Site of recurrence after primary radiation therapy

This is the first of a two-part commentary. In this part, we look at studies that identified the site of failure after primary radiation treatment. In next part, we will look at how SBRT is being used to treat such local recurrences.

Before any kind of treatment is given for a recurrence only detected via rising PSA, it’s important to assure that it is indeed only a local recurrence. If there are already distant metastases, local salvage treatment would only create side effects without cancer control. In the past, we only had bone scans and CT scans that could detect only the larger metastases. New imaging technologies are enabling us to better assess the recurrence site.

Hannequin et al. posted the results of their study (abstract 23) at last week’s Genitourinary Conference. The authors retrospectively looked at 89 patients treated in Paris, France between 2010 and 2014 with either brachytherapy (23 patients) or EBRT (66 patients) and who had a biochemical recurrence detected as a rising PSA of at least 2.0 ng/ml over its lowest level.  The patients were classified at diagnosis as favorable risk (28%), intermediate risk (39%) or unfavorable risk (33%). They all had an 18-FCH-PET scan and may have had a multiparametric MRI as well. In 20 patients (22.5%), no target lesion could be clinically identified. Among the 69 patients in whom a clinically detected recurrence was identified, the recurrence site was as follows:

• ·      Local recurrence in 35 patients (51%)
• ·      Lymph node recurrence in 22 patients (32%)
• ·      Distant metastases in 12 patients (17%)

All of those 57 patients with local or lymph node recurrence (83%) were deemed eligible for salvage radiation, but only 17 (25%) could have it. The reasons for not having salvage radiation included advanced age, poor performance status, extensive disease, and patient refusal.

Zumsteg et al. published a retrospective analysis of 2,694 patients treated with external beam RT (IMRT or 3D-CRT) at Memorial Sloan Kettering Cancer Center between 1991 and 2008. The patient diagnosis and treatment characteristics were as follows:

• ·      Risk category:

o   Low risk: 22%

o   Intermediate risk: 48%

o   High risk: 30%

• ·      Median age: 69
• ·      Adjuvant ADT received: 54%, median of 6 months
• ·      Radiation dose received:

o   75.6 Gy (17%)

o   79.2-82.8 Gy (43%)

o   86.4 Gy (40%)

Local recurrence (prostate and seminal vesicles) was clinically detected mostly (71%) via biopsy, the rest radiographically (MRI or PET scan). Lymph node recurrences were detected by CT scan, and distant recurrences were clinically detected via biopsy, by radiographic response to ADT, or by rapidly rising PSA during the castrate-resistant phase. After 83 months of followup overall, and 111 months of followup on clinically recurrent patients:

• ·      22.6% had a biochemical recurrence, defined as nadir+2
• ·      17.6% had a clinically detected recurrence.
• ·      Recurrence by risk category:

o   Low risk: 5.8%

o   Intermediate risk: 13.4%

o   High risk:  32.8%

• ·      Recurrence by radiation dose:

o   75.6 Gy: 29.1%

o   79.2-82.8 Gy: 14.6%

o   86.4 Gy:  15.8%

• ·      Recurrence was also higher in men under 70, higher stage, PSA>10, >50% positive cores.

Among those in whom a clinical recurrence was detected within 8 years of primary treatment, the site of the first recurrence was as follows:

• ·      Local recurrence in 55%

o   74% in low-risk recurrent patients

o   68% in intermediate-risk recurrent patients

o   45% in high-risk recurrent patients

o   in 87% of local recurrences, it was the only recurrence site

• ·      Pelvic lymph nodes (PLN) in 21%*

o   None in low-risk recurrent patients

o   in 38%, of PLN recurrences, it was the only recurrence site

• ·      Abdominal lymph nodes in 9%
• ·      Thoracic lymph nodes in 2%
• ·      Bone in 34%

o   40% in high-risk recurrent patients

o   in 66% of bone recurrences, it was the only recurrence site

• ·      Viscera in 2%

*Patients who presented with enlarged nodes were excluded, and no one received whole pelvic radiation.

The authors also note that a first isolated PLN recurrence was a rare event among all the men treated with EBRT, only occurring in 1.5% of them.

The site of recurrence was strongly correlated with prostate cancer-specific mortality. Compared to locally recurrent prostate cancer, the risk of prostate cancer death after a median of 111 months of followup was:

• ·      4.2 times higher for lymph recurrences
• ·      8.1 times higher for bone recurrences
• ·      9.6 times higher for multi-organ/visceral recurrences

In fact, after accounting for the site of recurrence, only the Gleason score, but none of the other risk factors (e.g., PSA kinetics, stage, age, time to recurrence), predicted prostate cancer mortality. This, and the fact that a first recurrence site was often the sole recurrence site, suggests that there are different types of prostate cancer (phenotypes) with characteristic patterns of spreading and characteristic virulence.

The authors draw 3 conclusions:

“1) The prostate is the most common initial site of recurrence in patients in all risk groups with an increasing absolute incidence that correlates with increasing NCCN risk group.  

2) Isolated PLN relapse is rare in all patients, including those at high risk treated without elective PLN irradiation, at least when using CT for detection.

3) Tumors in many patients display a tropism for specific anatomical compartments and these anatomical patterns of recurrence independently predict prostate cancer specific mortality after clinically detected recurrence.

Unfortunately, their report doesn’t show the time to first recurrence broken down by recurrence site. It may be that the much shorter followup in the French study (patients were treated 2-6 years ago) may explain the lower incidence of bone metastases in that study. Detection methods may explain the differences as well.

In both studies, more than half of the recurrences after primary radiation therapy were local and were at least potentially treatable with salvage therapies. That may not hold true for other kinds of radiation. There isn’t a lot of data on recurrence sites, but the higher biologically effective doses available with SBRT, HDR monotherapy, and multi-modal radiation may be better able to overcome the more radio-resistant cells. In a recent commentary, we saw that a novel kind of radiation, called Carbon Ion Radiotherapy, could kill cancer cells even in a low-oxygen (hypoxic) tumor environment.

The table below shows the range of biologically effective doses for various radiation modalities, and the percent of local failures in all treated patients (not just those with a recurrence), broken out by risk group where available.

Percent Local Failures by Risk Group

	EBRT	SBRT	HDR brachy monotherapy	EBRT+HDR brachy boost
Relative biologically effective dose*	.89-1.02	1.06-1.17	1.27-1.36	.97-1.17
Low Risk	4%	0.9%	2.5%	1%
Intermediate Risk	9%	2.6%		1%
High Risk	15%	7%		9%
Followup	9 years	6/7 years	10 years	4 years
Reference	Zumsteg et al.	Katz and Kang Katz and Kang	Hauswald et al.	Kamrava et al.

*Relative to 80 Gy of IMRT for cancer control

The local failure rates seem to be higher for EBRT than for SBRT, HDR brachy monotherapy, or HDR brachy boost therapy. Only a randomized comparative trial can decide what relative role biologically effective doses, radiation intensity, patient selection, and detection techniques play in determining the extent of local control. It would be useful to know as well whether genetic tests like Prolaris or Oncotype Dx can predict local response to radiation, and whether there are identifiable subtypes that metastasize to lymph nodes, bones or viscera. Better detection of local and distant recurrence is needed as well.

In the next commentary, we will look at how SBRT is being used in salvage treatment of those isolated local recurrences.

written January 13, 2016