AI-Based OCT Systems to Detect Diabetic Macular Edema, Low Back Pain Review, Healthy Dialogue on Early-Onset Cancer, and more

Today I'm joined by JAMA Deputy Editor, Dr. Preethi Malani. Let's jump straight into this week's research. So I have the first. For adults with pre-diabetes, We know that preventing not only diabetes, but also related comorbidities is essential for healthy aging. In a 21-year observational follow-up Study by Dr. Saleev and colleagues from the Diabetes Prevention Program or DPP research group, investigators tracked over a thousand U.S. Medicare participants with pre diabetes.

These participants from the original DPP study were randomly assigned to either an intensive lifestyle intervention, metformin, or placebo. The lifestyle intervention focused on reduced calories and fat, as well as 150 minutes of physical activity weekly. The team in this study specifically asked whether these early interventions could reduce the risk of chronic diseases like hypertension, kidney disease, COPD, depression, and others later in life.

At the end of the twenty one year follow up, eighty five percent of participants had developed at least two chronic diseases, with a median number of conditions at five per person. the risk of developing multimorbidity was twenty one percent lower in those in the lifestyle intervention group compared to placebo. But there was no difference in risk seen between metformin and placebo.

So, based on this study for adults with prediabetes, a sustained focus on healthy diet and exercise remains powerful tools for curbing the domino effect of chronic diseases as we age.

We need dedicated infrastructure and coverage to deliver lasting behavior change, especially for older, underserved populations who may be at higher risk. They also highlight that this benefit was not seen for metformin And while the door remains open for newer drug therapies, diet and exercise is still the core. Prevention pays off not just for diabetes, but also for the broader problem of multiple chronic illnesses.

Our next original investigation touches on the growing importance of AI in medicine, this time for diabetic macular edema, or DME, which is a leading cause of vision loss among adults with diabetes. Screening programs commonly use retinal photos, but those can have a high false positive rate, meaning many people get referred unnecessarily for confirmatory testing. This can create burdens for both specialty clinics and patients.

In this prospective trial conducted in Hong Kong, doctor Shang and colleagues tested whether an AI powered optical coherence tomography. or AIOCT, can be integrated into real world diabetic retinopathy screening to help triage patients for DME. After validating the system in more than six hundred patients,

The investigators conducted a randomized trial. Two hundred and seventy-six people with suspected DME from standard screening were either referred for specialist evaluation based only on the photographs. or using the AI O C T to help refine referral decisions. The key finding was that using AI O C T decreased unnecessary referrals from about 69% down to 24% without missing true disease.

In fact, no cases of DME occurred among non-referred participants in the AI pathway. The authors suggest that integrating AI OCT as a second step after photo based screening could meaningfully reduce unnecessary eye clinic visits. freeing up resources and saving patients the time and expense of clinic visits without compromising safety.

So let's switch gears now from adult outpatients to hospitalized infants, in this case with a study in acute bronchiolitis. Prone positioning has shown some benefits in adults with acute respiratory failure. But its utility in infants with respiratory distress is less clear. The propositist randomized trial led by doctor Bowden and colleagues enrolled four hundred and fifty one infants with acute bronchiolitis.

ages six months or younger, across fifteen French hospitals. Infants had moderate to severe respiratory distress, requiring high flow nasal cannula, and were assigned to prone versus the usual supine position during the first forty eight hours of their stay. The primary outcome was escalation to higher levels of respiratory support within seventy two hours.

Escalation occurred in about 18% overall, 15% in the prone group, and 21% in the supine group, a difference that did not reach statistical significance. Serious adverse events were extremely rare in both groups. So, prone positioning did not seem to significantly reduce escalations of care for infants with moderate to severe bronchiolitis.

However, it's important to note that a large proportion of infants in the study were not able to complete the proning intervention, something the investigator did not anticipate. and a factor that may have limited the study's ability to identify a significant impact for the intervention.

Next up, a question for oncologists. When patients with EGFR variant non small cell lung cancer progress after receiving tyrosine kinase inhibitors, Options are limited, and overall outcomes tend to be very poor. Well a new approach, a biphasic antibody called Ivanesimab that targets both GD1 and VEGF.

extend survival when added to standard chemotherapy. The Harmony A phase three trial addressed this by randomly assigning three hundred and twenty two patients in China with locally advanced or metastatic EGFR variant non squamous, non small cell lung cancer, Who had received prior EGFR tyrosine kinase inhibitors to chemotherapy with or without ivanesimab. Over a median follow-up of more than thirty two months.

Adding the antibody improved median overall survival from fourteen point one months to sixteen point eight months. a two point seven month absolute gain. The survival curves separated early and stayed apart at two years and beyond, with thirty percent of patients alive at thirty months in the Ivanessa MAB group versus eighteen percent with chemotherapy alone.

Importantly, about two thirds of patients receiving Ivanesimab experience severe treatment emergent adverse events, compared with just over half in the chemotherapy alone group. Bottom line, Ivanessim Mab plus chemotherapy produced a statistically significant survival benefit in this challenging to treat lung cancer population, although the absolute difference was modest.

Progression on EGFR TKI therapy while weighing benefits and risks carefully. Let's turn to two research letters that relate to distributions of healthcare resources in the US.

Although the laws aimed to boost placements in rural and primary care specialty, only two to six percent of the new slots actually went to rural institutions on any given round, and that proportion dropped over time. The proportion of new slots designated for primary care also fell over time.

The upshot? These federal expansions may not be closing the urban rural gap or addressing the primary care shortage as intended. The authors write that more targeted policy action likely is needed to truly address these distribution issues.

While pandemic era financial support was linked to fewer closures, The authors highlight an ongoing tension between striving for higher standards and preserving access to care, especially for patients in rural nursing home deserts.

Outpacing the US FDA's ability to monitor and enforce, the authors call for stronger federal oversight, coordinated digital platform enforcement, and better research to track harms, especially as influencer driven use grows among young people.

helps guide therapy, and improves shared decision making. The binary approach, by contrast, compresses a continuum and misses valuable cues for prognosis and patient counseling. These authors argue precision memory care means moving beyond binary results and interpreting the whole clinical context.

Risk factors include older age, obesity, depression, occupational strain, tobacco use, and other chronic illnesses. And a history of prior back pain. Most low back pain gets better on its own, with about 70% recovered at 12 months, but chronic pain is less likely to resolve.

First line management is reassurance, education, advice to stay active, and supporting self-management. Prolonged rest should be avoided. For acute pain, superficial heat, Spinal manipulation, massage, acupuncture, NSAIDs, and skeletal muscle relaxants offer short term relief. For chronic pain, structured exercise, psychological therapy, and integrated multidisciplinary care are the best evidence based options.

Medications for chronic pain are second line. Opioids, spinal injections, and surgery should be rare and reserved for severe refractory cases. Here's a bit more about medications for acute low back pain from this week's JAMA Clinical Reviews Podcast with Review Coauthor doctor James McAuley.

Acetamenophrine alone doesn't appear to be effective for acute low back pain, and I think this is a result of a trial that was published in the mid two thousand tens that's only on the basis of one trial, but all guidelines now. suggest that acetomenophrine is not effective on its own for low back pain. Benzodiazepines should be avoided.

And the recent Opel trial also demonstrated that opioids were not effective compared to placebo when added to guideline care. They also increased the risk of adverse events and later increased by a small amount opioid misuse.

Two, it's still important to follow a healthy diet and lifestyle, as Alana was saying, because even though I mentioned that so many of our patients do and they get diagnosed anyway, it is still really important to follow those healthy habits starting from early childhood. And three, and we didn't really touch on this a lot, but I think many people don't know their family history. And so it is really important for primary care doctors to ask.

for people to talk to their family members about whether there is a history of colorectal cancer, for example, or colon polyps, because that could change the age at which you begin screening and could be potentially life-saving.

A link to this JAMA content can be found in this episode's show notes. To follow this and other JAMA Network podcasts, please visit us online at JAMA NetworkAudio.com. Thank you for listening.