Matching Cancers With the Best Chemical Treatments

Roland Piquepaille writes "When oncologists meet a new patient affected by a cancer, they have to take decisions about the best possible treatment. Now, U.S. researchers have devised an algorithm which matches tumor profiles to best treatments. They've used a panel of 60 diverse human cancer cell lines from the National Cancer Institute — called NCI-60 — to develop their "coexpression extrapolation (COXEN) system." As said one researcher, "we believe we have found an effective way to personalize cancer therapy." Preliminary results have been encouraging and clinical trials are now planned."

Comments

Doctors generally won't like this

[syousef]

Most doctors won't even use computers to help them make diagnoses because they feel they should always be able to do better. What tends to happen is that if a rare condition presents they can miss it quite easily. I'm no doctor but I believe it has to do with the medical profession's heritage, culture and the politics of their licensing institutions. Doctors are taught that every diagnosis can be life or death. Using an aid like a computer to make the decision therefore is seen as a sign of weakness.

When you think about it that's insane. There's no way any doctor can know every medical condition that presents, even the rarer ones. What's needed is a system whereby the doctor can check his diagnosis against what comes up with a computer search against the same symptoms. There needs to be no stigma in doing this. If something comes up that's rare but could fit the doctor then needs to have a think about whether it's worth addressing. Systems like this have been rejected by the medical profession time and again which is unfortunate because to get good at diagnosis they'd need to be honed with a lot of feedback, particularly where multiple conditions present. However they have the potential to help pick up serious conditions earlier than what even the best doctor might without them.

Same goes for this system except we're talking treatment choice not diagnosis. One hurdle is getting other doctors to accept it. Another is making sure the control and final say remains with the doctor and patient not some machine. There'd be great temptation for the medical insurers to use such a system to avoid providing treatment that a doctor believes is necessary.

Re:Doctors generally won't like this

[bwen]

As a physician, I resent your inaccurate and uninformed response. "Doctors are taught that every diagnosis can be life or death" - where did you hear that? You are making sweeping generalizations and accusations. The ASSUMPTION that physicians resist using a computer to research a medical problem is ridiculous (at least in the US.) I do not know a MD that is not comfortable with a computer nor with researching a medical problem online. We often have resources that the general public does not use, and due to lack of an additional 7-11 years of post-grad training, would not understand. You seem to typify the person that turns to herbs from China that mostly consist of grass/dirt and expound how modern science is ignoring it. We very much appreciated you in the dark ages, thanks for your insight!

Algorithm kindly sponsored by Pfizer

[Anonymous Coward]

int drug_choice_algorithm(){
   int our_most_expensive_drug = 1;
   int other_cheaper_option = 0;

   if(patient_has_insurance()){
       return our_most_expensive_drug;
   }
   else {
       if (patient_is_rich()){
           return our_most_expensive_drug;
       }
   }
   return our_most_expensive_drug;
}