—  SPECIALTY CONFERENCE  —

Pulmonary Pathology

Case 5 - Nodular Lymphoid Hyperplasia

Michael Koss
Keck School of Medicine at USC
Los Angeles, California


Click on each slide thumbnail image for an enlarged view
Clinical History:
The patient is a 63 year-old woman with coronary artery disease who developed a progressive neurologic illness that resembled spinocerebellar ataxia. As her physician thought her neurologic illness may have been a manifestation of a paraneoplastic syndrome, she underwent chest and upper abdomen computed tomography (CT) looking for a primary lung malignancy. The chest CT showed a 2.0-cm in diameter nodule in the left lung apex. Contrast studies suggested it was likely malignant and she had the nodule resected. Sections are from this nodule.

Diagnosis: Consistent with nodular lymphoid hyperplasia (mantle cell/marginal zone B-cell hyperplasia); type 2 pneumocyte hyperplasia with "Dutcher Body-like inclusions.


Case 5 - Figure 1 - Low magnification view of the biopsy. It shows a localized, nodular lymphoid lesion that largely replaces the underlying lung architecture. The lesion is unencapsulated and shows mild extension into the surrounding lung around vessels.
Case 5 - Figure 2 - Scattered nodules composed of small lymphocytes are present within the lesion.


Case 5 - Figure 3 - This view shows the edge of a lymphocytic nodule and the intervening parenchyma. The nodule consist of small lymphocytes with round or oval nuclei and scant cytoplasm. The intervening parenchyma shows sheets of plasma cells of mature appearance with intermixed small lymphocytes.
Case 5 - Figure 4 - Another view of the parenchyma between nodules. This shows abundant plasma cells, small lymphocytes, and cuboidal cells (presumably type 2 cells) lining alveolar spaces. At least one of these alveolar lining cells contains an intranuclar inclusion, which can lead to its mistaken identification as a Dutcher body.

Follow-up:
The patient never developed signs or symptoms of hematolymphoid malignancy and died 3 years later of aspiration pneumonia secondary to her neurologic illness.

Light Microscopic Findings:
The sections show a lymphoid nodule whose border is unencapsulated and somewhat ill defined. The principal finding is effacement of much of the underlying architecture by a lymphoplasmacytic lesion. Specifically, there are scattered nodular aggregates of small lymphocytes with round uniform nuclei, occasional small germinal centers of reactive appearance, and intervening sheets of plasma cells with Russell bodies amid hyaline fibrous connective tissue. Scattered small germinal centers are present. While there appear to be abundant "Dutcher bodies" at medium magnification, on closer inspection the intranuclear inclusions lie within type 2 cells proliferating along the surface of residual alveolar spaces. Although a few lymphocytes infiltrate among bronchiolar cells, they are not sufficiently abundant in number to state that there are lympho-epithelial lesions present. The lymphoid cells do not invade pleura or bronchial cartilage; there are only a few perivascular aggregates of lymphoid cells at the margin of the lesion.

The lymphoid nodules are composed principally of B cells, as shown by staining with CD20, with only a small admixture of CD-3+ T cells. Somewhat greater numbers of CD3+ cells surround the nodules, and they are also present in the interstitium.

The key question is whether these nodular aggregates of B cells are follicular center cells, mantle cells, or marginal zone lymphocytes. Although a bcl-6 stain is not available, the lymphocytes do not have the indented or cleaved nuclei of follicular center cell lymphomas. Morphologically, the cells could be mantle cells, but an IgD immunohistochemical stain decorated only about 20-40% of them, which means that 60-80% of the cells do not stain as mantle cells. Thus, by elimination, these non-staining cells are probably marginal zone cells (unfortunately a CD21 stain is not available and the bcl-2 stain did not work), although a subset of mantle cells cannot be excluded. A CD5 stain was negative. Finally, scattered nodules of CD23+ dendritic cells were present, supporting the presence of residual germinal centers.

Antibodies against kappa and lambda light chains did not stain the small lymphocytes, but they did show the abundant plasma cells to stain in a polytypic manner. Molecular studies showed no evidence of clonal rearrangements of immunoglobulin genes (mu, lappa, IgH) or of T cell receptor genes (T-beta, T-gamma).

In sum, the nodular B cells are most likely marginal zone cells with an admixture of mantle cells. It is not clear whether the lesion is benign or malignant on histological grounds; this distinction requires immunohistochemical and molecular studies. Those results support a reactive lymphoid process or at least do not show a clonal process. Therefore, I favor a very unusual reactive lesion, that is, a form of nodular lymphoid hyperplasia with features of hyperplasia of mantle/marginal zone B cells.

Differential Diagnosis:
Pulmonary nodular lymphoid hyperplasia is a term first suggested by Kradin and Mark 9 to refer to one or more nodules or localized lung infiltrates consisting of reactive lymphoid cells. A less favored synonym is pseudolymphoma. The concept of masses of reactive lymphoid tissue in lung is controversial because most microscopically low-grade lymphoid proliferations in lung, including those with abundant germinal centers, are overwhelmingly extra-nodal marginal zone B-cell lymphomas of MALT type.2, 6, 11  This has led to suggestions, particularly from Europe, that the term "pseudolymphoma" be disregarded as redundant or inaccurate and that reactive lymphoid nodules or masses do not occur in lung.2, 5, 12  However, a recent immunohistochemical and molecular pathologic study has confirmed the existence of rare cases of pulmonary nodular lymphoid hyperplasia.1  This case provides further support to the existence of these lesions, but it is histologically different and indeed unique, as far as I can tell.

Cases of nodular lymphoid hyperplasia of lung (NLH) that we reported in the literature share some histologic features with the present case.1  Most lesions are solitary. The pulmonary nodules measure from 0.6 cm to 6 cm in greatest dimension, but most are 2-4 cm (the present case was almost 2 cm in maximum dimension).1 

The lymphoid lesion is usually subpleural, but it can be peribronchial. Of interest, despite the subpleural location of the lesions, the plaque-like invasion of the overlying pleura by the lymphoid infiltrate seen commonly in MALTomas is rarely seen.

NLH is by definition histologically localized. In classic cases, the most striking feature is the numerous reactive germinal centers with well-preserved mantle zones and sheets of interfollicular mature plasma cells associated with fibrosis. The germinal centers in classic cases of NLH stain for the B-cell marker CD20, while interfollicular lymphocytes are immunoreactive for CD3, CD43 and CD5.1  Stains for bcl-2 do not decorate the follicles. The CD20-positive lymphocytes in cases of NLH do not co-express either CD43 or CD5. The immunoglobulin light chain reactivity is typically polyclonal in NLH. Molecular genetic analyses show no rearrangement of the immunoglobulin heavy chain gene or for the T cell gene receptor in NLH.1  The plasma cells in NLH show Russell bodies, but as in this case, they don't contain Dutcher bodies.

Case 5 differs from these previously reported cases of NLH in that it shows only occasional small follicles and only scattered clusters of follicular dendritic cells, as shown by staining for CD23. Instead, there are the peculiar small lymphoid nodules. However, it does have the interfollicular sheets of mature plasma cells and interfollicular fibrosis that obliterates the underlying lung parenchyma typically seen in NLH. It is therefore reasonable to assume that it could be a related lesion, but it certainly differs histologically and is unique in the lung, in my experience. Has a similar lesion been described elsewhere? Hunt and associates 14 described reactive lesions of unusual histologic appearance in peripheral lymph nodes that they termed "hyperplasia of mantle/marginal zone B cells". These lesion shared some features with the present case, in that they consisted of nodules of monomorphic small B cells, identified as either marginal zone cells or mantle cells, with round nuclei, with few, small and scattered germinal centers present. These nodular B cells lacked monotypic staining for light chains and failed to show molecular evidence of clonality (except in one case). However, these lesions differed from case 5 in failing to show the extensive plasmacytosis so prominent in it.

Certainly, the most important disease in differential diagnosis is low-grade marginal zone B-cell lymphoma of MALT (MALToma).6, 8, 11  This low-grade B-cell lymphoma of lung is far more frequent than nodular lymphoid hyperplasia; hence, any solitary lymphoid mass in lung is much more likely to be lymphoma than not. The clinical presentation of patients with extra-nodal marginal zone B-cell lymphomas, including the age distribution, symptoms and chest radiographic appearance, greatly ovelaps that of individuals with nodular lymphoid hyperplasia (6, 8, 11). Typically, these lymphomas are of low histological grade and they often contain reactive germinal centers, nodules of small lymphocytes, monocytoid B cells, and admixed plasma cells.6, 8, 10  A subset show Dutcher bodies, which, if present, point towards a diagnosis of lymphoma. Case 5 shows nodules of small lymphocytes and sheets of plasma cells; histologically, it could be a MALT lymphoma.

Case 5 also shows numerous eosinophilic intra-nuclear inclusions that resemble Dutcher bodies. However, these nuclear inclusions, which are also PAS-positive, are present in replicating type 2 cells, not in lymphocytes or plasmacytoid lymphocytes. These type 2 cellular inclusions are well known to occur in bronchioloalveolar carcinomas derived from type 2 cells and of course can occur in their non-neoplastic counterparts.3  Misidentification of type 2 cell inclusions as Dutcher bodies is therefore an important pitfall in diagnosis of MALT lymphomas (and lymphoplasmacytic lymphomas) in the lung.

Pulmonary MALT lymphoma is far more invasive of surrounding lung than NLH. It can show penetration of bronchial cartilage plates (occurring in up to 66% of cases), plaque-like involvement of pleura over a low-power field (over 50% of cases), lymphoepithelial lesions, and extensive lymphangitic spread. These features are not prominent in case 5 and thus a diagnosis of MALT lymphoma cannot be made based using them.4, 8 

MALT lymphomas usually (about 75% of cases) show a monoclonal pattern for immunoglobulin light chains in paraffin sections and they consist of CD20+, CD43+ cellular infiltrates with smaller numbers of infiltrating T cells. Despite the absence of a CD43 stain in this case (the presence of numerous plasma cells that usually stain for CD43 can also cause confusion), the absence of a monotypic pattern of light chain staining in a case with a prominent plasmacytic component, and the absence of a clonal population by gene rearrangement serve to make a pulmonary MALT lymphoma unlikely. The same argument can be used to exclude lymphoplasmacytic lymphoma.

Giant lymph node hyperplasia or Castleman disease also needs to be considered in differential diagnosis. The disease is of two types: the hyaline vascular and plasma cell types, although mixed or transitional forms can also occur. It is the plasma cell and mixed types that must be considered in the differential diagnosis of this case because they show hyperplastic or involuted follicles, aberrant mantle zone lymphocytes and an interfollicular stroma that is both hyalinized and rich in polytypic plasma cells.13  However, Castleman disease rarely involves lung, and most cases that do so are of the hyaline vascular type. Typically, pulmonary lesions present as solitary hilar masses due to involvement of the hilar lymph nodes. In the plasma cell type, there is typically multicentric disease in the thorax and the lesions are systemic, with associated hepatosplenomegaly and massive systemic lymphadenopathy. Polyclonal hypergammaglobulinemia is typical, and patients with HIV seropositivity show an increased incidence of this variant. There are no clinical findings in this case to support this scenario, as far as I am aware. Further, these lesions typically occur within lymph nodes in the lung, although an LIP pattern in the lung has been reported in the plasma cell variant.7  Histologically, a concentric pattern of sclerosis is not seen in case 5, the lesion is not lymph node-based and there is no evidence of LIP.

References

  1. Abbondanzo SL, Rush W, Bijwaard KE, et al. Nodular lymphoid hyperplasia of the lung: A clinicopathologic study of 14 cases. Am J Surg Pathol 2000;24:587-97.
  2. Addis B, Hyjek E, Isaacson P. Primary pulmonary lymphoma: a re-appraisal of its histogenesis and is relationship to pseudolymphoma and lymphoid interstitial pneumonia. Histopathology 1988;13:1-17.
  3. Clayton F. Bronchoalveolar carcinomas. Cell types, patterns of growth and prognostic correlates. Cancer 1986: 57:1555-64.
  4. Colby T, Yousem S. Pulmonary lymphoid lesions. Semin Diagn Pathol 1985;2:183-.
  5. Corrin B. Pathology of the Lungs. London: Churchill Livingstone, 2000.
  6. Kennedy J, Nathwani B, Burke J, et al. Pulmonary lymphomas and lymphoid lesions. Cancer 1985;56:539-52.
  7. Johkoh T, Muller NL, Ichikado K, et al.: Intrathoracic multicentric Castleman disease: CT findings in 12 patients. Radiology 1998; 209:477-81.
  8. Koss M, Hochholzer L, Nichols P, et al. Primary non-Hodgkin's lymphoma and pseudolymphoma of lung: A study of 161 patients. Hum Pathol 1983;14:1024-38.
  9. Kradin R, Mark E. Benign lymphoid disorders of the lung, with a theory regarding their development. Hum Pathol 1983;14:857-967.
  10. L'Hoste R, Filippa D, Lieberman P, et al. Primary pulmonary lymphomas. A clinicopathologic analysis of 36 cases. Cancer 1984;54:1397-406.
  11. Li G, Hansmann M-L, Zwingers T, et al. Primary lymphoma of the lung: morphological, immunohistochemical and clinical features. Histopathology 1990;16:519-31.
  12. Nicholson A, Wotherspoon A, Diss T, et al. Reactive lymphoid disorders. Histopathology 1995;26:405-12.
  13. Weisenberger, DD, Nathwani, BN, Winberg CD, Rappoport, H. Multicentric angiofollicular lymph node hyperplasia: A clinicopathologic study of 16 cases. Hum Pathol 1985;16:162-
  14. Hunt JP, Chan JA, Samoszuk M, Brynes RK, et al. Hyperplasia of mantle/marginal zone cells with clear cytoplasm in peripheral lymph nodes. Am J Clin Pathol 2001; 116: 550-9.