Critical Issues About the Diagnosis of MPNs: Bone Marrow Histopathology (General Issues in the Management of MPNs) (Myeloproliferative Neoplasms) Part 2

Standardization of the WHO Morphological Criteria

The complex composition of BM tissue, particularly its changes associated with a malignant hematopoietic process warrants special care concerning the recognition of characteristic his-tological patterns that may be readily assigned to certain disease entities. Hematopathologists and clinicians are aware of the unwanted impact of subjectivity and therefore insist on reproducible morphological BM interpretations as a fundamental tool for an accurate discrimination of conditions that require elaborate up-to-date treatment regimens. A basal requirement is certainly the selection and definition of as many BM constituents as possible followed by a systematically conducted categorization of their most prominent alterations and semiquantitative grading of features (Florena et al. 2004; Thiele and Kvasnicka 2003a ; Thiele et al. 2005a, d). Regarding the situation in the three main entities of MPNs under discussion, standardized features of discriminating impact for generating a diagnostic histopathological pattern are shown in Table 3.1 ; In this context, it should be emphasized that apart from their incidence, these features exert a variable impact on differential diagnosis. While iron deposits or lymphoid nodules have little discriminative power, overall cel-lularity in concert with the listed alterations of granulo-, erythro-, and megakaryopoiesis as well as BM fiber content present the most important determinants to create a characteristic histologi-cal BM pattern as demonstrated for the conflicting differentiation between early/prefibrotic PMF and ET (Fig. 3.5). Concerning the crucial parameter of hematopoietic cellularity, changes associated with age have to be explicitly regarded (Table 3.2) including especially the subcortical marrow spaces (Thiele et al. 2005c). Altogether, Table 3.1 may serve as a checking list to create a constellation of morphological features exerting a discriminating impact in order to achieve an accurate diagnosis. In this scheme, exact grading of BM fibrosis is a crucial point because as reviewed by Kuter et al. (2007), essentially two different scoring systems exist (Table 3.3). In most studies, a four-graded scheme was used occasionally with slight modification of the most frequently applied scoring systems (Bauermeister 1971; Manoharan et al. 1979). On the other hand, the revised 2008 WHO classification adopted a more simplified three-graded system (Thiele et al. 2005c) that has been proven to be of clinical relevance (Thiele and Kvasnicka 2006a ; Vener et al. 2008). Concerning all these features, it has to be emphasized that not a single morphological constituent of the BM or its alteration characterizes a specific subtype of MPNs, but only a synoptical view of a large number of components creates a certain pattern. This approach resembles the recognition of the various colorful particles included in a mosaic pavement forming a complex picture.


Table 3.1 Discriminating features according to WHO morphological criteria generating histological patterns in initially performed bone marrow biopsy specimens as modified from Thiele and Kvasnicka (2009)

PV, %

ET, %

PMF, %

Prefibrotic/early

stage

Overt fibrotic stage

I. Cellularity (age-matched) increased

100

10-20

80-100

10-20

II. Neutrophil granulopoiesis

Increased quantity

80-100

tmp10E-5

50-80

0

Left shifting

50-80

tmp10E-6

20-50

10-20

III. Erythropoiesis

Increased quantity

100

tmp10E-7 tmp10E-8

0

Left shifting

80-100

tmp10E-9

10-20

tmp10E-10

IV. Megakaryopoiesis

Increased quantity

50-80

80-100

50-80

20-50

Size:

Small

20-50

0

20-50

20-50

Medium

20-50

10-20

10-20

10-20

Large

20-50

20-50

20-50

10-20

Giant

10-20

20-50

10-20

tmp10E-11

Histotopography

Endosteal

10-20

10-20

20-50

20-50

translocation

Cluster formation:

Size:

Small (at least 3)

10-20

10-20

50-80

50-80

Large (more than 7)

tmp10E-12

0

20-50

20-50

Quality:

Dense

tmp10E-13

0

20-50

50-80

Loose

20-50

20-50

50-80

10-20

Nuclear features:

Hypolobulation (bulbous)

10-20

tmp10E-14

50-80

50-80

Hyperlobulation (staghorn-like)

50-80

50-80

tmp10E-15

0

Maturation defects

0

0

50-80

80-100

Naked nuclei

20-50

20-50

50-80

80-100

V. Fibers

1. Increased reticulin

10-20

0

20-50

80-100

2. Increased collagen

0

0

0

50-100

VI. Osteosclerosis

0

0

0

20-50

VII. Iron deposits

0

20-50

10-20

tmp10E-16

VIII. Lymphoid nodules present

10-20

0

10-20

tmp10E-17

Finally, the possibility of unclassifiable entities has to be addressed. Their frequency ranges from 5% to 10%, and incidence significantly depends not only on the experience of the investigator but particularly on the examination of a representative, properly sized (at least 1.5 cm) biopsy specimen performed at clinical diagnosis, and before any relevant therapy (Vardiman et al. 2009).

Schematic presentation of characteristic histo-logical bone marrow patterns and morphological criteria in early/prefibrotic primary myelofibrosis (PMF) versus essential thrombocythemia (ET) generated from semiquantitative evaluations of prominent diagnostic features as listed in Table 3.1

Characteristic features:

• Marked to moderate increase in age-matched cellularity

• Pronounced proliferation of neutrophil granulopoiesis and reduction of erythroid precursors

• Dense or loose clustering and frequent endosteal translocation of medium sized to giant megakaryocytes with hyperchromatic, hypolobulated, bulbous, or irregularly folded nuclei and an aberrant nuclear/cytoplasmic ratio

• No or no significant increase in reticulin fibers

• No or only slight increase in age-matched cellularity

• No significant increase or left-shift in neutrophil granulopoiesis and erythropoiesis

• Dispersed or loosely clustered large to giant mature megakaryocytes with hyperlobulated or deeply folded nuclei

• No or very rarely minor increase in reticulin fibers

Fig. 3.5 Schematic presentation of characteristic histo-logical bone marrow patterns and morphological criteria in early/prefibrotic primary myelofibrosis (PMF) versus essential thrombocythemia (ET) generated from semiquantitative evaluations of prominent diagnostic features as listed in Table 3.1

Table 3.2 Normal ranges of bone marrow cellularity for selected age groups (years) as adapted from the literature (Thiele et al. 2005c)

Age

% Hematopoietic area

20-30

60-70

40-60

40-50

>70

30-40

>80

10-20

Reproducibility of the WHO Morphological Criteria

Concerning the reproducibility of the WHO-defined histological criteria regarding MPNs (Vardiman et al. 2009), a conflict of opinion has to be discussed. In this context, among several pathologists controversy is especially focused on the crucial discrimination between ET and early stage PMF presenting with an elevated platelet count (false ET). Independently from the Cologne group, an increasing number of authorities were able to confirm the reliability of morphological BM features as proposed by the WHO (Boiocchi et al. 2011; Koopmans et al. 2011; Florena et al. 2004; Gianelli et al. 2006, 2008; Kreft et al. 2005; Vener et al. 2008). Additionally, in subsequently performed central reviews, involving more than 1,400 patients and a large number of local pathologists from different institutions, a more than 80% concordance was revealed.This rate of agreement is comparable with or even better than results of blinded histological panel evaluations by international study groups subtyping malignant B- and T-cell lymphomas (Diebold et al. 2002; Lones et al. 2000; Rüdiger et al. 2002) and Hodgkin’s disease (Glaser et al. 2001) as well as cytological classifications of myelodysplastic syndromes (Mufti et al. 2008). Contrasting these confirmatory studies, a blinded semiquantitative evaluation of certain BM features among three hematologists/pathologists questions explicitly these results (Wilkins et al. 2008) . Noteworthy is that the 370 BM biopsy specimens investigated were derived from the UK-PT1 trial (Harrison et al. 2005) and that the study design and performance are unfortunately impaired by a number of inconsistencies. Among others, these impairments include a lack of a clear-cut standardization of the 16 evaluated BM parameters strictly according to WHO (Thiele and Kvasnicka 2003a, 2009; Thiele et al. 2005a) .

Table3.3 Standard (Bauermeister 1971; Manoharan et al. 1979) and European Consensus (Thiele et al. 2005c) criteria for grading of myelofibrosis

Fibrosis gradea

Standard grading of reticulin fibrosis

European Consensus grading of myelofibrosis

0

Occasional fine and coarse individual fibers only; or foci of perivascular fiber network associated with benign lymphoid nodules

Scattered linear reticulin with no intersections (crossovers) corresponding to normal bone marrow

1

Fine fiber network (with occasional coarse fibers) throughout most of the section; or focal increase in reticulin away from vessels and benign lymphoid nodules

Loose network of reticulin with many intersections, especially in perivascular areas

2

Diffuse fiber network with an increase in scattered fibers

Diffuse and dense increase in reticulin with extensive intersections, occasionally with only focal bundles of collagen, and/or focal osteosclerosis

3

Diffuse, often coarse fiber network with no evidence of collagenization (negative trichrome stain)

Diffuse and dense increase in reticulin with extensive intersections with coarse bundles of collagen, often associated with significant osteosclerosis

4

Diffuse, coarse fiber network with areas of collagenization (positive trichrome stain)

Not applicable

aFiber density should be assessed in hematopoietic (cellular) areas

Remarkable was further a failing intraobserver evaluation (self-assessment) during the long period of examination with the unwanted bias of a learning effect. Moreover, the small size of evaluated biopsy specimens (>0.5 cm), contrasting the minimally requested length of >1.5 cm (Thiele et al. 2005c; Vardiman et al. 2009) precludes an accurate recognition of localized features like clusters or a more exact grading of fibrosis and assessment of age-matched cellularity. Finally, the very poor reproducibility of basic BM features that may have served as controls for reliability like quantity of erythropoiesis raises serious concerns. Although this series of patients was explicitly defined to be consistent with ET according to the diagnostic criteria of the PVSG (Murphy et al. 1997; Pearson 1998;; in a conspicuously wide range from 37% to 76% among the panelists, higher levels of BM fibrosis (probably grades 3 and 4) on a four-graded scale (Table 3.3) and new bone formation (osteosclerosis) were found (Wilkins et al. 2008) . This result regarding the semiquantitative scoring of a basic parameter like BM fibrosis among the panelists is not compatible with a consistent evaluation strictly applying standardized parameters. Comparable findings were reiterated in a study of the same group on 361 patients with ET mostly (82%) derived from the UK-PT1 trial (Harrison et al. 2005) reporting that about 60% of cases showed an increased BM fibrosis at disease onset, including more than 20% with moderate to overt myelo-fibrosis (Campbell et al. 2009) . Although the PVSG criteria for ET (Murphy et al. 1997; Pearson 1998) do allow a certain amount of BM fibrosis at presentation, the 80 patients with outright myelofibrosis, i.e., grades 3 and 4 (Table 3.3), are hardly consistent with these diagnostic guidelines. Concerning outcome, Campbell et al. (2009) reported that 11 of 226 patients diagnosed with ET, who showed reticulin fibrosis grades > 2, developed overt myelofibrosis after 68 months of follow-up, contrasting none of 135 patients with reticulin grade < 1. Additionally, a greater proportion of patients with so-called ET and reticulin grades > 2 presented with splenomegaly, a higher white blood cell count and increased marrow granulocytic cellularity as compared to patients with reticulin grade < 1. According to a number of experts, at presentation, ET is not characterized by BM fibrosis or osteosclerosis, and contrasting PMF, progression into overt myelofibrosis or leukemia is a rather rare event occurring after many years.For this reason, it may be assumed that a considerable number of cases reported in the three papers of the English group (Campbell et al. 2009; Harrison et al. 2005; Wilkins et al. 2008) including overlapping cohorts of patients are more likely consistent with throm-bocythemic manifestations of PMF (Thiele et al. 2009, 2011).

In another study critical to the WHO classification, two pathologists tried to reproduce the relevant criteria on 127 BM biopsy specimens with ET originally diagnosed according to PVSG (Murphy et al. 1997; Pearson 1998) resulting in a significant discordance of 35% among the panelists (Brousseau et al. 2010). The overall conclusion of these authors was that a discrimination between ET and early/prefibrotic PMF according to WHO criteria is impaired by subjectivity and of questionable clinical relevance. This argument has not only been refuted by large multicenter studies.The selection of the specimens up to 3 years after clinical diagnosis, with no data on the biopsy size and particularly the disturbing finding that 54% of the patients claimed explicitly by the authors to present WHO-defined ET showed a minor to moderate reticulin fibrosis, is rather confusing. Following strictly the WHO classification of MPNs, in ET incidence of minor reticulin fibro-sis (grade 1, Table 3.3), is very rare and observed in less than 5%.Finally, statistical analysis of morphological features (semiquantitative scoring system) and clinical data including outcome did not take the significant disparity in the number of cases in both groups (ET 102 patients versus PMF 18 patients) into account (Brousseau et al. 2010) ; Altogether, there is a striking prevalence of authors and relevant studies that endorse the histological WHO criteria including an increasing number of reports in favor of a reproducibil-ity of entities, especially concerning ET and early/prefibrotic PMF to a substantial extent (Thiele et al. 2011).

Finally, to enhance the reliability of reproducing these features in accordance with the WHO classification, hematopathologists involved in this field should be provided with a training set of characteristic slides and/or attend corresponding tutorials/ workshops as is usually done in the case of other hematological malignancies (subtyping of malignant lymphomas, myelodysplastic syndromes, etc.).

Conclusion

Histopathological diagnosis, particularly discrimination of the different MPN entities, is significantly based on standardized morphological features and recognition of distinctive BM patterns generating concordant results among pathologists as first step toward clinical studies. By regarding these postulates on treatment-naive and representative BM biopsy specimens, early stages of PMF presenting with thrombocythemia (false ET) are clearly separated from true ET. Persuasive evidence has been provided that this distinction exerts a significant clinical impact on the outcome concerning progression to myelofibrosis, leukemic transformation, and survival. Histological data, particularly regarding BM fibrosis derived from the UK-PT1 study and related investigations, suggest that these partially overlapping cohorts consist of a heterogeneous series of patients with a significant prevalence of PMF.

Next post:

Previous post: