Machine learning outperformed logistic regression classification even with limit sample size: A model to predict pediatric HIV mortality and clinical progression to AIDS.

Logistic regression (LR) is the most common prediction model in medicine. In recent years, supervised machine learning (ML) methods have gained popularity. However, there are many concerns about ML utility for small sample sizes. In this study, we aim to compare the performance of 7 algorithms in the prediction of 1-year mortality and clinical progression to AIDS in a small cohort of infants living with HIV from South Africa and Mozambique. The data set (n = 100) was randomly split into 70% training and 30% validation set. Seven algorithms (LR, Random Forest (RF), Support Vector Machine (SVM), K-Nearest Neighbor (KNN), Naïve Bayes (NB), Artificial Neural Network (ANN), and Elastic Net) were compared. The variables included as predictors were the same across the models including sociodemographic, virologic, immunologic, and maternal status features. For each of the models, a parameter tuning was performed to select the best-performing hyperparameters using 5 times repeated 10-fold cross-validation. A confusion-matrix was built to assess their accuracy, sensitivity, and specificity. RF ranked as the best algorithm in terms of accuracy (82,8%), sensitivity (78%), and AUC (0,73). Regarding specificity and sensitivity, RF showed better performance than the other algorithms in the external validation and the highest AUC. LR showed lower performance compared with RF, SVM, or KNN. The outcome of children living with perinatally acquired HIV can be predicted with considerable accuracy using ML algorithms. Better models would benefit less specialized staff in limited resources countries to improve prompt referral in case of high-risk clinical progression.

Optimizing the World Health Organization algorithm for HIV vertical transmission risk assessment by adding maternal self-reported antiretroviral therapy adherence.

BACKGROUND: The World Health Organization (WHO) risk assessment algorithm for vertical transmission of HIV (VT) assumes the availability of maternal viral load (VL) result at delivery and early viral control 4 weeks after initiating antiretroviral treatment (ART). However, in many low-and-middle-income countries, VL is often unavailable and mothers' ART adherence may be suboptimal. We evaluate the inclusion of the mothers' self-reported adherence into the established WHO-algorithm to identify infants eligible for enhanced post-natal prophylaxis when mothers' VL result is not available at delivery. METHODS: We used data from infants with perinatal HIV infection and their mothers enrolled from May-2018 to May-2020 in Mozambique, South Africa, and Mali. We retrospectively compared the performance of the WHO-algorithm with a modified algorithm which included mothers' adherence as an additional factor. Infants were considered at high risk if born from mothers without a VL result in the 4 weeks before delivery and with adherence <90%. RESULTS: At delivery, 143/184(78%) women with HIV knew their status and were on ART. Only 17(12%) obtained a VL result within 4 weeks before delivery, and 13/17(76%) of them had VL ≥1000 copies/ml. From 126 women on ART without a recent VL result, 99(79%) had been on ART for over 4 weeks. 45/99(45%) women reported suboptimal (< 90%) adherence. A total of 81/184(44%) infants were classified as high risk of VT as per the WHO-algorithm. The modified algorithm including self-adherence disclosure identified 126/184(68%) high risk infants. CONCLUSIONS: In the absence of a VL result, mothers' self-reported adherence at delivery increases the number of identified infants eligible to receive enhanced post-natal prophylaxis.

Dolutegravir as First- or Second-Line Treatment for HIV-1 Infection in Children.

BACKGROUND: Children with human immunodeficiency virus type 1 (HIV-1) infection have limited options for effective antiretroviral treatment (ART). METHODS: We conducted an open-label, randomized, noninferiority trial comparing three-drug ART based on the HIV integrase inhibitor dolutegravir with standard care (non-dolutegravir-based ART) in children and adolescents starting first- or second-line ART. The primary end point was the proportion of participants with virologic or clinical treatment failure by 96 weeks, as estimated by the Kaplan-Meier method. Safety was assessed. RESULTS: From September 2016 through June 2018, a total of 707 children and adolescents who weighed at least 14 kg were randomly assigned to receive dolutegravir-based ART (350 participants) or standard care (357). The median age was 12.2 years (range, 2.9 to 18.0), the median weight was 30.7 kg (range, 14.0 to 85.0), and 49% of the participants were girls. By design, 311 participants (44%) started first-line ART (with 92% of those in the standard-care group receiving efavirenz-based ART), and 396 (56%) started second-line ART (with 98% of those in the standard-care group receiving boosted protease inhibitor-based ART). The median follow-up was 142 weeks. By 96 weeks, 47 participants in the dolutegravir group and 75 in the standard-care group had treatment failure (estimated probability, 0.14 vs. 0.22; difference, -0.08; 95% confidence interval, -0.14 to -0.03; P = 0.004). Treatment effects were similar with first- and second-line therapies (P = 0.16 for heterogeneity). A total of 35 participants in the dolutegravir group and 40 in the standard-care group had at least one serious adverse event (P = 0.53), and 73 and 86, respectively, had at least one adverse event of grade 3 or higher (P = 0.24). At least one ART-modifying adverse event occurred in 5 participants in the dolutegravir group and in 17 in the standard-care group (P = 0.01). CONCLUSIONS: In this trial involving children and adolescents with HIV-1 infection who were starting first- or second-line treatment, dolutegravir-based ART was superior to standard care. (Funded by ViiV Healthcare; ODYSSEY ClinicalTrials.gov number, NCT02259127; EUDRACT number, 2014-002632-14; and ISRCTN number, ISRCTN91737921.).

ODYSSEY clinical trial design: a randomised global study to evaluate the efficacy and safety of dolutegravir-based antiretroviral therapy in HIV-positive children, with nested pharmacokinetic sub-studies to evaluate pragmatic WHO-weight-band based dolutegravir dosing.

BACKGROUND: Dolutegravir (DTG)-based antiretroviral therapy (ART) is highly effective and well-tolerated in adults and is rapidly being adopted globally. We describe the design of the ODYSSEY trial which evaluates the efficacy and safety of DTG-based ART compared with standard-of-care in children and adolescents. The ODYSSEY trial includes nested pharmacokinetic (PK) sub-studies which evaluated pragmatic World Health Organization (WHO) weight-band-based DTG dosing and opened recruitment to children < 14 kg while dosing was in development. METHODS: ODYSSEY (Once-daily DTG based ART in Young people vS. Standard thErapY) is an open-label, randomised, non-inferiority, basket trial comparing the efficacy and safety of DTG + 2 nucleos(t) ides (NRTIs) versus standard-of-care (SOC) in HIV-infected children < 18 years starting first-line ART (ODYSSEY A) or switching to second-line ART (ODYSSEY B). The primary endpoint is clinical or virological failure by 96 weeks. RESULTS: Between September 2016 and June 2018, 707 children weighing ≥14 kg were enrolled; including 311 ART-naïve children and 396 children starting second-line. 47% of children were enrolled in Uganda, 21% Zimbabwe, 20% South Africa, 9% Thailand, 4% Europe. 362 (51%) participants were male; median age [range] at enrolment was 12.2 years [2.9-18.0]. 82 (12%) children weighed 14 to < 20 kg, 135 (19%) 20 to < 25 kg, 206 (29%) 25 to < 35 kg, 284 (40%) ≥35 kg. 128 (18%) had WHO stage 3 and 60 (8%) WHO stage 4 disease. Challenges encountered include: (i) running the trial across high- to low-income countries with differing frequencies of standard-of-care viral load monitoring; (ii) evaluating pragmatic DTG dosing in PK sub-studies alongside FDA- and EMA-approved dosing and subsequently transitioning participants to new recommended doses; (iii) delays in dosing information for children weighing 3 to < 14 kg and rapid recruitment of ART-naïve older/heavier children, which led to capping recruitment of participants weighing ≥35 kg in ODYSSEY A and extending recruitment (above 700) to allow for ≥60 additional children weighing between 3 to < 14 kg with associated PK; (iv) a safety alert associated with DTG use during pregnancy, which required a review of the safety plan for adolescent girls. CONCLUSIONS: By employing a basket design, to include ART-naïve and -experienced children, and nested PK sub-studies, the ODYSSEY trial efficiently evaluates multiple scientific questions regarding dosing and effectiveness of DTG-based ART in children. TRIAL REGISTRATION: NCT, NCT02259127 , registered 7th October 2014; EUDRACT, 2014-002632-14, registered 18th June 2014 ( https://www.clinicaltrialsregister.eu/ctr-search/trial/2014-002632-14/ES ); ISRCTN, ISRCTN91737921 , registered 4th October 2014.