OBJECTIVE: We measure the transcript levels of the proapoptotic GALIG, antiapoptotic MCL1 genes and those of the autophagy genes BECN1, MAP1LC3B, ATG9a, P62/SQSTM1, GABARAP, GABARAPL1 and GABARAPL2 to define if mRNA alteration can characterize HIV-infected patients effectively treated with combined antiretroviral therapy (cART). DESIGN: Monocentric pilot study conducted on peripheral blood mononuclear cell (PBMC) of 40 uninfected donors and 27 HIV-positive patients effectively treated by cART for at least 8.4 years. METHODS: Transcripts of the various genes were quantified by reverse transcription (RT)-quantitative PCR (qPCR) and RT-droplet digital PCR and compared using the standard statistical Mann-Whitney U test and machine learning algorithms. RESULTS: A concomitant overexpression of GALIG and MCL1 is detected in PBMC of effectively cART-treated patients. Overexpression of MAP1LC3B and GABARAPL1 is also measured, whereas BECN1 is underexpressed. Finally, accurate classification (94.5%) of our PBMC samples as HIV-negative donors or HIV-positive cART-treated is obtained in three separate machine-learning algorithms with GABARAPL1 and ATG9a as input variables. CONCLUSION: cART-treated HIV patients display altered transcript levels for three genes of basal autophagy. Some of these alterations may appear contradictory: BECN1 and ATG9a, both key actors in the formation of mammalian autophagosome, exhibit decreased amount of transcripts, whereas mRNA from the ATG8 family increase. Given the known role of impaired basal autophagy in immune senescence and chronic inflammation, the functional significance of our findings should be explored in larger studies.