CNS involvement predicts relapse but not survival in ARL, study shows

CHICAGO—Investigators have found evidence to suggest that identifying central nervous system (CNS) involvement at diagnosis does not impact overall survival for patients with AIDS-related lymphoma (ARL).

The research showed that ARL patients with CNS involvement at diagnosis were nearly 3 times as likely as their peers to have CNS relapse during cancer treatment.

But there was no significant difference between the 2 groups with regard to survival.

This may be due to the low overall incidence of CNS relapse, the use of insufficient treatments, and/or inadequate methods for identifying patients with CNS involvement, according to the investigators.

Stefan K. Barta, MD, of the Fox Chase Cancer Center in Philadelphia, Pennsylvania, and his colleagues conducted this research and presented the results at the 2014 ASCO Annual Meeting (abstract 8570).

In 2013, Dr Barta led the assembly of a database containing medical data from more than 1500 patients newly diagnosed with ARL who participated in clinical trials in Europe and the US from 1990 through 2010.

In the new study, he and his colleagues used the database to identify 880 patients with ARL whose data included complete information on CNS involvement at diagnosis and CNS relapse.

The team set out to find associations between CNS relapse and patient characteristics, including age, sex, CD4 count, lymphoma subtype, treatment history with combination antiretroviral therapies (cART), rituximab use, and the type of initial chemotherapy.

All of the patients had received either intrathecal therapy for CNS involvement or intrathecal prophylaxis with single-agent or 3-agent regimens. Sixty-nine percent of patients (n=607) had received cART, and 31% (n=276) had received rituximab-containing induction chemoimmunotherapy.

CNS involvement was identified in 13% of patients at diagnosis, including 27% of patients with Burkitt lymphoma or Burkitt-like lymphoma and 6% of patients with diffuse large B-cell lymphoma.

There was no difference in the prevalence of baseline CNS involvement between patients treated before and after the introduction of cART (13% each).

In all, 5.3% of patients experienced CNS relapse at a median of 4.2 months after diagnosis (range, 0.3-19.3 months). This included 12% of patients diagnosed with CNS involvement at baseline and 4% of patients who were not.

The median overall survival after CNS relapse was 1.6 months (range, 0-86.4 months). There was no significant difference in overall survival between patients with CNS involvement at diagnosis and those without it. The hazard ratio was 0.85 (P=0.32).

Multivariate analysis showed the only baseline characteristic significantly associated with the frequency of CNS relapse was CNS involvement, with a hazard ratio of 2.9 (P=0.01). None of the treatments had a significant impact on CNS relapse.

Dr Barta said these results suggest that current treatments are insufficient, and the approaches used to identify CNS involvement may be missing many patients who are at risk of CNS relapse.

“A lot of patients who relapsed probably had undetected CNS involvement at diagnosis,” he said. “We want to figure out if there are better strategies to identify patients at risk.”

CHICAGO—Investigators have found evidence to suggest that identifying central nervous system (CNS) involvement at diagnosis does not impact overall survival for patients with AIDS-related lymphoma (ARL).

The research showed that ARL patients with CNS involvement at diagnosis were nearly 3 times as likely as their peers to have CNS relapse during cancer treatment.

But there was no significant difference between the 2 groups with regard to survival.

This may be due to the low overall incidence of CNS relapse, the use of insufficient treatments, and/or inadequate methods for identifying patients with CNS involvement, according to the investigators.

Stefan K. Barta, MD, of the Fox Chase Cancer Center in Philadelphia, Pennsylvania, and his colleagues conducted this research and presented the results at the 2014 ASCO Annual Meeting (abstract 8570).

In 2013, Dr Barta led the assembly of a database containing medical data from more than 1500 patients newly diagnosed with ARL who participated in clinical trials in Europe and the US from 1990 through 2010.

In the new study, he and his colleagues used the database to identify 880 patients with ARL whose data included complete information on CNS involvement at diagnosis and CNS relapse.

The team set out to find associations between CNS relapse and patient characteristics, including age, sex, CD4 count, lymphoma subtype, treatment history with combination antiretroviral therapies (cART), rituximab use, and the type of initial chemotherapy.

All of the patients had received either intrathecal therapy for CNS involvement or intrathecal prophylaxis with single-agent or 3-agent regimens. Sixty-nine percent of patients (n=607) had received cART, and 31% (n=276) had received rituximab-containing induction chemoimmunotherapy.

CNS involvement was identified in 13% of patients at diagnosis, including 27% of patients with Burkitt lymphoma or Burkitt-like lymphoma and 6% of patients with diffuse large B-cell lymphoma.

There was no difference in the prevalence of baseline CNS involvement between patients treated before and after the introduction of cART (13% each).

In all, 5.3% of patients experienced CNS relapse at a median of 4.2 months after diagnosis (range, 0.3-19.3 months). This included 12% of patients diagnosed with CNS involvement at baseline and 4% of patients who were not.

The median overall survival after CNS relapse was 1.6 months (range, 0-86.4 months). There was no significant difference in overall survival between patients with CNS involvement at diagnosis and those without it. The hazard ratio was 0.85 (P=0.32).

Multivariate analysis showed the only baseline characteristic significantly associated with the frequency of CNS relapse was CNS involvement, with a hazard ratio of 2.9 (P=0.01). None of the treatments had a significant impact on CNS relapse.

Dr Barta said these results suggest that current treatments are insufficient, and the approaches used to identify CNS involvement may be missing many patients who are at risk of CNS relapse.

“A lot of patients who relapsed probably had undetected CNS involvement at diagnosis,” he said. “We want to figure out if there are better strategies to identify patients at risk.”

CHICAGO—Investigators have found evidence to suggest that identifying central nervous system (CNS) involvement at diagnosis does not impact overall survival for patients with AIDS-related lymphoma (ARL).

The research showed that ARL patients with CNS involvement at diagnosis were nearly 3 times as likely as their peers to have CNS relapse during cancer treatment.

But there was no significant difference between the 2 groups with regard to survival.

This may be due to the low overall incidence of CNS relapse, the use of insufficient treatments, and/or inadequate methods for identifying patients with CNS involvement, according to the investigators.

Stefan K. Barta, MD, of the Fox Chase Cancer Center in Philadelphia, Pennsylvania, and his colleagues conducted this research and presented the results at the 2014 ASCO Annual Meeting (abstract 8570).

In 2013, Dr Barta led the assembly of a database containing medical data from more than 1500 patients newly diagnosed with ARL who participated in clinical trials in Europe and the US from 1990 through 2010.

In the new study, he and his colleagues used the database to identify 880 patients with ARL whose data included complete information on CNS involvement at diagnosis and CNS relapse.

The team set out to find associations between CNS relapse and patient characteristics, including age, sex, CD4 count, lymphoma subtype, treatment history with combination antiretroviral therapies (cART), rituximab use, and the type of initial chemotherapy.

All of the patients had received either intrathecal therapy for CNS involvement or intrathecal prophylaxis with single-agent or 3-agent regimens. Sixty-nine percent of patients (n=607) had received cART, and 31% (n=276) had received rituximab-containing induction chemoimmunotherapy.

CNS involvement was identified in 13% of patients at diagnosis, including 27% of patients with Burkitt lymphoma or Burkitt-like lymphoma and 6% of patients with diffuse large B-cell lymphoma.

There was no difference in the prevalence of baseline CNS involvement between patients treated before and after the introduction of cART (13% each).

In all, 5.3% of patients experienced CNS relapse at a median of 4.2 months after diagnosis (range, 0.3-19.3 months). This included 12% of patients diagnosed with CNS involvement at baseline and 4% of patients who were not.

The median overall survival after CNS relapse was 1.6 months (range, 0-86.4 months). There was no significant difference in overall survival between patients with CNS involvement at diagnosis and those without it. The hazard ratio was 0.85 (P=0.32).

Multivariate analysis showed the only baseline characteristic significantly associated with the frequency of CNS relapse was CNS involvement, with a hazard ratio of 2.9 (P=0.01). None of the treatments had a significant impact on CNS relapse.

Dr Barta said these results suggest that current treatments are insufficient, and the approaches used to identify CNS involvement may be missing many patients who are at risk of CNS relapse.

“A lot of patients who relapsed probably had undetected CNS involvement at diagnosis,” he said. “We want to figure out if there are better strategies to identify patients at risk.”