.
In addition, this study has shown improved response rates (partial response [PR] and complete response [CR]) with high-dose chemotherapy when compared with low-dose chemotherapy. However, this did not translate into improved overall survival rates. Therefore, it would have been interesting to know the cause of death in these patients over the period of study, whether they died from the progression of the disease, infections related to chemotherapy, or HIV-related complications.
AIDS-related lymphoma
The suggestion by Singh et al to include the number of patients randomized in the low-dose chemotherapy arm and the high-dose chemotherapy arm based on their International Prognostic Index (IPI) score is interesting, since it would affect the survival results. The multivariate model failed to detect interaction between IPI and treatment within strata, but investigated that problem. In the low-risk group, 45% of patients had an IPI score of 0-1 in the control arm versus 55% in the high-dose arm (P = .38); in the intermediate-risk group, 39% versus 43% (P = .12); and in the high-risk group, 30% versus 24% (P = .44).
The same argument would hold good for the pathologic subtype. Again, the multivariate model failed to detect interaction with pathologic subtype and treatment within strata. In addition, the small sample size of each subset prevented us from making any conclusion.
Concerning the question of the cause of death, the database is not precise enough but we are able to draw some conclusions. In all risk groups, the leading cause of death was lymphoma (n = 214; 66%). In the low-risk group, death was due to lymphoma in 77 patients (67%), infection in 28 (24%), and toxicity or several factors in 11 patients (9%). There was no difference between ACVBP (doxorubicin, cyclophosphamide, vindesine, bleomycin, and prednisolone) and CHOP (doxorubicin, cyclophosphamide, vincristine, and prednisolone) treatment groups. In the intermediate-risk group, death was due to lymphoma in 46 patients (75%) on low-dose CHOP (Ld-CHOP) and 35 (54%) on CHOP (P = .003) and to infection in 11 (18%) and 20 patients (31%) in these groups, respectively. In the high-risk group, death was due to lymphoma in 56 patients (71%), infection in 16 (20%), and toxicity or several factors in 7 (9%). There was no difference between Ld-CHOP or vincristine and steroids (VS). There seems to be a trade-off between efficacy and toxicity, with more deaths due to infection on CHOP than Ld-CHOP (31% vs 18%). However, this hypothesis needs to be confirmed by competing risk analysis of the cause of death. If death due to infection on CHOP was early in the course of the trial, while subjects were on therapy and their absolute neutrophil count (ANC) was depressed, then the subject could not die due to lymphoma.
The author declares no competing financial interests.
Correspondence: Nicolas Mounier, Department of Onco-Hematology, Centre Hospitalier Universitaire l'Archet, 151 route Saint Antoine-Ginestière,Nice, 06200 France; e-mail: mounier.n@chu-nice.fr.
The authors declare no competing financial interests.
