Efficacy, Convenience, Safety and Durability of DTG-Based Antiretroviral Therapies: Evidence from a Prospective Study by the Italian MaSTER Cohort

Background: Dolutegravir (DTG) is recommended by international guidelines as a main component of an optimal initial regimen of cART (combination antiretroviral treatment) in people living with HIV (PLWH) and in case of switching for failure or optimization strategies. However, studies on the performance of DTG-containing regimens and indications for switching therapies in the long term are sparse. The purpose of this study was to evaluate prospectively the performance of DTG-based regimens, using the metrics of “efficacy”, “safety”, “convenience” and ‘’durability’’, among a nationally representative cohort of PLWH in Italy. Methods: We selected all PLWH in four centers of the MaSTER cohort who initiated a DTG-based regimen either when naïve or following a regimen switch between 11 July 2018 and 2 July 2021. Participants were followed until the outcomes were recorded or until the end of the study on 4 August 2022, whichever occurred first. Interruption was reported even when a participant switched to another DTG-containing regimen. Survival regression models were fitted to evaluate associations between therapy performance and age, sex, nationality, risk of HIV transmission, HIV RNA suppression status, CD4+ T-cell count, year of HIV diagnosis, cART status (naïve or experienced), cART backbone and viral hepatitis coinfection. Results: There were 371 participants in our cohort who initiated a DTG-based cART regimen in the time frame of the study. The population was predominantly male (75.2%), of Italian nationality (83.3%), with a history of cART use (80.9%), and the majority initiated a DTG-based regimen following a switch strategy in 2019 (80.1%). Median age was 53 years (interquartile range (IQR): 45–58). Prior cART regimen was based mostly on a combination of NRTI drugs plus a PI-boosted drug (34.2%), followed by a combination of NRTIs plus an NNRTI (23.5%). Concerning the NRTI backbone, the majority comprised 3TC plus ABC (34.5%), followed by 3TC alone (28.6%). The most reported transmission risk factor was heterosexual intercourse (44.2%). Total interruptions of the first DTG-based regimen were registered in 58 (15.6%) participants. The most frequent reason for interruption was due to cART simplification strategies, which accounted for 52%. Only 1 death was reported during the study period. The median time of total follow-up was 556 days (IQR: 316.5–722.5). Risk factors for poor performance of DTG-containing-regimens were found to be: a backbone regimen containing tenofovir, being cART naïve, having detectable HIV RNA at baseline, FIB-4 score above 3.25 and having a cancer diagnosis. By contrast, protective factors were found to be: higher CD4+ T-cell counts and higher CD4/CD8 ratio at baseline. Conclusion: DTG-based regimens were used mainly as a switching therapy in our cohort of PLWH who had undetectable HIV RNA and a good immune status. In this type of population, the durability of DTG-based regimens was maintained in 84.4% of participants with a modest incidence of interruptions mostly due to cART simplification strategies. The results of this prospective real-life study confirm the apparent low risk of changing DTG-containing regimens due to virological failure. They may also help physicians to identify people with increased risk of interruption for different reasons, suggesting targeted medical interventions.


Introduction
For people living with HIV (PLWH) with detectable viremia (HIV RNA), international guidelines recommend a combination antiretroviral therapy (cART) consisting of two nucleoside reverse transcriptase inhibitors (NRTIs) and one integrase strand transferase inhibitor (INSTI) as an optimal initial regimen unless otherwise indicated [1][2][3].An INSTI-based regimen is also recommended in case of switching strategies for those with an undetectable HIV RNA [3].Certain regimens may be contraindicated for some people who suffer from clinical conditions such as cardiovascular, kidney or liver diseases, opportunistic infections, or who are pregnant or planning a pregnancy, among other reasons [1,2].Dolutegravir (DTG), has demonstrated good efficacy and tolerability in clinical trials [4] and in real life [5,6]; however, studies on performance of DTG-containing regimens and indications for switching therapies in the long-term are sparse.Recently, a retrospective study conducted in the Management Standardizzato di TErapia antiRetrovirale (MaSTER) cohort [7], showed that INSTI-based regimens were interrupted mostly due to safety concerns and suggested that DTG performed better in terms of durability compared to raltegravir (RAL) and elvitegravir (EVG).However, the results of this work were limited due to its retrospective nature and the shorter length of follow-up for DTG-containing regimens compared to RAL and EVG.
For this reason, a prospective study was designed, aimed at better evaluating in a multidimensional manner the performance of DTG-based regimens over a longer follow-up, controlling for potential confounders.

Materials and Methods
This is a prospective study nested in a multicenter, hospital-based cohort established in the mid-1990s that has currently enrolled over 24,000 people with HIV infection, i.e., the MaSTER cohort [8].Data was collected from the enrolled persons' medical, prescription, and laboratory records at regular time intervals (baseline, six months, and twelve months).

Inclusion Criteria
PLWH from four Italian hospital centers of the MaSTER cohort (Fondazione Policlinico Universitario A. Gemelli IRCCS, Rome, accounting for 19.1% of the total cohort; Brescia ASST Spedali Civili Hospital, Brescia, accounting for 57.1% of the total cohort; Cremona ASST Hospital, Cremona, accounting for 22.4% of the total cohort; S. Maria Annunziata Hospital, Florence, accounting for 1.4% of the total cohort) who initiated a DTG-based regimen, either when cART naïve or following a regimen switch, were included in the study between 11 July 2018 and 2 July 2021.

Outcomes
Participants were followed until the study outcomes were recorded or until the end of the study on 4 August 2022, whichever occurred first.Interruption was also reported when a participant switched to another DTG-containing regimen.Performance of DTG-based regimens was evaluated using the metrics of "efficacy", "safety", "convenience" and ''durability".An event due to efficacy concerns was defined as virological failure (HIV RNA > 50 copies/mL at least six months after initiating therapy), whether or not therapy interruption occurred.Events due to safety concerns were defined as a newly occurring laboratory alteration of grades 3-4 and/or clinical progression whether or not there was a therapy interruption.Grade 3+ laboratory alterations were defined using the following measurements: aspartate transaminase (AST) > 260 units/L, alanine transaminase (ALT) > 235 units/L, bilirubin > 2.6 mg/dL, cholesterol ≥ 275 mg/dL, triglycerides > 750 mg/dL, glycemia > 250 mmol/L, creatinine > 1.9 mg/dL, calcium < 7 mg/dL, and iron < 2 g/dL.Conditions considered as clinical progression of HIV disease included a diagnosis of acquired immunodeficiency syndrome (AIDS), cancer, hepatic cirrhosis defined as a FIB-4 ≥ 3.25 [9], an ischemic cardiovascular event, kidney disease defined as eGFR (estimated glomerular filtration rate) < 89 mL/min, and any-cause death.An event due to convenience concerns was defined as having HIV RNA > 50 copies/mL and a laboratory alteration (composite outcome of efficacy and safety), whether or not there was a therapy interruption.Durability was defined as the total course of treatment until interruption regardless of tolerability, HIV RNA level, laboratory alteration, or clinical progression.

Statistical Approach
Kaplan-Meier estimators, univariate and multivariable Cox proportional hazards regression models were fitted to associate incidence of efficacy, convenience, safety and durability (interruption) events of DTG therapies with demographics (age, sex, and nationality), risk factors for HIV transmission (heterosexual intercourse, people who inject drugs (PWID), men who have sex with men (MSM), or other), cART status (naïve or experienced), CD4+ T-cell count, CD8 + T-cell count, CD4/CD8 ratio, HIV RNA at baseline, presence of hepatitis B infection or hepatitis C antibodies (HCVAb), cART backbone, eGFR, FIB-4 score [9], Framingham score [10], prior diagnosis of cancer, and year of starting DTG.Missing values in the multivariable analysis were imputed using random forests, while case deletion was used in univariate analysis.In the multivariable models, each exposure of interest above defined was adjusted by a set of covariates identified through the generalized adjustment criterion on a directed acyclic graph agreed by the co-authors (Figure S1).

Patient Characteristics
There were 371 participants who initiated a DTG-based cART regimen in four centers of the MaSTER cohort during the time frame of the study.Median age was 53 years (IQR: 45-58).The majority of the study population were male (75.2%), of Italian nationality (83.3%), with a history of cART use (80.9%) and the majority of them initiated a DTG-based regimen after switching from previous regimens in 2019 (80.1%) (Table 1).Among experienced people, prior cART regimens were mostly based on a combination of nucleoside/nucleotide reverse transcriptase inhibitors (NRTIs) in addition to a boosted protease inhibitor (PI) (34.2%) followed by a combination of NRTIs plus a non nucleoside reverse transcriptase inhibitor (NNRTI) (23.5%) (Table 1).Other prior cART regimens (18.3%) included a miscellaneous combination of drugs of which previous INSTI exposure accounted for 8.1% (Table S1).The most frequent NRTI backbone was represented by lamivudine (3TC) plus abacavir (ABC) (34.5%), followed by 3TC alone (28.6%), and FTC in addition to tenofovir disoproxil fumarate (TDF) (12.9%) (Table 1).Dual therapies were more frequent among experienced participants compared to naïve people (43.7% vs. 15.5%),while tenofovir containing backbones were more used by naïve participants compared to experienced people (59.2% vs. 12.7%) (Tables S2 and S3).
HBsAg chronic carriers and/or positive HCV Ab participants accounted for 4.3% of the population.A previous diagnosis of cancer was present in 33 participants (8.9%) (Table 1).

DTG Regimen Performance and Rates of Interruption
Regarding DTG performance in the study cohort, 81 efficacy events (25.2%), 107 convenience events (33.2%), 76 safety events (23.6%) and 58 durability events (18.0%) occurred.These events led to a DTG-containing regimen interruption or change (durability events) in 58 (15.6%) people overall, over a median follow-up of 556 days (IQR: 316.5-722.5).The most frequent reason for interruption was cART simplification, which accounted for 52%, followed by a 29.3%rate of interruption due to toxicity (mostly rash, hypersensitivity reactions and nephrotoxic effects) and a 6.9% interruption due to virological failure.Participants lost to follow-up were only 2 (3.5%), and only 1 patient died during the study period (Table 1).
As shown in the Kaplan-Meier estimate curves, lower performance of DTG-based regimens according to the metric of durability (Figure 1) and efficacy (Figure 2) were found in the following categories of participants: (i) those who were prescribed a NRTI backbone containing tenofovir; (ii) FIB4 score > 3.25; (iii) higher HIV RNA levels; (iv) CD4+ T-cell count less than 200 cells/mm 3 ; (v) previously cART naïve.In addition to the categories listed above, the Kaplan-Meier survival analysis showed lower performance of DTG-based regimens according to the metric of convenience (Figure 3) and safety (Figure 4) in the following participants: (i) those switched to a DTG-containing regimen from a previous cART containing another INSTI; (ii) those with eGFR below 60 mL/min.

Associations with Efficacy Events
With univariate analysis, greater hazards of efficacy event occurrence were observed among individuals with a backbone regimen containing emtricitabine plus tenofovir, naïve people at baseline, those with a detectable HIV RNA and those with FIB-4 above 3.25.Participants switched to a DTG-containing regimen from a previous cART not containing an INSTI had a lower hazard of efficacy event occurrence, as did people with higher CD4 + T-cell count and higher CD4/CD8 ratio at baseline (Table 2).Using the multivariable model, significantly higher hazards of efficacy event occurrence were found in participants who were prescribed emtricitabine plus tenofovir and with detectable HIV RNA at baseline (Table 3); experienced people switched from a cART regimen had a lower hazard of efficacy event occurrence compared to naïve participants (Table 3).

Associations with Convenience Events
With univariate analysis, higher hazards of a convenience event occurring were observed among individuals with a backbone regimen containing emtricitabine plus tenofovir, those who were naïve, those with a higher viral load at baseline, FIB-4 at baseline above 3.25, positive HBsAg, and eGFR at baseline below 60 mL/min (Table 2).Lower hazards of a convenience event occurring were observed among experienced people and those having higher CD4+ T-cell count and higher CD4/CD8 ratio at baseline.
Using the multivariable model, significantly higher hazards of a convenience event occurring were indicated among individuals with a backbone regimen containing emtric-itabine plus tenofovir and higher HIV RNA at baseline.Experienced people had lower hazards of occurrence of an event due to convenience (Table 3).

Associations with Safety Events
With univariate analysis, higher hazards of occurrence of events due to safety concerns were observed among individuals with a backbone regimen containing emtricitabine plus tenofovir, among participants with a higher viral load at baseline, FIB-4 above 3.25 and eGFR at baseline below 60 mL/min.Experienced people, having higher CD4+ T-cell count and higher CD4/CD8 ratio at baseline and those having an increased Framingham score at baseline had lower hazards of occurrence of events due to safety (Table 2).
Using the multivariable model, significantly higher hazards of occurrence of a safety event were maintained only among individuals with a backbone regimen containing emtricitabine plus tenofovir, while lower hazards were maintained in experienced people (Table 3).

Associations with Durability Events
With univariate analysis, hazards of interruption or therapy change (durability) were significantly greater in the following participants: those receiving a backbone regimen containing emtricitabine plus tenofovir, cART naïve participants, those with higher HIV RNA, FIB4 at baseline above 3.25, and cancer.Lower hazards of interruption were observed among people with higher CD4+ T-cell count, higher CD4/CD8 ratio at baseline, switching from cART not containing an INSTI compared to naïve people and having an increased Framingham score at baseline (Table 2).
Using the multivariable model, significant higher hazards of interruption were maintained among individuals prescribed a backbone regimen containing emtricitabine plus tenofovir and with higher HIV RNA at baseline.Experienced participants prescribed cART regimens not containing an INSTI had lower hazards of interruption compared with cART naïve people for the metric of durability (Table 3).
The backbone regimen containing emtricitabine plus tenofovir was mainly interrupted due to simplification strategies as we observed in a competing risk analysis (Figure 5).

Associations with Efficacy Events
With univariate analysis, greater hazards of efficacy event occurrence were observed among individuals with a backbone regimen containing emtricitabine plus tenofovir, naïve people at baseline, those with a detectable HIV RNA and those with FIB-4 above 3.25.Participants switched to a DTG-containing regimen from a previous cART not containing an INSTI had a lower hazard of efficacy event occurrence, as did people with higher CD4 + T-cell count and higher CD4/CD8 ratio at baseline (Table 2).
Using the multivariable model, significantly higher hazards of efficacy event occurrence were found in participants who were prescribed emtricitabine plus tenofovir and with detectable HIV RNA at baseline (Table 3); experienced people switched from a cART regimen had a lower hazard of efficacy event occurrence compared to naïve participants (Table 3).With univariate analysis, hazards of interruption or therapy change (durabili significantly greater in the following participants: those receiving a backbone r containing emtricitabine plus tenofovir, cART naïve participants, those with high RNA, FIB4 at baseline above 3.25, and cancer.Lower hazards of interruption w served among people with higher CD4 + T-cell count, higher CD4/CD8 ratio at b switching from cART not containing an INSTI compared to naïve people and ha increased Framingham score at baseline (Table 2).
Using the multivariable model, significant higher hazards of interruption wer tained among individuals prescribed a backbone regimen containing emtricitab tenofovir and with higher HIV RNA at baseline.Experienced participants pre cART regimens not containing an INSTI had lower hazards of interruption compar cART naïve people for the metric of durability (Table 3).
The backbone regimen containing emtricitabine plus tenofovir was mainl rupted due to simplification strategies as we observed in a competing risk analysis 5).

Discussion
In this study, we prospectively evaluated the performance of DTG-based regimens in a multi-dimensional way, using the metrics of "efficacy", "convenience", "safety" and ''durability", in a nested study of the Italian MaSTER cohort.The population examined had a median age of 53 years and was mainly composed by PLWH experienced for cART, since naïve participants were only 19% of the overall population.The majority of participants had undetectable HIV RNA and a high CD4+ T-cell count (median of 640 cells/mm 3 ), reflecting that DTG-containing regimens were mainly prescribed (particularly in the year 2019) in order to optimize cART in those already responding to treatment.
The results of the study in terms of independent predictors of the outcomes for each of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated the performance of DTG-b in a multi-dimensional way, using the metrics of "efficacy", "convenience ''durability'', in a nested study of the Italian MaSTER cohort.The popula had a median age of 53 years and was mainly composed by PLWH experie since naïve participants were only 19% of the overall population.The maj pants had undetectable HIV RNA and a high CD4+ T-cell count (median of reflecting that DTG-containing regimens were mainly prescribed (particul 2019) in order to optimize cART in those already responding to treatment.
The results of the study in terms of independent predictors of the out of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated the performa in a multi-dimensional way, using the metrics of "efficacy", ''durability'', in a nested study of the Italian MaSTER coho had a median age of 53 years and was mainly composed by P since naïve participants were only 19% of the overall popula pants had undetectable HIV RNA and a high CD4+ T-cell cou reflecting that DTG-containing regimens were mainly prescr 2019) in order to optimize cART in those already responding The results of the study in terms of independent predic of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated in a multi-dimensional way, using the metrics o ''durability'', in a nested study of the Italian M had a median age of 53 years and was mainly co since naïve participants were only 19% of the o pants had undetectable HIV RNA and a high CD reflecting that DTG-containing regimens were m 2019) in order to optimize cART in those alread The results of the study in terms of indepe of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectivel in a multi-dimensional way, using ''durability'', in a nested study of t had a median age of 53 years and w since naïve participants were only pants had undetectable HIV RNA a reflecting that DTG-containing regi 2019) in order to optimize cART in The results of the study in term of the study metrics are summarize

Discussion
In this study, we prospectively evaluated the performance of DTG-b in a multi-dimensional way, using the metrics of "efficacy", "convenience ''durability'', in a nested study of the Italian MaSTER cohort.The popula had a median age of 53 years and was mainly composed by PLWH experie since naïve participants were only 19% of the overall population.The maj pants had undetectable HIV RNA and a high CD4+ T-cell count (median of reflecting that DTG-containing regimens were mainly prescribed (particul 2019) in order to optimize cART in those already responding to treatment.
The results of the study in terms of independent predictors of the out of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated the performa in a multi-dimensional way, using the metrics of "efficacy", ''durability'', in a nested study of the Italian MaSTER coho had a median age of 53 years and was mainly composed by P since naïve participants were only 19% of the overall popula pants had undetectable HIV RNA and a high CD4+ T-cell cou reflecting that DTG-containing regimens were mainly prescr 2019) in order to optimize cART in those already responding The results of the study in terms of independent predic of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated in a multi-dimensional way, using the metrics o ''durability'', in a nested study of the Italian M had a median age of 53 years and was mainly co since naïve participants were only 19% of the o pants had undetectable HIV RNA and a high CD reflecting that DTG-containing regimens were m 2019) in order to optimize cART in those alread The results of the study in terms of indepe of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectivel in a multi-dimensional way, using ''durability'', in a nested study of t had a median age of 53 years and w since naïve participants were only pants had undetectable HIV RNA a reflecting that DTG-containing regi 2019) in order to optimize cART in The results of the study in term of the study metrics are summarize

Discussion
In this study, we prospectively evaluated the performance of DTG-b in a multi-dimensional way, using the metrics of "efficacy", "convenience ''durability'', in a nested study of the Italian MaSTER cohort.The popula had a median age of 53 years and was mainly composed by PLWH experie since naïve participants were only 19% of the overall population.The maj pants had undetectable HIV RNA and a high CD4+ T-cell count (median of reflecting that DTG-containing regimens were mainly prescribed (particul 2019) in order to optimize cART in those already responding to treatment.
The results of the study in terms of independent predictors of the out of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated the performa in a multi-dimensional way, using the metrics of "efficacy", ''durability'', in a nested study of the Italian MaSTER coho had a median age of 53 years and was mainly composed by P since naïve participants were only 19% of the overall popula pants had undetectable HIV RNA and a high CD4+ T-cell cou reflecting that DTG-containing regimens were mainly prescr 2019) in order to optimize cART in those already responding The results of the study in terms of independent predic of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated in a multi-dimensional way, using the metrics o ''durability'', in a nested study of the Italian M had a median age of 53 years and was mainly co since naïve participants were only 19% of the o pants had undetectable HIV RNA and a high CD reflecting that DTG-containing regimens were m 2019) in order to optimize cART in those alread The results of the study in terms of indepe of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectivel in a multi-dimensional way, using ''durability'', in a nested study of t had a median age of 53 years and w since naïve participants were only 1 pants had undetectable HIV RNA a reflecting that DTG-containing regi 2019) in order to optimize cART in The results of the study in term of the study metrics are summarize

Discussion
In this study, we prospectively evaluated the performance of DTG-b in a multi-dimensional way, using the metrics of "efficacy", "convenience ''durability'', in a nested study of the Italian MaSTER cohort.The popula had a median age of 53 years and was mainly composed by PLWH experie since naïve participants were only 19% of the overall population.The maj pants had undetectable HIV RNA and a high CD4+ T-cell count (median of reflecting that DTG-containing regimens were mainly prescribed (particul 2019) in order to optimize cART in those already responding to treatment.
The results of the study in terms of independent predictors of the out of the study metrics are summarized in Table 4. Concerning the metric of efficacy, our results suggest that potential the occurrence of an efficacy event were: (a) the backbone regimen contai bine plus tenofovir.Regarding this finding, however, we did not direct efficacy of the backbone regimens with each other and we are not able to fu results.They most likely reflect the extensive use of tenofovir in cART com pecially in naïve patients, therefore increasing the possibility of its statist with detectable viraemia; (b) the condition of being cART naïve, probab

Discussion
In this study, we prospectively evaluated the performa in a multi-dimensional way, using the metrics of "efficacy", ''durability'', in a nested study of the Italian MaSTER coho had a median age of 53 years and was mainly composed by P since naïve participants were only 19% of the overall popula pants had undetectable HIV RNA and a high CD4+ T-cell cou reflecting that DTG-containing regimens were mainly prescr 2019) in order to optimize cART in those already responding The results of the study in terms of independent predic of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated in a multi-dimensional way, using the metrics o ''durability'', in a nested study of the Italian M had a median age of 53 years and was mainly co since naïve participants were only 19% of the o pants had undetectable HIV RNA and a high CD reflecting that DTG-containing regimens were m 2019) in order to optimize cART in those alread The results of the study in terms of indepe of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectivel in a multi-dimensional way, using ''durability'', in a nested study of t had a median age of 53 years and w since naïve participants were only pants had undetectable HIV RNA a reflecting that DTG-containing regi 2019) in order to optimize cART in The results of the study in term of the study metrics are summarize

Discussion
In this study, we prospectively evaluated the performance of DTG-b in a multi-dimensional way, using the metrics of "efficacy", "convenience ''durability'', in a nested study of the Italian MaSTER cohort.The popula had a median age of 53 years and was mainly composed by PLWH experie since naïve participants were only 19% of the overall population.The maj pants had undetectable HIV RNA and a high CD4+ T-cell count (median of reflecting that DTG-containing regimens were mainly prescribed (particul 2019) in order to optimize cART in those already responding to treatment.
The results of the study in terms of independent predictors of the out of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated the performa in a multi-dimensional way, using the metrics of "efficacy", ''durability'', in a nested study of the Italian MaSTER coho had a median age of 53 years and was mainly composed by P since naïve participants were only 19% of the overall popula pants had undetectable HIV RNA and a high CD4+ T-cell cou reflecting that DTG-containing regimens were mainly prescr 2019) in order to optimize cART in those already responding The results of the study in terms of independent predic of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated in a multi-dimensional way, using the metrics o ''durability'', in a nested study of the Italian M had a median age of 53 years and was mainly co since naïve participants were only 19% of the o pants had undetectable HIV RNA and a high CD reflecting that DTG-containing regimens were m 2019) in order to optimize cART in those alread The results of the study in terms of indepe of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectivel in a multi-dimensional way, using ''durability'', in a nested study of t had a median age of 53 years and w since naïve participants were only pants had undetectable HIV RNA a reflecting that DTG-containing regi 2019) in order to optimize cART in The results of the study in term of the study metrics are summarize

Discussion
In this study, we prospectively evaluated the performance of DTG-b in a multi-dimensional way, using the metrics of "efficacy", "convenience ''durability'', in a nested study of the Italian MaSTER cohort.The popula had a median age of 53 years and was mainly composed by PLWH experie since naïve participants were only 19% of the overall population.The maj pants had undetectable HIV RNA and a high CD4+ T-cell count (median of reflecting that DTG-containing regimens were mainly prescribed (particul 2019) in order to optimize cART in those already responding to treatment.
The results of the study in terms of independent predictors of the out of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated the performa in a multi-dimensional way, using the metrics of "efficacy", ''durability'', in a nested study of the Italian MaSTER coho had a median age of 53 years and was mainly composed by P since naïve participants were only 19% of the overall popula pants had undetectable HIV RNA and a high CD4+ T-cell cou reflecting that DTG-containing regimens were mainly prescr 2019) in order to optimize cART in those already responding The results of the study in terms of independent predic of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated in a multi-dimensional way, using the metrics o ''durability'', in a nested study of the Italian M had a median age of 53 years and was mainly co since naïve participants were only 19% of the o pants had undetectable HIV RNA and a high CD reflecting that DTG-containing regimens were m 2019) in order to optimize cART in those alread The results of the study in terms of indepe of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectivel in a multi-dimensional way, using ''durability'', in a nested study of t had a median age of 53 years and w since naïve participants were only 1 pants had undetectable HIV RNA a reflecting that DTG-containing regi 2019) in order to optimize cART in The results of the study in term of the study metrics are summarize

Discussion
In this study, we prospectively evaluated the performance of DTG-b in a multi-dimensional way, using the metrics of "efficacy", "convenience ''durability'', in a nested study of the Italian MaSTER cohort.The popula had a median age of 53 years and was mainly composed by PLWH experie since naïve participants were only 19% of the overall population.The maj pants had undetectable HIV RNA and a high CD4+ T-cell count (median of reflecting that DTG-containing regimens were mainly prescribed (particul 2019) in order to optimize cART in those already responding to treatment.
The results of the study in terms of independent predictors of the out of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated the performa in a multi-dimensional way, using the metrics of "efficacy", ''durability'', in a nested study of the Italian MaSTER coho had a median age of 53 years and was mainly composed by P since naïve participants were only 19% of the overall popula pants had undetectable HIV RNA and a high CD4+ T-cell cou reflecting that DTG-containing regimens were mainly prescr 2019) in order to optimize cART in those already responding The results of the study in terms of independent predic of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated in a multi-dimensional way, using the metrics o ''durability'', in a nested study of the Italian M had a median age of 53 years and was mainly co since naïve participants were only 19% of the o pants had undetectable HIV RNA and a high CD reflecting that DTG-containing regimens were m 2019) in order to optimize cART in those alread The results of the study in terms of indepe of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectivel in a multi-dimensional way, using ''durability'', in a nested study of t had a median age of 53 years and w since naïve participants were only 1 pants had undetectable HIV RNA a reflecting that DTG-containing regi 2019) in order to optimize cART in The results of the study in term of the study metrics are summarize

Discussion
In this study, we prospectively evaluated the performance of DTG-b in a multi-dimensional way, using the metrics of "efficacy", "convenience ''durability'', in a nested study of the Italian MaSTER cohort.The popula had a median age of 53 years and was mainly composed by PLWH experie since naïve participants were only 19% of the overall population.The maj pants had undetectable HIV RNA and a high CD4+ T-cell count (median of reflecting that DTG-containing regimens were mainly prescribed (particul 2019) in order to optimize cART in those already responding to treatment.
The results of the study in terms of independent predictors of the out of the study metrics are summarized in Table 4. Concerning the metric of efficacy, our results suggest that potential the occurrence of an efficacy event were: (a) the backbone regimen contai

Discussion
In this study, we prospectively evaluated the performa in a multi-dimensional way, using the metrics of "efficacy", ''durability'', in a nested study of the Italian MaSTER coho had a median age of 53 years and was mainly composed by P since naïve participants were only 19% of the overall popula pants had undetectable HIV RNA and a high CD4+ T-cell cou reflecting that DTG-containing regimens were mainly prescr 2019) in order to optimize cART in those already responding The results of the study in terms of independent predic of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectively evaluated in a multi-dimensional way, using the metrics o ''durability'', in a nested study of the Italian M had a median age of 53 years and was mainly co since naïve participants were only 19% of the o pants had undetectable HIV RNA and a high CD reflecting that DTG-containing regimens were m 2019) in order to optimize cART in those alread The results of the study in terms of indepe of the study metrics are summarized in Table 4.

Discussion
In this study, we prospectivel in a multi-dimensional way, using ''durability'', in a nested study of t had a median age of 53 years and w since naïve participants were only pants had undetectable HIV RNA a reflecting that DTG-containing regi 2019) in order to optimize cART in The results of the study in term of the study metrics are summarize Concerning the metric of effic the occurrence of an efficacy event w had a median age of 53 years and was mainly composed by PLWH experie since naïve participants were only 19% of the overall population.The maj pants had undetectable HIV RNA and a high CD4+ T-cell count (median of reflecting that DTG-containing regimens were mainly prescribed (particul 2019) in order to optimize cART in those already responding to treatment.
The results of the study in terms of independent predictors of the out of the study metrics are summarized in Table 4. Concerning the metric of efficacy, our results suggest that potential the occurrence of an efficacy event were: (a) the backbone regimen contai bine plus tenofovir.Regarding this finding, however, we did not direct efficacy of the backbone regimens with each other and we are not able to fu results.They most likely reflect the extensive use of tenofovir in cART com pecially in naïve patients, therefore increasing the possibility of its statist with detectable viraemia; (b) the condition of being cART naïve, probab "selection of the fittest" effect, because treatment experienced participan undetectable HIV RNA) were those on stable treatment and therefore wit adherence to cART.In addition, naïve participants with higher baseline HI S2 and S3) may need more time on cART to obtain undetectable HIV RNA dition of being naïve was associated with a higher probability of having det after six months of follow-up compared to experienced people who alread treatment and in whom cART was previously optimized; (c) having highe baseline, as already found in a previous work of the MaSTER cohort [7], p importance of carefully monitoring these people for HIV RNA, even tho had a median age of 53 years and was mainly composed by P since naïve participants were only 19% of the overall popula pants had undetectable HIV RNA and a high CD4+ T-cell cou reflecting that DTG-containing regimens were mainly prescr 2019) in order to optimize cART in those already responding The results of the study in terms of independent predic of the study metrics are summarized in Table 4. Concerning the metric of efficacy, our results suggest t the occurrence of an efficacy event were: (a) the backbone re bine plus tenofovir.Regarding this finding, however, we d efficacy of the backbone regimens with each other and we ar results.They most likely reflect the extensive use of tenofov pecially in naïve patients, therefore increasing the possibilit with detectable viraemia; (b) the condition of being cART "selection of the fittest" effect, because treatment experienc undetectable HIV RNA) were those on stable treatment and adherence to cART.In addition, naïve participants with high S2 and S3) may need more time on cART to obtain undetecta dition of being naïve was associated with a higher probability after six months of follow-up compared to experienced peop treatment and in whom cART was previously optimized; (c) baseline, as already found in a previous work of the MaSTE importance of carefully monitoring these people for HIV R had a median age of 53 years and was mainly co since naïve participants were only 19% of the o pants had undetectable HIV RNA and a high CD reflecting that DTG-containing regimens were m 2019) in order to optimize cART in those alread The results of the study in terms of indepe of the study metrics are summarized in Table 4. Concerning the metric of efficacy, our res the occurrence of an efficacy event were: (a) the bine plus tenofovir.Regarding this finding, h efficacy of the backbone regimens with each oth results.They most likely reflect the extensive u pecially in naïve patients, therefore increasing with detectable viraemia; (b) the condition of "selection of the fittest" effect, because treatme undetectable HIV RNA) were those on stable t adherence to cART.In addition, naïve participa S2 and S3) may need more time on cART to obta dition of being naïve was associated with a high after six months of follow-up compared to expe treatment and in whom cART was previously o baseline, as already found in a previous work o importance of carefully monitoring these peop had a median age of 53 years and w since naïve participants were only pants had undetectable HIV RNA a reflecting that DTG-containing regi 2019) in order to optimize cART in The results of the study in term of the study metrics are summarize Concerning the metric of effic the occurrence of an efficacy event w bine plus tenofovir.Regarding th efficacy of the backbone regimens w results.They most likely reflect the pecially in naïve patients, therefore with detectable viraemia; (b) the co "selection of the fittest" effect, beca undetectable HIV RNA) were those adherence to cART.In addition, naï S2 and S3) may need more time on dition of being naïve was associated after six months of follow-up comp treatment and in whom cART was baseline, as already found in a prev importance of carefully monitoring Concerning the metric of efficacy, our results suggest that potential the occurrence of an efficacy event were: (a) the backbone regimen contai bine plus tenofovir.Regarding this finding, however, we did not direct efficacy of the backbone regimens with each other and we are not able to fu results.They most likely reflect the extensive use of tenofovir in cART com pecially in naïve patients, therefore increasing the possibility of its statist with detectable viraemia; (b) the condition of being cART naïve, probab "selection of the fittest" effect, because treatment experienced participan undetectable HIV RNA) were those on stable treatment and therefore wit adherence to cART.In addition, naïve participants with higher baseline HI S2 and S3) may need more time on cART to obtain undetectable HIV RNA dition of being naïve was associated with a higher probability of having det after six months of follow-up compared to experienced people who alread treatment and in whom cART was previously optimized; (c) having highe baseline, as already found in a previous work of the MaSTER cohort [7], p importance of carefully monitoring these people for HIV RNA, even tho Concerning the metric of efficacy, our results suggest t the occurrence of an efficacy event were: (a) the backbone re bine plus tenofovir.Regarding this finding, however, we d efficacy of the backbone regimens with each other and we ar results.They most likely reflect the extensive use of tenofov pecially in naïve patients, therefore increasing the possibilit with detectable viraemia; (b) the condition of being cART "selection of the fittest" effect, because treatment experienc undetectable HIV RNA) were those on stable treatment and adherence to cART.In addition, naïve participants with high S2 and S3) may need more time on cART to obtain undetecta dition of being naïve was associated with a higher probability after six months of follow-up compared to experienced peop treatment and in whom cART was previously optimized; (c) baseline, as already found in a previous work of the MaSTE importance of carefully monitoring these people for HIV R since naïve participants were only 19% of the o pants had undetectable HIV RNA and a high CD reflecting that DTG-containing regimens were m 2019) in order to optimize cART in those alread

Efficacy
The results of the study in terms of indepe of the study metrics are summarized in Table 4. Concerning the metric of efficacy, our res the occurrence of an efficacy event were: (a) the bine plus tenofovir.Regarding this finding, h efficacy of the backbone regimens with each oth results.They most likely reflect the extensive u pecially in naïve patients, therefore increasing with detectable viraemia; (b) the condition of "selection of the fittest" effect, because treatme undetectable HIV RNA) were those on stable t adherence to cART.In addition, naïve participa S2 and S3) may need more time on cART to obta dition of being naïve was associated with a highe after six months of follow-up compared to expe treatment and in whom cART was previously o baseline, as already found in a previous work o importance of carefully monitoring these peop since naïve participants were only 1 pants had undetectable HIV RNA a reflecting that DTG-containing regi 2019) in order to optimize cART in The results of the study in term of the study metrics are summarize Concerning the metric of effic the occurrence of an efficacy event w bine plus tenofovir.Regarding th efficacy of the backbone regimens w results.They most likely reflect the pecially in naïve patients, therefore with detectable viraemia; (b) the co "selection of the fittest" effect, beca undetectable HIV RNA) were those adherence to cART.In addition, naï S2 and S3) may need more time on dition of being naïve was associated after six months of follow-up comp treatment and in whom cART was baseline, as already found in a prev importance of carefully monitoring Concerning the metric of efficacy, our results suggest that potential the occurrence of an efficacy event were: (a) the backbone regimen contai bine plus tenofovir.Regarding this finding, however, we did not direct efficacy of the backbone regimens with each other and we are not able to fu results.They most likely reflect the extensive use of tenofovir in cART com pecially in naïve patients, therefore increasing the possibility of its statist with detectable viraemia; (b) the condition of being cART naïve, probab "selection of the fittest" effect, because treatment experienced participan undetectable HIV RNA) were those on stable treatment and therefore wit adherence to cART.In addition, naïve participants with higher baseline HI S2 and S3) may need more time on cART to obtain undetectable HIV RNA dition of being naïve was associated with a higher probability of having det after six months of follow-up compared to experienced people who alread treatment and in whom cART was previously optimized; (c) having highe baseline, as already found in a previous work of the MaSTER cohort [7], p importance of carefully monitoring these people for HIV RNA, even tho Concerning the metric of efficacy, our results suggest t the occurrence of an efficacy event were: (a) the backbone re bine plus tenofovir.Regarding this finding, however, we d efficacy of the backbone regimens with each other and we ar results.They most likely reflect the extensive use of tenofov pecially in naïve patients, therefore increasing the possibilit with detectable viraemia; (b) the condition of being cART "selection of the fittest" effect, because treatment experienc undetectable HIV RNA) were those on stable treatment and adherence to cART.In addition, naïve participants with high S2 and S3) may need more time on cART to obtain undetecta dition of being naïve was associated with a higher probability after six months of follow-up compared to experienced peop treatment and in whom cART was previously optimized; (c) baseline, as already found in a previous work of the MaSTE importance of carefully monitoring these people for HIV R Concerning the metric of efficacy, our res the occurrence of an efficacy event were: (a) the bine plus tenofovir.Regarding this finding, h efficacy of the backbone regimens with each oth results.They most likely reflect the extensive u pecially in naïve patients, therefore increasing with detectable viraemia; (b) the condition of "selection of the fittest" effect, because treatme undetectable HIV RNA) were those on stable t adherence to cART.In addition, naïve participa S2 and S3) may need more time on cART to obta dition of being naïve was associated with a high after six months of follow-up compared to expe treatment and in whom cART was previously o baseline, as already found in a previous work o importance of carefully monitoring these peop in a multi-dimensional way, using ''durability'', in a nested study of t had a median age of 53 years and w since naïve participants were only pants had undetectable HIV RNA a reflecting that DTG-containing regi 2019) in order to optimize cART in

Baseline cancer diagnosis
The results of the study in term of the study metrics are summarize Concerning the metric of effic the occurrence of an efficacy event w bine plus tenofovir.Regarding th efficacy of the backbone regimens w results.They most likely reflect the pecially in naïve patients, therefore with detectable viraemia; (b) the co "selection of the fittest" effect, beca undetectable HIV RNA) were those adherence to cART.In addition, naï S2 and S3) may need more time on dition of being naïve was associated after six months of follow-up comp treatment and in whom cART was baseline, as already found in a prev importance of carefully monitoring of the study metrics are summarized in Table 4. Concerning the metric of efficacy, our results suggest that potential risk factors for the occurrence of an efficacy event were: (a) the backbone regimen containing emtricitabine plus tenofovir.Regarding this finding, however, we did not directly compare the efficacy of the backbone regimens with each other and we are not able to fully explain our results.They most likely reflect the extensive use of tenofovir in cART combinations, expecially in naïve patients, therefore increasing the possibility of its statistical association with detectable viraemia; (b) the condition of being cART naïve, probably because of a "selection of the fittest" effect, because treatment experienced participants (mostly with undetectable HIV RNA) were those on stable treatment and therefore with consolidated adherence to cART.In addition, naïve participants with higher baseline HIV RNA (Tables S2 and S3) may need more time on cART to obtain undetectable HIV RNA.Thus, the condition of being naïve was associated with a higher probability of having detectable viremia after six months of follow-up compared to experienced people who already responded to treatment and in whom cART was previously optimized; (c) having higher HIV RNA at baseline, as already found in a previous work of the MaSTER cohort [7], pointing out the importance of carefully monitoring these people for HIV RNA, even those using DTG-Not a statistically significant association; ''durability'', in a nested study of the Italian MaSTER cohort.The population e had a median age of 53 years and was mainly composed by PLWH experienced f since naïve participants were only 19% of the overall population.The majority o pants had undetectable HIV RNA and a high CD4+ T-cell count (median of 640 ce reflecting that DTG-containing regimens were mainly prescribed (particularly in 2019) in order to optimize cART in those already responding to treatment.
The results of the study in terms of independent predictors of the outcomes of the study metrics are summarized in Table 4. Concerning the metric of efficacy, our results suggest that potential risk fa the occurrence of an efficacy event were: (a) the backbone regimen containing e bine plus tenofovir.Regarding this finding, however, we did not directly com efficacy of the backbone regimens with each other and we are not able to fully exp results.They most likely reflect the extensive use of tenofovir in cART combinat pecially in naïve patients, therefore increasing the possibility of its statistical as with detectable viraemia; (b) the condition of being cART naïve, probably bec "selection of the fittest" effect, because treatment experienced participants (mo undetectable HIV RNA) were those on stable treatment and therefore with cons adherence to cART.In addition, naïve participants with higher baseline HIV RNA S2 and S3) may need more time on cART to obtain undetectable HIV RNA.Thus, dition of being naïve was associated with a higher probability of having detectable after six months of follow-up compared to experienced people who already resp treatment and in whom cART was previously optimized; (c) having higher HIV baseline, as already found in a previous work of the MaSTER cohort [7], pointin importance of carefully monitoring these people for HIV RNA, even those usi protective factor; In this study, we prospectively evaluated the performanc in a multi-dimensional way, using the metrics of "efficacy", "co ''durability'', in a nested study of the Italian MaSTER cohort.T had a median age of 53 years and was mainly composed by PLW since naïve participants were only 19% of the overall populatio pants had undetectable HIV RNA and a high CD4+ T-cell count ( reflecting that DTG-containing regimens were mainly prescribe 2019) in order to optimize cART in those already responding to The results of the study in terms of independent predictor of the study metrics are summarized in Table 4. Concerning the metric of efficacy, our results suggest tha the occurrence of an efficacy event were: (a) the backbone regim bine plus tenofovir.Regarding this finding, however, we did efficacy of the backbone regimens with each other and we are no results.They most likely reflect the extensive use of tenofovir i pecially in naïve patients, therefore increasing the possibility o with detectable viraemia; (b) the condition of being cART naï "selection of the fittest" effect, because treatment experienced undetectable HIV RNA) were those on stable treatment and th adherence to cART.In addition, naïve participants with higher b S2 and S3) may need more time on cART to obtain undetectable dition of being naïve was associated with a higher probability of after six months of follow-up compared to experienced people w treatment and in whom cART was previously optimized; (c) ha baseline, as already found in a previous work of the MaSTER c importance of carefully monitoring these people for HIV RNA risk factor; cART, combination antiretroviral treatment; INSTI, integrase strand inhibitors; HBsAg, hepatitis B antigen; eGFR, estimated Glomerular Filtration Rate.
Concerning the metric of efficacy, our results suggest that potential risk factors for the occurrence of an efficacy event were: (a) the backbone regimen containing emtricitabine plus tenofovir.Regarding this finding, however, we did not directly compare the efficacy of the backbone regimens with each other and we are not able to fully explain our results.They most likely reflect the extensive use of tenofovir in cART combinations, expecially in naïve patients, therefore increasing the possibility of its statistical association with detectable viraemia; (b) the condition of being cART naïve, probably because of a "selection of the fittest" effect, because treatment experienced participants (mostly with undetectable HIV RNA) were those on stable treatment and therefore with consolidated adherence to cART.In addition, naïve participants with higher baseline HIV RNA (Tables S2 and S3) may need more time on cART to obtain undetectable HIV RNA.Thus, the condition of being naïve was associated with a higher probability of having detectable viremia after six months of follow-up compared to experienced people who already responded to treatment and in whom cART was previously optimized; (c) having higher HIV RNA at baseline, as already found in a previous work of the MaSTER cohort [7], pointing out the importance of carefully monitoring these people for HIV RNA, even those using DTG-containing regimens, who may run the risk of low level viremia after six months of treatment [14]; (d) FIB4 score > 3.25.Regarding this finding, the correlation between poor cART adherence and increased values of FIB4 score is known to grow in the eventuality of there being detectable viremia in people with liver fibrosis [15].These findings provide further support to improve medical interventions and counseling with the aim of improving cART adherence in this population, minimizing the risk of virological failure.In contrast, protective factors for the occurrence of an efficacy event were: (a) previous cART not containing an INSTI.This finding seems to confirm the appropriateness of switching to an INSTI-based regimen to simplify regimens containing alternative drugs without compromising the virological control in line with the conclusions by Raffi F. et al. [6], who reported good efficacy outcomes after switching to INSTI regimens in virologically suppressed PLWH who were taking PI/r or NNRTI containing-regimens; (b) higher CD4+ T-cell count and higher CD4/CD8 ratio at baseline, in line with the findings of a study nested in a cohort of South African PLWH in which lower baseline CD4+ T-cell count was associated with a greater propensity toward virological failure, underlining the importance of preserving the immune system even in people prescribed well tolerated and effective drugs [16].
Concerning the metric of safety, we found as potential risk factors: (a) the backbone regimen containing emtricitabine plus tenofovir, suggesting the importance of taking into consideration the potential toxicity of NRTIs in composing the backbone of the cART regimens.For example, the risk of renal and bone toxicity associated with the use of TDF is well known [17,18]; (b) the condition of being cART naïve, probably, for the same reasons already mentioned for the metric of efficacy; (c) having higher HIV RNA at baseline, confirming the importance of maintaining HIV RNA control for prevention of the clinical events, even independently from the immunological status of the patient.For example, a previous study performed in the MaSTER cohort demonstrated how a longer delay from HIV diagnosis to HAART was an independent predictor of new AIDS-defining events and deaths.Thus, a longer time spent with lower HIV RNA plasma levels was found to be protective from clinical progression [19]; (d) eGFR at baseline below 60 mL/min and FIB4 score > 3.25, reflecting the importance of care for PLWH with kidney failure and advanced liver disease, ensuring regular checkups and medical interventions to avoid an unfavorable outcome.In contrast, protective factors for occurrence of safety events were: (a) previous cART not containing an INSTI, in line with what already discussed for the metric of efficacy and the findings of a review by Raffi F. et al., in which switching to INSTI regimens, in virologically suppressed PLWH, was associated with improved tolerability and greater reported patient satisfaction and outcomes compared to other ARV drug classes [6]; (b) higher CD4+ T-cell count and higher CD4/CD8 ratio at baseline, according to previous study showing the importance of having higher basal CD4+ T-cell count and a good CD4/CD8 ratio to avoid AIDS-defining events and deaths [19,20]; (c) an increased Framingham score at baseline, probably supporting the overall safety profile of using INSTI in PLWH with high risk of cardiovascular adverse events.A decreased risk of cardiovascular disease with the use of INSTI containing regimens compared to other ARV drug classes is described in a retrospective study of a large cohort of PLWH [21].By contrast, the findings from the RESPOND cohort consortium showed how INSTI initiation was associated with an early onset and excess incidence of cardiovascular disease in the first 2 years of exposure [22].These preliminary findings require analysis in larger randomized controlled trials to clarify the relationship between INSTI use and cardiovascular adverse events.
Since the metric of convenience constitutes a composite outcome of efficacy and safety, in general it shares the same risk and protective factors.However, the status of chronic HBsAg carrier at baseline was also found to be a risk factor for occurrence of a convenience event.It is well known that people with HIV and HBV co-infection with controlled plasma viral load still have an increased risk of liver disease progression, liver-related mortality, and overall mortality compared to people with either HIV or HBV alone [23][24][25].Moreover, a recent study demonstrated liver as an HIV reservoir in PLWH on cART, showing persistence of HIV DNA in hepatocytes of PLWH on antiretroviral therapy, resulting in a status of chronic inflammation that could lead to adverse liver outcomes [26].In addition to this, a positive association between liver fibrosis and suboptimal cART was observed in previous studies [15].Thus, close attention should be paid to co-infected patients in order to optimize therapy and implement appropriate medical interventions.
In the present cohort, durability appeared to work in favor of DTG based regimens with 58/371 (15.6%) participants stopping the initial regimen, especially if one considers that changing any drugs in the regimen was considered as an outcome.Moreover, durability events were mostly due to cART simplification which explained half of the interruptions, followed by treatment modification or interruption due to drug regimen toxicity, occurring in 29.3% of interruptions; this is in line with the results of previous studies [27] and suggests the effectiveness of a proactive treatment switch in the modern antiretroviral treatment era.Importantly, interruptions due to virological failure were few, accounting for only 1.1% (4/371) of total interruptions, confirming low rates of virological failure of DTG-based regimens.
When exploring potential risk factors associated with the metric of durability, we found that the following factors were associated with an increased risk of treatment interruption: (a) the backbone regimen containing tenofovir, mainly due to simplification strategies as shown in the KM estimates resulting from competitive risk analysis (Figure 5).We can hypothesize that this reflects the choice of removing tenofovir disoproxil fumarate to improve renal and bone toxicities [17,18].An alternative explanation could be related to treatment optimization by changing treatment to tenofovir alafenamide in those already on tenofovir disoproxil fumarate to prevent renal and bone deterioration, given the comparative renal safety advantage of TAF versus TDF in PLWH [28]; (b) the condition of being cART naïve, in apparent accordance to the findings of a multicenter Italian cohort in which treatment in naïve people showed a lower probability of maintaining DTG containing regimens at three and five years compared to treatment-experienced PLWH [27].Since, in our cohort, DTG containing regimens were mostly interrupted due to simplification strategies, we can suppose that naïve participants (as well as experienced ones) interrupted their regimen for simplification reasons, either to reduce pill burden or because of drug related toxicity or tolerability reasons; (c) having higher HIV RNA at baseline, confirming and supporting the findings already discussed about the metrics of efficacy, safety and convenience; (d) a FIB-4 at baseline above 3.25 compared to those whose FIB-4 was in the range from 1.45 to 3.25, confirming results of previous studies in which a positive association between liver fibrosis and suboptimal cART was observed [15] and encouraging physicians to start earlier cART [29,30]; (e) a baseline cancer diagnosis was associated with higher hazards of interruptions of DTG-containing therapies.It may be that antiretroviral therapy was optimized to minimize drug interactions and chemotherapy-related toxicities.It is known that PLWH are at increased risk of cancer when compared to the general population, particularly for malignancies driven by viral and bacterial co-infections, although they also have excess risk of infection-unrelated malignancies [31,32].Recent randomized data (START Trial) indicated that immediate cART initiation reduces risk of cancer during early HIV infection before the development of overt immunosuppression [33].Despite the increased toxicity and drug-drug interactions during cancer treatment, deferring cART during chemotherapy is unfavorable in PLWH with cancer and can lead to a poorer outcome; in general, any treatment interruption is not recommended during cancer treatment [34], with INSTI drugs appearing to be a good therapeutic choice in people with hematological malignances or those receiving various chemotherapeutic agents [35].In contrast, protective factors for durability of DTG containing regimens were found to be: (a) previous cART not containing an INSTI and previous cART containing other prior cART regimens that were composed of single class, triple class, non-boosted PI regimens and combinations including both INSTI and PI in which previous INSTI exposure accounted for only 8.1% (Table S1); (b) higher CD4+ T-cell count and CD4/CD8 ratio at baseline, confirming results observed for the metrics of efficacy, convenience and safety, and reflecting the importance of earlier diagnosis and cART introduction to restore or maintain a better immune status; (c) an increased Framingham score at baseline, probably indicating a good durability of DTG based regimens in participants with high cardiovascular risk, as we have already discussed for the metric of safety.
Lastly, in this cohort, no statistically significant associations were observed between the outcomes analyzed and the demographic (age, gender, and nationality) or epidemiological (heterosexual, PWID, MSM, or other risk factors for HIV transmission) patient characteristics.
This study is affected by several limitations.Firstly, for some parameters analyzed, there was a large confidence interval, due to fragmented data in the follow-up period; this was probably due to the COVID-19 pandemic which occurred concomitantly with enrollment and follow-up of participants, reducing the precision of our estimates.Secondly, since the study is not randomized, it is affected by the intrinsic limitations in any observational datasets, including possible confounding by indication biases.Third, although INSTI drugs, particularly DTG, showed better efficacy, safety and durability compared to PI drugs in improving lipid profiles [5,36,37], we did not evaluate metabolic and cardiological toxicities linked to this drug class [22,38].These adverse events need to be evaluated in the near future by larger studies, in order to implement educational and multidimensional interventions to prevent metabolic alterations in PLWH, especially for those with particular risk factors such as alcohol abuse, osteoporosis, previous AIDS events, and polypharmacy [39].Further, we did not assess changes of specific drugs in the study regimens, types of drugs in the regimens to which the participants were switched, or the outcome of these regimens after the switch; more studies are needed to address these points.Notwithstanding the above limitations, we feel that the prospective design, the prolonged follow-up (median: 556 days), the "real-life" conditions, and the multi-dimensional evaluation with several outcomes may provide interesting observations on the way DTG was used and how the regimens containing this drug should be optimized.
Not a statistically significant association;protective factor; antiretroviral treatment; INSTI, integrase strand inhibitors; HBsAg mated Glomerular Filtration Rate.

Table 1 .
Characteristics of the full study population.

Table 2 .
Hazards of treatment interruption in relation to each of the outcome events (univariate analysis).

Table 4 .
Summary of the risk and protective factors associated with the study metrics.

Table 4 .
Summary of the risk and protective factors associated with the study metr

containing regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis
Not a statistically significant association; protective factor; risk factor; cA antiretroviral treatment; INSTI, integrase strand inhibitors; HBsAg, hepatitis B ant mated Glomerular Filtration Rate.Concerning the metric of efficacy, our results suggest that potential Viruses 2023, 15, x FOR PEER REVIEW

Table 4 .
Summary of the risk and protective factors associated with

containing regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis
Not a statistically significant association; protective factor; antiretroviral treatment; INSTI, integrase strand inhibitors; HBsAg mated Glomerular Filtration Rate.Concerning the metric of efficacy, our results suggest t Viruses 2023, 15, x FOR PEER REVIEW

Table 4 .
Summary of the risk and protective factors a

containing regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis
Not a statistically significant association; prote antiretroviral treatment; INSTI, integrase strand inhi mated Glomerular Filtration Rate.Concerning the metric of efficacy, our resViruses 2023, 15, x FOR PEER REVIEW

Table 4 .
Summary of the risk and prote

Efficacy Tenofovir-containing regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis
Not a statistically significant associa antiretroviral treatment; INSTI, integra mated Glomerular Filtration Rate.

Table 4 .
Summary of the risk and protective factors associated with the study metr

containing regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis
Not a statistically significant association;protective factor; risk factor; cA antiretroviral treatment; INSTI, integrase strand inhibitors; HBsAg, hepatitis B ant mated Glomerular Filtration Rate.

Table 4 .
Summary of the risk and protective factors associated with

containing regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis
Not a statistically significant association;protective factor; antiretroviral treatment; INSTI, integrase strand inhibitors; HBsAg mated Glomerular Filtration Rate.Viruses 2023, 15, x FOR PEER REVIEW

Table 4 .
Summary of the risk and protective factors a

containing regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis
Not a statistically significant association; prote antiretroviral treatment; INSTI, integrase strand inhi mated Glomerular Filtration Rate.Viruses 2023, 15, x FOR PEER REVIEW

Table 4 .
Summary of the risk and prote

Table 4 .
Summary of the risk and protective factors associated with the study metr

containing regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis
Not a statistically significant association;protective factor; risk factor; cA antiretroviral treatment; INSTI, integrase strand inhibitors; HBsAg, hepatitis B ant mated Glomerular Filtration Rate.Concerning the metric of efficacy, our results suggest that potential

Table 4 .
Summary of the risk and protective factors associated with

containing regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis
Not a statistically significant association; protective factor; antiretroviral treatment; INSTI, integrase strand inhibitors; HBsAg mated Glomerular Filtration Rate.Concerning the metric of efficacy, our results suggest t Viruses 2023, 15, x FOR PEER REVIEW

Table 4 .
Summary of the risk and protective factors a

containing regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis
Not a statistically significant association; prote antiretroviral treatment; INSTI, integrase strand inhi mated Glomerular Filtration Rate.Concerning the metric of efficacy, our resViruses 2023, 15, x FOR PEER REVIEW

Table 4 .
Summary of the risk and prote

Table 4 .
Summary of the risk and protective factors associated with the study metr

Table 4 .
Summary of the risk and protective factors associated with

Table 4 .
Summary of the risk and prote

Table 4 .
Summary of the risk and protective factors associated with the study metr

regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis
Viruses 2023, 15, x FOR PEER REVIEW

Table 4 .
Summary of the risk and protective factors associated with

regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis
Viruses 2023, 15, x FOR PEER REVIEW

Table 4 .
Summary of the risk and protective factors a

regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis
Viruses 2023, 15, x FOR PEER REVIEW

Table 4 .
Summary of the risk and prote

regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis Baseline CD4+ T-cell count
Viruses 2023, 15, x FOR PEER REVIEW

Table 4 .
Summary of the risk and protective factors associated with the study metr

Table 4 .
Summary of the risk and protective factors associated with

Table 4 .
Summary of the risk and protective factors a

Table 4 .
Summary of the risk and prote

regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score Baseline cancer diagnosis
Viruses 2023, 15, x FOR PEER REVIEW

Table 4 .
Summary of the risk and protective factors associated with the study metr

containing regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score
Viruses 2023, 15, x FOR PEER REVIEW

Table 4 .
Summary of the risk and protective factors associated with

containing regimens Naïve status Prior cART not containing an INSTI HBsAg positive carriers Baseline HIV RNA Baseline CD4+ T-cell count Baseline CD4/CD8 ratio eGFR < 60 mL/min FIB4 score > 3.25 Baseline Framingham score
Viruses 2023, 15, x FOR PEER REVIEW

Table 4 .
Summary of the risk and prote

Table 4 .
Summary of the risk and protective factors associated with the study metr

Table 4 .
Summary of the risk and protective factors associated with

Table 4 .
Summary of the risk and protective factors a

Table 4 .
Summary of the risk and prote

Table 4 .
Summary of the risk and protective factors associated with the study metr

Table 4 .
Summary of the risk and protective factors associated with

Table 4 .
Summary of the risk and protective factors a

Table 4 .
Summary of the risk and prote

Table 4 .
Summary of the risk and protective factors a

Table 4 .
Summary of the risk and prote

Table 4 .
Summary of the risk and protective factors associated with the study metr

Table 4 .
Summary of the risk and protective factors associated with

Table 4 .
Summary of the risk and protective factors a

Table 4 .
Summary of the risk and prote

Table 4 .
Summary of the risk and protective factors associated with the study metrics.

Table 4 .
Summary of the risk and protective factors associated with the study metrics.

Table 4 .
Summary of the risk and protective factors associated with th