PML causing immune reconstitution problems when ARVs begin

Many of the cases of progressive multifocal leukoencephalopathy (PML) now being reported are actually related to immune reconstitution inflammatory syndrome (IRIS), according to several presentations made last month at the Second HIV Infection and Central Nervous System: Developed and Resource-Limited Settings meeting in Venice.

IRIS is a delayed but potent inflammatory immune response that can develop after beginning antiretroviral therapy (ART) and that can “unmask” previously subclinical infections. (For background to PML, see related article).

According to intensive pathological investigations presented by Dr Françoise Gray of the University of Paris, and Dr Avindra Nath of Johns Hopkins University, Baltimore, Maryland, USA, the mechanisms causing PML-IRIS may be somewhat different that what is seen with classical PML — where the damage is believed to be primarily caused by JCV itself. In the case of PML-IRIS, most of the damage appears to be caused by the immune system.

For instance, according to Dr Gray, brain biopsies of patients with progressive PML-IRIS demonstrated massive inflammation with infiltration by macrophages and CD8 lymphocytes.

In some cases, this immune response had completely wiped out any evidence of active JCV infection but the immune cells continued reacting, perhaps to latent antigens or inactive infectious particles.

Dr Nath presented evidence from an examination of one patient with IRIS that suggested that a subset of memory CD8 cells was not just causing demylelination but also bystander damage to neurons as well— possibly by releasing a neurotoxic chemical called granzyme B that triggers neuronal apoptosis.

The differences between classical and IRIS-PML are often detectable by MRI. “With PML in IRIS, the MRI may show contrast enhancement usually at the periphery of lesions,” said Dr Simonetta Gerevini also of the San Raffaele Scientific Institute in Milan, Italy. “It’s not well known, but it could be a marker of improved immune status usually associated with good outcomes, but in some cases it is associated with clinical and radiological deterioration.”

She also added that sometimes the enhancement can be diffuse or within the lesion and that patients with PML-IRIS may even develop vasculitis-type lesions more typical of HIV encephalopathy. However, for each type of lesion, when steroids are added, the patient generally improves, and when he or she does, the contrast enhancement, and eventually the entire lesion, disappears on MRI — leaving behind only signs of atrophy where the lesions once were.

From her pathology studies, Dr Gray believes the problem could be especially pronounced when the CD8 immune response occurs before the CD4 cell responses have fully recovered in patients.

“In our experience, the only factor constantly associated with an unfavourable outcome is a dysregulation of the CD4/CD8 balance in brain tissue, with an absence of CD4+ lymphocytes in brain tissue,” she said. “This inflammatory syndrome is usually reversible spontaneously or following treatment [HAART and steroids]. However, in some cases it may have fatal outcome.”

Indeed, according to a study by Dr Annamaria Pazzi, of the San Raffaele Scientific Institute in Milan, Italy, and other Italian colleagues, IRIS-PML seems no less severe than classical PML.

"The similarities between variables at baseline, survival and predictors of prognosis in [PML-IRIS] and [classical] PML suggest that although the mechanisms related to PML onset may be different, the presentation and the course of the disease are apparently the same," she said.

The study involved 17 patients who were classified as PML-IRIS (defined as PML that occurred after a response to ART — which, in this study meant that patients had to have experienced at least 0.5 log or greater drop in HIV load) and 48 with classical PML. Thirteen of 17 with PML-IRIS and 42 of 48 with untreated PML had a confirmed diagnosis.

All patients received ART and two-year survival was essentially the same. A total of 50% of people diagnosed with PML-IRIS survived for two years, compared to 47% of those who had developed PML before going onto anti-HIV therapy.

In each group, better immune responses on HAART were associated with better outcomes.

The study had limited power to reach some conclusions because of the small sample sizes of the groups (the PML-IRIS group especially). For example, Kaplan-Meier curves seemed to suggest an advantage in survival for patients with PML-IRIS who had higher CD4 cell counts and CD4% at baseline, but this did not reach statistical significance.

However, a higher CD4% at PML onset was associated with better survival in patients with classical PML (p=0.04). In addition, a better CD4 response at three months after diagnosis was a significant predictor of a better outcome (in the classical IRIS group this included improved CD4%).

The role of CD8 reconstitution was more ambiguous. The trend on the Kaplan-Meier curve seemed to suggest that patients with PML-IRIS and higher CD8 counts at baseline had a lower rate of survival but this was not statistically significant. However, better CD8 cell responses (actually a less marked loss (-70 CD8 cells) was significantly associated with better outcomes in these same patients (p=0.017). In patients with classical PML, neither CD8 cell count at baseline (which was somewhat higher on average) nor changes in CD8 cell count seemed to affect outcomes.

Finally JCV DNA levels at baseline were not predictive of better outcome in the patients with PML. But after diagnosis, an increase in JCV DNA levels of more than 0.07 log at three months was strongly associated with poorer survival (p=0.01, log rank test) in the patients with classical PML. Unfortunately, they could not calculate the effect of JCV DNA changes of survival of the PML-IRIS patients because they had too few measurements in that group of patients.

See also related article on the development of immune-based therapies for PML