Time to First Progression Is the Best Predictor of Survival Duration in SWOG Standard Dose Chemotherapy Studies

Reviewer: Ryan Smith, MD
Last Modified: December 10, 2002

Presenter: Brian G.M. Durie
Presenter's Affiliation: SWOG
Type of Session: Scientific


    Multiple myeloma has been treated with combination chemotherapy for years. Various regimens have various response rates and have been used in the past to determine efficacy. However, reponse (by % regression of disease) has absolutely no correlation with subsequent survival. This statistical study was done to determine what predictors could be used for survival in patients with multiple myeloma.

Materials and Methods

  • 1555 multiple myeloma (MM) patients from 4 different SWOG studies with long follow up were used in the analysis.
  • These 4 studies used VMCP/VBAP and VAD chemotherapy with interferon +/- prednisone +/- verapamil/quinine
  • Median age of patients was 62 and median B2Microglobulin was 4.8. 20% had poor risk factors.
  • Patients were evaluated to determine response and magnitude of response, time to first progression, and survival


  • Overall, 62% of patients had some response, with 43% with a greater than 75% response.
  • Overall survival was 33 months (overall, and in each individual study) and event free survival was 18 months (overall and in each individual study).
  • 20% of patients remain in remission at 4 years
  • 20% of patients remain alive at 6 years
  • Median survivals for varying levels of regression were 28-32 months-not significantly different from eachother.
  • Using data from 6 months after diagnosis, initial regression had no impact on eventual overall survival. Patients who progressed at 6 months had a worse survival.
  • Using data from 1 year after diagnosis, it was shown that the group with a good initial response but progression by one year had a poor eventual overall survival. Conversely, patients with no initial regression but with no progresion by one year actually had the highest survival of any subset.

Author's Conclusions

  • Amount of initial regression should be interpreted with caution, as it showed no bearing on patients' overall survivals.
  • Time to first progression is the best predictor for survival, regardless of initial response.
  • The establishment of stable disease for > 6 months is the best short term predictor of survival

Clinical/Scientific Implications

    Amount of regression is classically reported in studies of patients with MM. However, there has never been a correlation with overall outcome with amount of regression. According to this study, the % regression is not at all important. From their data, if an immediate predictor is needed, stability of disease for > 6 months should be used. In other words, the durability of response is much more important than the magnitude of response. What is not known from these data is what impact a complete response has on this entire picture. Complete response rate was not reported, and it is conceivable that patients with a true complete response may be in a different category altogether. This could be a future point of study. Also, this is to some extent, a self-fulfilling phophecy, as it is not a huge jump to predict that patients who fail early will fail eventually. This is somewhat like saying that, in a group where individual patients die, the group as a whole will have a lower survival. Regardless, this study demonstrates that initial response should not be used to modify therapy or predict outcome, but rather the time to progression should be the important data viewed when reviewing studies on patients with multiple myeloma.

Oncolink's ASH Coverage made possible by an unrestricted Educational Grant from Ortho Biotech.