S9900: A phase III trial of surgery alone or surgery plus preoperative (preop) paclitaxel/carboplatin (PC) chemotherapy in early stage non-small cell lung cancer (NSCLC): Preliminary results

Reviewer: Walter Sall, MD
Last Modified: May 16, 2005

Presenter: K. Pisters
Presenter's Affiliation: MD Anderson Cancer Center
Type of Session: Scientific


  • Five year Overall Survival (OS) for stages T2NO-T3N1 NSCLC are low, ranging from 9-38%.
  • When this study was designed in the late 1990s, the benefit of adjuvant chemotherapy in resected lung cancer was not yet confirmed.
  • The BLOT trial was a phase II trial showing safety and activity of pre-operative chemotherapy in early stage NSCLC.
  • This prompted development of this phase III trial of surgery with and without pre-operative chemotherapy to determine the benefit of neoadjuvant chemotherapy.

Materials and Methods

  • Entry criteria included NSCLC patients, stage T2NO, T1-2N1 and T3N0-1, excluding superior sulcus tumors.  Staging included CT scan and mediastinoscopy (for nodes >1cm).  PET was not required but was obtained in many patients.  Expected post-operative FEV1 was > 1 liter.
  • Patients were randomized to surgery alone vs surgery preceeded by three cycles of carboplatin (AUC 6)
     and paclitaxel (225mg/m2) every three weeks.
  • Planned accrual was 600 patients to detect a 33% increase in median OS.


  • 354 patients were accrued ffrom 10/1999 through 7/2004.  Accrual was closed early because of the release of adjuvant chemotherapy data showing a survival benefit.
  • The two arms were well balanced with regard to age, stage, sex and histology as well as all other factors.
  • 77% of patients completed all 3 cycles of chmotherapy.  Response rate (RR) was 41%.  Rates of grade II/IV neutropenia were 50%, grade I/II neuropathy 50% with very low rates of nausea and vomiting.  2% of patients died during chemotherapy.
  • Exploratory surgery was performed on 89% of chemotherapy and 97% of surgery alone patients.  By intention to treat, 84% of patients in each arm achieved an R0 resection.  Pathologic complete response rate was 10% in the chemotherapy arm.  Seven post-operative deaths occurred in the chemotherapy arm vs 4 in the surgery alone arm.
  • Median progression free survival (PFS) was 31 months (chemotherapy)  vs 20 months ( surgery alone) with a nonsignificant p value.  Five year OS was 47 vs 40 months, respectively (p=0.32).

Author's Conclusions

  • A trend towards improved PFS and OS was seen in this trial.  Unfortunately, it was underpowered to more convincingly demonstrate these potential benfits.  The benefit of chemotherapy was also overshadowed by better than expected survival in the control arm.
  • This trial supports the use of chemotheraoy in resectable NSCLC.
  • A meta-analysis of pre-operative chemotherapy trials is being planned to better evaluate the value of pre-operative chemotherapy.
  • Trials comparing pre-operative vs. post-operative chemotherapy are warranted.

Clinical/Scientific Implications
Because  of three recent large trials showing a significant survival benefit to adjuvant chemotherapy in resected NSCLC, the pre-operative approach has recently received little attention.  Pre-operative chemotherapy may improve patient compliance compared to post-operative therapy, potentially improving outcomes.  It is unclear how much pre-operative chemotherapy increases the risk of surgery, however.  The only way to determine the best method of administering chemotherapy is through a randomized trial comparing the two strategies.  Until that occurs, post-operative chemotherapy will remain the standard of care.