The tragedy of autism is compounded by one fact that makes desperate parents wish they could turn back the hands of time: symptoms of the neurodevelopmental disorder typically show up when a child is 2 or 3 or even older, but by then it may be too late to prevent or reverse whatever glitches in brain development (still pretty much a mystery) underlie the disease. It is even on the late side for getting a child the behavioral interventions and special education that might mitigate some of the worst symptoms.

If scientists at the M.I.N.D. Institute of the University of California, Davis, are right, however, there may be a reliable warning sign of autism much earlier: how a child plays with his or her toys at the tender age of 12 months. In particular, scientists led by Sally Ozonoff will report in the journal Autism (it’s the October issue, but not out yet; keep checking the web site), children who were later diagnosed with autism were more likely to spin, repetitively rotate, stare at and look out of the corners of their eyes at toys such as a rattle.

There is a big research effort aimed at picking up the earliest harbingers of autism. One of the most promising discoveries came in 2003, when researchers led by neuroscientist Eric Courchesne of the University of California, San Diego, concluded that an odd pattern of skull growth might be a tip to autism, as they described in a paper in the Journal of the American Medical Association. Children with autism, the scientists found, had a smaller head circumference at birth than healthy babies, and by 6 to 14 months their head circumference was in the 84th percentile, a huge increase and greater than the rate of increase in healthy children. “The clinical onset of autism appears to be preceded by 2 phases of brain growth abnormality: a reduced head size at birth and a sudden and excessive increase in head size between 1 to 2 months and 6 to 14 months,” the scientists wrote. “Abnormally accelerated rate of growth may serve as an early warning signal of risk for autism.” Still, the correlation wasn’t perfect: 6% of healthy infants in the study also showed abnormal head growth from birth to 6 to 14 months, and 41% of babies later diagnosed as autistic didnot show that pattern.

The American Academy of Pediatrics recommends that all infants be screened for autism twice before they are 2. Pediatricians look for language delays and lack of interest in people, such as not responding to their name and failing to make eye contact. But these can be present even when autism is not. The latest findings are not perfect either, but they are something parents can watch for every day rather than relying on—and waiting for—a short visit to the doctor. “There is an urgent need to develop measures that can pick up early signs of autism, signs present before 24 months,” Ozonoff says. “The finding that the unusual use of toys is also present early in life means that this behavior could easily be added to a parent check-list.”

For the study, Ozonoff recruited 66 1-year-olds; 9 were later diagnosed with autism. The children were given a metal lid, a round plastic ring, a rattle and a baby bottle, one at a time for 30 seconds each while being videotaped. Seven of the 9 later diagnosed with autism were more likely to repeatedly spin and rotate the objects. They were also more likely to look at them in unusual ways, like glancing sideways at them or staring intently at them for a long time—behaviors that were rare in babies not later diagnosed with autism. “About a third of parents notice signs [of autism] before a child’s first birthday,” Ozonoff said. “We felt that our field could do a better job at early diagnosis. Our results suggest that these particular behaviors might be useful to include in screening tests. The earlier you treat a child for autism, the more of an impact you can have on that child’s future.”

It’s no secret that oncologists and cancer researchers have made pitifully little progress against lung cancer, even compared to the less-than-stellar progress against other cancers, as I explained in a recent story. But as the Lung Cancer Alliance, a patient-advocacy group, details in its annual report card—filled with F’s—the news is so grim that it is impossible to avoid the conclusion that we need a whole new approach to this tragic disease.

Some of the highlights—or lowlights—include the fact that lung cancer remains the nation’s #1 cause of cancer death: it will claim another 162,000 lives by the time 2008 ends, triple the number of deaths from prostate cancer, nearly twice as many deaths as from breast cancer and more than from breast, prostate, colon, leukemia, ovarian and cervical cancercombined . Nor has there been any progress on five-year survival: only 15% of patients make it to this milestone, compared to 89% for breast cancer, 99% for prostate cancer and 65% for colon cancer. Because there is no good early-screening technique, only 16% of lung cancers are diagnosed at an early, potentially-curable stage. That compares to 61% for breast cancers, 91% of prostate cancers and 39% of colon cancers.

Why is progress essentially nonexistent? Lung cancer is particularly tricky in terms of its genetics, with malignant cells amassing countless mutations that, among other things, allow them to outfox both standard chemotherapies and newer targeted molecular drugs including angiogenesis inhibitors. There’s nothing we can do about that basic biology.

But oncologists tell me there is another reason: there is such stigma attached to lung cancer, because of the view that only people stupid enough to smoke develop it and it is therefore their own dumb fault, that research funding remains criminally tight, especially compared to “celebrity” cancers such as breast and prostate. Lung cancer doesn't have the celebrity spokespeople that these cancers do, so less money is raised from the public. it doesn't have anything like the pink ribbons and "runs for the cure" of breast cancer, for the same reason. In other words, it's not even a fair fight.