Tracking the Extramedullary PML-RARα-Positive Cell Reservoirs in a Preclinical Model of APL: Biomarker of Long-Term Drug Efficacy

Relapses are now relatively rare in APL and occur mainly in the bone marrow (BM) within 3 years of complete response achievement, but later BM relapses can also occur (Kelaidi, Leukemia 2006), while a few cases of extramedullary (EM) relapse are observed, particularly in the central nervous system (CNS) (De Botton, Leukemia 2006). The transplantable transgenic mouse model of APL is a well characterized preclinical model which mimics human APL in its biological characteristics. We have previously reported the response of these mice to ATRA and a combination of ATRA with a DNA vaccine (Padua, Nat Med 2003, Fuguraki, Blood 2010). We analyzed in those treated mice the presence of PML-RARα transcripts in the peripheral blood (PB), BM and various organs and tissues.

The reproducible APL development was obtained, as previously described (Padua Nat Med 2003) by intravenous injection of 10⁴ spleen cells from APL transgenic mice expressing the human PML-RARα cDNA (bcr1) into syngeneic recipients. ATRA and the DNA vaccine (a plasmid containing PML-RARα cDNA (bcr1) sequence fused to the tetanus toxin fragment C) were administered as described (Padua Nat Med 2003). In this model, ATRA alone gives transient remissions, while about 30 % of APL mice treated with ATRA combined with a DNA vaccine achieve long-term remissions. A standardized RT-qPCR MRD monitoring protocol was applied to assess PML-RARα-positive cell clearance in PB and BM. In this assay, the number of PML-RARα transcripts was normalized (normalized copy number or NCN) to 10⁶ copies of 18S rRNA transcripts, allowing us to detect up to 1 PML-RARα-positive cell among 10⁴ negative cells. Taking advantage of the presence of PML-RARα cDNA transgene in the transplanted leukemia cells, we next used genomic DNA as template for a qPCR assay, allowing us to use 10 times more template and increased sensitivity (1 in 10⁵) in order to examine the presence of PML-RARα-positive cells in various organs and tissues of long-term survivors (LTS, ie with survival > 120 days).

APL mice treated with ATRA alone (n=55), ATRA combined with PML-RARαFrC DNA (n=94) or untreated (n=65) were analyzed. Untreated APL mice always remained positive in the PB and BM for PML-RARα transcripts, and in organs and tissues positive for PML-RARα cDNA. On day 60, in the surviving ATRA-treated mice (n=21), 15 (71%) had PB PML-RARα normalized copy number (NCN)>100, 6 (29 %) an NCN between 10–100 and none an NCN<10, while in ATRA+DNA-treated mice (n=35), 11 (31%) had NCN>100, 9 (26%) NCN between 10–100 and 15 (43%) NCN<10 (p<0.01). ATRA+DNA-treated mice achieving NCN<10 (43%) constituted the group with the best survival (p<0.0001). To further assess tumour burden, LTS were sacrificed at different time intervals. No PML-RARα transcripts were detected in PB and BM of any LTS (n=15) suggesting complete molecular remission. On the other hand, while PML-RARα cDNA, analyzed by qPCR in skin, salivary glands, thymus, kidney, muscle, heart, spleen, liver, lung and CNS was negative in all tissues in 10 (67%) LTS, it was positive in 5 (33%) LTS, including 4 in the CNS (2 of them were also positive in the spleen) and 1 in the spleen only.

In this preclinical model, analyzed with sensitive molecular assays, while two thirds treated long-term survivors were in molecular remission in PB, BM and other organs studied, about one third still had leukemic cells, mainly in the CNS and to a lesser extent in the spleen. This model could be of interest to better understand relapses in APL patients, especially late and CNS relapses and to evaluate drugs aimed at eliminating those reservoirs of residual cells.

No relevant conflicts of interest to declare.