We describe the case of a R263K mutation detected in the cerebrospinal fluid (CSF) of an HIV-1-infected patient with HIV-associated neurocognitive disorder and receiving the integrase inhibitor dolutegravir.
A 45-year-old female patient, HIV-1 infected since 1989 with a subtype B virus, with a history of illicit drug use, presented in September 2013 with sudden onset of anxiety and ataxia. She had no history of any AIDS-defining event and the CD4+ nadir cell count was 198 cm3/μl (18%). She had previously received treatment for chronic hepatitis C in 2011, achieving sustained virological response.
She had been treated with combination antiretroviral therapy (ART) since 1996, initially with two nucleoside analog reverse transcriptase inhibitors (NRTIs), then with two NRTIs and one non-NRTI (NNRTI), then with two NRTIs and one protease inhibitor. At the time of admission, in September 2013, she had been taking tenofovir, emtricitabine, and boosted 800-mg darunavir once daily for 2 years. Treatment adherence had not always been satisfactory, as proved by multiple occurrences of detectable plasma viral loads, including a measurement a few months before presentation that showing 240 copies/ml, with genotyping revealing multiple mutations on NRTIs (M41L, T215Y, L74I/V, and M184V), NNRTIs (Y181C), and protease inhibitors (M46I, I62V, A71V, V77I, L76V, and L89I).
At the time of admission, the patient presented with cerebellar syndrome, whereas MRI revealed diffuse demyelination in frontal and parietal areas. Plasmatic viral load was below 40 copies/ml and CD4+ cell count 500 cm3/μl. The CSF was clear with low cellular content (7 cells/μl), moderate elevation of protein concentration (0.66 g/l), and normal glucose CSF/plasma ratio. Gram stain was negative, and bacterial culture pursued for 5 days remained sterile. HIV RNA concentration in the CSF was 2600 copies/ml; however, viral genotyping was not performed. Testing for JC virus was negative. Tropism testing was in favor of a C-C chemokine receptor type 5 (CCR5)-expressing virus.
HIV-associated encephalopathy was suspected and antiretroviral treatment was changed to raltegravir 400 mg twice daily (b.i.d.), maraviroc 150 mg b.i.d., in addition to boosted darunavir 800 mg once daily.
The patient initially improved, but 6 months later, she relapsed. Plasma viral load was confirmed undetectable, whereas the CD4+ cell count revealed 726 cm3/μl. CSF viral load measured 170 copies/ml, with genotyping showing the same mutations found in the plasma HIV genotyping described above.
No mutations were found for the integrase site. ART was modified as follows: boosted darunavir 600 mg, dolutegravir 50 mg, and maraviroc 150 mg, all b.i.d.
She rapidly improved and plasma viral load was confirmed undetectable on multiple samples.
However, 15 months later, she presented with recurring cerebellar symptoms. Plasma viral load was confirmed below 40 copies/ml, whereas CSF viral load revealed 200 copies/ml. Genotyping showed the same mutations for NRTIs, NNRTIs, and protease inhibitors, whereas the integrase site analysis revealed the R263K mutation. A tropism test confirmed a CCR5-expressing virus.
This is a typical case of viral compartmentalization associated with severe neurological complications. The R263K mutation has been detected in HIV-1-infected patients who experience treatment failure with dolutegravir [1,2]. Even though this substitution only modestly decreases sensitivity to dolutegravir and viral fitness , especially for subtype B viruses , it is considered as a signature of resistance to dolutegravir in both tissue culture and individuals . This observation confirms what has previously been shown by Pham et al., who found that the combination of this mutation with those for reverse transcriptase only moderately reduces viral infectiousness (about 70% compared with wild-type viruses).
Continued exposure to dolutegravir could in theory increase chances to develop the H51Y mutation for the integrase site, which could significantly decrease viral replication, DNA integration capacity, and therefore the size of the viral reservoir , thus leading HIV into an evolutionary dead end . However, we believe patients with such clinical severity should be managed differently, and the choice of combination ART should be based on the global genotyping score rather than on speculative fitness costs. Indeed, in this patient, we replaced dolutegravir with raltegravir, as its pathway of substitutions is generally incompatible with the dolutegravir-associated R263K mutation [1,5,6]. Maraviroc was continued unchanged, whereas darunavir was increased to 800 mg b.i.d.
1. Anstett K, Mesplede T, Oliveira M, Cutillas V, Wainberg MA. Dolutegravir resistance mutation R263K cannot coexist in combination with many classical integrase inhibitor resistance substitutions
. J Virol
2. Cahn P, Pozniak AL, Mingrone H, Shuldyakov A, Brites C, Andrade-Villanueva JF. Dolutegravir versus raltegravir in antiretroviral experienced, integrase inhibitor naïve adults with HIV: week 48 results from the randomized, double-blind, non inferiority SAILING study
3. Pham HT, Mesplède T, Wainberg MA. Effect on HIV-1 viral replication capacityof DTG-resistance mutations in NRTI/NNRTI resistant viruses
4. Mesplede T, Quashie PK, Hassounah D, Osman N, Han Y, Liang J, et al The R263K substitution in HIV-1 subtype C is more deleterious for integrase enzymatic function and viral replication than in subtype B
5. Anstett K, Fusco R, Cutillas V, Mesplède T, Wainberg MA. Dolutegravir-selected HIV-1 containing the N155H and R263K resistance substitutions does not acquire additional compensatory mutations under drug pressure that lead to higher-level resistance and increased replicative capacity
. J Virol
6. Oliveira M, Mesplède T, Moïsi D, Ibanescu RI, Brenner B, Wainberg MA. The dolutegravir R263K resistance mutation in HIV-1 integrase is incompatible with the emergence of resistance against raltegravir