Long-term survival of men initially diagnosed with advanced (Stage IV) prostate cancer


An article in the July issue of the journal Radiation Oncology, along with a commentary on the UroToday web site, provides us with some interesting insights about the long-term survival of patients initially diagnosed with more advanced forms of prostate cancer.

The full text of the original paper in Radiation Oncology (by Muralidhar et al.) can be accessed on line. The additional commentary by Faltas on the UroToday web site requires readers to register as members of the UroToday web site.

Muralidhar et al. used data from the Surveillance, Epidemiology, and End Results (SEER) database to identify men initially diagnosed with AJCC Stage IV prostate cancer between 1973 and 2011 and treated with . (AJCC Stage IV prostate cancer encompasses all patients initially diagnosed with clinical stages T4N0Mo, TxN+M0, and TxNxM+.) Their goal was then to evaluate the long-term survival of these patients and to calculate their conditional, prostate cancer-specific mortality at 5, 10, and 15 years after diagnosis. So what do we mean by “conditional, prostate cancer-specific mortality” (cPCSM)?

cPCSM is the rate of death of men with prostate cancer from prostate cancer for a prior specific point in time. Thus, the authors set out to evaluate three specific mortality rates:

  • The mortality rate of this group of men at 5 years after diagnosis (the initial 5-year PCSM)
  • The mortality rate of the subset of these men who were alive at 5 years after diagnosis over the next 5 years (the 5-year cPCSM rate)
  • The mortality rate of the subset of these men who were alive at 10 years after diagnosis over the next 5 years (the 10-year cPCSM rate)
  • The mortality rate of the subset of these men who were alive at 15 years after diagnosis over the next 5 years (the 15-year cPCSM rate)

We should note that the authors were aware at the outset that the risk of prostate cancer-specific mortality (PCSM) after an initial diagnosis of prostate cancer appear to improve when patients have survived for  a significant period of time. In conducting this study, the authors used a form of competing risk analysis competing risks that adjusted all data for tumor grade, age, income level, and marital status.

Here are the key study findings:

  • 66,817 eligible patients could be identified in the SEER database.
    • 13,345/66,817 patients (20.0 percent) were initially diagnosed with T4 disease.
    • 12,450/66,817 patients (18.6 percent) were initially diagnosed with N1 disease.
    • 41,022/66,817 patients (61.4 percent) were initially diagnosed with M1 disease.
  • Median follow-up among survivors was
    • 123 months (range, 0 to 382 months) for men initially diagnosed with T4 disease
    • 61 months (range, 0 to 410 months) for men initially diagnosed with N1 disease
    • 30 months (range, 0 to 370 months) for men initially diagnosed with M1 disease
  • cPCSM rates improved in all three groups over time.
  • For the men initially diagnosed with T4 disease
    • Initial 5-year PCSM was 13.9 percent at diagnosis.
    • The 5-year cPCSM rate was 11.2 percent (p < 0.001).
    • The 10-year cPCSM rate was 8.1 percent (p < 0.001).
    • The 15-year cPCSM rate was 6.5 percent (p < 0.001).
  • For the patients initially diagnosed with N1 disease
    • Initial 5-year PCSM was 18.9 percent at diagnosis.
    • The 5-year cPCSM rate was 21.4 percent (p < 0.001).
    • The 10-year cPCSM rate was 17.6 percent (p < 0.055).
    • The 15-year cPCSM rate was 13.8 percent (p < 0.001).
  • For the patients initially diagnosed with M1 disease
    • Initial 5-year PCSM was 57.2 percent at diagnosis.
    • The 5-year cPCSM rate was 41.1 percent (p < 0.001).
    • The 10-year cPSM rate was 28.8 percent (p < 0.001).
    • The 15-year cPCSM rate was 20.8 percent (p < 0.001).

Muralidhar et al were also able to show that there were some differences among cPCSM rates for non-white as compared to white patients, as follows:

  • For patients initially diagnosed with T4 disease
    • Initial 5-year PCSM rate was higher for non-white patients compared to white patients.
    • Improvements in cPCSM rates at 5, 10, and 15 years for non-white patients were 20 to 39 percent lower than those for white patients.
  • For patients initially diagnosed with N1 disease
    • White patients had similar cPCSM rates at 5, 10, and 15 years to the overall cohort.
    • Non-white patients had no significant reduction in conditional PCSM rates at 5, 10, or 15 years of survival compared to their PCSM rate at diagnosis.
  • For patients initially diagnosed with M1 disease,  cPCSM rates were similar between white and non-white patients.

The authors conclude that:

While patients with T4, N1, or M1 prostate cancer are never “cured,” their odds of cancer-specific survival increase substantially after they have survived for 5 or more years. Physicians who take care of patients with prostate cancer can use this data to guide follow-up decisions and to counsel newly diagnosed patients and survivors regarding their long-term prognosis.

We have known for a very long time that some men initially diagnosed with advanced prostate cancer can have long survival times. (Note that at least one patient identified in this study as being initially diagnosed with metastatic prostate cancer survived for 370 months after diagnosis. That’s just over 30 years.) However, this analysis provides us with an interesting way to be able to discuss potential for such long-term survival with patients over time. Clearly, for men who make it through the first 5 years after diagnosis with Stage IV prostate cancer, in general, the chances for surviving for another 5, 10 or even 15 years start to go up significantly.

On the other hand, these data do come from a retrospective analysis, and so a certain degree of caution needs to be used in interpreting these data for individual patients. The authors are careful to observe that improvements in prostate cancer care over time might well introduce overestimates of PCSM rates compared to those found among patients diagnosed more recently. However, in the Discussion section of their paper, Muralidhar et al. offer some concrete suggestions about sound ways that these data could be applied to subgroups of patients in the clinical setting.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: