Science, Numbers, Journalism and Critical Thinking

I believe it was on a Merlin Mann podcast, where someone jokingly mentioned the concept of an iColon website which in typical Web 2.0 fashion people could share the images and films of their colonoscopy procedures. This week the New York Times managed to demonstrate where their heads were at regarding simple mathematics and statistics. An article by one of their most popular reporters reviewed a new study that showed that colonoscopies were not all they were cracked up to be as far as picking up colon cancer and in their ability to prevent future cancers from developing from detection and removal of polyps.

The American Cancer Society still strongly recommends the colonoscopy as a useful  screening procedure. and the physicians I know who both perform this procedure and recommend it all strongly feel that it is one procedure that can actually save lives. We will however take a quick and simple look at the use of math and see a lack of simple math skills and the failure of editorial checking at the New York Times.

The NY Times reported on a study that demonstrated that colonoscopy missed virtually all polyps on the right side of the colon. The  article stated that 40% of all cancers and polyps originated on this side. Of those on the left side of the colon up to 30% of those were missed. The article then went on to quote and conclude that rather than preventing up to 90% of colon cancer from devleloping it might only be useful in preventing 60% to 70%. Now, this seems to be a fairly simple math problem. And a quick glance by anyone even one with only minimal math skills just looks and sounds wrong on the surface of it all.

If 40% of the lesions occur on the left side but are undetectable, than you have remaining 60% of the lesions to deal with that you can find. If 1/3 of these ones are not found, by splitting them into 3 equal parts of 20% 20% and 20%, and you remove 1/3 you are then left with 40% of the lesions that you would be able to find.
Hence, it is clear that it is not 60%-70% number of lesions that you’ll expect to find, but a number significantly less than 50% and which clearly looks to be 40%. The physician interviewed in the New York Times encourages people to have the colonoscopy as a screening procedure, but to not necessarily expect to be safe for the next 10 years. It is strongly recommnended that you follow the colon cleaning procedures prescribed prior to undergoing the procedure.

There will be more studies in the future and likely ones that will be quite positive on colonoscopy. Techniques will likely be modified that also result in improved detection of right sided lesions. Until then, we now and always will need clear heads to do the math and keep evidence based medicine, based on evidence and not on an easily obtained quote.

Up soon we may dissect an article published within the last few months that looked at the longevity of runners and non-runners. If you know the study, here’s a hint: check out the disparate population groups studied and see if it looked like a well designed and controlled study.

New Math From the New York Times:

“In the new study, the test missed just about every cancer in the right side of the colon, where cancers are harder to detect but about 40 percent arise. And it also missed roughly a third of cancers in the left side of the colon.

Instead of preventing 90 percent of cancers, as some doctors have told patients, colonoscopies might actually prevent more like 60 percent to 70 percent.”

Evidence Based Medicine vs. Sensationalism
The New York Times, today, in a fairly sensational manner, discusses Non-Hodgkins Lymphoma and medications in vogue in reference to two medicines, including Zevalin, which at this time is used in cases of failure of other therapies. The accusastion being made is that physicians are not using Zevalin, a radioactive drug, that can only be administered in a hospital using a complex protocol, because of incentives to use more established and well studied protocols including Rituxan in conjunction with “CHOP”.

They highlight the essentially anecdotal evidence of 3 patients who have done well after failing other protocols. That is both the manner in which Bexxar and Zevalin are to be used and the manner in which they have been studied so far. While studies that will indicate if survival time is increased should be complete soon, there is no study that by the principles of Evidence Based Medicine (EBM) do this yet. EBM requires evidence of a high order in well conducted studies that effectively, objectively and statistically demonstrate the utility of a treatment.

This is a higher standard than media generated sensationalism or pharmaceutical company based marketing offer. The studies must stand on their own merit. So far the medications are being used as studied and demonstrated. They are used after failure of other treatment. They are complicated and difficult to administer. They should be used though where needed. Hopefully, the studies that will reportedly be completed soon will demonstrate their superiority, effectiveness, and utility for Non-Hodgkins Lymphoma. Knowledge is increasing fast in this area, but of course not fast enough.

A sensational article like this one, discussing treatment, witholding of treatment, and implying decicsion making is arbitrary and even selfish without even the mention of what EBM is, is not responsible.

You can do your own literature search at PUBMED for more information.

U.S. Clinical Trials – searchable database. You may find active studies using Zevalin or other products here.

Selected Literature Search ResultsExpert Rev Anticancer Ther. 2002 Oct;2(5):485-93
The current indications are: low-grade or follicular lymphoma refractory to rituximab, and relapsed or refractory, low-grade, follicular or transformed lymphoma. Additional studies have been initiated to further define the role of this new therapy in the treatment of patients with B-cell non-hodgkin’s lymphoma.

Semin Oncol. 2003 Dec;30(6 Suppl 17):17-22.
Gregory, SA. Selecting patients for treatment with 90Y ibritumomab tiuxetan (Zevalin).

Yttrium 90 ibritumomab tiuxetan (Zevalin; Biogen Idec Inc, Cambridge, MA) was the first radioimmunotherapeutic agent approved by the US Food and Drug Administration. It is indicated for treating patients with relapsed or refractory low-grade, follicular, or transformed B-cell non-Hodgkin’s lymphoma, including patients with rituximab-refractory follicular non-Hodgkin’s lymphoma. Proper patient selection is essential for optimizing the efficacy and safety of treatment with (90)Y ibritumomab tiuxetan. It may be advisable to use (90)Y ibritumomab tiuxetan relatively early in a patient’s course of treatment because overall and complete response rates, and the estimated median duration of response, are higher among patients who have had fewer median prior antineoplastic regimens than among those who have had a greater median number of such regimens. Furthermore, the myeloablative effect of multiple courses of chemotherapy can preclude the later use of (90)Y ibritumomab tiuxetan. In contrast, other therapies, including chemotherapy and rituximab, can be used safely and successfully after (90)Y ibritumomab tiuxetan without concerns about increased hematologic toxicity from the previous radioimmunotherapy. The main adverse event associated with (90)Y ibritumomab tiuxetan therapy is hematologic toxicity and, as a result, only patients with adequate bone marrow reserves and less than 25% lymphoma marrow involvement should currently be considered for clinical therapy.

Cancer Biother Radiopharm. 2005 Apr;20(2):185-8.

Yttrium-90 (90Y) ibritumomab tiuxetan (Zevalin) induces long-term durable responses in patients with relapsed or refractory B-Cell non-Hodgkin’s lymphoma.
Wiseman GA, Witzig TE.
Mayo Clinic, Rochester, MN, USA. gwiseman@mayo.edu

AIM: Yttrium-90 ((90)Y) ibritumomab tiuxetan (Zevalin) radioimmunotherapy is an effective treatment for relapsed or refractory B-cell non-Hodgkin’s lymphoma (NHL), with overall response rates ranging from 74% to 82%. This retrospective analysis was conducted to determine the number of patients achieving long-term durable responses with (90)Y ibritumomab tiuxetan treatment. MATERIALS AND METHODS: The medical records of patients (n = 211) with relapsed, refractory, or transformed indolent CD20+ B-cell NHL who were treated with 90Y ibritumomab tiuxetan were reviewed. Time to progression (TTP) of > or =12 months was noted in 78 patients (37%), who were identified as long-term responders and were further characterized. RESULTS: Median age of the long-term responders was 58 years (range, 24-80 years) with 44% over 60 years, and 55% were male. Notably, 59% of patients had received > or =2 prior regimens, 33% had received > or =3 prior regimens, and 37% had failed to respond to immediate prior therapy. Median response duration was 28.1 months (range, 10.5-80.3+ months). Median TTP was 29.3 months (range, 12.1-81.5+ months). In patients with ongoing response, median TTP was 53.9 months (range, 49-82+ months). CONCLUSIONS: (90)Y ibritumomab tiuxetan produces durable long-term responses in patients with relapsed/refractory B-cell NHL. Failure to respond to prior therapy does not preclude achieving a long-term response with 90Y ibritumomab tiuxetan.