Improved Survival & Chemotherapy Response among Patients with AIDS-related NHL
Receiving Highly Active Antiretroviral Therapy
Catherine Diamond1, Thomas H. Taylor1, Mohammed Miradi2, Mark Wallace3, Theresa Im1, Hoda Anton-Culver1
1University of California Irvine Department of Medicine, Epidemiology Division, Irvine, CA,2Scripps Health, San Diego, CA, 3Naval Medical Center, San Diego, CA USA
Characteristics Associated with Complete Response to Characteristics of Patients with AIDS & CNS NHL & Univariate
Characteristics of AIDS Patients with Systemic NHL & Univariate & Chemotherapy among AIDS Patients with Systemic NHL, Univariate & Multivariate Associations with Survival by Cox Proportional Results Contact Information
Abstract Multivariate Associations with Survival by Cox Proportional •Among 233 patients with systemic NHL, the
& Multivariate Odds Ratios by Logistic Regression, San Diego & Hazards Model, San Diego & Orange County, CA, 1994-1999 Catherine Diamond
Background: Highly active antiretroviral therapy (HAART) became available Hazards Model, San Diego & Orange County, CA, 1994-1999 Orange County, CA, 1994-1999 median survival was 16 months for patients who
in the United States in 1996. We estimate the effect of HAART on survival received HAART at the time of NHL diagnosis or 101 The City Drive S
among patients with acquired immunodeficiency syndrome (AIDS) & non- thereafter vs. 3 months for patients who did not
Orange CA 92866
Hodgkin’s lymphoma (NHL). Methods: We used the population-based cancer receive HAART at NHL diagnosis or thereafter
registry to identify patients with AIDS-related lymphoma diagnosed in San •HAART, chemotherapy, high performance status (714) 456-7612
Diego or Orange County 1994-1999. After separating patients into systemic & NHL stage I-III were associated with improved
Fax (714) 456-7169
or primary central nervous system (CNS) lymphoma categories, we performed survival in systemic NHL
Kaplan-Meier analyses to compare survival between patients who received diamondc@uci.edu
HAART at NHL diagnosis or thereafter versus untreated patients & Cox
proportional hazard models for adjusted survival. We used logistic regression Results
to determine if concomitant HAART changed the probability of complete •Among 94 patients with CNS NHL, the median
response to chemotherapy for systemic NHL patients. We used the Mann survival was 8 months for patients who received
Whitney U test to compare the median number of chemotherapy cycles HAART at the time of NHL diagnosis or thereafter
between patients with systemic NHL who received HAART during vs. 1 month among patients who did not receive Study Strengths
chemotherapy versus patients who did not receive HAART with HAART at NHL diagnosis or thereafter •Our data are unique in being population-
chemotherapy. Results: Among 168 AIDS patients diagnosed with systemic •HAART, radiation therapy & high performance based, rather than from cohorts, clinical
or CNS NHL in 1996-1999, only 47 (28%) were taking HAART at the time of status were associated with improved survival in trials, or a single institution
NHL diagnosis. Among 233 patients with systemic NHL, the median survival CNS NHL •The number of systemic cases is large
was 16 months for patients who received HAART at the time of NHL
diagnosis or thereafter, versus 3 months for patients who did not receive
HAART at NHL diagnosis or thereafter. HAART, chemotherapy, high
performance status & NHL stage I-III were associated with improved survival Results
in systemic NHL. HAART, completion of at least 6 chemotherapy cycles & •Among 135 patients who received chemotherapy
Study Limitations
NHL stage I-III were associated with complete response to & had a charted chemotherapy response, 43 (32%)
•Few CNS cases, reflecting the dwindling
chemotherapy. The median number of chemotherapy cycles was 5 among had a complete response, 63 (47%) had a partial
incidence of this malignancy
HAART-treated patients vs. 3 among patients who did not receive response, & 29 (21%) had no response
•Incomplete chart information for LDH
HAART. Among 94 patients with CNS NHL, the median survival was 8 •Since 65% of patients with a complete response to
and CD4
months for patients who received HAART at the time of NHL diagnosis or chemotherapy were alive at most recent follow-up
•Do not know whether deceased patients
thereafter, versus 1 month among patients who did not receive HAART at (median follow up time: 51 months), we could not
died of NHL, chemotherapy-related
NHL diagnosis or thereafter. HAART, radiation therapy & high performance estimate median survival for these patients
toxicity, or opportunistic infection
status were associated with improved survival in CNS NHL. Conclusion: •The median survival was 5 months for those who
•No data regarding relapse rates or
HAART should be initiated or continued after NHL diagnosis. had a partial or no response to chemotherapy
disease-free survival
•HAART, completion of at least 6 chemotherapy
•Potential underreporting to cancer
cycles & NHL stage I-III were associated with
registry
Background & Objective complete response to chemotherapy
•By analyzing whether patients received
•The median number of chemotherapy cycles was 5
•HAART became available in the United States in 1996 HAART after their NHL diagnosis, we
among HAART-treated patients vs. 3 among
bias our results toward an improved
•We sought to estimate the effect of HAART on survival and chemotherapy patients who did not receive HAART
survival with HAART, because patients
response among patients with AIDS & NHL would have to survive & present for
treatment after their NHL diagnosis
Discussion Points •Physicians may have been more likely to
•The median survival among HAART-treated start HAART, chemotherapy, or radiation
Methods
patients in our population-based study probably is in patients who they perceived as having a
•We used the population-based cancer registry to identify patients with AIDS-
shorter than what is currently achievable chance of survival
related lymphoma diagnosed in San Diego or Orange County 1994-1999
•Unknown whether the increased likelihood of •Physicians might be more likely to stop
•We used routinely collected cancer registry data & performed a special chart
chemotherapy response among patients who HAART in patients with poor tolerance,
review
received HAART concurrent with chemotherapy is response, or adherence with HAART
•We separated patients into systemic or primary central nervous system
due to the influence of improved virologic control
(CNS) lymphoma
& immune restoration
•We performed Kaplan-Meier analyses to compare survival between patients
•HAART-treated patients had better performance
who received HAART at NHL diagnosis or thereafter vs. untreated patients
status & fewer B symptoms & thus were better able
•We performed Cox proportional hazard models for adjusted survival
to tolerate chemotherapy
•We used logistic regression to determine if concomitant HAART changed the
•We were unable to detect a significant difference
probability of complete response to chemotherapy for systemic NHL patients Conclusion
in the frequency of chemotherapy-related toxicity
•We used the Mann Whitney U test to compare the median number of HAART should be initiated or continued
between patients who received HAART concurrent
chemotherapy cycles between patients with systemic NHL who received after NHL diagnosis, including during
with chemotherapy & those who did not, but limited
HAART during chemotherapy vs. patients who did not receive HAART with chemotherapy
power to detect a difference
chemotherapy
Survival among AIDS patients with systemic NHL who received Survival among AIDS patients with systemic NHL who had a complete Survival among AIDS patients with CNS NHL who received
HAART at NHL diagnosis or thereafter vs. untreated patients, San response to chemotherapy vs. patients with partial or no response, HAART at NHL diagnosis or thereafter vs. untreated patients,
Diego & Orange County, CA, 1994-1999 San Diego & Orange County, CA, 1994-1999 San Diego & Orange County, CA, 1994-1999