Five PubMed papers for August 4–11: PSMAddition leads on early evidence

Five papers newly entered PubMed between August 4, 09:00 and August 11, 09:00 (UTC-05:00). The order combines an early OpenAlex cited-by snapshot retrieved on August 11 with journal tier; because these papers are only days old, study design, endpoint strength, sample size, and clinical usefulness break ties. The early citation counts are a signal, not a stable leaderboard.

The common thread is practical rather than thematic: each paper tests a decision clinicians or health systems could act on, but each leaves a different uncertainty—overall survival, implementation, toxicity, durability, or external validity.

Journal and design. The Lancet published the second interim analysis of PSMAddition, an ongoing open-label phase 3 randomized superiority trial conducted at 169 sites in 20 countries. It enrolled 1,144 men with PSMA-positive metastatic androgen-pathway-modulator-naive or -sensitive prostate cancer. Patients received ADT plus an androgen-receptor pathway inhibitor, with or without up to six cycles of intravenous lutetium-177 PSMA-617. The PubMed record entered the database on August 7, 2026, and lists Novartis as the funder. 5

Primary result. Radiographic progression or death occurred in 24% versus 30% of patients, favoring the radioligand combination: HR 0.72 (95% CI 0.58–0.90; P=0.0021). Median radiographic progression-free survival was not reached in either group. The trade-off was more grade 3 or higher adverse events with lutetium-177 PSMA-617, 51% versus 43%; dry mouth occurred in 46% versus 4%, although all reported cases were grade 1 or 2. 6

Why it matters. This is a large, biomarker-selected phase 3 signal that radioligand therapy can move radiographic disease control earlier in metastatic hormone-sensitive disease. It is not yet an overall-survival answer: the reported endpoint is radiographic progression-free survival, the trial is ongoing, and control-arm crossover can complicate later comparisons.

People and funding. First author Scott T. Tagawa is affiliated with Weill Cornell Medicine; the corresponding author, Michael J. Morris, is affiliated with Memorial Sloan Kettering Cancer Center. The study was funded by Novartis. 5

Read the original paper on PubMed or via the DOI landing page.

Journal and design. ICECAP was a multicenter adaptive-allocation randomized clinical trial across 71 US hospitals. It randomized 1,158 comatose adults after out-of-hospital cardiac arrest to therapeutic hypothermia at 33 °C for durations ranging from 6 to 72 hours. The primary endpoint was 90-day neurological function measured with a weighted modified Rankin Scale and analyzed with a Bayesian duration-response model. PubMed indexed the JAMA paper on August 5, 2026. 7

Primary result. In patients with nonshockable rhythms, the posterior probability that 6 hours was the shortest duration achieving the best mean neurological outcome was 0.51; the shockable-rhythm findings were similar. The trial stopped at a prespecified interim rule. No differences were observed in secondary outcomes or mortality across cooling durations. 8

Why it matters. The clinically useful conclusion is narrower than "six hours is the new standard": extending cooling beyond the shortest duration that performed best did not improve outcomes in this trial, and the 0.51 posterior probability is not a definitive optimum. The result directly challenges the assumption that more prolonged hypothermia automatically adds neuroprotection.

People and funding. First author William J. Meurer is affiliated with the University of Michigan and its SIREN Clinical Coordinating Center. The PubMed record does not state a trial funder; its disclosure section reports NIH and NINDS support to several investigators for related or outside work, so that should not be read as proof that NIH funded the trial itself. 7

Read the original paper on PubMed or via the DOI landing page.

Journal and design. This multicenter, open-label phase 3 randomized trial enrolled 294 patients with treatment-naive stage IV or recurrent nonsquamous NSCLC carrying both an EGFR-sensitizing mutation and a TP53 mutation at 17 sites in China. Patients received osimertinib plus pemetrexed-carboplatin followed by osimertinib plus pemetrexed, or osimertinib alone. PubMed entered the paper on August 10, 2026. 9

Primary result. Median progression-free survival was 34.0 versus 15.6 months with combination therapy versus osimertinib alone, a difference of 18.4 months (95% CI 9.9–22.3); HR 0.44 (95% CI 0.32–0.60; P<0.001). The effect was consistent in prespecified subgroups, including patients with brain metastases and those with L858R mutations. Overall-survival data were immature at 30.6% maturity, and grade 3 or higher treatment-related adverse events were more frequent with combination therapy. 10

Why it matters. The trial gives a concrete rationale for intensifying first-line treatment in the molecular subgroup most likely to relapse on osimertinib alone. It does not yet establish an overall-survival benefit, and the higher toxicity burden means the PFS gain still has to be weighed against patient fitness, competing risks, and the value of chemotherapy.

People and funding. First author Ting Zhou is affiliated with Sun Yat-sen University Cancer Center in Guangzhou. The PubMed record retrieved for this issue does not report a funding source. 9

Read the original paper on PubMed or via the DOI landing page.

Journal and design. This multicenter, double-blind, placebo-controlled randomized trial assigned 171 adults with PTSD and recurrent nightmares to dronabinol 2.5–15 mg or placebo once daily before bedtime for 10 weeks. The primary endpoint was the CAPS-IV B2 item for nightmare frequency and intensity. PubMed indexed the Nature Medicine paper on August 6, 2026. 11

Primary result. At week 10, dronabinol improved the CAPS-IV B2 score more than placebo by −1.50 points (95% CI −2.28 to −0.71; Cohen's d=0.65; P<0.001). Adverse events occurred in 89.7% versus 77.1%; discontinuation because of adverse events was similar at 5.7% versus 7.2%. Serious adverse events occurred in 8.0% of the dronabinol group and 0% of the placebo group. 12

Why it matters. The effect size is large enough to justify a closer look at the full paper, but the serious-adverse-event imbalance makes this a treatment signal rather than a practice shortcut. The study lasted 10 weeks; it does not answer whether benefit persists, whether dose reduction changes tolerability, or how dronabinol compares with established nightmare treatments.

People and funding. First and corresponding author Stefan Roepke is affiliated with Charité–Universitätsmedizin Berlin and Oberberg Fachkliniken in Berlin. The PubMed record does not identify a single trial funder; disclosures list support from German and European public research bodies for several authors. 11

Read the original paper on PubMed or via the DOI landing page.

Journal and design. This pragmatic, two-arm randomized trial enrolled 555 people aged 16 years or older with newly diagnosed drug-susceptible tuberculosis across four public hospitals in Buenos Aires. The intervention combined a Spanish-language mobile app, daily adherence reporting, bidirectional messaging, education, and weekly urine-based isoniazid verification. Participants and caregivers knew their assignment, while data collectors and analysts were masked. PubMed indexed the BMJ paper on August 7, 2026. 13

Primary result. In the intention-to-treat analysis, treatment success was 82% versus 74% with TB-TST plus standard care versus standard care alone: risk difference 7.1% (95% CI 1.1%–14.1%), risk ratio 1.10 (95% CI 1.00–1.20; P=0.04). Loss to follow-up was 17% versus 24% (risk ratio 0.70, 95% CI 0.50–0.98; P=0.04). 14

Why it matters. This is one of the week's most immediately actionable health-system studies: a digital intervention that links self-reporting to human treatment support, rather than relying on reminders alone. The trial excluded people with drug-resistant disease, HIV coinfection, severe illness, or no practical smartphone access. Recruitment and care were also disrupted by COVID-19, so implementation in other settings needs a cost and workflow assessment rather than a simple app rollout.

People and funding. First author Fernando Rubinstein is affiliated with the Instituto de Efectividad Clinica y Sanitaria in Buenos Aires; the University of Washington contributed the biobehavioral nursing and health-informatics team. The study was funded by NIH/NIAID grant 1R01AI147129-01, and the authors state that the funder had no role in design, implementation, or the decision to submit. 14

Read the original paper on PubMed or via the DOI landing page.

For a clinician, PSMAddition and ICECAP offer the clearest near-term changes to how treatment intensity is chosen, but neither closes its main long-term endpoint. For a thoracic-oncology reader, the osimertinib paper is the most direct molecularly stratified treatment comparison. The dronabinol trial is worth reading with its safety table open. The TB-TST paper is the one to open if the question is implementation: its intervention depends as much on staff response and patient access as on the app itself.