Prediction of relapse in stage I nonseminomatous germ cell tumors by CXCL12

Share :
Published: 25 Feb 2013
Views: 6032
Dr Duncan Gilbert - Sussex Cancer Centre, Brighton , UK

Dr Gilbert talks to ecancer at the 2013 ASCO GU symposium about prediction of relapse in stage I nonseminomatous germ cell tumours (NSGCT) by CXCL12. Results from the MRC TE08 and TE22 clinical trials showed that patients who need to get chemotherapy can be picked out and those who don't can be spared. CXCL12 intensity is prognostic for relapse independently of VI. This additional prognostic information can identify groups at very low and very high risk of relapse with potential clinical utility.

ASCO GU 2013

Prediction of relapse in stage I nonseminomatous germ cell tumours by CXCL12

Dr Duncan Gilbert – Sussex Cancer Centre, Brighton, UK


We were looking at patients with stage I nonseminomas, which is a sub-type of testicular cancer. These are patients who have had their testicles removed at orchidectomy and they have an option either to have either adjuvant chemotherapy, to reduce the risk of relapse, or surveillance and then treatment when their cancer comes back. Either those patients, patients who have adjuvant chemotherapy, the vast majority of them, 97-98%, never relapse again but those patients, even if the patients who are surveyed, those that do relapse and need adjuvant chemotherapy, still 97-98% are cured. So they have an excellent outcome either way, what we’re looking at is whether they need adjuvant treatment up front or whether they can be safely surveyed and treated at relapse.


What are the existing methods of discovering metastasis?

Yes, there are fairly strict surveillance protocols – patients have scans, but at the outset we try and predict their risk of recurrence and at the moment the only clinically useable factor is something called vascular invasion, so evidence that the cancer cells have begun to invade into the blood vessels within the tumour.


What did you do in the study with CXCL12?

We have some previous data that showed that CXCL12 was independent and additive in terms of its predictive features over vascular invasion so we took clinical sample sets from two large MRC trials, the TE08 trial of surveillance and the TE22 trial of testing PET scans. That gave us 500 patients, we were able to get the blocks from a little under half of those, so from 182 patients, and we tested whether CXCL12 had independent predictive value on top of vascular invasion. We found that using CXCL12 we could further refine a patient’s chances of recurrence and we could define a group with a very low rate, so using the two markers together, patients who over 90% of them were cured by an operation alone and didn’t need any further treatment, but also a group at very high risk where over 75% of patients subsequently recurred.


Can you target use of the therapy?

Absolutely, that group of patients would be very eligible to have chemotherapy up front that should get rid of their cancer once and for all. And the other thing is that when the patients recur they need three cycles of chemotherapy; at the outset, at the moment in the UK, they would get two but potentially in the future they’ll only need one, but that trial is on-going. And the difference between one cycle of chemotherapy now and three cycles later is significant to a patient and will allow them to get on with their life knowing that they were cured.

The aim of this, as we said at the beginning, the patients have a good rate of cure whatever, what we’re trying to do is minimise unnecessary treatment and give them the best chance of long-term health.

What would be the final message to doctors?


CXCL12 is an independent marker on top of vascular invasion and we’d propose, it’s up for discussion, but we’d propose that this is now ready for clinical use.