|Values are valid only on day of printing.|
Published: November 2008Print Record of Viewing
This presentation will open in a new tab or window.
Dr. Hanson will discuss the various test options for risk stratification in chronic lymphocytic leukemia (CLL). This is the second of two presentations on CLL and laboratory testing.
Presenter: Curtis A. Hanson, MD of the Division of Hematopathology at Mayo Clinic
Welcome to Mayo Medical Laboratories' Hot Topics. These presentations provide short discussions of current topics and may be helpful to you in your practice.
Our speaker for this program is Dr. Curtis Hanson, from the Division of Hematopathology at Mayo Clinic. Dr. Hanson will discuss the various test options for risk stratification in chronic lymphocytic leukemia (CLL). This is the second of two presentations on CLL and laboratory testing. View Part 1.
Our goals today will be to understand the various prognostic risk factors that are used in evaluating chronic lymphocytic leukemia (CLL), to realize how clinicians use prognostic risk factors in the management of patients, and to understand how risk factor analysis is used in conjunction with diagnostic testing in CLL.
Please note that there is a previous Hot Topics on CLL: “Chronic Lymphocytic Leukemia: New Approaches for a Common Disease” That particular Hot Topic dealtwith the diagnostic criteria of CLL and monoclonal B-cell lymphocytosis as well as minimal residual disease detection in patients with CLL.
This slide highlights the three key points that a laboratory needs to be aware of in evaluating patients with CLL.
First, one needs to reach an accurate and correct diagnosis.
Second, in today’s environment, risk assessment is critical in order to help clinicians provide the appropriate therapy and follow-up for these patients.
Finally, there is the assessment of outcome and potentially the detection of minimal residual disease. As I mentioned previously, the diagnostic and outcome assessment were previously discussed, and today we will be focusing on risk assessment.
There are various ways to assess risk in patients with CLL. Clinically, staging methodologies have been around for a long time and have been used quite successfully. The staging techniques used have either been Rai or Binet with the Rai stage being the most common one used in this country. Morphology also offers some information as there are marrow patterns of infiltration that may give an idea for risk assessment. I will also briefly talk about fragile cells and lymphocyte doubling time. It is important to note that other aspects, such as the detection of prolymphocytes and Richter’s transformation, will not be discussed today.
Immunophenotypic markers are increasingly being used for prognostic risk factor assessment in CLL, including CD38, ZAP-70, and most recently CD49d, and molecular and other genetic-type assays, including mutational status of immunoglobulin variable heavy chain region (IgVH) as well as metaphase and FISH studies, looking for specific gene anomalies.
This slide shows a Kaplan-Meyer graph for patients with CLL and distributed by Rai stage. One can see that those patients with low-risk Rai stage disease don’t reach a 50% treatment-free point until approximately 8 years after diagnosis.
This falls to approximately 2-3 years for those with an intermediate risk, for as one can see that virtually all patients in high risk disease based on Rai staging require treatment early on in the course of their disease.
This slide demonstrates the typical nodular pattern of bone marrow infiltration that one can see in CLL.
On the left hand side of this slide is the nodular and interstitial pattern of bone marrow involvement. On the right hand side is the diffuse pattern of marrow infiltration, defined as having less than 5% of the cellularity as adipose tissue. Several studies have pointed out the association of this diffuse pattern of infiltration, if strictly defined, with more rapid progression of disease. Recent studies, however, are showing that this pattern of infiltration may not be independent of other prognostic risk factors.
In the blood, doubling time (DT) is the length of time it takes to double the white count and has been used as an indirect marker of progression of disease.
I also want to bring up the topic of fragile cells. We all know that fragile cells are a common finding in a peripheral blood smear of a patient with CLL. The lab techs often consider fragile cells as an irritant to be dealt with. The big question that remains is why do we see fragile cells in some cases, but not in others?
This slide shows examples of four different CLL cases. If we look at the two LH photos, the blue arrows point out the numerous fragile cells that are present, while being only rarely seen in the two cases demonstrated on the RH side of this slide.
A very interesting observation by Nowakowski and colleagues was that the number of fragile cells was actually related to the time of treatment in these patients. Those cases with numerous fragile cells failed to reach a 50% treatment mark up to 12 years post-diagnosis, whereas those who had few fragile cells progressed more rapidly to treatment, i.e., reaching a 50% point at approximately 6 years.
Using Immunophenotyping to diagnose CLL is a well-established methodology. Recently the use of immunophenotyping to identify various proteins that are associated with outcome in CLL have become established within the clinical laboratory.
The initial one identified was CD38. CD38 was developed originally as an early T-cell marker and then found to be present in mature T cells, T-cell ALL, AML, and in some cases B-cell lymphoproliferative disorders. In addition, it was identified in a subset of patients with CLL.
As it turns out, CD38 is stable in untreated CLL patients but may vary when patients undergo chemotherapy. Initially, CD38 was found to correlate with the IgVH mutation status, i.e., a poorer prognosis. Greater than 30% expression of CD38 has been used as the “cut-point” between calling a case positive or negative.
This Kaplan-Meyer curve points out the difference in time-to-treatment in early-stage CLL based on CD38 expression. Again, CD38 negativity is associated with a much longer time-to-treatment that those are that are CD38 positive.
ZAP-70 is an intracellular tyrosine kinase marker, normally expressed in T cells and NK cells, but has also been found to be associated as a risk factor marker in CLL. We know that ZAP-70 is involved with the signaling response, activation and proliferation of T-cells in particular.
In CLL, it has been discovered that B-cell receptor signaling is certainly much more efficient in ZAP-70+ cells thus implying an association with activation and proliferation of those B-cells.
ZAP-70 expression has also been associated with unmutated IgVH status, which is a worse prognostic finding. ZAP-70 appears to be stable over time regardless of treatment, as opposed to CD38.
The problem with ZAP-70 comes in some of the technical aspects in detecting this marker. It is certainly technically challenging within the laboratory. Various antibodies are available, and they all have some variation in their ability to detect ZAP-70. Since this is an intracellular marker, it requires making the cytoplasmic membrane permeable. Controls are needed to use to set both positive and negative staining and appropriate gating strategies must be employed.
Likewise, there are different methods analytically that have been used to calculate the percent positivity, thus a lot of problems have probably prevented this marker from being uniformly adapted in all flow cytometry laboratories.
ZAP-70 can also be detected by immunohistochemistry and tissue sections. However, this technique is certainly associated with a subjective interpretation and relatively few studies have been done showing its clinical utilization.
Several questions remain for ZAP-70 in CLL.
When validation studies for ZAP-70 are done, should they be done versus clinical outcome or IgVH status? Clearly clinical outcome is the gold standard to be used.
Do peripheral blood and bone marrow give comparable results? Bone marrow clinically validation studies have not been done, and thus is not the preferred specimen type.
For tissue stains is ZAP-70 fixation-dependent? We don’t know the impact that various fixatives and decalcification may have on ZAP-70. We do know that ZAP-70 is not stable and begins to deteriorate after 48 hours and is temperature dependent. It also must be realized that as of now, ZAP-70 does not appear to have any prognostic role in non-CLL B-cell lymphoproliferative cases.
This slide demonstrates an example case of ZAP-70 expression detection by flow in a patient with CLL. The top histograms are from the control. The upper left hand histogram quad stats are set such that the upper right hand quadrant contains approximately 0.10% ZAP-70+ B-cells.
In the upper right hand histogram, we see by using this gating technique in the control specimen that the majority of T-cells and all the NK cells are indeed positive as they should be for ZAP-70. This will serve as an internal positive control.
Once that is established, we can then look at the patient’s results in the bottom histograms by using the same quad stat settings as were established in the control. You can see that 9.2% of the total lymphocyte population is positive for both ZAP-70 and CD19 and 80.1% are ZAP-70 negative.
If one looks in the lower right hand histogram from this particular patient, one sees that the majority of T-cells and all the NK cells are appropriately ZAP-70+, which serves as an internal positive patient control.
To then calculate the percent positivity, the percent ZAP-70+, or 9.2%, is divided into the total B-cell percentage, or 9.2% + 80.1%, which would then equal 10% in this case. This would be interpreted as negative for this patient, remembering that 20% is used as the cut-off point for ZAP-70 positivity.
Here’s another example. Again, in the upper left histogram, quad stats are set based on 0.10% ZAP-70 positive B-cells. In the upper right histogram, one sees that the T-cells and NK-cells are appropriately positive.
In the lower left histogram, we see the patient of interest and indeed, in contrast to the previous patient, we see here if we take 62.5% divided by the total B-cell percentage, that we end up with 76% of the B-cells expressing ZAP-70, which would be interpreted as positive for this particular patient.
It should be noted that one needs to pay close attention to the T-cells in this particular analysis. This is a separate study of a control with quad stats being appropriately set based on the ZAP-70 B-cell expression of 0.10%, but note that the T-cells remain negative indicating that the technique to make the membrane permeable did not work, or possibly the cell stability was not acceptable. Based on this finding, the control should be rejected and the study reset.
As I mentioned earlier, immunohistochemistry for ZAP-70 can be done. On the right hand side is a demonstration of a positive stain; on the left, negative. I think you can appreciate the difficulty in interpreting ZAP-70 by this immunostain methodology.
This Kaplan-Meyer curve shows the separation of ZAP-70 positive and negative cases in early stage CLL, and again a much shorter time-to-treatment in those cases that are ZAP-70 positive.
CD49d is the newest prognostic marker in CLL. It is also known as “alpha 4 integrin” and acts as an adhesion molecule for various extracellular matrix components. It is thought to mediate cell-cell interaction through the binding of VCAM-1, and also is felt to serve as a signaling receptor that influences B-cell survival through the Bcl-2 family.
CD49d has two established clinical roles. One is actually as a therapeutic agent. Anti-CD49d has been available as a therapeutic agent for patients with multiple sclerosis and inflammatory bowel disease. Clinical trials are ongoing for using anti-CD49d in various hematologic disorders.
For the purpose of this talk, it should be noted that CD49d is gaining momentum as a prognostic marker in patients with CLL.
This slide from data that we have at Mayo Clinic shows that CD49d positive expression is associated with both a decreased time-to-treatment and a decreased overall survival as compared to patients that are CD49d negative.
Here’s an example flow histogram of CD49d in a patient with CLL. On the left graph, one can see that the CD19 positive cells are clearly CD49d negative, whereas on the right hand histogram, one sees the T-cells as expected showing CD49d expression.
This slide demonstrates a case of CD49d being positively expressed on B-cells in a patient with CLL.
Immunoglobulin variable region (IgVH) mutation status has been established as an important marker for prognosis in patients with CLL. This is a very complicated technique that requires PCR amplification of the IgVH region, followed by DNA sequencing. A mutation percentage is determined by dividing the number of mutations identified by the total base pairs analyzed. If the number of mutations is less than 2%, it is then considered as unmutated, which is a poor prognostic finding. If it is greater than 2%, it is considered as mutated which is associated with a good prognosis.
IgVH mutation status has been a very good marker when used in population studies. In our experience, it may not necessarily have the best reproducibility across laboratories, and without a doubt, it is a very sophisticated technique which is not easily adaptable to the clinical laboratory.
Here is an example of an IgVH sequencing histogram.
Again, another Kaplan-Meyer curve showing the separation between IgVH mutated and unmutated cases in early-stage CLL patients. And again, a much shorter time-to-treatment in those cases that are unmutated.
Genetic factors are also a very important finding in patients with CLL. Here we see both FISH and as well as metaphase studies in patients with CLL highlighting some of the more common anomalies that can be found in these patients. Since CLL cells do not easily divide in-vitro, FISH studies are the preferred methodology to be used in evaluating the CLL patient.
As noted on this pie graph, 13q- is the most common abnormality that can be detected by FISH studies in patients with CLL, with trisomy 12 and 11q- being the next most common anomalies identified.
13q- is a microdeletion in CLL that cannot be easily identified by metaphase analysis and can only be consistently identified using FISH techniques.
This Kaplan-Meyer curve points out the differences in overall survival based on the particular FISH anomaly identified. If we look in the upper right hand, both 13q- and normal FISH studies are associated with more prolonged survivals, with trisomy 12 being in the middle, whereas 11q- and 17p- are both associated with much shorter survival times.
To summarize risk assessment in CLL, there are various techniques ranging from clinical, laboratory and pathology studies to flow cytometry assays to genetic studies that all contribute in determining risk assessment in patients with CLL. The goal obviously, then, is to put these patients into a high, intermediate, or low risk assessment category.
This table attempts to do that by using the risk factors we have talked about today.
If we look first at the right hand column, for low-risk CLL, we see that it is associated with early Rai stage, 13q- as the sole FISH abnormality, a mutated IgVH and negative for CD38 and ZAP-70. One exception, those cases with an IgVH mutation involving the 3-21 region are considered high-risk. Certainly these low-risk patients should not require any therapy or at best very low toxic or early intervention trials that are now being studied.
If I go to the left hand column for those high-risk patients, these patients may be at any Rai stage, associated with the worst FISH prognostic markers of 11q- and 17p-, typically being unmutated IgVH or having a mutation of the 3-21 region, and typically positive, but perhaps negative with CD38 and ZAP-70. These patients typically require standard chemotherapy and certainly transplant for those patients with 17p- will likely be considered.
The remaining patients would fall into an intermediate risk category.
To summarize, when a patient with potential CLL comes to the laboratory, think of it as a three-step process.
First of all, the diagnosis is made based on the peripheral blood CBC and differential count as well as the immunophenotype. Second, risk assessment based on clinical, hematology, immunophenotyping and genetic and molecular studies.
Finally, assessing outcome post-therapy based on immunophenotype, potential immunohistochemistry and certainly clinical and MRI staging.
Today we have tried to cover this middle segment, or risk assessment.
Our goals today were to understand the various prognostic risk factors that are used in evaluating CLL, to realize how clinicians use prognostic risk factors in the management of patients, and to understand how risk factors analysis is used in conjunction with diagnostic testing in CLL. Hopefully over these two Hot Topic sessions, I’ve been able to answer some of your basic questions regarding that very common hematologic malignancy, chronic lymphocytic leukemia.