College of American Pathologists
Printable Version

  Watching the bone marrow story unfold




November 2007
Feature Story

William Check, PhD

All of life evolves, so why not the field of bone marrow evaluation? There have been significant changes in the past several decades in how bone marrow biopsy procedures are done, which patients they are done on, and who does them. With regard to the most substantial change—who does bone marrow biopsy—just as in all of biology, diversity is the key. And, just as in the world of nature, different species of medical personnel—pathologists, hematology-oncology specialists, or nonphysicians—are deemed the fittest to survive in various medical niches.

"Who performs bone marrows is extremely variable across the country," says Rebecca Johnson, MD, chair of pathology and clinical laboratories at Berkshire Health Systems, Pittsfield, Mass. The current situation, she says, is "a reflection of changing practice patterns.

"When I was a medical student, many bone marrows were performed by pathologists. Practice has evolved over time to be quite different."

Evolution in equipment has contributed to this change. "Early needles were very difficult to use," says Robert W. McKenna, MD, professor and vice chair for academic affairs and consultant in hematopathology in the Department of Laboratory Medicine and Pathology at the University of Minnesota, Minneapolis. "With the newer needles it is much easier to obtain a biopsy."

Most important, at some institutions, nonphysicians—technologists, nurse practitioners, or nurses—are doing the bone marrow sampling. A nurse practitioner was brought into Dr. McKenna's department recently to do bone marrow procedures. "We do about the same number of marrows as in the past," Dr. McKenna says. "But the number of residents and staff has decreased. We thought that residents and faculty were spending too much time doing biopsies." The biopsy service remains based in the hematology/pathology laboratory.

It may be the Mayo Clinic that has used nonphysicians in this capacity for the longest time. "For several decades a group of dedicated nurses have had as their primary job to obtain bone marrow specimens, whether in the outpatient or inpatient setting," says Curtis A. Hanson, MD, professor of laboratory medicine and pathology at Mayo Clinic, where about 4,000 bone marrow procedures are done annually. "The nurses work with a laboratory technician who manages specimen flow." This approach "works great," Dr. Hanson says. "The quality of material that comes from the nurses here is high. Seldom if ever do we have anything to complain about."

Though Dr. McKenna admits to bias, it's his view that pathologists are in the best position to perform biopsies. "We are the ones who are going to be interpreting the biopsy, so we are more sensitive to the importance of obtaining good specimens." But, he adds, "the most important thing is that the individual, whether clinician or pathologist, be sufficiently trained so they can do it well."

Who performs bone marrow procedures can have an impact on the ability of pathology residents to meet training standards. In the Northeastern United States, for example, the extensive penetration of hematology-oncology practitioners presents an obstacle to meeting the requirement that pathology residents demonstrate proficiency in bone marrow sampling. "Almost every hospital and every town [in the Northeast] has heme-onc specialists," says Dr. Johnson. "In fact," she notes, "this is pretty much true across the country now as well."

Evolution in the indications for bone marrow biopsy also has had an impact. "Most bone marrows nowadays are done for staging and followup of malignancies," Dr. Johnson says. "By far most bone marrows are actually performed in heme-oncs' offices, not in the hospital." Notwithstanding, Dr. Johnson says, "We have a program that can offer that experience to our residents." At the University of Minnesota, the same is true. "We still have residents and fellows perform a sufficient number [of bone marrow biopsies] so they develop comfort and some level of expertise in the technique," Dr. McKenna says.

Evaluation of marrow specimens has also evolved. "Historically, hematologists obtained bone marrow specimens and examined them," says Russell K. Brynes, MD, professor of clinical pathology and director of the hematopathology fellowship program in the Department of Pathology at the University of Southern California Keck School of Medicine, Los Angeles. That was when a bone marrow consisted primarily of an aspirate. "With the rise of the trephine or core biopsy in the late 1960s and early 1970s, evaluation moved to the pathologist," Dr. Brynes says. "Now virtually all evaluation goes to the pathologist."

Since the introduction of the core biopsy, it's not clear that evaluation strategy has evolved much. One can argue that this basic plan—like the basic body plan of vertebrates—has remained the same for quite a while, in this case about 30 years. In 1978, when Dr. Brynes was a hematopathology fellow and Dr. McKenna was an attending, they published a paper, "Bone Marrow Aspiration and Trephine Biopsy. An Approach to a Thorough Study" (Brynes RK, McKenna RW, Sundberg RD. Am J Clin Pathol. 1978;70:753-759). "That description is still pertinent," Dr. McKenna says. "Our procedure is pretty much the same."

Allocating bone marrow procedures among specialties is partly an economic issue, because it is a billable procedure, but it also reflects practice patterns. "Most people trained in heme-oncology practice mostly oncology in the community setting," Dr. Brynes says. "In a busy office practice it may not be practical to spend an hour going to the hospital, reading a chart, and doing a procedure." This has been his experience in the Los Angeles area as well as in metropolitan Atlanta. This approach creates a need for people to perform bone marrow biopsies in community hospitals. Enter the community pathologist. (Agreeing with Dr. Johnson, Dr. Brynes notes that the Northeast is an exception to this pattern.)

Carla S. Wilson, MD, PhD, professor of pathology at the University of New Mexico Health Science Center, Albuquerque, has experienced the different models of bone marrow sampling. When Dr. Wilson was at the University of Arkansas for Medical Sciences in Little Rock, there was a large transplant program in the cancer center that did about 20 bone marrows per day. "We had two or three hematology technologists perform them for extra pay. The technologists were extremely good and efficient," Dr. Wilson says. "Patients actually preferred them. The only problem was that you couldn't charge for their work, since we didn't have the luxury to assign supervising physicians."

When Dr. Wilson came to the University of New Mexico, she says, "At first some pulls were being done by the hematopathology team and the majority by hematology-oncology fellows and attendings. Performing bone marrow procedures was very disruptive to our hematopathology sign-out sessions," she says, "because the procedures were often either previously unscheduled or delayed by the clinical service.

"Now hematology-oncology physicians primarily train and supervise our residents and fellows. They do a great job of overseeing this. In return we train their fellows in reading peripheral blood and bone marrow specimens."

(As for resident training at the University of Arkansas, Dr. Wilson says, "It was easier for us to have our residents go to the cancer center where the techs would train them. They did so many that a resident could do two or three in one day.")

Dr. Wilson also works in a reference laboratory where her coworkers are pathologists in private practice. "In that situation pathologists are doing all the bone marrow pulls," she says. As Dr. Brynes described, hematology-oncology physicians in private practice are too busy with office work, so it is easier for them to send patients to the pathologists.

Amy S. Gewirtz, MD, medical director of clinical laboratories and associate professor of pathology at The Ohio State University Medical Center, Columbus, has also experienced different ways of handling the procedures. At OSU, nurse practitioners trained by hematology-oncology physicians and based in that service do the vast majority of bone marrow procedures. "They perform the procedures," Dr. Gewirtz says, "but we send a technologist to the bedside to process slides." Dr. Gewirtz agrees that quality doesn't depend on who does the procedure but on the person's expertise. However, she notes, "When there is turnover in nurse practitioners, there can be problems in specimen biopsy quality."

In some institutions hematology-oncology fellows do bone marrow procedures on their patients, Dr. Gewirtz continues. "So there is a training curve." She previously worked at a hospital where a medical technologist employed by the laboratory did the procedure and made slides.

One advantage of having a technologist at the bedside is that, if the aspirate is not of good quality, he or she can assess it right there and make a touch preparation. "Many people immediately put the biopsy material into formalin," Dr. Gewirtz says. "Then you can't do a touch prep later. Whereas if you put the material onto saline-soaked gauze, you could do a touch prep in the lab."

LoAnn C. Peterson, MD, has also experienced diversity in who performs bone marrow procedures. "In some places where I have been the clinicians and pathologists split them," says Dr. Peterson, professor of pathology at Northwestern Memorial Hospital, Chicago. At Hennepin County Medical Center in Minneapolis, for example, pathologists did procedures on patients in the clinic, and heme-onc physicians did them on patients on the floor. At Northwestern, nurse practitioners do many of the bone marrow biopsies. "Most of the remaining procedures are done by heme-onc fellows," she says, "although the heme-path fellow also performs several bone marrow biopsies. Heme-onc attendings are very busy seeing patients, so lots of patient contact activities here are done by nurse practitioners."

Regardless of who does the procedure at Northwestern, someone from the pathology bone marrow laboratory is always present. "Getting a bone marrow biopsy is not just getting a specimen for morphology; it can also be for flow cytometry or cytogenetics or other special testing," Dr. Peterson says. "The technologist assists with that and makes slides on the spot to see whether it is an adequate specimen. What we pathologists are interested in is an excellent quality biopsy. If I get a core biopsy and I can't interpret it, that's part of my report. And if I see a trend that biopsies are not as good as they should be, I will discuss this with the people doing them."

Steven H. Kroft, MD, has encountered less variety in who performs bone marrow biopsies. "In all the institutions that I have been in, the large bulk of marrows have been performed strictly by clinicians," says Dr. Kroft, professor of pathology and director of hematology/pathology at the Medical College of Wisconsin, Milwaukee. "I have never been in an institution where pathologists did bone marrows, outside of training. At the last three institutions where I've been, they have been done exclusively by heme-onc faculty or fellows. Here they are done by physician assistants under the supervision of heme-onc physicians."

Dr. Kroft speculates that a lot of pathology departments doing bone marrow biopsies would "be happy to give it up.

"It is a tremendous use of resources to have pathologists on wards doing bone marrows."

Training and jurisdictional issues go hand-in-hand. At the Mayo Clinic, training of the nurses who do bone marrow procedures used to be under the pathology department. However, because of overall nursing competency issues, "it was important that nurses report to other nurses," Dr. Hanson says. "Yet medically they are still under direction from pathologists for quality of material and clinically under heme-onc physicians." When new nurses begin in this position along with other nurses, they train themselves.

Training of residents must also be considered. "All residents here as part of their heme/path rotation have to obtain 10 or 20 bone marrows," Dr. Hanson says. "Who knows what situation they will end up in when they get a job?" Expert nurses train the residents. "That has worked well," Dr. Hanson says.

At the University of Minnesota, the majority of bone marrow sampling procedures migrated not long ago to a nonphysician. "Because of the large volume, which has required more than the optimal amount of time performing biopsies by residents and faculty," Dr. McKenna says, "we have hired a nurse practitioner who will perform many of the biopsies." However, he emphasizes, "Pathology introduced the program and we have maintained control of it. The problem in many programs," he notes, "is that heme-onc physicians had control, so pathologists had to go to them and ask for training. And some programs thought their residents were getting inadequate training."

Having the program under the control of the pathology department facilitates training of pathology residents. "All of our residents end up doing 20 to 25 procedures," Dr. McKenna says. "If they do that many, they will have the technique down reasonably well." Pathology also provides training for hematology-oncology house staff.

The Accreditation Council for Graduate Medical Education Pathology Review Committee, which accredits pathology training programs, has recognized the training-related issues. Dr. Johnson, who is chair of the committee, says, "We found that many large academic programs had great difficulty meeting a requirement for training in bone marrow aspiration and biopsy techniques, because they were competing with heme-onc fellows in training. Internal medicine residents are not required to perform bone marrow, so if internal medicine residents don't have to know the technique," Dr. Johnson asks, "should we maintain a requirement that pathology residents have to learn how to do it?"

About two years ago the requirements for pathology residents to perform bone marrow aspirates and biopsies were relaxed. "There is still an expectation that residents will learn and become familiar with those techniques," Dr. Johnson says, "but there are many ways to accomplish that, such as videos or simulated bone marrow performance when taking bone marrow samples at autopsy." Hematopathology fellows, however, are still expected to get hands-on training.

Once you have a bone marrow aspirate and core biopsy, how do you evaluate them? In their 1978 paper, Drs. Brynes, McKenna, and Sundberg wrote: "A thorough bone marrow morphologic study involves examination of peripheral blood smears, direct, particle, and buffy coat bone marrow smears, trephine biopsy imprints, particle and trephine biopsy sections, and marrow volumetric data." This list still applies, Dr. McKenna says. "Each form of preparation has advantages in different situations," he says. "You never know what you are going to encounter." Contemporary bone marrow analysis may also include flow cytometry and cytogenetic analysis.

By and large, other pathologists agree with Dr. McKenna's estimate of what is needed for morphologic analysis of bone marrow, though the list may be tailored to the diagnosis. "Not all preparations are required in all scenarios," Dr. Kroft says. "For primary evaluation, you would want to see all that. In a patient being followed for acute myelogenous leukemia post-treatment, core biopsy is not required."

Dr. Gewirtz says there are some disease processes for which the pathologist could think both core biopsy and aspirate are not needed. "In acute leukemias you may or may not need a biopsy; in lymphomas you may or may not need an aspirate. But you don't know what you are going to find, so it's better to do both." Adding flow cytometry and cytogenetic analysis to the standard bone marrow workup bears heavily on this consideration. "For acute leukemias, you can evaluate more immunophenotypic antigens with flow cytometric analysis on an aspirate than on a bone marrow biopsy. Aspirate material is also needed for cytogenetic evaluation," Dr. Gewirtz says. For lymphomas, morphologic review of the biopsy often demonstrates marrow involvement when the aspirate and even ancillary studies such as flow cytometric analysis are negative. On the other hand, she says, "Ancillary studies that can be done on an aspirate to subclassify non-Hodgkin's lymphoma are far superior to what you can do on the biopsy."

"Sometimes you think you are going in for lymphoma and the patient has secondary myelodysplasia. If you didn't get an aspirate, you wouldn't be able to tell that," Dr. Gewirtz says.

Despite this general agreement, there is one important exception to Dr. McKenna's list: buffy coats. Dr. Brynes, who trained with Dr. McKenna, is the only other pathologist interviewed who uses them. In a comment typical of what others said, Dr. Wilson says, "They are very useful but a lot of work for their yield."

Dr. Gewirtz emphasizes the value of the simplest element in this menu: the peripheral blood smear. "A peripheral blood smear can be quite helpful depending on the diagnosis," she says. "In myelodysplasia, the granulocytic dyspoiesis is sometimes better seen in a blood smear than in the aspirate. And knowing CBC values can be helpful when looking at bone marrow."

A touch prep or core biopsy imprint is essential, in Dr. Wilson's view. "Often you don't get a good aspirate," she says—a so-called dry tap. Dr. Wilson initially leaves some of the touch preparations unstained. They can be used to assess cytology, to perform cytochemical or iron stains, to do FISH for specific cytogenetic abnormalities, or to scrape for molecular analyses.

Sections of the formalin-fixed aspirate—so-called clot sections—can be helpful if a good core is not obtained. "You can see cellularity and do stains or molecular studies on a clot section," says Dr. Wilson. "Everything you can do on a biopsy except that you may miss focal lesions or infiltrates associated with fibrosis."

Flow cytometry and cytogenetics may be ordered on virtually all samples or triaged. Dr. Wilson's group does the latter: "Our clinicians trust us to make those decisions, so usually pull an extra syringe as a hold specimen. After we look at the initial smears, we decide whether to send that extra material for FISH, flow cytometry, or cytogenetic and/or molecular analyses." They do one or the other test about 75 percent of the time.

At Northwestern Memorial Hospital, hematopathologists and clinicians together developed guidelines. For example, at diagnosis of a new acute leukemia both flow and cytogenetics would always be done. Says Dr. Peterson: "If we are diagnosing a new lymphoproliferative disorder, such as CLL, hairy cell leukemia, or a new lymphoma, we would always do flow. There is more variability in followup cases. We repeat cytogenetic analysis for acute leukemia at the remission biopsy if there was a clonal abnormality on the diagnostic biopsy."

With regard to staining bone marrows for iron, Dr. Kroft expresses the most common opinion: "We do it on all marrows, but it is not essential." Dr. Wilson says: "It does not always need to be done, but it's a problem for the people in the histology lab to figure out when it is needed and when not. The logistics are such that it is easiest to do it all the time to make sure you have it when you need it."

Dr. Gewirtz agrees: "We do them on all cases," she says. Iron stains are best performed on aspirates, Dr. Gewirtz says, since decalcification can leach out iron. She notes that an iron stain is the best method to demonstrate ringed sideroblasts and that it is needed for an anemia or myelodysplasia workup.

Unanimity also prevails for PAS staining, though in the opposite direction. "We almost never do PAS stains," Dr. Kroft says. "Though Europeans do PAS on marrow cores."

Though it is important to understand which evaluations are best to get the most information from a bone marrow specimen, remembering the basics is essential. "We see a lot of consults here at Mayo," Dr. Hanson says. "Bone marrows can be difficult to evaluate. Some people run into problems either from not good material or not good stains of material. As pathologists, we need to make sure we are getting good material and that the material is being stained in a high-quality manner." Dr. Hanson talks to many pathologists among whom there is a lot of frustration about hematology-oncology physicians getting bone marrows and not doing a good job of obtaining quality marrow specimens, "which puts pathologists in a very difficult position," he notes. Of course, all hematologists don't do a bad job, Dr. Hanson is quick to make clear, but "there are situations in which that is encountered," he says.

The University of New Mexico Health Science Center also has a large consult service. The biggest problem Dr. Wilson finds is that fixation and staining are not optimized. Difficult-to-read specimens are one reason that consults get sent. "In any lab it is important to optimize processing," Dr. Wilson says. "We can end up doing a number of immunohistochemical stains to figure out what is what. Erythroid cells, myeloid cells, and blasts are hard to distinguish without special stains in these difficult preps." Dr. Wilson finds that, as people have tried to find alternatives to B-5 fixative, problems have arisen. "We use a version of zinc formalin fixative," she says, "as do many other hospitals. But some people are still using B-5 even though safety and disposal requirements are problematic in the laboratory because of its mercury content."

Dr. Peterson, too, sometimes has trouble evaluating consult material that does not have good quality staining. Recently, she saw a bone marrow aspirate sent to Northwestern from another state that was stained with Diff-Quik. She says, "It is quick but not good for bone marrows, and I could not distinguish cellular details. Adequate staining is as important as getting an adequate biopsy."

When all the data from a bone marrow evaluation have been assembled, Dr. Peterson says, "I think the best approach is for one person—the pathologist—to integrate everything, including flow cytometry, cytogenetics, molecular tests, special stains, and immunostaining, into a single diagnosis."

Dr. Peterson is a member of the International Council for Standardization in Hematology Working Party for Bone Marrow Standardization that is looking at many of these questions (but not who should perform the bone marrow procedures). Szu-hee Lee, MD, of Australia, chairs the committee. The preliminary conclusions were to be presented last month to the ICSH meeting.

William Check is a medical writer in Wilmette, Ill.