Showing posts with label patient-reported outcomes. Show all posts
Showing posts with label patient-reported outcomes. Show all posts

Monday, December 5, 2016

SBRT vs. moderate hypofractionation: same or better quality of life

We have seen in several randomized clinical trials of external beam treatment of primary prostate cancer that moderately hypofractionated IMRT (HypoIMRT) treatment (accomplished in 12-26 treatments or fractions) is no worse than conventionally fractionated IMRT treatment (in 40-44 fractions).  We recently saw in a randomized clinical trial from Scandinavia that SBRT (in 5 fractions) is no worse than conventional IMRT (see this link) in long-term quality-of-life outcomes, even though they used inferior technology. The missing piece of the puzzle is to answer the question of whether SBRT is any worse than HypoIMRT.

We don’t yet have a definitive answer (which would require a randomized clinical trial), but an analysis of pooled data from 5 different clinical trials, suggests that SBRT is no worse and may be better than HypoIMRT in its urinary, rectal, and sexual outcomes. Johnson et al. pooled SBRT data from clinical trials among 534 men at 3 institutions (UCLA, Georgetown, and 21st Century Oncology) and HypoIMRT data from clinical trials among 378 men at Fox Chase Cancer Center and the University of Wisconsin. All patients were treated between 2002 and 2013 at those top institutions, with state-of-the-art equipment in the context of carefully controlled clinical trials. Because of this, all outcomes are probably better than those achieved in everyday community practice. The only significant difference in patient characteristics was that SBRT patients were about 5 years older (69 vs. 64 years of age for HypoIMRT). We expect older men to have more natural deterioration in urinary and sexual function.

The following table shows the percent of men receiving each treatment who suffered from at least the minimally detectable difference in patient-reported scores on validated quality-of-life questionnaires with respect to urinary, rectal, and sexual function. Numbers in bold typeface represent a statistically significant difference.


SBRT
HypoIMRT
Odds Ratio (adjusted)
Urinary
14%
33%
0.24
Rectal
25%
37%
0.66
Sexual
33%
39%
0.73

The data support the following conclusions:
  • Urinary and rectal problems at 2 years were experienced by fewer of the men who had SBRT.
  • Urinary and rectal problems improved after 2 years compared to 1 year post-treatment. For SBRT, they approached baseline values.
  • Sexual issues did not improve at 2 years.
  • While we expected the SBRT patients to experience greater deterioration owing to their age, the opposite occurred.
(update: 4/11/2020) Kwan et al. reported on 78 patients randomized to SBRT (36.25 Gy in 5 weekly treatments) or moderate hypofractionation (70 Gy in 28 treatments). After at least 6 months of follow-up:
  • there were no statistically significant differences in grade 2+ or grade 3 toxicities
  • there were no minimally important differences in patient-reported quality of life on incontinence, irritative/obstructive urinary issues or bowel issues.


Why were the SBRT outcomes better?

SBRT is not just a high-dose-per-fraction version of IMRT, although it is that too. When the linear accelerator is delivering only 2 Gy per fraction, missing the beam target by a little bit is not likely to make much difference – it will average out in the long run. Because a geographic “miss” of the beam target has much greater consequence for SBRT, where the dose per fraction can be 8 Gy, much more care is taken to achieve pinpoint accuracy. This includes such steps as:
  • Fiducials/transponders aligned within each treatment and not just between treatments.
  • Fast linear accelerators that minimize the time during which the prostate can move.
  • No treatment if the bowel is distended or the bladder is not full.
  • Tighter margins: as low as 0 mm on the rectal side and 2 mm on the front side. This compares to margins of 0.5-1 cm for IMRT.
  • Narrower dose constraints for organs at risk, including the bladder, rectum, urethra, femurs and penile bulb.
  • More care taken to find a plan that optimizes prostate dose relative to organs at risk.


It is entirely possible that IMRT outcomes might be equivalent to SBRT outcomes if the same factors were incorporated into IMRT planning and delivery. But fractionation probably has an effect as well. To understand why, we must look at the radiobiology of prostate cancer. Prostate cancer has been found to respond remarkably well to fewer yet higher doses of radiation. This is reflected in a characteristic called the “alpha/beta ratio (α/β).” The α/β of prostate cancer is very low, at about 1.5. It is lower, in fact, than that of surrounding healthy tissues. Many of those healthy tissues have an early response, which is responsible for acute toxicity, typically within 3 months of treatment (α/β = 10.0). Rectal mucosal tissue is an example. This means that a hypofractionated dosing schedule will kill relatively more cancer cells, while preserving more of the cells in the nearby organs.

There are fewer types of tissue in the pelvic area that have a delayed response to radiation, and those tissues, like nerve cells, tend to be radio-resistant. This is why late-term toxicity is relatively low. Some of the late-term effects we do see are due to cumulative responses to radiation, like the buildup of scar tissue and other reactive responses in vasculature, along the urethra, and in the rectum. Late responding tissue has an α/β of about 3.5

We can compare the biologically effective dose (BED) of the various dosing schedules to see the effect that hypofractionation would theoretically have in killing cancer cells and preserving healthy tissue.



BED for cancer control
Relative BED for cancer control
BED for acute side effects
Relative BED for acute side effects
BED for late side effects
Relative BED for late side effects
80 Gy in 40 fractions
187 Gy
1.00
96 Gy
1.00
126 Gy
1.00
60 Gy in 20 fractions
180 Gy
0.96
78 Gy
0.81
111 Gy
0.89
40 Gy in 5 fractions
253 Gy
1.35
72 Gy
0.75
131 Gy
1.05

So the kind of fractionation used in SBRT theoretically has about 35% more effective cancer-killing power than conventional fractionation, while its ability to generate acute toxic side effects is reduced by 25%, and its late-term side effects would be similar.

Why isn’t everyone who elects to have primary treatment with external beam radiation treated with SBRT?

It’s one thing to make predictions based on theory, but it’s quite another to determine whether it works as well in clinical practice. So far, non-randomized trials like the ones examined in this study have shown excellent oncological and quality-of-life outcomes for SBRT with up to 9 years of follow-up. We await the oncological results of randomized trials comparing SBRT to IMRT. The oncological outcomes from the randomized Scandinavian trial are expected any time now. There are several others that are ongoing.

With SBRT, the patient enjoys the obvious benefits of appreciably lower cost and a more convenient therapy regimen. Medicare and most (but far from all) insurance companies now cover SBRT. There is considerable resistance from radiation oncologists in private practice who would get reduced revenues, and would have to learn the new techniques and gain adequate experience in using them.



Wednesday, September 28, 2016

Toxicity equal for SBRT and conventional external beam radiation


There has been some question as to whether the toxicity of delivering very high doses of external beam radiation per treatment (or fraction) in fewer fractions (called “extreme hypofractionation” or SBRT) would be high compared to conventional dose rates per fraction. While SBRT practitioners have reported very low toxicity rates (see table in this link), there has always been some doubt because there may have been some bias in how patients were selected in the various studies.

The HYPO-RT-PC trial was the first trial ever to randomly assign patients to one kind of radiation or the other. Between 2005 and 2015, they enrolled 1200 intermediate-risk patients in Scandinavia to receive either:
  1.  Conventional fractionation: 78 Gy in 39 fractions
  2. SBRT: 42.7 Gy in 7 fractions

The biologically effective dose is 19% higher for SBRT in terms of cancer control. The biologically effective doses are equivalent in terms of toxicity.

There were a few differences from some US practices:
  • “Intermediate risk” was defined as one or two of the following 3 risk factors:

  1. Stage T1c-T3a (T3a is a high risk factor in the commonly used US definition)
  2. PSA> 10 ng/ml (PSA> 20 ng/ml is a high risk factor in the commonly used US definition)
  3. Gleason score ≥7 (Gleason scores greater than 7 are a high risk factor in the commonly used US definition)

  • 80% of the men were treated with a technology called 3D-CRT, which is seldom used for external beam therapy anymore at major tertiary care centers. It is never used for SBRT in the US because it is considered not precise enough, and too toxic.
  • SBRT is usually delivered in 4 or 5 fractions in the US. CyberKnife and VMAT are the most common technologies in use, and use of sophisticated image guidance throughout each treatment is a common practice.

The toxicity results are based on 866 patients who had 2-year follow-up results. There were some differences in acute toxicity:
  • Acute urinary toxicity was 27.6% for the SBRT group and 22.8% for the conventional fractionation group, but the difference was not statistically significant.
  • Acute rectal toxicity was 9.4% for the SBRT group and 5.3% for the conventional fractionation group. The difference was statistically significant, but narrowed by 3 or 6 months.

Neither physician-reported toxicity nor patient-reported late-term toxicity differed by the fractionation schedule they received. By two years:
  • Late-term urinary side effects were reported by 5.4% of the SBRT group and 4.6% of the conventional fractionation group. The difference was not statistically significant.
  • Late term rectal side effects were reported by 2.2% of the SBRT group and 3.7% of the conventional fractionation group. The difference was not statistically significant.
  • Impotence was reported by 34% of both groups, up from 16% at baseline.
  • Patient-reported bother from urinary, rectal and sexual side effects were not different.

Given their use of the largely outmoded 3D-CRT technology, it was not surprising that acute toxicity would be elevated. I’m frankly surprised that late-term toxicity was not higher for SBRT.

They plan to present their findings on oncological outcomes at a future time.

Monday, September 19, 2016

Hypofractionated radiation therapy using IMRT has a clear advantage

I was reticent to write about hypofractionation yet again after writing about it so often in the last year. See this link for my latest summary. In a sea of randomized trials demonstrating that hypofractionated radiation therapy (i.e., it is delivered in fewer treatments or fractions) was no worse in cancer control or in toxicity to conventionally fractionated (40-44 treatments), there was one study, the Dutch HYPRO study, where the toxicity was a bit worse. At the time (see this link), I speculated that that was because they included an older radiation technique called 3D-CRT rather than the IMRT technology that is now prevalent in the US. A new study from MD Anderson suggests that may indeed be the case.

Hoffman et al. presented the patient-reported outcomes of 173 men with localized prostate cancer who were treated at M.D. Anderson in Houston. They were randomized to receive either:
  1. 75.6 Gy in 42 fractions (conventional fractionation) via IMRT
  2. 72 Gy in 30 fractions (hypofractionation) via IMRT
The men filled out questionnaires at baseline, and at 2, 3, 4, & 5 years after treatment. Patients were probed on their urinary, rectal and sexual status. Patient-reported outcomes on validated questionnaires is a more reliable source of toxicity data because it does not rely on the patient volunteering information to the doctor or the doctor assessing or recording that information. Analysis of the two groups showed that:
  • there was no difference with regard to rectal issues (urgency, control, frequency, or bleeding).
  • there was no difference with regard to urinary issues (pain, blood in urine, waking to urinate at night, or leakage)
  • there was no difference with regard to sexual issues (erections firm enough for intercourse)
  • there were no differences at 2, 3, 4, or 5 years.
This should dispel any concerns that completing IMRT in less time may be more toxic. Just as with all forms of radiation, the technology has improved greatly over the years. In the hands of an experienced and careful radiation oncologist, there is no reason that external beam therapy cannot be completed in less time and at lower cost.

Saturday, September 17, 2016

Patient-reported outcomes from ProtecT - the first randomized trial comparing surgery, radiation, and active surveillance

While there were no differences in 10-year mortality when patients were randomized to surgery (RP), external beam radiotherapy (EBRT), and active surveillance (AS) (see this link), the side effects patients suffered from those treatments differed markedly. Johnson et al. have published the patient-reported outcomes of the ProtecT trial in the New England Journal of Medicine (see this link).

In the ProtecT trial, all participants (or actually 85% of them) filled out a series of validated questionnaires (EPIC and others) that probed issues of urinary function, rectal function, sexual function, and general health. I will ignore the overall health, vitality and mental status questions for now. Suffice it to say that they did not differ among therapies, nor were they very much affected by them. Questionnaires were filled out before the biopsy (the baseline), and at 6 and 12 months after randomization, and annually thereafter until 6 years from the initial biopsy.

What is especially interesting is seeing how equivalent patients (they are equivalent because they were randomized to the 3 treatments) did over the 6 years after receiving each treatment. This means that, for the first time, the side effect profiles are completely comparable (well, almost) and almost without bias.

Some messy data

I say "almost" because there was some switching of treatments that did occur. 22% of the men did not get the therapy they were originally randomized to, and they self-selected some other therapy or no therapy. However, in the analysis they are treated as if they got therapy that they were originally intended to get. Strange, huh?

In addition, they may have received salvage therapy after biochemical failure, and 55% of those assigned to AS did get a radical therapy eventually. So for each intended therapy:
  • Among those 291 men who started on AS but got radical treatment: 49% had surgery, 33% had radiation as specified, and 18% had another kind of radiation or HIFU
  • Among those 391 men who started on RP, 14 (4%) had adjuvant or salvage radiation, and 1 went on lifelong androgen deprivation therapy (ADT) within a year.
  • Among those 405 men who started on EBRT,  3 had salvage RP, 14 (3%) went on lifelong ADT, and 1 had HIFU.
Whether men had the assigned therapy or not, and irrespective of any other therapy they had, they are included with the group they were originally assigned to. It's messy.

Fortunately, there's hope in sight. In one of the Appendices (Section S3), they added the note:
"In future analyses, we intend to present patient-reported outcomes according to treatment received and an economic evaluation including assessment of therapies received for treatment impacts, as well as details about the reasons for change of management in the active monitoring arm to further inform individual and clinical decision-making"

That will give us a much truer picture of the side effects associated with the treatments they actually received.

The treatments

RP was open and nerve sparing. While most men now have robotic surgery rather than open surgery, it seems to make little difference, except for some higher incidents of issues arising during the operation (see this link).

AS did not have required follow-up biopsies, so their side effects may be a little better than on contemporary AS programs. Biopsy complications are never long-lasting anyway.

EBRT was different from contemporary standards. The dose was lower (74 Gy vs. 80 Gy), so there may have been fewer complications due to dose. They used an older delivery technique (3D-CRT vs IMRT) which had higher rates of side effects. And it was given together with short term (3-6 months) of ADT, which would certainly increase the early sexual side effects. ADT is seldom given to favorable risk patients today.

note: all of the patient-reported outcomes include the effect of whatever remedies they used to treat them.

1. Urinary Adverse Outcomes

a. Incontinence

This was a big issue for RP, of course, but not for AS or EBRT. The percent using one or more pads per day is one commonly used measure. As one can see in the following table, incontinence was highest at the 6-month time point, but had gotten somewhat better by the end of the first year. 20% were incontinent by the end of two years, with little improvement from that point.

For EBRT, incontinence peaked at 5% at 6 months. Remember, this was 3D-CRT - a technology that has greater toxicity than the IMRT predominantly in use now. It hovered around 3-4% thereafter.

For AS, incontinence also peaked at 4% at 6 months, and stayed at that level for the next couple of years. From then, it steadily rose to 8% by the end of the 6 year study. Remember that for the purposes of this trial, men were still included in the AS cohort whether they were eventually treated or not. By 6 years, more than half the men had been treated, primarily with surgery.

Table 1. Incontinence: The percent who used one or more pads per day

Time point
AS
RP
EBRT
Baseline
0%
2%
0%
6 months
4%
46%
5%
1 year
4%
26%
4%
2 years
4%
20%
4%
3 years
5%
20%
3%
4 years
7%
17%
4%
5 years
7%
17%
3%
6 years
8%
17%
4%


b. Urinary Irritation/Obstruction

The researchers examined the question of whether urination become more difficult or more frequent after therapy. One way to look at this is a set of questions on the EPIC questionnaire asking about urinary frequency and retention. On that questionnaire, a score of 100% means that function is perfect in that regard, no issues whatever.

On this dimension only EBRT had a clinically detectable effect, and it was only at the 6 month mark. EPIC score dropped from 93% to 84%. After that, it returned quickly to baseline levels.

Table 2. Urinary Irritation/Obstruction EPIC scores, where 100% would be the best possible score.

Time point
AS
RP
EBRT
Baseline
93%
92%
94%
6 months
92%
89%
84%
1 year
93%
93%
93%
2 years
92%
93%
93%
3 years
91%
93%
93%
4 years
91%
94%
93%
5 years
92%
94%
93%
6 years
92%
94%
93%

2. Rectal Adverse Outcomes

The researchers asked the trial participants about their bowel function at baseline and after treatment. There were no discernable effects of AS or RP. Bowel function among the men enrolled for EBRT declined by 6 months (from a score of 93% to 86%). Thereafter, bowel function scores returned to near baseline levels. Other than the 6 month time point, there were no significant differences among the 3 treatments.

Table 3. Bowel function EPIC scores, where 100% would be the best possible score.

Time point
AS
RP
EBRT
Baseline
92%
91%
93%
6 months
91%
92%
86%
1 year
92%
93%
90%
2 years
92%
93%
90%
3 years
92%
93%
91%
4 years
92%
93%
91%
5 years
92%
93%
90%
6 years
92%
92%
91%

3. Sexual Adverse Outcomes

This is one of the few trials that asked men detailed questions about their sexual function at baseline and for 6 years thereafter. One of the key measures of sexual function is the ability to have erections firm enough for intercourse. At baseline, about two-thirds of these 62 year old men (range 50-69), some with other comorbidities like diabetes, cardiovascular disease, and smoking, had suitable erectile function. 

None of the questionnaires asked about perceptions of penile shrinkage in length and girth, climacturia (urination at orgasm), or Peyronie's (abnormal penile curvature), which are often symptoms that affect sexual function post-prostatectomy. Nor do they ask about how the loss of ejaculate has affected sex. That is a certainty with surgery, a near-certainty after radiation, and is not affected by AS. Their definition of erectile function includes the effect of any erectile function aids (e.g. ED meds, injections, pumps, or implants) they may have been using.

For those randomized to RP, erectile function was reduced to 12% at 6 months (remember: they all had nerve-sparing surgery). It recovered somewhat to as much as 21% at 3 years but did not recover beyond that. At every time point, their erectile function was significantly worse than the other treatment cohorts.

For the AS cohort, erectile function declined by 6 months and continued to deteriorate thereafter as they elected to have radical therapies, predominantly surgery. 11% of this cohort had already elected to have radical treatment by the 6-month mark.

For the EBRT cohort, erectile function had dropped to a minimum value of 22% at 6 months. This may be largely attributable to the fact that all of the men in the EBRT cohort had 3-6 months of ADT. It is unknown how much, if any, of their testosterone came back after that and how long it took to recover. Erectile function snapped back a bit post-ADT, getting as high as 38% at 1 year, and declined to 27% by 6 years. Again, this was based on the 3D-CRT technology, and is below the rates usually seen for this age group with IMRT, brachytherapy, or SBRT.

Table 4. Erectile function - the percent who had erections firm enough for intercourse

Time point
AS
RP
EBRT
Baseline
68%
65%
68%
6 months
52%
12%
22%
1 year
49%
15%
38%
2 years
47%
19%
34%
3 years
41%
21%
34%
4 years
37%
20%
32%
5 years
35%
20%
27%
6 years
30%
17%
27%


Myths Exploded by this study:

Myth #1: The side effects end up about the same for surgery or radiation

That's clearly not true for incontinence or erectile function, It is true for urinary irritation and rectal function, which are at baseline levels and similar in all cohorts at 6 years.

Myth #2: With surgery, you get the side effects all at once and steadily recover; with radiation, the side effects come on steadily and may hit you many years later.

What we've seen here belies that myth. There is some recovery of continence up to two years later, but not thereafter. After radiation, incontinence was a minor symptom (except to those who had it, of course), but it did not increase over the years. Urinary irritation/obstruction increased at 6 months for EBRT, but returned to baseline permanently thereafter. Rectal function scores also permanently returned to baseline levels after the 6-month time point.

Myth #3: Over time, erectile function is about the same for surgery and radiation. 

As we've just seen, erectile function is much worse after surgery, and it never recovers much beyond 2 years. It is worth tracking this myth down to its source. I have even heard John Mulhall, the eminent Memorial Sloan Kettering sex specialist quote this myth.

I believe this myth started with the PROSTQA study published in 2008. Until the ProtecT trial, it was our best source of patient-reported outcomes after the various treatments. The patients were not randomly assigned, however, and differed markedly in their characteristics. Those characteristics, especially age, varied greatly with the treatments they chose. In the following table, hidden in an appendix (all the good stuff is usually back there), we can extract the following table:

Table 5. Percent with preserved erectile function sufficient for intercourse 2 years after treatment, broken down by age at treatment 



Age
RP
EBRT
BT
<50
55
100*
75*
50-59
43
52
67
60-69
27
39
44
70+
8*
30
24
Total
35
37
43
Median Age
60 years
70 years
66 years

small sample size

Although the potency doesn’t seem to vary much between treatments in total (range 35% to 43%), it is only because the men who received EBRT and BT were older than the men who were treated with RP. Within every age group, potency preservation was higher with radiation.

There are other differences between the two studies, such as: 
• this table only includes men who were potent before therapy, which would exclude about a third of men in the ProtecT trial. This would lower all the percentages in Table 5 relative to Table 4. 
• ProtecT included only men in the 50-69 age range, while half of the findings in ProstQA came from men treated with radiation over the age of 70. 
• Finally, ProtecT didn't yet report erectile function according to the therapy or therapies they actually received.

It is gratifying to see these myths shattered. Patients are the beneficiary.