Soft tissue sarcomas (STS) are relatively rare and represent about 1% of all solid tumours. There are more than 50 subtypes of STS, making it an extremely broad and diverse cancer type.

What are they?

Simply put, they are malignant tumours that arise in the soft tissues of the body, such as the muscle, fat, tendons, nerves, even the synovial tissues in joints are not exempted. As such, with the exception of those originating in muscle (leiomyosarcomas) and gastrointestinal tumours (GIST), they tend to be small and difficult to reach cancers. They differ from bone related sarcomas such as osteosarcoma and Ewings Sarcoma, which are usually considered separate categories.

The most common types of soft tissue sarcoma include malignant fibrous histiocytoma (MFH, more recently renamed by leading sarcoma pathologists as high grade undifferentiated pleomorphic sarcoma or HGUPS), liposarcoma and leiomyosarcoma.

The challenges of treating STS

As you can imagine, with such a diverse group of tumours, in many ways they each represent different cancers whose only commonality is the original site of origin. I actually struggle to think of a more heterogeneous group of tumours. Each one may have a different driver e.g. we know that GIST is largely KIT driven and is therefore responsive to KIT inhibitors such as imatinib (Gleevec), sunitinib (Sutent), pazopanib (Votrient) and regorafenib (Stivarga).

Some STS are known to be chemo-sensitive, at least for a few months, such as angiosarcoma, leiomysarcoma and synovial sarcoma, while many (most) are sadly highly chemo-resistant and do not respond well to cytotoxic therapy. Surgery is usually the standard of care for many of these subtypes as a result.

This means that while common cytotoxics such as ifosfamide and doxorubicin might be helpful in treating angiosarcoma, ifosfamide in synovial sarcoma and gemcitabine with or without docetaxel giving a benefit of 6.3 months vs 3 months in leiomysarcoma, many will see litle or no effect from therapy. In essence, a broad catch-all trial of STS needs to be carefully thought out viz inclusion criteria because if too many non-chemosensitive types are included then they may well negate any positive gains seen with chemo-sensitive subtypes.

It is likely that such allcomer trial designs didn’t help trabectedin (Yondelis), ridaforolimus (mTOR) in the SUCCEED trial or palifosfamide, which Ziopharm reported failed to meet the primary endpoint of PFS in the PICASSO 3 trial this morning. While there was no doubt that some patients did well on ridaforolimus with extended periods of stable disease (SD), others didn’t and without a biomarker from phase II trials to predict which patients would do well with mTOR therapy in the phase III registration study, the odds become little better than Russian roulette for registrability.

All of these results were most unfortunate for patients living with the disease, since there are so few options and effective new approaches would be most welcome.

What about palifosfamide?

Palifosfamide is essentially a different formulation of ifosfamide with a few extra excipients also added. This type of approach is usually something that makes me wary – new formulations might improve toxicities or drug delivery (in theory) but they rarely move the efficacy needle in a big way. By that, I mean one might reasonably expect small increments at very best, certainly not paradigm shifting change in the way we see from new approaches with targeted agents such as treating a very clearly defined subtype such as GIST with a KIT inhibitor.

Ultimately, the lack of a clear target means that chemotherapy should be given to chemo-sensitive subtypes of STS. An examination of the PICASSO 3 registration trial, however, highlighted that while a diagnosis of STS was the key inclusion critiera, only GIST and Ewings Sarcoma were excluded.

This is an extremely broad allcomers trial and probably doomed the study to predictable failure from the start – little was learned from the ridaforolimus experience, unfortunately.  It’s well known that I’m not a fan of either chemotherapy or catch-all studies and this throw-it-at-wall-and-hope-it-sticks approach was no exception.

The chance of success would likely have increased by excluding known chemo-insenstive subtypes and including only known chemosensitive subtypes such as angiosarcomas, synovial sarcomas and leiomyosarcomas. Yes, recruitment would be much slower, but a smaller N number in a more sensitive population would be less risky and also less likely to have the treatment arm full of patients who were unlikely to respond.

Where do we go from here?

The development of $ZIOP’s palifosfamide is likely headed for dog drug heaven at this point, given the press release from Ziopharm this morning gave no concrete details about the PFS numbers, stating that the:

“Phase 3 trial of palifosfamide (ZIO–201) for the treatment of metastatic soft tissue sarcoma in the first-line setting (PICASSO 3) did not meet its primary endpoint of progression-free survival (PFS). The study’s independent data monitoring committee (IDMC) has recommended that patients be followed for overall survival (OS), the study’s secondary endpoint, however the Company does not expect to continue follow up for OS.”

Until we more clearly define the biology of the various subtypes of STS and develop targeted agents for each subgroup based on what’s actually driving the tumour, we aren’t likely to see much progress in the short term.  Given the recent phase III failures from trabectedin, ridaforolimus and now palifosfamide, I find it hard to see how Threshold’s TH-302 will fare any better.


Well, I do think pharma and biotech companies need to stop treating STS as one disease and instead, focus on the different subtypes, much in the same way that imatinib revolutionalised the treatment of GIST and encouraged other companies to enter what previously was considered a tiny market – it can be done.  Doing yet another all comer/catch-all trial in a heterogeneous population without a clearly defined mechanism of action specific to a well tested subgroup, accompanied by more overly broad inclusion and exclusion criteria is sadly doomed to fail.