This coverage is supported by an educational grant from

Early Clinical Experience with Yttrium-Labeled Radioimmunotherapy in Patients with Low-Grade Non-Hodgkin's Lymphoma

 

 

 

Presented by Leo I. Gordon, MD, at the Radioimmunotherapy of Non-Hodgkin's Lymphoma symposium held at the American Society of Hematology 42nd Annual Meeting, December 1, 2000, in San Francisco, California.

When we begin to consider the subject of early clinical experiences with yttrium-labeled radioimmunotherapy in patients with lymphomas, it is useful to review certain information about the diseases themselves, and then begin to think about how the radiolabeled monoclonal antibodies may play a role in treatment.

The incidence of lymphoma is increasing. There are about 50,000-60,000 new cases per year, with 240,000 cases at any given time. Lymphomas can, in most classifications, be divided almost equally, with low-grade, intermediate-grade, and high-grade each occurring in about one-third of cases.

This discussion of the role of radioimmunotherapy in the treatment of lymphomas will center on low-grade and intermediate-grade lymphomas, because the initial trials of this therapy have been conducted in patients with those subtypes.

In the case of low-grade lymphomas, most patients present with advanced stages of the disease, and we think of them as incurable by conventional therapy. This status, however, makes them candidates for novel therapy.

With intermediate-grade lymphomas, about 40% of patients are thought to be curable with conventional chemotherapy. Median survival is about 2 to 2.5 years, and those 60% of patients that are not curable by conventional therapy are candidates, again, for novel approaches.

In both low-grade and intermediate-grade lymphomas, we are able to take advantage of the fact that the diseases are predominantly of B-cell origin, and for the most part they express the CD-20 antigen, which lends itself to targeting with antibodies.

Keep in mind that in all of the trials discussed in this report, yttrium is actually linked to the murine antibody, not to the humanized antibody. This is the case because there was concern that since half-life is longer in the humanized antibody, there would be extensive exposure to the radiolabel, and, therefore, more toxicity.

Zevalin is yttrium linked to the murine monoclonal antibody C2B8 by means of a novel chelate -- the so-called MX-DTPA chelate -- which is conjugated to the antibody, forming a strong urea-type bond. This bonding results in very stable retention of yttrium.

This antibody targets the CD-20 antigen, which is hydrophobic, and, most importantly, does not shed, internalize, or modulate. This means that once the antigen is exposed to the antibody, inside or outside the cell it is not lost.

The initial Phase I/II trial of Zevalin, launched about three years ago, was a dose-escalating study of 0.2, 0.3, or 0.4 millicuries per kilogram (mCu/kg) of yttrium-labeled monoclonal antibody. Patients with low-grade, intermediate-grade, or mantle-cell non-Hodgkin's lymphoma were eligible. Patients in all of the trials discussed must have had less than 25% lymphoma involvement of the marrow, baseline platelet count of greater than 100,000, at least in the initial studies, no prior stem cell transplant, and, in the initial trial, no prior treatment with Rituxan.

On day zero patients first received a dose of Rituxan -- so the humanized antibody first -- followed by an injection of indium-labeled C2B8 for imaging purposes. The rationale for giving the Rituxan first was based on Phase I data from Susan Knox and others, which suggested strongly that the binding of a radioimmunoconjugate to lymphoma was enhanced if there was pre-treatment with cold antibody. The reasons for this aren't clear, but the prevailing opinion is that you can take up binding sites in the periphery and allow more of the radioimmunoconjugate to get to the tumor through this method. There is some thought from others, however, that by giving the Rituxan first a localized tumor response, in which cytokines are released and binding sites for the radiolabeled antibody are freed up, is induced.

On days zero through six, scans and dosimetry calculations were conducted. After the indium injection, the dose to normal organs was determined. Then, on day seven, another dose of cold antibody was administered, followed by the yttrium-labeled C2B8 given in about a 10-15 minute injection.

In all of these trials treatment was administered on an outpatient basis.

Looking at the original group of 57 patients, the median age was 60, most were male, and each had undergone prior therapy. The number of prior therapies ranged from one to eight, with a median of two. Most patients had been diagnosed with lymphoma about four years previously, and there were clearly displayed variations in terms of resistance to chemotherapy in the cohort. About 40% of patients had bone marrow involvement. About 25-27% of patients had extra-nodal involvement. Fifteen percent had splenomegaly, and 37% of patients had bulky disease, defined as greater than a seven centimeter mass.

The overall response rate in all 51 evaluable patients was 67%. If you look, however, at the 34 patients who had low-grade lymphoma, the response rate was 82%. Interestingly, 27% of patients had complete remission.

In the 14 patients with intermediate-grade lymphoma the response rate was lower at about 43%, although this patient subset showed a surprisingly high complete remission rate.

Of interest, there were three patients in this initial trial with mantle cell lymphoma. And while the response rate was zero in these patients, at least two of them had massive splenomegaly, both of whom showed evidence of dramatic reduction in the spleen size, but no response in their indicator lesions following therapy. What this appears to indicate is that we have what has been referred to as a 'sink phenomenon' with the radioimmunotherapy; that the dose of radiation will go to the bulkiest site of disease. This, then, gives us pause about future studies and how to approach this problem to see if we can do something about the spleen prior to radioimmunotherapy.

In terms of response and time to progression there really was no significant difference between the 0.2 mCu/kg, 0.3 mCu/kg, and 0.4 mCu/kg doses. Also in these studies it was determined that the 0.4 mCu/kg dose was the maximum tolerated dose, although it should be noted that that dose was not surpassed. So while the traditional scheme for determining maximum tolerated dose wasn't used in these trials, 0.4 mCu/kg seemed to be the dose that was well tolerated, with blood counts recovering usually within about two weeks.

The toxicity was primarily hematologic and transient, and there was no major organ dysfunction in these initial studies. The mean serum immunoglobulin levels remained within the normal range and only 4% of patients had a decrease of 50% from baseline. Over a one-year period 6% of patients developed infections which required hospitalizations, and all of these recovered. The human anti-mouse antibody (HAMA) and the human anti-chimeric antibody (HACA) response occurred in about 2% of patients.

If you look at the hematologic toxicity, specifically in the 0.4 mCu/kg cohort, the median platelet nadir was about 50,000, and the median time to recovery was 14 days. The median neutrophil nadir was 1,100, with recovery in about 10 days. And the median hemoglobin nadir was about 9.9 grams/deciliter. In short, these toxicity data indicate a fairly acceptable toxicity in the 0.4 mCu/kg cohort in this Phase I,II trial.

According to indium scans taken at 144 hours, there was fairly impressive uptake of the antibodies at tumor sites. These scans can also be used as a measure of hepatic or other organ uptake, and one of the focuses of this Phase I/II study was to get an assessment of dosimetry to try and correlate dosimetry, if possible, with toxicity. So in the initial group of patients, the dose of yttrium was measured in two ways: it was measured directly and it was also measured as a predicted value from the indium scans. There was fairly good correlation between the two measures, so in subsequent patients the dosimetry was only done on the indium scans.

There was no correlation of toxicity with absorbed radiation doses. If you look at absolute neutrophil count nadirs and you measure them against the median red marrow dose derived by scans or by blood samples, and against the median total body dose, there was no correlation. In fact, surprisingly, the best correlation with neutrophil toxicity and platelet toxicity was the extent of marrow involvement. So patients with between 20 and 25% involvement had more toxicity than those between with between 15 and 20% involvement, etc.

A follow-up study conducted after the initial Phase I,II trial involved 35 patients with platelet counts between 100,000 and 150,000, who were treated with a lower dose of Zevalin at 0.3 mCu/mg. The design was similar so that Rituxan was first given, indium given, and then Rituxan followed by Zevalin. Bone marrow was involved in about 65% of patients, 27% of patients had splenomegaly, 50% had bulky disease, and again, there was no major organ dysfunction. Toxicity was, again, primarily hematologic and reversible. In these patients with lower platelet counts, the median absolute neutrophil count nadir was 600 instead of 1,100, and median platelet nadir was about 31,000. The overall response rate was about 67%. So, there is similar data in a slightly higher risk cohort of patients.

In summary, these trials have shown overall response rates on the order of 80% for yttrium-labeled monoclonal antibody therapy in the treatment of low- and intermediate-grade non-Hodgkin's lymphoma. The toxicity profiles showed that primary toxicity was hematologic and reversible. The maximum tolerated dose of the radionuclide was 0.4 mCu/kg, and treatment was administered on an outpatient basis.

It is hoped that these trials will pave the way for future studies using radioimmunotherapy in patients with non-Hodgkin's lymphoma.