jump to navigation

Survival of Patients with Stage I Lung Cancer Detected on CT Screening (NEJM, October 26, 2006) November 12, 2006

Posted by rajkmd in Oncology, Primary Care, Pulmonary/Critical Care.
trackback

 

For the actual article, click here

Background

  • In 1993, the International Early Lung Cancer Action Project (I-ELCAP) studied the use of CT cancer as a screening tool in the early detection of lung cancer .
    • The study found that eighty percent of lung cancers detected by annual screening were stage 1 cancer.
  • But does this actually save lives?
  • Using ELCAP protocal, this study investigates whether early detection leads to a mortality benefit.
  • This study was supported with grants from National Institute of Health

Methods

  • For a flowchart of the actual protocol, click here
  • Pt’s initially underwent baseline screening.  If the patient had a negative CT scan, then they underwent annual screening at 12 month intervals
  • This is a prospective, observational trial with no control group.
  • For baseline screening, a positive result on the initial low-dose CT scan was defined as
    • At least one solid (completely obscuring lung paranchyma) or partly solid (obscuring part of paranchyma) non calcified pulmonary nodule 5 mm or more in diameter
    • At least one non solid, non calcified pulmonary nodule 8mm or more in diameter, or
    • a solid endobronchial nodule
  • For positive results at the baseline screen:
    • Patients with nodules 5 mm to 14 mm were either re-imaged with CT scan or underwent a PET scan.  If the PET scan was positive or after 3 months the CT scan showed nodules that were growing, then a biopsy was performed
    • For lesions greater or equal to 15 mm, biopsy was immediately performed.
    • Lesions that were suspected to be infections underwent a 2wk period of antibiotics
  • For annual screening, a positive results was any newly identified noncalcified nodule, regardless of size.
    • Repeat CT scan performed for nodules less than 5 cm at 3 months or 6 months depending on the size of the nodule.  If there was no growth in the nodules, then the workup was stopped and pt was re-screened at 12 month interval from the previous CT
    • For pts with nodules greater than 5 cm, a 2wk course of antibiotics was prescribed followed by a CT scan one month later.  If the nodules did not resolve, then a biopsy was performed
  • Patients were enrolled from 1993 to 2995
  • Inclusion criteria- 40 yo or older, at risk for lung cancer b/c of cigarette smoking, occupational exposure, or exposure to second hand smoking, pts were “considered fit” to undergo thoracic surgery
  • 31,567 pts underwent baseline screening and of these, 27,456 underwent annual screening
  • The average age 61 yo and median pack year history was 30

Results

  • Among the 31,567 asymptomatic patients that underwent baseline screening, 405 patients were diagnosed with lung cancer.
  • Among the 27,456 patients who underwent annual screening 74 patients were diagnosed with lung cancer.
  • Of the total of 484 patients that developed lung cancer, 411 (85 percent) underwent resection, 57 (12 percent) underwent radiation and/or chemotherapy, and 16 recieved no treatment.
    • Of the 484 patients diagnosed 85% had stage I lung ca 
  • The estimated 10 year survival for all participants was 80% (95%CI, 74 to 85), but in patients with stage I disease the survival rate was 88% (95% CI, 84 to 91).  Moreover, in patients with stage 1 disease who underwent resection within 0ne month of diagnosis the survival rate was 92 % (95% CI, 88 to 95).
  • The operative mortality rate was 0.5%.

Discussion

  • The authors contend that using CT scans to screen for lung cancer could prevent 80% of death from lung cancer.  Currently the death rate of lung cancer in US is 95%.
  • The rate of detection in this study was 1.3% on baseline screen and 0.3% on annual screen, which is comparable to breast cancer screening with mammograms.

THE BOTTOM LINE:  In high risk patients patients, using the I-ELCAP protocol which includes a baseline CT scan and subsequent annual CT scans, 85 percent of the cancers identified in this study were stage I disease.  Moreover, the ten year survival was 80 percent in all patients diagnosed with lung cancer, 88 percent in patients with stage I disease, and 92% in patients with stage I disease and resection within one month of diagnosis.

Comments»

1. Geoffrey Wittig, M.D. - November 12, 2006

This study has to be seen as a first crack at the problem, but by no means definitive. Lead-time bias is an obvious and huge problem by its very design. There is also no discussion of the risks and costs versus purported benefit. For instance, even accepting the absurdly optimistic CT cost of $150-200 per study, this translates to $8 billion per year just to scan every American smoker, ignoring the inevitable follow-on costs of biopsies and repeat studies. Then there is the estimated 1:1200 risk of future cancer per CT scan, which quickly becomes a very significant number when you’re scanning millions of smokers. Finally, there is no recognition of the opportunity cost of sinking that $8 billion per year into post-hoc screening, rather than zero-risk smoking cessation.

2. George D. Bussey, M.D. - November 12, 2006

Re the issue of cost/benefit: Any CT operations folks out there that could address the question of the true marginal cost of adding this type of screening to existing CT capacity? If you have a scanner that has free time in off hours – what would it cost to bring in the staff and pay the other associated costs to run a screening clinic during “off hours?” Medicare and other insurers would have to agree to not force similar rates onto other CT services, but that approach might allow screening to be economially feasible.

3. Michael LoCurcio - November 13, 2006

I very much agree with the comments made by Drs. Wittig and Bussey. In addition, any screening test will have a tendency to find disease that is relatively more benign when compared to historical data because historically the workups were driven by symptoms (“length bias”). This combined with the lead-time bias makes the comparison of this group’s mortality to estimated mortality very troublesome. When considering these biases, cost, and CT availability, I believe this screening modality should not be employed based on the evidence provided in this study. In the meantime we should continue to use the following tool for lung cancer screening: “Do you smoke” and “Are you exposed to smoke?”

4. A Medical Discussion on the Use of CT Scans to Screen for Lung Cancer « Medical Discussions - November 24, 2006

[...] This in relation to the following article.  [...]

5. brooklyn center real estate - May 31, 2007

brooklyn center real estate

ka-ka-sh-ka 4739694 brooklyn center real estate info

6. java script image preloader - May 31, 2007

java script image preloader

ka-ka-sh-ka 4739694 This site is about java script image preloader.

7. 1000 apc - June 3, 2007

1000 apc

ka-ka-sh-ka 4739694 The 1000 apc

8. code nm roswell zip - June 6, 2007

code nm roswell zip

code nm roswell zip moves

9. coed need cash - June 7, 2007

coed need cash

Actual articles about coed need cash and more.

10. 44l kiton - June 7, 2007

44l kiton

44l kiton foundation

11. Christopher - July 23, 2007

Very interesting article and good links

http://beat-cancer.blogspot.com