Side effects

Recently, there has been some discussion among CLL patients about the side effects of drugs used to treat the disease. Are there any treatments that are side effect free?

Unless there were a perfectly targeted drug that killed silently and without fuss, it is inevitable that all effective drugs will have side effects. Whether the side effects matter, is another question.

It is possible to classify side effects in a number of different ways: subjective versus objective; transient versus permanent; short or long lived; rare or common; intrinsic to the treatment or to the patient; inevitable or idiosyncratic.

Let's take the individual drugs that are used in CLL.

Chlorambucil has been around the longest and has relatively few side effects. It is an alkylating agent and as such it interferes with the integrity of DNA, causing breaks that often can't be repaired. These go on to induce apoptosis through the mitochondrial pathway. Its activity is not confined to the CLL cells, but it kills any dividing cell. Unlike many alkylating agents it doesn't cause hair loss, and it can be given quite safely by mouth. It does kill normal bone marrow cells but at the doses used in CLL this effect is not too severe. It kills normal lymphocytes too, but again the effect is much less than with other drugs. Its effect on the bowel is also not marked, though this is the effect most reported by patients. They often mention a metallic taste in the mouth and in about a quarter of patients nausea is mentioned. In a very small number of patients an allergic rash is seen.

Chlorambucil can be used as a pattern for the other drugs used in CLL. Many of them work like chlorambucil by targeting dividing cells so all dividing cells will be damaged, not only CLL cells but also hair, skin, bowel, red cells, platelets and neutrophils, T and B lymphocytes. Because these cells are dividing they are usually rapidly replenished, so although with all chemotherapy there is a tendency for the patient to become anemic, thrombocytopenic and neutropenic, these usually recover within a week or so. When hair loss occurs it is also transient, but recovery takes longer, usually not occurring until all courses of treatment have finished. Gut and bowel symptoms, on the other hand, are usually much shorter-lived, lasting only a few days.

On these transient symptoms, chlorambucil scores very well. Nausea may occur but it is seldom sufficient to cause the use of anti-nausea pills. Hair loss hardly ever occurs. Transient suppression of the bone marrow is mild and does not require the use of G-CSF to stimulate the neutrophils, Epo to stimulate the red cells or transfusions of either red cells or platelets. Damage to the skin only occurs in the rare case of allergy. There is some suppression of T lymphocytes, but this is seldom a problem unless the patient is severely immunosuppressed by his CLL or there have been many courses of treatment beforehand. The commonest T cell effects are triggering of shingles (or Zoster) and triggering of autoimmune hemolytic anemia (AHA).

Can there be permanent damage? Probably, though this is seldom a problem. Bone marrow stem cells can be damaged so that either the patient has long term bone marrow cytopenias or develops myelodysplastic syndrome which can transform to acute myeloid leukemia. It is always hard to be sure of these complications because if the marrow remains full of CLL it is difficult to know whether the CLL is suppressing the bone marrow or the chemotherapy. Of all the chemotherapeutic drugs that we use chlorambucil is least likely to cause these problems. I have never seen MDS or AML develop in CLL because of chlorambucil, though I have seen both develop in untreated patients.

The other alkylating agents that we use in CLL are cyclophosphamide and bendamustine. Compared to chlorambucil, cyclophosphamide is more likely to make you sick, much more likely to make your hair fall out, but less likely to cause bone marrow suppression. However, it is more likely to cause long term damage to the bone marrow and myelodysplastic syndrome and acute leukemia definitely occur. It seems less likely to cause AHA, perhaps because it targets B cells more than T cells. Cyclophosphamide can be given either by mouth or intravenously.

Bendamustine is still a bit of an unknown quantity, but it seems more toxic than chlorambucil under all headings, though we have to suspend judgment a little. It is likely that many of these drugs have very similar effects as long as they are given in equally effective doses. It then becomes a matter of which drug has the fewer side effects at equally effective doses. Unfortunately, manufacturers want their drug to look good so they compare it with suboptimal doses of chlorambucil. At these doses chlorambucil has fewer side effects, but how would these compare if enough chlorambucil were given? With fludarabine, we know that it is more toxic than chlorambucil when the latter is given in an equally effective dose, but we don't know about bendamustine.

Although fludarabine is not an alkylating agent it still makes use of the same mitochondrial apoptosis pathway to kill CLL cells and in in vitro testing it has the same sensitivity pattern as chlorambucil. It has roughly the same pattern of toxicity as chlorambucil only more so. However, it is much more toxic to T cells, knocking the T helper cell levels to about the same level as is seen in AIDS patients for two years or more. As a consequence shingles is commoner and AHA more severe. There are also occasional cases of CMV reactivation and pneumocystis pneumonia occurs. A transfusion into someone who has had fludarabine can allow the donor lymphocytes to remain alive and attack the bone marrow causing transfusion associated graft versus host disease, which is uniformly fatal. For this reason, we recommend that patients on fludarabine receive prophylactic Bactrim and acyclovir and any transfusion they require should be irradiated. Fludarabine was originally just intravenous, but an oral preparation is now available even in the US, which was last to see the convenience of this.

Fludarabine is often used with cyclophosphamide. It is certainly more effective when they are used together, but also more toxic. We insist that a blood test is taken two weeks after the combination starts because the incidence of severe neutropenia is so great that G-CSF or prophylactic ciprofloxacin might be needed to protect the patient from gram negative septicemia (which can kill in a few hours). However this combination is less likely to trigger AHA.

Prednisolone (or other steroids like dexamethasone and methyl prednisolone) is not a cytotoxic drug though it can kill CLL cells by a pathway different from the mitochondrial apoptosis pathway. In high doses it kills even cells with the del17p lesion. But it has a whole set of different side effects. Transiently it causes fluid retention, disturbances in the control of blood sugar, hypertension and labile mood, but when continued for longer periods steroids cause muscle wasting and thinning of the bones as well.

Vincristine is a cytotoxic drug that is not very toxic to the bone marrow, and it is not much use against CLL either even though it is used in the CVP and CHOP regimens. It does have quite unusual side effects though. It causes a peripheral neuropathy. Everyone loses their ankle jerks, but some people are paralyzed. More troubling is autonomic neuropathy, which causes problems with bowel movement leading to abdominal pains and constipation.

Doxorubicin, also known as Adriamycin, causes hair loss and vomiting that are much worse than with any other CLL drug, but it is also toxic to the heart in a cumulative way. No more than 9 courses can be given and fewer if there is pre-existing heart disease. Mitoxantrone is a drug with similar qualities and similar side effects.

The monoclonal antibody rituximab is not marrow suppressive and does not cause AHA. Its main side effect is an infusion reaction with flushing, shivering attacks, low blood pressure leading to fainting or collapse, changes in the pulse rate and possibly breathlessness. These symptoms are caused by the consumption of complement and the release of complement fragments that are vaso-active peptides. By giving the infusion more slowly this side effect can be avoided. Occasionally, patients become allergic to rituximab, but this is very, very rare.

Campath is less well directed that rituximab since CD52 is also on T cells. It too can cause infusion reactions, but these can be avoided by subQ administration. Instead it causes local skin reactions at the injection sites. These can be unpleasant, but they are not serious. The main side effect is suppression of T cells so that infections with viruses and fungi become a problem. Bactrim and acyclovir prophylaxis is necessary, but because reactivation of CMV is a real possibility, it is necessary to do weekly tests for CMV antigen by PCR. Ganciclovir is given if two successive tests are raised, even if there are no symptoms. Untreated CMV can be fatal. The T cells recover about 6 months after stopping Campath.

Any effective treatment can cause tumor lysis syndrome (TLS). this means that release of toxins from the dying CLL cells are not cleared quickly enough and can damage or even kill the patient. The main toxins are potassium and uric acid. Prevention of TLS is by allopurinol for the uric acid and lots of fluids for the potassium.

Revlimid is one of the drugs that cause TLS. It can also cause marrow suppression and it has a strange complication known as tumor flare, where a lymph node mass increases in size and may become painful. This can be suppressed by steroids, but it usually means that the treatment is working.

If there are any other side effects that I have not mentioned, please ask.