[gdpawel]

There isn't one paper, or two, which by itself, makes a case for or against cell-based assays. Nor does the proposition that the whole thing depends on one study or even one review. You've got to consider the body of literature as a whole.

The fact that none of this exists as one neat, convenient paper in the New England Journal of Medicine does not, in any way, negate the existence of this body of information. It has been found that newer methods of "cell-death" assays have an overall predictive accuracy of 98.2% concerning treatment response, which compares favorably with older, previously published data ranging from 75% to 92%. (Staib,P.et al. Br J Haematol 128 (6):783-781, March 2005)

We have tests such as estrogen receptor, progesterone receptor, Her2/neu, BCR-ABL, C-KIT, CD-20, etc., and panels of immunohistochemical stains for subclassifying tumors. All of these tests are used to select chemotherapy in precisely the same manner as cell culture assay tests are used.

Also, we have the use of additional medical tests, such as serial CT, MRI, and PET scans, performed for the purpose of monitoring the size of the tumor to determine if it is shrinking or growing with chemotherapy. The purpose of this testing is to determine if chemotherapy with specific drugs should be continued or changed to different drugs. These radiographic tests are also used as an aid in making clinical decisions about the choice of chemotherapy.

So yes, there is precedent for using cell culture assays.

The June issue of Oncology News International (June 2010, V 19, No 6) quotes a Duke University study of the use of high-tech cancer imaging, with one representative finding being that the average Medicare lung cancer patient receives 11 radiographs, 6 CT scans, a PET scan, and MRI, two echocardiograms, and an ultrasound, all within two years of diagnosis. A study co-author (Dr. Kevan Schulman) asks: "Are all these imaging studies essential? Are they all of value? Is the information really meaningful? What is changing as a result of all this imaging?"

Why is it that oncologists are so accepting of high tech, expensive imaging studies, yet so reluctant to consider the use of cell culture diagnostic tests? For one thing, clinical trials virtually always have time to disease progression as a primary endpoint. Without the imaging studies, one can't get accurate time to progression data. So these are tests performed for the benefit of drug companies seeking new drug approval, for clinical investigators seeking contracts and publications, and for clinicians seeking an easy way to make clinical decisions (and, occasionally, seeking income enhancement).

In the absence of information provided by cell culture testing, oncologists have complete freedom to choose between a myriad of drug regimens. The proven basis on which they make these selections, by and large, is on the benefit a given regimen provides to the oncologist (or academic institution). Cell culture testing threatens this freedom of choice. There's absolutely nothing in it for the oncologist or academic medical center (unlike, for example, imaging studies).