Lymphoid neoplasms: Steven Swerdlow on classification revisions
William Check, PhD
December 2016—In his CAP16 talk, “Lymphoma Diagnosis and Classification: My Search for the Holy Grail,” Steven H. Swerdlow, MD, acknowledged that the quest has been long and contentious and the resulting classification complex. Why, he asked, is lymphoma classification so complex?
It reflects an explosion of knowledge about the immune system and lymphomas and an increasing number of therapeutic targets requiring increasingly precise diagnoses. In other words, it reflects the complexity of the disease itself. As oncologist Alan C. Aisenberg, MD, PhD, wrote in 1995, “The complexity of non-Hodgkin’s lymphoma reflects the complexity of the lymphoid system” (J Clin Oncol. 1995;13:2656-2675).
One of Dr. Swerdlow’s main objectives at CAP16, and in an interview with CAP TODAY, was to help pathologists recognize changes in practice related to the 2016 revision of the WHO lymphoma classification, with a focus on non-Hodgkin lymphoma. The way to best serve patients with lymphoma, he says, is to use an up-to-date multiparameter diagnostic approach that incorporates clinical data, morphology, immunophenotyping, and cytogenetic and molecular findings, as needed and appropriate. “This does not mean that all studies need to be performed in all cases,” he says, “as we all aim to practice cost-effective medicine.” Even today, “histopathology remains critical,” says Dr. Swerdlow, a professor of pathology and director of the Division of Hematopathology at the University of Pittsburgh School of Medicine and UPMC Health System.
To characterize the difficulties in categorizing lymphomas, Dr. Swerdlow refers to his decades of involvement in this field as “my journey through the minefields of lymphoma classification.” For pathologists and other interested parties, he acknowledges, it has been “a source of frustration with constantly shifting sands and a field littered with the remains of academic battles.”
Ultimately, the winners of these past conflicts are the patients. Five-year relative survival of non-Hodgkin lymphoma patients increased significantly over the past 30 years. Better treatments and supportive care are critical contributors, but precise classification and diagnosis by pathologists have made it possible to use the new treatments optimally, Dr. Swerdlow says. Investigators cannot develop enhanced therapies and clinicians cannot apply those therapies to entities or specific targetable abnormalities that are either undiscovered or not diagnosed, he points out, emphasizing the critical role of pathologists in this process.
The famed hematopathologist Henry Rappaport, MD, established a widely used and important classification of lymphoma in the mid-1950s that was solely based on histopathology, like those that had preceded it. “The big revolution in lymphoma classification came in the early 1970s,” Dr. Swerdlow says, with the recognition that lymphomas were malignant neoplasms of the immune system, primarily of B cells and T cells, and should be classified as such. This formed the basis of the Lukes-Collins and Kiel “functional” lymphoma classifications.
As with most revolutions, it took time to be accepted. In 1977 oncologist Vincent DeVita, MD, later director of the National Cancer Institute, would still write, “There seems to be a consensus among clinicians that pathologists everywhere should adhere to the Rappaport classification” (Cancer Treat Rep. 1977;61:1223–1237).
Out of the back-and-forth claims and counterclaims and following the 1994 publication of the REAL classification by the International Lymphoma Study Group emerged the first of the modern World Health Organization classifications in 2001. Dr. Swerdlow calls it “not just a ‘cell of origin’ classification” but one recognizing distinct clinicopathologic entities, based on all the parameters noted above.
Newer findings included the notion that size of the neoplastic cells doesn’t always correlate with aggressiveness. “The older way of thinking about things could mislead you,” he says. “Lymphomas composed of large cells used to be thought of as necessarily aggressive and those with small cells thought to be indolent.” To illustrate this is not always the case, he showed an ALK-positive anaplastic large cell lymphoma that is not considered to be among the most aggressive lymphomas and a mantle cell lymphoma composed of much smaller cells that is. “Now nobody would have an issue with this idea,” he notes, but it took a long time for all to realize the importance of precise diagnoses no longer just based on cell size or growth pattern.
As the WHO classification was revised, first in 2008 and then in 2016, it became more complex. The number of lymphoid neoplasms went from fewer than 30 to more than 80.
Technological advances evolved along with the increasingly complex lymphoma classification. Dr. Swerdlow calls them “two parallel but intersecting universes.” He notes increasing numbers of monoclonal antibodies, available fluorochromes, and lasers for enhanced flow cytometric immunophenotyping studies; an explosion of cytogenetic FISH studies that can be performed on formalin-fixed, paraffin-embedded (FFPE) tissue; and the extensive use of PCR to detect clonal B and T cell populations even when only FFPE is available. Initially most monoclonal antibodies worked only in frozen sections or on cell suspensions. “Now innumerable antibodies have been developed along with improved antigen retrieval methods leading to our great reliance these days on FFPE immunohistochemistry,” he says.
Next-generation sequencing has led to the discovery of important gene mutations that have had a major impact on diagnosis and classification of lymphomas. Many can also be detected using older methods, Dr. Swerdlow notes, such as real-time PCR with mutant-specific primers for MYD88 L265P mutations.
Publication of the 2016 edition of the WHO classification of hematologic malignancies is expected in early 2017. Originally it was intended to be only a Web-based update, but it will now be published as a hard-copy monograph. The revised classification of lymphoid neoplasms has already been published, emphasizing the changes from the 2008 WHO classification (Swerdlow SH, et al. Blood. 2016;127:2375–2390).
“How do you deal with such an admittedly daunting classification?” Dr. Swerdlow asks. He rejects two extreme strategies: memorize the whole list or give up. His approach is to understand the basis for what is there and something about normal lymphoid cells and their development and how they can “go bad” as an aid in feeling more comfortable diagnosing lymphomas. “We need to understand basic immunology and what happens when normal lymphocytes become neoplastic,” he says. “You can’t memorize the whole classification but you can have the monograph handy where you’re working. Many people, including me, sit with that monograph by their side when they’re signing out their cases. I use it so much that the cover falls off.”
One important aspect of the updated WHO classification that pathologists should understand is that, as with the former WHO classifications, the entities it includes and the criteria for their recognition are based on published data and often simply serve to codify evolving preexisting practices. “There shouldn’t be any shockers for those who keep up on the lymphoma literature,” Dr. Swerdlow says.
One important concept of the new document is that it reflects growing conservatism in the diagnosis of lymphomas. “Instead of trying to diagnose as many lymphomas as possible, we try not to overdiagnose things as lymphomas that aren’t going to behave as true neoplasms,” Dr. Swerdlow says. “Our modern technologies have led to our ability to recognize tiny proportions or very focal neoplastic-appearing lymphoid populations.”
One example relates to what used to be considered “chronic lymphocytic leukemia (CLL) with a low count.”
“The name has been changed and we don’t even call it a leukemia or lymphoma anymore,” Dr. Swerdlow says. It is now designated as monoclonal B-cell lymphocytosis (MBL), with greater acceptance of this terminology. “The new thing for 2016 is that we need to distinguish low count [
The possibility of a nodal equivalent of MBL is also recognized. “With our current rules, even with subtle invasion in the lymph node, we have to call it small lymphocytic lymphoma,” Dr. Swerdlow says. A retrospective study from Dr. Swerdlow’s group and colleagues at Massachusetts General Hospital (Gibson SE, et al. Haematologica. 2011;96:1144–1152) suggested that lymph node infiltration by CLL-type cells without proliferation centers and with lymph nodes < 1.5 cm on imaging might be better considered as tissue-based MBL.
Similarly, in situ follicular lymphoma is now to be called in situ follicular neoplasia. And in situ mantle cell lymphoma (MCL) is getting a name change to in situ mantle cell neoplasia. Another significant change relating to MCL is that “we now view MCL as developing along two different routes,” Dr. Swerdlow explains. One pathway leads to classic MCL, which is considered an aggressive neoplasm that can progress to an even more aggressive form, and the other, usually where the B cells appear to have traversed the germinal center environment, leads to the leukemic, non-nodal MCL that involves the blood, bone marrow, and sometimes spleen and which is less aggressive but can also progress.
Discoveries made by next-generation sequencing are having an impact on diagnosis of small B-cell lymphoid neoplasms, such as lymphocytoplasmacytic lymphoma (LPL; Waldenstrom’s macroglobulinemia). About 90 percent of LPL/WM have been found to harbor MYD88 L265P mutations. Dr. Swerdlow calls this mutation “not specific but characteristic” for this neoplasm, saying it has already had an effect on how we define nodal LPL.
Changes are also occurring in handling the aggressive end of the B-cell lymphoma spectrum, including diffuse large B-cell lymphoma, NOS (DLBCL, NOS). There is a new requirement to distinguish the germinal center (GC) B cell from the non-GC/activated B-cell type. Dr. Swerdlow notes the “well-known prognostic implications” of making this distinction, as well as newer data suggesting potential therapeutic implications of making this distinction and mitigating the generally more adverse prognosis of the non-GC/ABC group (Molina TJ, et al. J Clin Oncol. 2014;32:3996–4003). “Immunohistochemical methods are currently the most widely used technique for making this distinction in our daily practices, but molecular methods using FFPE tissues have been reported and are on the horizon for wider usage,” he says.
A “hot topic” in the 2016 revision, Dr. Swerdlow says, is dealing with “B-cell lymphoma, unclassifiable, with features intermediate between DLBCL and Burkitt lymphoma [BL]” and the “double hit” lymphomas.
Double or triple hit lymphomas are those that have MYC and BCL2 and/or BCL6 rearrangements. Double or triple hit lymphomas that are in the blastoid/BL-like to DLBCL spectrum are included in one category. “These are very aggressive lymphomas that probably need to be treated with something other than R-CHOP,” Dr. Swerdlow says.
“Double expressor” lymphomas are distinct from double hit lymphomas, he stresses, even if there is some overlap. “Double expressor cases are defined by IHC finding more than a certain percentage of neoplastic cells positive for MYC and BCL2,” Dr. Swerdlow says. Double expressor status doesn’t carry the weight of a double hit lymphoma, although in many series it is associated with an adverse prognosis.
Changes on the T-cell lymphoma side include “great interest” in grouping the non-cutaneous T-cell lymphomas of T follicular helper (TFH) cell origin, Dr. Swerdlow says. This spectrum of lymphomas includes angioimmunoblastic T-cell lymphoma and the provisional entity follicular T-cell lymphoma. These malignancies are grouped based largely on clinical features, histology, and immunophenotyping. “There are associations with certain mutations,” Dr. Swerdlow says, but the mutations are not specific and their identification not yet a part of standard practice. The existence of more indolent T-cell lympho-proliferative disorders, specifically indolent T-cell lymphoproliferative disorder of the GI tract, will now be recognized.
Greater complexity of the ALK-negative anaplastic large cell lymphomas (ALCL) is also recognized. Cytogenetic subtypes of ALK-negative ALCL with widely disparate prognoses have been reported. Those with DUSP22 rearrangements have a better survival, similar to the ALK-positive cases (Parrilla Castellar ER, et al. Blood. 2014;124:1473–1480). A distinctive breast implant-associated ALCL is also newly recognized.
In addition to imparting information in his CAP16 talk, Dr. Swerdlow’s aim was to inspire younger pathologists and those in training to be curious, to get involved, and to not just accept the status quo. “Our field is rapidly changing,” he says. “Younger people will be better trained in newer technologies. It is going to be important that people recognize that continuous education is critical if you are going to do a good job as a pathologist.”
William Check is a writer in Ft. Lauderdale, Fla.