The importance of radiotherapy dose escalation and long-term ADT for success

 Localized prostate cancer (PCa) is highly curable. We usually divide localized PCa into 3 risk categories: low-risk, intermediate-risk, and high-risk of recurrence after treatment. Even high-risk PCa is highly curable - 80+% of patients are cured in clinical trials of various radiation therapy regimes (see this link, for example). With new PET scans recently approved for high-risk patients, patients who truly have localized PCa have every hope of achieving even better cure rates.

This begs the question: what do we mean by "cured." What most patients mean is that no recurrence will ever be detected. The first sign of recurrence is a rising PSA more than 2.0 ng/ml over the lowest PSA achieved (nadir). This is called a "biochemical recurrence" (BCR). Other deleterious events may happen. An undetected ("occult") metastasis may grow. The patient may die due to some other cause. If the former never happens, it is called "metastasis-free survival (MFS)." It is highly dependent on the technology used to detect occult metastases. If the latter never happens within the time patients are tracked after treatment, it is called "overall survival (OS)." It is highly dependent upon other diseases ("comorbidities"), treatments given, and the length of follow-up. Often, there are undetermined variables (called "confounders") that tilt OS in one direction or another. Only BCR is relevant for the patient making a therapy choice for his localized prostate cancer.

As we saw previously (at this link), the MARCAP consortium has found that the duration of androgen deprivation therapy (ADT) given along with ("adjuvant to") radiation therapy depends on how the radiation is delivered to high-risk patients - either 12 months for brachy boost therapy or 26 months for external beam radiation therapy. Kishan et al. has analyzed a large number of clinical trials to answer the following questions:

1. What is the role of radiation dose escalation in minimizing BCR?
2. What is the role of long-term vs short-term ADT in minimizing BCR?

• They defined "high dose" radiation as any dose equivalent to greater than or equal to 74 Gy (or its equivalent)
• They defined "long-term" (LTADT) as any duration longer than 18 months, while "short-term" (STADT) was defined as 3-6 months.

For high-risk patients, compared to treating them with low-dose RT without ADT:

• Adding high dose RT (without ADT) reduced BCR by 26%
• Adding short-term ADT reduced BCR by 36%
• Adding high dose RT and STADT reduced BCR by 55%
• Adding low dose RT and LTADT reduced BCR by 61%
• Adding high dose RT and LTADT reduced BCR by 69%

Intermediate risk patients were treated before NCCN distinguished "favorable" intermediate-risk from "unfavorable" intermediate-risk (see this link). For intermediate-risk patients, taken as a whole, compared to treating them with low-dose RT without ADT:

• Adding high dose RT (without ADT) reduced BCR by 21%
• Adding short-term ADT reduced BCR by 32%
• Adding high dose RT and STADT reduced BCR by 46%
• Adding low dose RT and LTADT reduced BCR by 55%
• Adding high dose RT and LTADT reduced BCR by 74%

In both risk groups, long-term ADT provided greater benefit than high dose RT, but combining LTADT with high dose RT provided the best cure rates. 

There are some seeming contradictions between this meta-analysis and the DART 01/05 randomized clinical trial. The purpose was to see if there was a difference in biochemical disease-free survival (bDFS) among intermediate and high-risk patients treated with high-dose radiation and either 28 months or 4 months of ADT. At 5 years of follow-up (see this link), the LTADT group had a significantly higher bDFS than the STADT group. The difference was particularly noticed among the high-risk subgroup. However, with 10 years of follow-up, the difference was no longer significant. 

• For the total, the bDFS was 70% for LTADT vs 62% for STADT (not statistically significant)
• For the high-risk subgroup, the bDFS was 67% for LTADT vs 54% for STADT (not statistically significant)

At least for the high-risk subgroup, the difference was large but not statistically significant. What happened?

What happened was a quarter of the men in the study died in the interim (median age was 72 at the start). Only 3% died of prostate cancer. Many of the men who would have shown no biochemical progression had they lived were eliminated from the trial because they died of other causes. This is called "survivorship bias." The high dropout rate due to death from other causes tells us that follow-up of such trials beyond 5 years will introduce bias into our most important endpoint. It is also another reason that "overall survival" is not a useful endpoint when patients are older. Men with less than 10 years of expected survival due to age or comorbidities should consider watchful waiting rather than any kind of radical treatment. Patients can determine their actuarial expected survival with this calculator: (scroll down to "Male Life Expectancy").