Blinded by the light: from Jack O'Lantern mushroom to chemotherapy with risk of retinal damage

December 16, 2005. Irofulven is a synthetic version of a chemical found in the fungi family that includes the poisonous jack o'lantern mushroom (Omphalotus illudens). The jack o'lantern contains two chemicals, illudin S and illudin M, which extremely toxic to cells. Researchers made irofulven (6-hydroxymethylacylfulvene) from illudin S.

Illudins belong to the sesquiterpenoid plant family of resinous plants containing essential oils called terpenes (the old spelling for turpentine).

Irofulven (also known as MGI 114, hydroxymethylacylfulvene, or HMAF) is classed as an alkylating agent, like cyclophosphamide and mytomycin C. Rapidly absorbed by tumor cells, Irofulven binds to DNA and protein targets. This interferes with DNA replication and cell division -- genetic process which which tumor cells must do to multiply. The drug causes tumor-specific apoptosis ("cell suicide").

This compound looked extremely promising at the stage of lab testing on tumors in mice, when it was known simply as MG1 114. One published report said it showed "high activity against a wide spectrum of human cancers, including lung, gastric, breast, colon and nasopharyngeal tumours, as well as against melanomas. Daily administrations were found to be more effective than intermittent regimens, but intravenous (i.v.) and intraperitoneal (i.p.) routes resulted in a similar efficacy. In a comparison with other anti-tumour drugs, including paclitaxel, cisplatin, irinotecan, 5-fluorouracil (5-FU), doxorubicin, etoposide and vindesine, MGI-114, in general, showed a higher or equivalent efficacy."

MGI PHARMA, the company that makes irofulven, has clinical trials underway for people with refractory or recurrent advanced epithelial ovarian cancer, hormone-refractory prostate cancer, recurrent malignant glioma and inoperable liver cancer. Currently, the company is recruiting patients for a randomized phase II study comparing irofulven to mitoxantrone plus prednisone in patients with hormone-refractory prostate cancer (HRPC). Only patients "whose disease has progressed following Taxotere® treatment" are eligible. This study is recruiting worldwide in locations from the USA, Canada, Brazil, Chile and Peru to France, Croatia, Romania and the Russian Federation.

Before entering this trial patients may need to know, firstly, that three years ago MGI PHARMA halted the Phase III irofulven clinical trial for refractory pancreatic cancer patient when it became clear that the standard drug worked better.

Secondly, published results from an earlier Phase II single-agent trial in Texas for men with hormone-refractory prostate cancer are not all that impressive.

Thirdly, irofulven has one relatively rare and possibly disturbing side effect -- it interferes with vision and may damage the retina of the eye.

Some may decide that this risk is worth taking given that it is usually a temporary visual disturbance (a few hours or days of flashing lights etc.) and some other drugs carry fatal risks such as risk of a stroke or a heart attack. But so far it has affected about 1 out of 4 patients, and some have suffered permanent damage.

As it happens, this earlier trial for hormone refractory prostate cancer apparently was designed to enroll patients who had not previously taken any chemotherapy (below). This is a fair way of testing whether a drug will do as well as other first line chemotherapies. But it tends to target patients with more restricted options or waylay patients from available options.

How so? In this case, patients with good enough health insurance and good enough doctoring probably would have been able to access Taxotere by the time of this trial, even though FDA approved Taxotere and prednisone for prostate cancer only in May, 2004, after this trial had closed. Two large randomized Phase III trials with Taxotere-based chemotherapy showed prolonged survival and a positive influence on pain and quality of life for men with advanced prostate cancer. At the time of the Texas Phase II trial for advanced prostate cancer, Taxotere was already available off-label, though access to it was limited by access to a medical oncologist who would prescribe it and a health insuarance company or private fund to pay for it.

Now that Taxotere is approved and available for men covered by Medicare, MGI Pharma is looking to recruit from the pool of patients who have already taken Taxotere. This is only ethical given the considerable median survival advantage demonstrated for Taxotere, yet perhaps still not ethical enough.

The timeline and the geography of trial locations may send up a couple of red flags for people concerned about people in clinical trials:

• Will patients who have already taken Taxotere respond better or worse to irofulven than the chemo-naive patients?
• Were the patients who entered the earlier Texas trial (in compliance with its demand of no previous chemotherapy) well-advised to do so when Taxotere was so close to FDA approval and evidently was working quite well in off-label use for men with advanced prostate cancer?
• Compared to entry into a clinical trial of a drug whose toxicities are just emerging and possibly not well explained to patients from Canada to Chile to Croatia, do certain other FDA and European approved drugs commonly used off-label for prostate cancer -- such as Taxol and Carboplatin -- present as good or better choices today for second-line chemotherapy after Taxotere fails?

Companies like Cancer Consultants are stating that "Irofulven can be added to the list of chemotherapeutic agents active in prostate cancer. It will be of interest to see the outcomes of combination chemotherapy studies."

Active is a relative word in chemotherapy. Results from the Phase II single-agent trial in Texas for men with hormone-refractory prostate cancer were published in February 2005 issue of American Journal of Clinical Oncology. This trial recruited 42 men (median age of 73 years) with hormone-refractory prostate cancer who had not received prior cytotoxic chemotherapy. The forty-two patients received a median of 3 courses of Irofulven. Thirty-two patients received at least 2 courses of therapy "and were evaluable for efficacy." This means 10 men dropped out of the study before completing two courses. What happened to those men?

Of note, the trial administrators (doctors and the company) based their estimate of drug efficacy on only the patients who were able to tolerate at least 2 courses on drug. Effectively, this removes one quarter of the patients from consideration. Although this way of presenting data is quite standard for clinical trials (not unduly massaged) it nonetheless cloaks part of the reality of a trial.

In such a small study, de-selection of 10 men has a warping effect on result percentages. "Four patients (13%) had a partial response, with a median duration of 2.9 months (range, 2.6-5.8 months)." the authors write. "Twenty-seven patients (84%) had disease stabilization and 1 patient (3%) progressed on study. Median progression-free survival was 3.2 months (range 2.3-4.2 months), with a median progression-free survival of 4.2 months (range, 3.5-6.9 months) for responders."

These kinds of calculations are allowed in clinical trials, yet in the real world, if 42 men sign up and 10 drop out after less than two rounds of treatment, you would say, 4 out of 42 had a partial response and 27 out of 42 got disease stabilization (not 4 and 27 out of 32) and 10 out of 42 for some reason got nothing or couldn't stand it. Maybe those 10 moved on quickly to more promising treatment, maybe nothing could help them. Maybe some of them were diverted from drugs with higher probablility of therapeutic effect, and wasted their time.

A strikingly favorable result that leaps out from Cancer Consultants' summary is that "Neutropenia or thrombocytopenia occurred in only two patients." It is true that Grade 4 toxicities (severe side effects) of lowered platelets (thrombocytopenia), anemia, and low white blood cells (neutropenia) occurred in only 1 patient each. The most common side effects were asthenia (weakness, lack of energy, fatigue), vomiting, nausea and infection. Rather troubling is the finding that "Ten per cent of the patients developed infection without grade 3/4 neutropenia."

Looking a little closer it emerges that in this trial as well as others, side effects included "visual disturbances." Indeed, University of Texas Health Science Center at San Antonio has reported a case of acute retinal toxicity from irofulven.

Hope springs eternal and irofulven might turn out to be a useful drug for advanced patients if oncologists dose it carefully and combine it well with other chemotherapies and supportive drugs. Is there any evidence, though, to encourage men to consent to enter into the ongoing Phase II trial and expected Phase III trials for this drug?

Would an informed patient choose to be randomized to take

• Irofulven + prednisone
• Irofulven + capecitabine (Xeloda®) + prednisone
• Mitoxantrone + prednisone

before taking other better established combination chemotherapy, especially when patients in trials don't receive optimum supportive care and dose adjustment?

Why, when it comes down to it, go through the hassle of entering a clinical trial if you know ahead of time that:

• patients who had taken no previous chemotherapy had a 1 in 4 chance (25%) of dropping out before taking 2 rounds of the drug.
• even if you can tolerate the drug for longer than that, and attain stable disease or better, you probably will not receive supportive medications sufficient to alleviate common chemotherapy side effects like fatigue, vomiting, nausea and infection.
• you might experience visual disturbances, which currently no one knows how to avoid or treat except by stopping the drug or lowering the dose and seeing if the symptoms abate.

Therapeutic index (see sidebar above) is key in new drug development. For all the romance of extracts from plants collected in bog, swamp or forest, drugs obtained from distilled plant parts may be intolerable when used at cancer-killing levels. Synthetics may actually turn out to be a little kinder and gentler. This was true for docetaxel, the lab-made form of a taxane extracted from the yew. Or take ptaquiloside, a toxin in bracken which has been tested for anti-cancer activity.

Bracken might seem harmless to humans -- people in Japan and parts of Canada eat bracken shoots, but only the youngest shoots before they unfold, known as fiddleheads. Traditionally, fiddleheads are eaten seasonally, not every day, and cooked in changes of boiling water. Which is just as well, for as science writer David Bradley explains, bracken ptaquiloside can cause "severe leukopenia - a form of white blood cell anemia, - thrombocytopenia - an abnormally low blood platelet count - and hemorrhagic syndrome." These sound like side effects of intensive chemotherapy to treat cancer.

A cancer patient probably wouldn't want to take that kind of chemotherapy without some supportive medications to get through the rough patches. But before a drug is approved it has to go through a string of clinical trials. At the start, tolerable dose is unknown and efficacy (whether the drug will work as well or better than existing drugs) has not been demonstrated. Also, in many clinical trials, supportive medications to reduce side effects if they emerge are not always offered or allowed.

In the effort to lower side effects while boosting therapeutic power, biochemists and medical oncologists take account of how long the body tissues will be exposed to a chemical. Iiludins are toxic to most cells after prolonged exposure (a couple of days or more). Used for shorter periods (no more than 2 hours) they show "selective toxicity." They target cancer cells of various types and do less harm to normal cells. Irofulven's toxic index is potentially manageable. The effect on cancer cells, though, may not yet be strong enough to bring about "durable responses" in patients. McMorris says, "We are now seeking ways to increase the toxicity of acylfulvenes in malignant cells relative to normal cells."

If irofulven is shown to be effective at doses which allow minimal or tolerable visual side effects, and if no more serious side effects emerge, then some patients and their doctors might compare this drug's profile quite favorably with side effects associated with taxanes and platin drugs. Although nothing like the median survival rates for Taxotere are on record for irofulven, possibly combining it with prednisone or with that plus capecitabine will pan out.

Preclinical studies of irofulven with or without mitoxantrone or Taxotere have been done recently on human prostate cancer tumors implanted in mice. "These studies demonstrate," the researchers write, "that irofulven displays strong activity as monotherapy and in combination with mitoxantrone or docetaxel against androgen-independent prostate cancer in vitro and in vivo; thus, supporting the clinical investigation of irofulven against hormone-refractory prostate cancer. " Unfortunately, this research is unable to assess the side effects or toxicities likely to surface from that combination if/when used on significant numbers of men.