In advanced cancer, anti-tumor therapies often work only partially or not at all, and tumors progress following treatment. Scientists have now linked a treatment-induced growth factor to the cancer’s future spread.

A team at Vanderbilt-Ingram Cancer Center team reports in the May issue of Journal of Clinical Investigation that radiation and chemotherapy increase circulating levels of the growth factor TGF-beta, circulating cancer cells, and tumor metastases in mice implanted with metastatic breast cancer.

Radiation therapy and the chemotherapeutic agents doxorubicin (Adriamycin) and docetaxel (Taxotere) all increased TGF-beta levels and accelerated metastasis, an effect that was blocked by neutralizing antibodies directed against TGF-beta.

Blocking TGF-beta in the model prevented tumor metastases, suggesting that drugs designed to block TGF-beta may be clinically useful in combination with primary therapies.
(full story…)

For an overview of 3 main types of immunotherapies, or therapeutic vaccines, irials for prostate cancer see Round Table meeting held on April 1, 2005 in Boston, Massachusetts. This meeting was paid for by Cell Genesys. Contents (with charts and explanatory cartoons):

FUNDAMENTALS OF TUMOR IMMUNOLOGY
RELEVANT TO PROSTATE CANCER IMMUNOTHERAPY
Charles G. Drake, MD, PhD
The Sidney Kimmell Comprehensive Cancer Center Johns Hopkins Medical Center

POXVIRUS VACCINES
Robert S. DiPaola, MD
The Cancer Institute of New Jersey

TUMOR CELL VACCINES
Jonathan Simons, MD
Emory University School of Medicine

DENDRITIC CELL VACCINES
Johannes Vieweg, MD
Duke University Medical Center

Cell Genesys says the 22 patients in their Phase 2 trial of GVAX who received the most effective tolerable dose — “the dose that is comparable to that being employed in the company’s ongoing Phase 3 program”– reached median survival of 35.0 months.

This trial enrolled men with hormone refractory prostate cancer. Cell Genesys previously reported final median survival results from its first multi-center Phase 2 trial of GVAX as 26.2 months.

“The survival results from the two, independent multi-center Phase 2 clinical trials compare favorably to the previously published median survival of 18.9 months for metastatic hormone-refractory prostate cancer patients treated with Taxotere(R) (docetaxel) chemotherapy plus prednisone, the current standard of care for these patients. The company’s ongoing Phase 3 program is designed to confirm this potential survival benefit for GVAX immunotherapy for prostate cancer.”

Read the full press release

In study 1 (D9901), median time to cancer progression in patients on Provenge was 11 weeks, and for patients taking placebo, 9.1 weeks. That’s a difference of a week and a half.

Let’s remember Stephen Jay Gould’s mantra “the median is not the message” and ask what might this median perhaps conceal and what might show up in clinical use if FDA OK’s Provenge (which I believe they will, given the FDA official’s intervention to encourage yes-voting and with Andrew von Eschenbach, M. D., a prostate cancer survivor, former head of both NCI and FDA, still in charge of FDA and calling for a cure for cancer by 2015).

Weel 8 and week 16 would have been time for patients in this trial to receive the first of their CT and bone scans likely to show progression of measurable disease.

Some men have such slow doubling time that they would show no progression by 12 - 16 weeks or more. Even in this trial, which enrolled men with advanced prostate cancer, some may have had slowly progressing, long doubling time disease. The entry criteria at some points in the series of trials excluded patients above Gleason 7.

In addition, as the company says, it may take longer than8, 12 or 16 weeks for Provenge to have any impact on progression, in which case, some men with rapid doubling times may have been kicked out to take Taxotere before the vaccine kicked in for them.

Presumably the distribution of rapid and slow progressing patients between the study arms was fairly even, I’ve not seen data on this. nor mention of rate of doubling, velocity, and so on.

Eric Small M.D. in his slide show at the Milken retreat Oct 2006 said Provenge:

• Raises questions about TTP as an appropriate endpoint for biologic/immunotherapy trials in which progression can occur before biologic effect.
• Probably not appropriate for rapidly progressing patients, given time course of immune response (2 to 3 months)

He recommends, if Provenge receives FDA approval, for doctors to consider selection of appropriate patients (ye old cherry-picking, always a good idea) based on a nomogram.

The slide show includes this statement:

PSA fall of > 50% in 6/62 (10%) of AiPCa patients

(i.e., 90% of patients did not have that fall. Well, is PSA fall an expected signal that Provenge is working? Maybe not, but if, say, bone mets don’t slow rate of growth, Provenge isn’t working).

Small’s “cherry-picked” graph on slide 4 is from a patient with 10-12 week doubling time who entered the trial with a fast-rising PSA but below PSA 20. In weeks, his PSA fell below 5, then immediately began bobbling and climbing with a slow, flattish slope.

Slide 7 graph seems to indicate that slightly more that 50% of both Provenge and placebo patients progressed by week 16. The steep dive in the slope to that time may indicate that just about as soon as the patients were scanned (or the second time) it was seen that those 50% were still progressing (at which time slow-progressors showed nothing?).

This may be where Small’s point applies, “Probably not appropriate for rapidly progressing patients, given time course of immune response (2 to 3 months). But it could also be that the vaccine just doesn’t do anything much.

(Could it be that if such patients received the vaccine earlier, in biochemical only progression stage, they could afford to wait for a slow time course of immune response?)

Beyond 16 weeks the two lines on the graph begin to divide and, with some 50% of both sets of men still in the trial, those on vaccine begin to show a slower rate of progression.

Even at the this stage, though, the difference in time to progression amounts to just a few weeks. In study 1, as the number of strong responders dwindled to less than 5%, at least one man who did NOT receive Provenge was still hanging in progression-free till 72 weeks,
just shy of the last men progression-free on Provenge.

Isn’t this more or less how median TTP advantage for Provenge turned out, in the final analysis, to be the statistically insignificant sliver of difference between Provenge 11 weeks, placebo arm 9.1 weeks?

That couldn’t be statistically significant given that nothing can predict which few men will make it beyond the median to this extent. If one man lives 72 weeks without disease progression and another one or two or 3 men live 75 weeks before their scans show progression, does that mean treatment did it? Not if 50% of men on that treatment progressed by 16 weeks. To be on that long tail end of the chart is the patient’s hope, the patient’s aim. It’s the place where “the
median is not the message.” In this trial it was not shown that Provenge made the difference in getting there. Perhaps a different analysis of the TTP data would have shown this but instead Dendreon focused on survival data.