Erythematous nodule of the face in 1 of the patients included in the study. Divergence of clinical appearance is unreliable for differential diagnosis between cutaneous lymphoid hyperplasia and B-cell lymphoma.
Dense dermal infiltrate of small lymphoid cells without atypia and polymorphism interpreted as pseudolymphoma in 1 of the patients included in the study (hematoxylin-eosin ×25).
Presence of a mixed T- and B-cell infiltrate and polytypic B cells by immunophenotypic studies. A, Pan B; B, λ chains (L); and C, κ chains (K).
Electrophoresis of IgH polymerase chain reaction products. Lane 1, B-cell lymphoma used as positive control; lane 4, patient 7 showing a dominant band; lanes 2, 3, and 5 through 10, polyclonal cases of pseudolymphoma; and lane 11, water used as negative control. bp indicates base pair.
Bouloc A, Delfau-Larue M, Lenormand B, Meunier F, Wechsler J, Thomine E, Revuz J, Farcet J, Joly P, Bagot M. Polymerase Chain Reaction Analysis of Immunoglobulin Gene Rearrangement in Cutaneous Lymphoid Hyperplasias. Arch Dermatol. 1999;135(2):168-172. doi:10.1001/archderm.135.2.168
The differential diagnosis of cutaneous lymphoid hyperplasia and B-cell lymphoma may be difficult. Whether the detection of clonal immunoglobulin gene rearrangement in the cutaneous lesion is predictive of a malignant outcome remains controversial. We therefore studied cases of cutaneous lymphoid hyperplasia by polymerase chain reaction analysis.
Retrospective study of patients seen between 1988 and 1996.
Two dermatology university departments.
Twenty-four patients with cutaneous lymphoid hyperplasias were included according to clinical, histopathological, and immunophenotypic criteria.
Main Outcome Measures
Clinical, histopathological, and laboratory findings.
There were 13 men and 11 women (mean age, 49 years) who presented with erythematous or violaceous papules or nodules. The lesions were unique in 13 cases and multiple in 11 cases. All patients had immunochemical evidence of a mixed T- and B-cell infiltrate with polytypic B cells. Polyclonality was demonstrated in 23 patients, whereas a dominant B-cell clone was detected in 1 patient. No lymphoma developed during the follow-up (median, 4 years). In the same period, we studied 53 cases of B-cell lymphomas. Thirty-five (66%) of the 53 cases had a detectable clonal immunoglobulin gene rearrangement.
In the majority of our cases, polyclonality demonstrated by polymerase chain reaction analysis was in accordance with the diagnosis of cutaneous lymphoid hyperplasia. In 1 of the 24 patients, the presence of a B-cell clone could be evidenced. This fact did not modify the treatment as there were no histological or immunophenotypic signs suggestive of a lymphoma.
THE DIFFERENTIAL diagnosis of B-cell lymphoma and cutaneous lymphoid hyperplasia may be difficult. Divergence of clinical appearance is unreliable for differential diagnosis.1 Lymphoid hyperplasias mimic lymphomas histologically and it is often impossible for the dermatopathologist to make a conclusive diagnosis.2,3 Immunophenotyping with leukocyte monoclonal antibodies is helpful to detect monoclonal B-cell proliferation by light chain restriction.4,5 However, a discrete monotypic B-cell infiltrate can be masked by a dense reactive infiltrate.6 Molecular analysis of lymphoid gene rearrangements by Southern blot analysis has proved to be useful for the detection of clonal populations and the determination of the lineage of lymphomas.7 The polymerase chain reaction (PCR) technique has more recently been used for the rapid detection of gene rearrangements in lymphoid proliferations as an alternative to traditional Southern blot hybridization.8 We studied 24 patients with cutaneous lesions in which a diagnosis of B-cell pseudolymphoma was made after clinical, histological, and immunophenotypic evaluations. Immunoglobulin gene rearrangement analysis was performed by PCR in the cutaneous lesions to determine the interest of this technique for the differential diagnosis between cutaneous lymphoid hyperplasias and B-cell lymphomas. Whether the detection of a clonal rearrangement is predictive of a malignant outcome remains controversial.
Twenty-four patients seen in 2 departments of dermatology between 1988 and 1996 were studied retrospectively. Criteria for inclusion were (1) clinical: erythematous papules and nodules, without any extracutaneous involvement ruled out by repeated complete physical examinations, standard biological tests including lactate dehydrogenase determinations, chest x-ray films, and abdominal ultrasound; (2) histopathological: biopsy specimen of lesional skin with dense dermal infiltrates of mixed centrocytes and centroblasts without atypia; (3) immunophenotypic: mixed T- and B-cell infiltrates with the presence of at least few clusters of B cells and polytypic B cells5; and (4) availability of frozen material of lesional skin biopsy specimen to perform PCR analysis.
Follow-up information was obtained by chart review or telephone call to the patient and the primary physician.
The charts of the patients diagnosed as having a B-cell lymphoma (defined by a typical histological evaluation and a monotypic B-cell infiltrate9) between 1988 and 1996 were also reviewed.
DNA was extracted from freshly frozen involved tissue according to standard procedures by a proteinase K digestion and a phenol/chloroform extraction. Polymerase chain reaction was performed by using 2 consensus oligonucleotides for the framework 3 region (FR3)/ACACGGC(C/T)(G/C)TGTATTACTGT and the framework 4 region (FR4)/ACCTGAGGAGACGGTGACC of the IgH chain genes.8 The reaction mixture included 250 ng of genomic DNA; 30 pmol of each primer; 1.5 U of Taq polymerase; 1 U of uracil DNA glycosylase (Gibco BRL, Gaithersburg, Md); 5 mL of 10× reaction buffer; 2.5-mmol/L magnesium chloride; 200 mmol/L each of deoxyadenosine triphosphate, deoxyguanosine triphosphate, and deoxycytidine triphosphate; and 400-mmol/L deoxyuridine triphosphate in a final volume of 50 mL. Samples were overlaid with mineral oil and transferred on the thermal cycler (model 480, Perkin Elmer Cetus, Norwalk, Conn). To avoid contamination by the amplification products, they were first held at 50°C for 10 minutes to allow uracil DNA glycosylase to destroy any deoxyuridine triphosphate containing amplified product that could have been carried over from previous reactions.10This incubation was followed by a 10-minute inactivation of uracil DNA glycosylase at 94°C before cycling. Each cycle included a denaturation step (94°C for 1 minute), an annealing step (56°C for 1 minute), and an elongation step (72°C for 1 minute). After 40 cycles, residual uracil DNA glycosylase was inactivated by the addition of 50 mL of chloroform. The amplified products were analyzed by electrophoresis on a 10% polyacrylamide gel with ethidium bromide staining.
This PCR technique allows the detection of 1% to 10% of tumor cells as demonstrated by dilution of a tumoral B-cell line (data not shown). A dominant band was considered significant when its intensity was equivalent to the presence of 10% of tumor cells in the control sample.
Twenty-four patients with cutaneous lymphoid hyperplasia were included in our study. There were 13 men and 11 women, with a mean age of 49 years. They all presented with erythematous or violaceous papules or nodules. The lesions were unique in 13 cases and multiple in 11 cases, and were localized on the head in 10 patients (Figure 1), the trunk in 10, the limbs in 7, and the genitalia in 1. The duration of the cutaneous disease before evaluation and diagnosis ranged from 2 weeks to 5 years (mean, 1.4 years). Treatment consisted of topical corticosteroids (7 patients), antibiotic therapy (6 patients), thalidomide (3 patients), radiotherapy (1 patient). The lesions resolved over a period of several months. One patient had recurrent disease after 3 years. The 23 remaining patients had no recurrence of disease. No extracutaneous involvement was detected during the follow-up period, which ranged from 1 to 9 years (median, 4 years).
All the biopsy specimens of lesional skin showed dense dermal infiltrates of mixed centrocytes and centroblasts without atypia and polymorphism (Figure 2, Table 1). The majority of the cases showed infiltrates with a diffuse and perivascular pattern throughout the reticular dermis. The infiltrate expanded in the superficial dermis in 9 cases. The topography was perivascular in 13 cases and/or periadnexal in 13 cases. Most of the cases (21/24) had no germinal centers, even though nests of B cells could be evidenced by immunohistochemistry. Epidermal and/or adnexal changes were observed in 16 cases. Immunophenotypic studies showed a mixed T- and B-cell infiltrate and the presence of polytypic B cells in all cases ( Figure 3).
Polymerase chain reaction analysis of lesional skin DNA showed a polyclonal pattern in 23 cases, in accordance with the diagnosis of cutaneous lymphoid hyperplasia. A clonal rearrangement of B-cell lineage was demonstrated in 1 patient (Figure 4). This patient (patient 7) was a 40-year-old man who presented with a unique lesion of the head; he was treated with antibiotics (amoxicillin) and the lesion disappeared within 6 months without any recurrence; he has been followed up for 9 years and no lymphoma has developed. This patient did not have any specific clinical or histological features to discriminate his disease from the that of the other patients.
During the same period, we studied 53 cases of B-cell lymphomas. Thirty-five (66%) of these cases had a clonal immunoglobulin gene rearrangement within the lesion as demonstrated by PCR (P<.001). There was no difference in the clinical appearance, histological findings, and course of the disease in the 2 subgroups of patients with lymphoma delineated by the PCR monoclonality status.
We included in our study 24 cases considered as benign cutaneous B-cell infiltrates according to clinical, histopathological, and immunohistochemical criteria. In 23 of 24 cases, polyclonality demonstrated by PCR analysis was in accordance with the diagnosis of cutaneous lymphoid hyperplasia. In 1 of the 24 patients, the presence of a B-cell clone could be evidenced. However, the lesion disappeared in 6 months without any recurrence and no sign of malignancy could be detected after 9 years of follow-up. This patient did not have any specific clinical or histological features of disease to discriminate him from the other patients studied.
Several studies have reported genotypic findings in cutaneous lymphoid hyperplasia. It is often considered that the best evidence for B- or T-cell malignancy is the presence of rearrangements of genes encoding immunoglobulin chains or the β chain of the T-cell receptor detected by Southern blot analysis.11 This technique has been used for the study of pseudolymphomas and reactive infiltrates. The significance of the detection of clonal gene rearrangements in patients with cutaneous lymphoid hyperplasia remains controversial. Wood et al12 studied by Southern blot analysis 14 patients with cutaneous lymphoid hyperplasia. Immunoglobulin gene rearrangement was found in 5 patients; 1 of these patients developed a large cell lymphoma after 2 years. The discovery of clonal rearrangement was interpreted as predictive of the former development of the lymphoma. Similarly, Hammer et al,13 in a study of 11 patients with cutaneous lymphoid hyperplasia, detected an immunoglobulin gene rearrangement in 2 patients who later developed a B-cell lymphoma. Landa et al14 also performed molecular rearrangement analysis by the Southern blot technique in 21 cases of lymphoproliferative lesions. Clonal immunoglobulin gene rearrangements were detected in 7 patients and systemic lymphoma developed in 1 patient. In contrast, Rijlaarsdam and Bakels,15 in their study of 7 patients with cutaneous lymphoid hyperplasia, found 4 immunoglobulin gene rearrangements, but the course of the disease was benign (follow-up, 18-66 months). We previously studied 18 patients with cutaneous lymphoid hyperplasia, who were different from the patients included in this study. One clonal immunoglobulin gene rearrangement was found, with a benign outcome after 7 years of follow-up.16
Using PCR, Fumiaki et al17 studied 3 patients with cutaneous lymphoid hyperplasia. In 1 of them, a clonal configuration was detected but there was no evidence of malignancy after 15 months. Similarly, among 19 patients described by Dubus et al,18 6 had polytypic benign cutaneous B-cell infiltrates. Two of them had a monoclonal gene rearrangement with no signs of lymphoma after 29 months. To summarize these published results, a dominant B-cell clone has been be found in some cases of cutaneous lymphoid hyperplasia, but whether the rearrangement is predictive of a malignant outcome was unresolved.
The advantages of PCR are the rapidity, the low cost, and the possibility to amplify small amounts of DNA.19 In contrast, Southern blot analysis requires large amounts of high-molecular-weight DNA.20 Polymerase chain reaction seems to be as sensitive and specific as Southern blot for the analysis lymphoid infiltrates.21 The absence or presence of a dominant clone must be interpreted carefully. In some cases of B-cell lymphoma, no clone could be demonstrated.22 In our laboratory, we studied by PCR 53 B-cell lymphomas between 1988 and 1996, and in 18 (34%) cases, no immunoglobulin gene rearrangement could be evidenced. Several reasons can be proposed to explain the discordance between clinicopathological and PCR results. The loss of signal from malignant cells overshadowed by a background population is always a potential problem. In addition, PCR may not detect all rearrangements due to somatic hypermutations, partial gene rearrangements, and variability in the V region.
However, clonal rearrangements have been documented in benign lymphoproliferative diseases with increased susceptibility to the development of lymphoma, such as Sjögren disease.23 The presence of a monoclonal population might be a prerequisite to neoplastic growth, but its presence does not imply in all cases a malignant outcome. At least 1 additional event may be required to lead to malignant transformation. Our patient with cutaneous lymphoid hyperplasia and a clonal immunoglobulin gene rearrangement was followed up for 9 years and there was no evolution to lymphoma. In most cases of cutaneous lymphoid hyperplasia, the results of PCR are an additional argument for the diagnosis of a benign infiltrate. In a minority of cases, the presence of a dominant B-cell clone is found, which should not modify the treatment if there are no other histological or immunophenotypic signs suggestive of a lymphoma.
In conclusion, the result of PCR analysis does not greatly contribute to patient care. If a systemic lymphoma can be ruled out, clonality has no significant implications for prognosis and treatment.
Accepted for publication November 20, 1998.
Members of the French Study Group for Cutaneous Lymphomas include: Liliane Laroche, MD (Bobigny); Michèle Delaunay, MD, Béatice Vergier, MD, Jean-Philippe Merlio, MD, Pierre Dubus, MD (Bordeaux); Pierre Souteyrand, MD, Michel D'Incan, MD, Pierre Dechelotte, MD (Clermont-Ferrand); Florent Grange, MD, Marie-Claire Tortel, MD (Colmar); Martine Bagot, MD, Janine Weshsler, MD, Marie-Hélène Delfau-Larue, MD (Creteil); Sophie Dalac, MD, Tony Petrella, MD (Dijon); Marie-Francoise Avril, MD, Jacques Bosq, MD (Villejuif); Anne Colson, MD (Mulhouse); Brigitte Dreno, MD (Nantes); Christine Bodemer, MD, Sylvie Fraitag, MD, Elizabeth MacIntyre, MD (Necker); Eric Estève, MD, Frédéric Maitre, MD (Orléans); Marie Beylot-Barry, MD (Pessac); Cécile Grosieux, MD, Anne Durlach, MD, Pascale Cornillet, MD (Reims); Pascal Joly, MD, Philippe Courville, MD, Elisabeth Thomine, MD, Bernard Lenormand, MD (Rouen); Nathalie Franck, MD, Agnès Carlotti, MD (Tarnier); and Loïc Vaillant, MD, Anne de Muret, MD (Tours).
Reprints: Anne Bouloc, MD, Department of Dermatology, Hospital Henri-Mondor 94010 Créteil, France (e-mail:firstname.lastname@example.org).