- Email: [email protected]
Lymphocyte-rich classical Hodgkin lymphoma (LRCHL): clinico-pathological characteristics and outcome of a rare entity
original article
Annals of Oncology 17: 141–145, 2006 doi:10.1093/annonc/mdj037 Published online 10 November 2005
Lymphocyte-rich classical Hodgkin lymphoma (LRCHL): clinico-pathological characteristics and outcome of a rare entity D. de Jong1*, J. Bosq2, K. A. MacLennan3, J. Diebold4, J. Audouin4, J. Chasle5, A.-M. Mandard5, J. Marnay5 & M. Henry-Amar5 On behalf of the European Organization for Research and Treatment of Cancer (EORTC) Lymphoma Group and the Groupe d’E´tude des Lymphomes de l’Adulte (GELA) The Netherlands Cancer Institute, Amsterdam, the Netherlands; 2Institut Gustave Roussy, Villejuif, France; 3St. James’s University Hospital, Leeds, UK; Hoˆtel-Dieu, Paris, France; 5Centre Francxois Baclesse, Caen, France
Received 7 July 2005; revised 6 September 2005; accepted 9 September 2005
Background: To investigate the proportion, clinical characteristics and outcome of lymphocyte-rich classical Hodgkin lymphoma (LRCHL) in relation to nodular lymphocyte predominant HL (NLPHL) and classical HL (cHL). Patients and methods: A series of 2743 HL patients of all stages enrolled into three EORTC trials (H7, H8, H34) conducted between 1988 and 2000 and forming an unbiased series of HL patients was studied. Results: Detailed histological classification after panel review was available in 96% of the cases to allow selection of all cases with features potentially compatible with the WHO-definition of LRCHL for this study. Cases with dominance of lymphocytic infiltrate and relative paucity of eosinophils and fibrosis could be selected for re-classification. Twenty-one (0.8%) LRCHL cases were identified of which three were originally classified as NLPHL, seven as nodular sclerosis HL (NSHL) and 11 as mixed cellularity (MCHL), indicating that LRCHL is a rare disease. Conclusions: Clinical evaluation of the unselected series of patients (n = 2743) showed that LRCHL and NLPHL cases more often presented with favorable features. Clinical outcome adjusted on ab initio patient prognosis did not differ between the three histological entities. These results strongly suggest that LRCHL corresponds to an early stage in the spectrum of cHL rather than a biologically different disease entity. Key words: lymphocyte-rich classical Hodgkin lymphoma, lymphoma classification, clinical trial, prognosis
introduction In the World Health Organization (WHO) classification for lymphoid malignancies, five subtypes of Hodgkin lymphoma (HL) are recognized [1]. An important dichotomy is made between nodular lymphocyte predominant HL (NLPHL) and the group of classical HL (cHL). These two entities differ essentially in their clinical course with a significantly more indolent behavior associated with NLPHL, less frequent adverse clinical features, late relapses and deaths more often due to late effects of treatment than to the disease itself. Therefore, patients with NLPHL are currently treated with protocols that fundamentally differ from those applied to patients with cHL. Four different subtypes of cHL are now recognized with the introduction of the novel entity lymphocyte-rich classical HL (LRCHL) [1]. This variant was only recently defined as a subtype of cHL [2–4].
*Correspondence to: Dr D. de Jong, The Netherlands Cancer Institute, Department of Pathology, Plesmanlaan 1211066 CX Amsterdam, The Netherlands. Tel: +31-20-512-752; Fax: +31-20-512-2759; E-mail: [email protected]
ª 2005 European Society for Medical Oncology
In a combined study based on data from 17 European and American centers, diagnostic criteria were set for LRCHL: scattered Hodgkin-Reed-Sternberg cells with a classical immunophenotype in a background of small lymphocytes without admixture of eosinophils and neutrophils and without sclerosis [4]. Two morphological variants are identified (nodular versus interfollicular and diffuse). Clinical characteristics and outcome were analyzed based on 512 patients diagnosed as NLPHL between 1970 and 1994 and heterogeneously treated [3]. The results suggested that LRCHL may be associated with early stage disease and favorable prognostic factors compared with nodular sclerosis (NS) and mixed cellularity (MC) subtypes, leading to better overall survival. In this series, LRCHL patients more often presented with stage I than NS and MC patients (46% versus 10% and 21%, respectively), infrequent B-symptoms (11% versus 54%/ 40%) and infrequent mediastinal involvement (15% versus 80%/40%). Data was originally collected to study the differential diagnosis of NLPHL and T cell-rich B-cell lymphoma. This selection may have resulted in a significant bias, precluding strong conclusions on its clinical relevance.
Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on April 26, 2015
4
original article
1
original article
Annals of Oncology
Few well-documented, well-collected and uniformly treated patient series are available that may be less biased. The European Organization for Research and Treatment of Cancer (EORTC) HL Pathology Review Panel had the opportunity to study the histological material of three series of patients enrolled onto three prospective controlled clinical trials to study the incidence and clinical features of the WHO-HL entities [5–8].
patients and methods
results and discussion Of the 2743 available cases, 21 (0.8%) fulfilled the criteria for LRCHL as set in the WHO classification (Table 1). Three cases were originally classified as NLPHL, seven as NSHL and 11 as MCHL. A nodular growth pattern was seen in 18/21 cases with a varying extent of the presence of eccentrically located atrophic germinal centers in nodi. This growth pattern was combined with a component with a diffuse pattern in two cases, while three cases showed only the diffuse growth pattern. The distribution of the two growth patterns is comparable to previously reported findings [4]. Cases diagnosed as LRCHL showed a classical immunophenotype with positivity for CD30 in all cases and CD15 in 15/21 (71%). CD20 expression was observed in one of 18 cases. No signs of progressive transformation of germinal centers (PTGC) were observed. In addition, nine cases originally classified as NLPHL were re-classified as cHL (NSHL and MCHL) and two cases of cHL as NLPHL. A majority of these 11 cases were originally diagnosed without immunohistochemical data and additional immunophenotypical information collected for the present study was the dominant parameter for re-classification. The European Taskforce on Lymphoma first described LRCHL in a series of cases that were originally diagnosed as NLPHL or considered as possible mimics [4]. The majority of LRCHL, therefore, were originally classified as NLPHL
Table 1. Reclassification of 851 cases according to the criteria set by the WHO classification for lymphoid malignancies EORTC trial
Original diagnosis
Selected for review
Reclassification NLPHL
LRCHL, nodular
LRCHL, diffuse
cHL
H7 (#20881) early stages
NLPHL NS MC NLPHL NS MC NLPHL NS MC
27 86 135 44 205 152 12 95 95 851
21 0 1 40 1 0 10 0 0 73
3 0 4 0 5 1 0 1 3 17
0 0 0 0 0 1 0 1 2 4
3 86 130 4 199 150 2 93 90 757
H8 (#20931) early stages
H34 (#20884) advanced stages
Total
NLPHL, nodular lymphocyte predominant Hodgkin lymphoma; LRCHL, lymphocyte-rich classical Hodgkin lymphoma; cHL, classical Hodgkin lymphoma; NS, nodular sclerosis subtype of Hodgkin lymphoma; MC, mixed cellularity subtype of Hodgkin lymphoma.
142 | de Jong et al.
Volume 17 | No. 1 | January 2006
Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on April 26, 2015
In all EORTC HL trials since 1964, pathology review by an international panel of expert pathologists has been part of the protocol. Cases have been diagnosed according to the Rye classification using a standardized protocol and criteria. Histological material from patients treated in two trials on supradiaphragmatic stages I and II (H7, 1988–1993, 833 patients; H8, 1993–1998, 1578 patients) and in one trial on stages III and IV (H34, 1989–2000, 739 patients) was re-analyzed. Material was collected for 2869 cases and the diagnosis of HL confirmed in 2743 (96%) including 83 (3%) NLPHL, 2267 (83%) NSHL and 386 (14%) MCHL; seven (<1%) cases were unclassified. At the initial central pathology review, detailed morphological and immunohistochemical parameters were recorded. Detailed classification in NSHL cases included information on the extent of fibrosis (minimal sclerosis/cellular phase, well-formed collagen bands and extensive sclerosis) and subclassification according to the BNLI-guidelines (BNLI-type 1 and type 2) in all reviewed cases. MCHL cases are subclassified as classical, interfollicular and as ‘by exclusion of other subtypes’ according to set criteria in all reviewed cases. For NLPHL cases, the extent of nodularity and diffuse components was recorded as well as the presence of a component of so-called progressive transformation of germinal centers (PTGC). According to the current definition of LRCHL the presence of extensive fibrosis and/or a component of eosinophils in the background pattern excludes this classification. Cases that were previously classified as NSHL with well-formed collagen bands or extensive fibrosis and cases classified as BNLI-type 2, as well as cases classified as classical type MCHL that includes an eosinophilic infiltrate as a defining criterion, do not fit the definition of LRCHL and were therefore excluded from re-review in this study. According to morphological parameters assessed by the panel, cases classified as ‘NSHL with minimal sclerosis’ (n = 386), MCHL interfollicular and MCHL by exclusion of other subtypes (n = 382) and NLPHD (n = 83) may include LRCHL. These 851 cases were selected for re-classification and corresponding material retrieved. Immunohistochemical stainings on
paraffin-embedded biopsy samples was performed according to standard antigen retrieval procedures to include CD20, CD3, CD30 and CD15 (all from DAKO, Glostrup, Denmark) if lacking, or repeated if of insufficient quality. Each case was evaluated by the full panel of pathologists (DdJ, JB, KAML, JD, JA) without knowledge of corresponding clinical data. Data was prospectively collected and centrally computerized at Centre Franc xois Baclesse, Caen, France. Data was updated by 1 June 2002, at the latest. Relapse-free survival (RFS), event-free survival (EFS; including treatment failure, relapse or death of any cause) and overall survival (OS) were estimated using the Kaplan–Meier method and compared using the log-rank test. Ninety-five per cent confidence limits (95% CL) of rates were estimated using the Rothman method [9]. The Stata statistical software was used to analyze data [10]. Cases classified as LRCHL or NLPHL on review and cHL cases (reviewed and non-reviewed) were compared.
original article
Annals of Oncology
Table 2. Clinical features of NLPHL, LRCHL and cHL subtypes in EORTC H7, H8 and H34 trials LRCHL
cHL reviewed
cHL not reviewed
cHL all cases
P value
73
21 45.6 33%
757 37.8 22%
1892 32.8 12%
2649 34.215%
<0.001
4.3
1.9
0.9
1.1
63%/22% 11%/14% 4% 3% 16% 0% 11%
43%/24% 19%/14% 19% 19% 24% 5% 14%
37%/38% 15%/10% 32% 35% 48% 17% 35%
17%/60% 13%/10% 37% 49% 82% 31% 40%
23%/54% 13%/10% 36% 45% 72% 27% 38%
<0.001 <0.001 <0.001 <0.001 <0.001
18%/71%/11% 86%/14%
14%/50%/36% 67%/33%
2%/45%/53% 51%/49%
2%/32%/65% 63%/37%
3%/36%/61% 60%/40%
<0.001 <0.045
18% 41% 32% 9%
29% 29% 13% 29%
2% 21% 50% 27%
6% 18% 42% 34%
5% 19% 44% 32%
0.003
17% 37% 34% 5% 5% 2%
0% 29% 43% 14% 14% 0%
2% 23% 21% 20% 18% 16%
2% 14% 15% 23% 23% 23%
2% 17% 17% 22% 21% 21%
<0.001
20% 30% 10% 40%
29% 13% 29% 29%
26% 28% 10% 36%
21% 21% 11% 47%
23% 23% 11% 43%
0.282
32.8 11% 2.7
<0.001 <0.001
NLPHL, nodular lymphocyte predominant Hodgkin lymphoma; LRCHL, lymphocyte-rich classical Hodgkin lymphoma; cHL, classical Hodgkin lymphoma; EORTC prognostic groups: VF, very favorable (female aged <40 with clinical stage I and no B symptoms and ESR <50 and no bulky mediastinal involvement and lymphocyte predominant or nodular sclerosis histologic type); U, unfavorable (age ‡50 or no B symptoms and ESR ‡50 or B symptoms and ESR ‡30 or ‡4 nodal areas involved or bulky mediastinal involvement); F, favorable (all other patients); IPI score: International Prognostic Index; Mantle RX, mantle field irradiation; STNI, subtotal nodal irradiation; IF RX, involved field irradiation; EBVP, epirubicin, bleomycin, vinblastine, prednisone chemotherapy; M/A, i.e. MOPP/ABV hybrid (mechlorethamine, vincristine, procarbazine, prednisone, doxorubicin, bleomycin, vinblastine) chemotherapy.
(67/115, 58%). Immunohistochemical information at the time of diagnosis was mostly lacking. Since in the absence of immunohistochemical information NLPHL is the major morphological mimic of LRCHL, a relatively high percentage of LRCHL may be expected in this selection. Our selection of cases is essentially different, but obviously also resulted in a cohort that is enriched for cases with dominance of a lymphocytic infiltrate and paucity of eosinophils and well-formed collagen bands. However, since in our series immunohistochemical data was available at the time of original classification in 81% of cases, a classical immunophenotype could be recognized and, therefore, the majority of our cases of LRCHL were already recognized as cHL and classified as NS or MC (18/21, 86%). This further underlines the importance of immunohistochemical information as an integral component of
Volume 17 | No. 1 | January 2006
the modern approach to lymphoma classification according to the WHO classification. The relatively unbiased EORTC series of all stage disease can be used to estimate the actual frequency of LRCHL. It shows that among all assessable cases, this variant is significantly less frequent than previously suggested with <1% LRCHL compared with 3% NLPHL and 96% cHL. The much lower frequency of LRCHL in our series compared with the findings of the European Taskforce may be fully explained by the strong bias in their series of difficult NLPHL cases and the above-mentioned effects of immunohistochemical information. It should be noted that the distribution of clinical stage is highly comparable in both series (for stage I/II versus II/IV the distribution was 76.6%/23.3% and 76.5%/23.4%, respectively) and will not account for the different findings.
doi:10.1093/annonc/mdj037 | 143
Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on April 26, 2015
Number of cases Age Mean (years) ‡50 years Male/female ratio Stage I/II III/IV B symptoms A, ESR ‡50 or B, ESR ‡30 Mediastinal involvement Enlarged mediastinum Bulky disease Prognostic group VF/F/U (2106) IPI 0–2/3+ (481) Treatment H7 Mantle RX STNI EBVP-IF RX M/A-IF RX H8 Mantle RX STNI 3 M/A-IF RX 6 M/A-IF RX 4 M/A-IF RX 4 M/A-STNI H34 CR, no RT CR, RT CR, no random PR, fail
NLPHL
original article
Annals of Oncology
Table 3. Clinical outcome of NLPHL, LRCHL and cHL subtypes in EORTC H7, H8 and H34 trials
Number of cases Median follow-up (range) in months Total failures Extra nodal failures 8-year RFSa (95% CL) Total deaths Deaths from second cancer 8-year OSa (95% CL) Failures or deaths (events) 8-year EFSa (95% CL)
NLPHL
LRCHL
cHLreviewed
cHLnot reviewed
cHL all cases
73 89 (1–133) 12% 4% 85% (73–92) 5% 1% 95% (86–97) 15% 82% (71–90)
21 74 (1–135) 14% 5% 80% (53–94) 14% 5% 84% (50–97) 24% 68% (39–86)
757 83 (1–156) 15% 5% 83% (81–87) 13% 3% 92% (89–95) 22% 79% (76–82)
1892 80 (1–160) 15% 4% 84% (82–86) 10% 2% 93% (91–95) 19% 82% (80–84)
2649 81 (1–160) 15% 4% 84% (83–85) 11% 2% 93% (91–94) 20% 81% (80–83)
P value 0.214
0.785
0.438 0.467
Clinical features of patients corresponding to all reviewed and non-reviewed cases are listed in Table 2. Compared with patients having other types of cHL, LRCHL patients presented significantly less frequently with mediastinal involvement, enlarged mediastinum, bulky disease and elevated ESR. Consequently, patients with LRCHL early stage disease more frequently belong to the (very) favorable prognostic group despite their male predominance and older age. Reviewed cases were selected for dominance of lymphocytes and lack of eosinophils and fibrosis, while non-reviewed cHL, per definition, showed higher numbers of eosinophils and wellformed fibrotic bands. With respect to clinical parameters, reviewed cases showed significant differences with male dominance, less elevated ESR and less frequent mediastinal involvement and mediastinal enlargement compared with the non-reviewed cases (Table 2). In contrast, no differences in age distribution, stage, proportion of B-symptoms and bulky disease were seen. Von Wasielewski and co-workers described a new grading system for nodular sclerosing HL, in which eosinophilia and lymphocyte depletion constitute dominant parameters and are predictive for adverse clinical course [11]. This is fully in line with the distribution of prognostic features of the present series of cHL that showed a gradient of the eosinophilic component and inverse gradient of the lymphocytic component with established clinical prognostic stratification, including mediastinal involvement. Since in these EORTC trials, treatment was stratified on ab initio prognosis, this is not reflected in clinical outcome in terms of RFS, EFS and OS for LRCHL as well as for NSHL and MCHL (Table 3). Taken together, a spectrum of patterns with nodular and interfollicular disease and varying density of Hodgkin cells, components of hyaline fibrosis and eosinophilic granulocytes are recognized. Within this spectrum, the different subtypes of cHL can be placed morphologically as well as clinically as stated above. Therefore, one may suggest that cHL is one single disease that evolves through a spectrum of morphological changes, that largely also follows the progression of clinical features. In the initial phase, Hodgkin cells start to appear within the periphery of the follicular mantle zone with subsequent expansion of the mantles. This pattern has been
144 | de Jong et al.
previously recognized as ‘cellular phase of nodular sclerosis’ (although this term may have included many cases that would now be recognized as different entities of both HL and nonHodgkin lymphoma) [12]. During the next phase, increasing numbers of proliferating Hodgkin cells either migrate into the follicular nodules or expand interfollicularly. At this stage, early fibrosis and induction of eosinophils and other constituents such as epitheloid cells are first appreciated. The first pattern results in nodular disease, diagnosed as NS according to classical and WHO criteria; the second pattern fits the criteria of MC disease. The earlier stages are fully compatible with LRCHL, nodular and interfollicular, and diffuse variants, respectively. Others have shown a similar infection rate with EBV and similar molecular features with the lack of ongoing immunoglobulin gene mutations and the presence of crippling mutations compatible with pre-apoptotic germinal center derived B-cells in LRCHL and the other types of cHL [4, 13]. We conclude that LRCHL is a rare variant of cHL that fully fits into the morphological and biological spectrum of cHL.
acknowledgements The authors thank Drs E. M. Noordijk, P. Carde, C. Ferme´, H. Eghbali, B. M. Aleman and J. M. Raemaekers for their support to this study as principal clinical investigators of the H7, H8 and H34 trial. Dr F. Berger has contributed to the diagnostic reviewing procedure for these trials.
references 1. Jaffe S, Harris N Lee, Stein H, Vardiman JW (eds).Tumours of Haematopoietic and Lymphoid Tissues. World Health Organization Classification of Tumours Pathology and Genetics. Lyon, France: IARC Press 2001. 2. Ashton-Key M, Thorpe PA, Allen JP, Isaacson PG. Follicular Hodgkin’s disease. Am J Surg Pathol 1995; 19: 1294–1299. 3. Diehl V, Sextro M, Franklin J et al. Clinical presentation, course and prognostic factors in lymphocyte-predominant Hodgkin’s disease and lymphocyte-rich classical Hodgkin’s disease: Report from the European task force on lymphoma project on lymphocyte predominant Hodgkin’s disease. J Clin Oncol 1999; 17: 776–783.
Volume 17 | No. 1 | January 2006
Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on April 26, 2015
NLPHL, nodular lymphocyte predominant Hodgkin lymphoma; LRCHL, lymphocyte-rich classical Hodgkin lymphoma; cHL, classical Hodgkin lymphoma; CR, complete remission after six to eight cycles of M/A chemotherapy; PR, partial remission after six cycles of M/A chemotherapy; 95% CL, 9% confidence limits assuming a binomial distribution of rates. a Adjusted on ab initio prognosis, i.e. based on prognostic groups (early stages VF + F or advanced stages IPI 0–2 versus early stages U or advanced stages IPI 3.
original article
Annals of Oncology
4. Anagnostopoulos I, Hansmann ML, Franssila K et al. European task force on lymphoma project on lymphocyte predominant Hodgkin’s disease: histologic and immunohistologic analysis of submitted cases reveals 2 types of Hodgkin disease with a nodular growth pattern and abundant lymphocytes. Blood 2000; 96: 1889–1899. 5. Noordijk EM, Carde P, Mandard AM et al. Preliminary results of the EORTC-GPMC controlled clinical trial H7 in early stage Hodgkin’s disease. Ann Oncol 1994; 5(Suppl 2): S107–S112. 6. Ferme´ C, Eghbali H, Hagenbeek A et al. MOPP/ABV hybrid and irradiation in unfavorable supradiaphragmatic clinical stages I–II Hodgkin’s disease: Comparison of three treatment modalities. Preliminary results of the EORTC-GELA H8-U randomized trial in 995 patients. Proceedings of the 42nd Annual Meeting of the American Society of Hematology, San Francisco, CA, 1–5 December 2000. Blood 2000; 96: 576a. 7. Hagenbeek A, Eghbali H, Ferme´ C et al. Three cycles of MOPP/ABV hybrid and involved-field irradiation is more effective than subtotal nodal irradiation in favorable supradiaphragmatic clinical stages I–II Hodgkin’s disease: Preliminary results of the EORTC-GELA H8-F randomized trial in 543 patients. Proceedings
8. 9. 10. 11.
12. 13.
of the 42nd Annual Meeting of the American Society of Hematology, San Francisco, CA, 1–5 December 2000. Blood 2000; 96: 575a. Aleman BM, Raemaekers JM, Tirelli U et al. Involved-field radiotherapy for advanced Hodgkin’s lymphoma. N Engl J Med 2003; 348: 2396–2406. Rothman KJ, Boice JD. Epidemiologic Analysis with a Programmable Calculator. Boston, MA: Epidemiology Resources Inc. 1982. StataCorp. Stata Statistical Software. Release 7.0. College Station, TX: Stata Corporation 2001. Von Wasielewski S, Franklin J, Fischer R et al. Nodular sclerosing Hodgkin disease: new grading predicts prognosis in intermediate and advanced stages. Blood 2003; 101: 4063–4069. Lukes RJ. Criteria for involvement of lymph node, bone marrow spleen and liver in Hodgkin’s disease. Cancer Res 1971; 31: 1755–1767. Brauninger A, Wacker HH, Rajewski K, Kuppers R, Hansman ML. Typing the histiogenetic origin of the tumor cells of lymphocyte-rich classical Hodgkin’s lymphoma in relation to tumor cells of classical and lymphocyte predominance Hodgkin’s lymphoma. Cancer Res 2003; 63: 1644–1651.
Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on April 26, 2015
Volume 17 | No. 1 | January 2006
doi:10.1093/annonc/mdj037 | 145
Annals of Oncology 17: 141–145, 2006 doi:10.1093/annonc/mdj037 Published online 10 November 2005
Lymphocyte-rich classical Hodgkin lymphoma (LRCHL): clinico-pathological characteristics and outcome of a rare entity D. de Jong1*, J. Bosq2, K. A. MacLennan3, J. Diebold4, J. Audouin4, J. Chasle5, A.-M. Mandard5, J. Marnay5 & M. Henry-Amar5 On behalf of the European Organization for Research and Treatment of Cancer (EORTC) Lymphoma Group and the Groupe d’E´tude des Lymphomes de l’Adulte (GELA) The Netherlands Cancer Institute, Amsterdam, the Netherlands; 2Institut Gustave Roussy, Villejuif, France; 3St. James’s University Hospital, Leeds, UK; Hoˆtel-Dieu, Paris, France; 5Centre Francxois Baclesse, Caen, France
Received 7 July 2005; revised 6 September 2005; accepted 9 September 2005
Background: To investigate the proportion, clinical characteristics and outcome of lymphocyte-rich classical Hodgkin lymphoma (LRCHL) in relation to nodular lymphocyte predominant HL (NLPHL) and classical HL (cHL). Patients and methods: A series of 2743 HL patients of all stages enrolled into three EORTC trials (H7, H8, H34) conducted between 1988 and 2000 and forming an unbiased series of HL patients was studied. Results: Detailed histological classification after panel review was available in 96% of the cases to allow selection of all cases with features potentially compatible with the WHO-definition of LRCHL for this study. Cases with dominance of lymphocytic infiltrate and relative paucity of eosinophils and fibrosis could be selected for re-classification. Twenty-one (0.8%) LRCHL cases were identified of which three were originally classified as NLPHL, seven as nodular sclerosis HL (NSHL) and 11 as mixed cellularity (MCHL), indicating that LRCHL is a rare disease. Conclusions: Clinical evaluation of the unselected series of patients (n = 2743) showed that LRCHL and NLPHL cases more often presented with favorable features. Clinical outcome adjusted on ab initio patient prognosis did not differ between the three histological entities. These results strongly suggest that LRCHL corresponds to an early stage in the spectrum of cHL rather than a biologically different disease entity. Key words: lymphocyte-rich classical Hodgkin lymphoma, lymphoma classification, clinical trial, prognosis
introduction In the World Health Organization (WHO) classification for lymphoid malignancies, five subtypes of Hodgkin lymphoma (HL) are recognized [1]. An important dichotomy is made between nodular lymphocyte predominant HL (NLPHL) and the group of classical HL (cHL). These two entities differ essentially in their clinical course with a significantly more indolent behavior associated with NLPHL, less frequent adverse clinical features, late relapses and deaths more often due to late effects of treatment than to the disease itself. Therefore, patients with NLPHL are currently treated with protocols that fundamentally differ from those applied to patients with cHL. Four different subtypes of cHL are now recognized with the introduction of the novel entity lymphocyte-rich classical HL (LRCHL) [1]. This variant was only recently defined as a subtype of cHL [2–4].
*Correspondence to: Dr D. de Jong, The Netherlands Cancer Institute, Department of Pathology, Plesmanlaan 1211066 CX Amsterdam, The Netherlands. Tel: +31-20-512-752; Fax: +31-20-512-2759; E-mail: [email protected]
ª 2005 European Society for Medical Oncology
In a combined study based on data from 17 European and American centers, diagnostic criteria were set for LRCHL: scattered Hodgkin-Reed-Sternberg cells with a classical immunophenotype in a background of small lymphocytes without admixture of eosinophils and neutrophils and without sclerosis [4]. Two morphological variants are identified (nodular versus interfollicular and diffuse). Clinical characteristics and outcome were analyzed based on 512 patients diagnosed as NLPHL between 1970 and 1994 and heterogeneously treated [3]. The results suggested that LRCHL may be associated with early stage disease and favorable prognostic factors compared with nodular sclerosis (NS) and mixed cellularity (MC) subtypes, leading to better overall survival. In this series, LRCHL patients more often presented with stage I than NS and MC patients (46% versus 10% and 21%, respectively), infrequent B-symptoms (11% versus 54%/ 40%) and infrequent mediastinal involvement (15% versus 80%/40%). Data was originally collected to study the differential diagnosis of NLPHL and T cell-rich B-cell lymphoma. This selection may have resulted in a significant bias, precluding strong conclusions on its clinical relevance.
Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on April 26, 2015
4
original article
1
original article
Annals of Oncology
Few well-documented, well-collected and uniformly treated patient series are available that may be less biased. The European Organization for Research and Treatment of Cancer (EORTC) HL Pathology Review Panel had the opportunity to study the histological material of three series of patients enrolled onto three prospective controlled clinical trials to study the incidence and clinical features of the WHO-HL entities [5–8].
patients and methods
results and discussion Of the 2743 available cases, 21 (0.8%) fulfilled the criteria for LRCHL as set in the WHO classification (Table 1). Three cases were originally classified as NLPHL, seven as NSHL and 11 as MCHL. A nodular growth pattern was seen in 18/21 cases with a varying extent of the presence of eccentrically located atrophic germinal centers in nodi. This growth pattern was combined with a component with a diffuse pattern in two cases, while three cases showed only the diffuse growth pattern. The distribution of the two growth patterns is comparable to previously reported findings [4]. Cases diagnosed as LRCHL showed a classical immunophenotype with positivity for CD30 in all cases and CD15 in 15/21 (71%). CD20 expression was observed in one of 18 cases. No signs of progressive transformation of germinal centers (PTGC) were observed. In addition, nine cases originally classified as NLPHL were re-classified as cHL (NSHL and MCHL) and two cases of cHL as NLPHL. A majority of these 11 cases were originally diagnosed without immunohistochemical data and additional immunophenotypical information collected for the present study was the dominant parameter for re-classification. The European Taskforce on Lymphoma first described LRCHL in a series of cases that were originally diagnosed as NLPHL or considered as possible mimics [4]. The majority of LRCHL, therefore, were originally classified as NLPHL
Table 1. Reclassification of 851 cases according to the criteria set by the WHO classification for lymphoid malignancies EORTC trial
Original diagnosis
Selected for review
Reclassification NLPHL
LRCHL, nodular
LRCHL, diffuse
cHL
H7 (#20881) early stages
NLPHL NS MC NLPHL NS MC NLPHL NS MC
27 86 135 44 205 152 12 95 95 851
21 0 1 40 1 0 10 0 0 73
3 0 4 0 5 1 0 1 3 17
0 0 0 0 0 1 0 1 2 4
3 86 130 4 199 150 2 93 90 757
H8 (#20931) early stages
H34 (#20884) advanced stages
Total
NLPHL, nodular lymphocyte predominant Hodgkin lymphoma; LRCHL, lymphocyte-rich classical Hodgkin lymphoma; cHL, classical Hodgkin lymphoma; NS, nodular sclerosis subtype of Hodgkin lymphoma; MC, mixed cellularity subtype of Hodgkin lymphoma.
142 | de Jong et al.
Volume 17 | No. 1 | January 2006
Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on April 26, 2015
In all EORTC HL trials since 1964, pathology review by an international panel of expert pathologists has been part of the protocol. Cases have been diagnosed according to the Rye classification using a standardized protocol and criteria. Histological material from patients treated in two trials on supradiaphragmatic stages I and II (H7, 1988–1993, 833 patients; H8, 1993–1998, 1578 patients) and in one trial on stages III and IV (H34, 1989–2000, 739 patients) was re-analyzed. Material was collected for 2869 cases and the diagnosis of HL confirmed in 2743 (96%) including 83 (3%) NLPHL, 2267 (83%) NSHL and 386 (14%) MCHL; seven (<1%) cases were unclassified. At the initial central pathology review, detailed morphological and immunohistochemical parameters were recorded. Detailed classification in NSHL cases included information on the extent of fibrosis (minimal sclerosis/cellular phase, well-formed collagen bands and extensive sclerosis) and subclassification according to the BNLI-guidelines (BNLI-type 1 and type 2) in all reviewed cases. MCHL cases are subclassified as classical, interfollicular and as ‘by exclusion of other subtypes’ according to set criteria in all reviewed cases. For NLPHL cases, the extent of nodularity and diffuse components was recorded as well as the presence of a component of so-called progressive transformation of germinal centers (PTGC). According to the current definition of LRCHL the presence of extensive fibrosis and/or a component of eosinophils in the background pattern excludes this classification. Cases that were previously classified as NSHL with well-formed collagen bands or extensive fibrosis and cases classified as BNLI-type 2, as well as cases classified as classical type MCHL that includes an eosinophilic infiltrate as a defining criterion, do not fit the definition of LRCHL and were therefore excluded from re-review in this study. According to morphological parameters assessed by the panel, cases classified as ‘NSHL with minimal sclerosis’ (n = 386), MCHL interfollicular and MCHL by exclusion of other subtypes (n = 382) and NLPHD (n = 83) may include LRCHL. These 851 cases were selected for re-classification and corresponding material retrieved. Immunohistochemical stainings on
paraffin-embedded biopsy samples was performed according to standard antigen retrieval procedures to include CD20, CD3, CD30 and CD15 (all from DAKO, Glostrup, Denmark) if lacking, or repeated if of insufficient quality. Each case was evaluated by the full panel of pathologists (DdJ, JB, KAML, JD, JA) without knowledge of corresponding clinical data. Data was prospectively collected and centrally computerized at Centre Franc xois Baclesse, Caen, France. Data was updated by 1 June 2002, at the latest. Relapse-free survival (RFS), event-free survival (EFS; including treatment failure, relapse or death of any cause) and overall survival (OS) were estimated using the Kaplan–Meier method and compared using the log-rank test. Ninety-five per cent confidence limits (95% CL) of rates were estimated using the Rothman method [9]. The Stata statistical software was used to analyze data [10]. Cases classified as LRCHL or NLPHL on review and cHL cases (reviewed and non-reviewed) were compared.
original article
Annals of Oncology
Table 2. Clinical features of NLPHL, LRCHL and cHL subtypes in EORTC H7, H8 and H34 trials LRCHL
cHL reviewed
cHL not reviewed
cHL all cases
P value
73
21 45.6 33%
757 37.8 22%
1892 32.8 12%
2649 34.215%
<0.001
4.3
1.9
0.9
1.1
63%/22% 11%/14% 4% 3% 16% 0% 11%
43%/24% 19%/14% 19% 19% 24% 5% 14%
37%/38% 15%/10% 32% 35% 48% 17% 35%
17%/60% 13%/10% 37% 49% 82% 31% 40%
23%/54% 13%/10% 36% 45% 72% 27% 38%
<0.001 <0.001 <0.001 <0.001 <0.001
18%/71%/11% 86%/14%
14%/50%/36% 67%/33%
2%/45%/53% 51%/49%
2%/32%/65% 63%/37%
3%/36%/61% 60%/40%
<0.001 <0.045
18% 41% 32% 9%
29% 29% 13% 29%
2% 21% 50% 27%
6% 18% 42% 34%
5% 19% 44% 32%
0.003
17% 37% 34% 5% 5% 2%
0% 29% 43% 14% 14% 0%
2% 23% 21% 20% 18% 16%
2% 14% 15% 23% 23% 23%
2% 17% 17% 22% 21% 21%
<0.001
20% 30% 10% 40%
29% 13% 29% 29%
26% 28% 10% 36%
21% 21% 11% 47%
23% 23% 11% 43%
0.282
32.8 11% 2.7
<0.001 <0.001
NLPHL, nodular lymphocyte predominant Hodgkin lymphoma; LRCHL, lymphocyte-rich classical Hodgkin lymphoma; cHL, classical Hodgkin lymphoma; EORTC prognostic groups: VF, very favorable (female aged <40 with clinical stage I and no B symptoms and ESR <50 and no bulky mediastinal involvement and lymphocyte predominant or nodular sclerosis histologic type); U, unfavorable (age ‡50 or no B symptoms and ESR ‡50 or B symptoms and ESR ‡30 or ‡4 nodal areas involved or bulky mediastinal involvement); F, favorable (all other patients); IPI score: International Prognostic Index; Mantle RX, mantle field irradiation; STNI, subtotal nodal irradiation; IF RX, involved field irradiation; EBVP, epirubicin, bleomycin, vinblastine, prednisone chemotherapy; M/A, i.e. MOPP/ABV hybrid (mechlorethamine, vincristine, procarbazine, prednisone, doxorubicin, bleomycin, vinblastine) chemotherapy.
(67/115, 58%). Immunohistochemical information at the time of diagnosis was mostly lacking. Since in the absence of immunohistochemical information NLPHL is the major morphological mimic of LRCHL, a relatively high percentage of LRCHL may be expected in this selection. Our selection of cases is essentially different, but obviously also resulted in a cohort that is enriched for cases with dominance of a lymphocytic infiltrate and paucity of eosinophils and well-formed collagen bands. However, since in our series immunohistochemical data was available at the time of original classification in 81% of cases, a classical immunophenotype could be recognized and, therefore, the majority of our cases of LRCHL were already recognized as cHL and classified as NS or MC (18/21, 86%). This further underlines the importance of immunohistochemical information as an integral component of
Volume 17 | No. 1 | January 2006
the modern approach to lymphoma classification according to the WHO classification. The relatively unbiased EORTC series of all stage disease can be used to estimate the actual frequency of LRCHL. It shows that among all assessable cases, this variant is significantly less frequent than previously suggested with <1% LRCHL compared with 3% NLPHL and 96% cHL. The much lower frequency of LRCHL in our series compared with the findings of the European Taskforce may be fully explained by the strong bias in their series of difficult NLPHL cases and the above-mentioned effects of immunohistochemical information. It should be noted that the distribution of clinical stage is highly comparable in both series (for stage I/II versus II/IV the distribution was 76.6%/23.3% and 76.5%/23.4%, respectively) and will not account for the different findings.
doi:10.1093/annonc/mdj037 | 143
Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on April 26, 2015
Number of cases Age Mean (years) ‡50 years Male/female ratio Stage I/II III/IV B symptoms A, ESR ‡50 or B, ESR ‡30 Mediastinal involvement Enlarged mediastinum Bulky disease Prognostic group VF/F/U (2106) IPI 0–2/3+ (481) Treatment H7 Mantle RX STNI EBVP-IF RX M/A-IF RX H8 Mantle RX STNI 3 M/A-IF RX 6 M/A-IF RX 4 M/A-IF RX 4 M/A-STNI H34 CR, no RT CR, RT CR, no random PR, fail
NLPHL
original article
Annals of Oncology
Table 3. Clinical outcome of NLPHL, LRCHL and cHL subtypes in EORTC H7, H8 and H34 trials
Number of cases Median follow-up (range) in months Total failures Extra nodal failures 8-year RFSa (95% CL) Total deaths Deaths from second cancer 8-year OSa (95% CL) Failures or deaths (events) 8-year EFSa (95% CL)
NLPHL
LRCHL
cHLreviewed
cHLnot reviewed
cHL all cases
73 89 (1–133) 12% 4% 85% (73–92) 5% 1% 95% (86–97) 15% 82% (71–90)
21 74 (1–135) 14% 5% 80% (53–94) 14% 5% 84% (50–97) 24% 68% (39–86)
757 83 (1–156) 15% 5% 83% (81–87) 13% 3% 92% (89–95) 22% 79% (76–82)
1892 80 (1–160) 15% 4% 84% (82–86) 10% 2% 93% (91–95) 19% 82% (80–84)
2649 81 (1–160) 15% 4% 84% (83–85) 11% 2% 93% (91–94) 20% 81% (80–83)
P value 0.214
0.785
0.438 0.467
Clinical features of patients corresponding to all reviewed and non-reviewed cases are listed in Table 2. Compared with patients having other types of cHL, LRCHL patients presented significantly less frequently with mediastinal involvement, enlarged mediastinum, bulky disease and elevated ESR. Consequently, patients with LRCHL early stage disease more frequently belong to the (very) favorable prognostic group despite their male predominance and older age. Reviewed cases were selected for dominance of lymphocytes and lack of eosinophils and fibrosis, while non-reviewed cHL, per definition, showed higher numbers of eosinophils and wellformed fibrotic bands. With respect to clinical parameters, reviewed cases showed significant differences with male dominance, less elevated ESR and less frequent mediastinal involvement and mediastinal enlargement compared with the non-reviewed cases (Table 2). In contrast, no differences in age distribution, stage, proportion of B-symptoms and bulky disease were seen. Von Wasielewski and co-workers described a new grading system for nodular sclerosing HL, in which eosinophilia and lymphocyte depletion constitute dominant parameters and are predictive for adverse clinical course [11]. This is fully in line with the distribution of prognostic features of the present series of cHL that showed a gradient of the eosinophilic component and inverse gradient of the lymphocytic component with established clinical prognostic stratification, including mediastinal involvement. Since in these EORTC trials, treatment was stratified on ab initio prognosis, this is not reflected in clinical outcome in terms of RFS, EFS and OS for LRCHL as well as for NSHL and MCHL (Table 3). Taken together, a spectrum of patterns with nodular and interfollicular disease and varying density of Hodgkin cells, components of hyaline fibrosis and eosinophilic granulocytes are recognized. Within this spectrum, the different subtypes of cHL can be placed morphologically as well as clinically as stated above. Therefore, one may suggest that cHL is one single disease that evolves through a spectrum of morphological changes, that largely also follows the progression of clinical features. In the initial phase, Hodgkin cells start to appear within the periphery of the follicular mantle zone with subsequent expansion of the mantles. This pattern has been
144 | de Jong et al.
previously recognized as ‘cellular phase of nodular sclerosis’ (although this term may have included many cases that would now be recognized as different entities of both HL and nonHodgkin lymphoma) [12]. During the next phase, increasing numbers of proliferating Hodgkin cells either migrate into the follicular nodules or expand interfollicularly. At this stage, early fibrosis and induction of eosinophils and other constituents such as epitheloid cells are first appreciated. The first pattern results in nodular disease, diagnosed as NS according to classical and WHO criteria; the second pattern fits the criteria of MC disease. The earlier stages are fully compatible with LRCHL, nodular and interfollicular, and diffuse variants, respectively. Others have shown a similar infection rate with EBV and similar molecular features with the lack of ongoing immunoglobulin gene mutations and the presence of crippling mutations compatible with pre-apoptotic germinal center derived B-cells in LRCHL and the other types of cHL [4, 13]. We conclude that LRCHL is a rare variant of cHL that fully fits into the morphological and biological spectrum of cHL.
acknowledgements The authors thank Drs E. M. Noordijk, P. Carde, C. Ferme´, H. Eghbali, B. M. Aleman and J. M. Raemaekers for their support to this study as principal clinical investigators of the H7, H8 and H34 trial. Dr F. Berger has contributed to the diagnostic reviewing procedure for these trials.
references 1. Jaffe S, Harris N Lee, Stein H, Vardiman JW (eds).Tumours of Haematopoietic and Lymphoid Tissues. World Health Organization Classification of Tumours Pathology and Genetics. Lyon, France: IARC Press 2001. 2. Ashton-Key M, Thorpe PA, Allen JP, Isaacson PG. Follicular Hodgkin’s disease. Am J Surg Pathol 1995; 19: 1294–1299. 3. Diehl V, Sextro M, Franklin J et al. Clinical presentation, course and prognostic factors in lymphocyte-predominant Hodgkin’s disease and lymphocyte-rich classical Hodgkin’s disease: Report from the European task force on lymphoma project on lymphocyte predominant Hodgkin’s disease. J Clin Oncol 1999; 17: 776–783.
Volume 17 | No. 1 | January 2006
Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on April 26, 2015
NLPHL, nodular lymphocyte predominant Hodgkin lymphoma; LRCHL, lymphocyte-rich classical Hodgkin lymphoma; cHL, classical Hodgkin lymphoma; CR, complete remission after six to eight cycles of M/A chemotherapy; PR, partial remission after six cycles of M/A chemotherapy; 95% CL, 9% confidence limits assuming a binomial distribution of rates. a Adjusted on ab initio prognosis, i.e. based on prognostic groups (early stages VF + F or advanced stages IPI 0–2 versus early stages U or advanced stages IPI 3.
original article
Annals of Oncology
4. Anagnostopoulos I, Hansmann ML, Franssila K et al. European task force on lymphoma project on lymphocyte predominant Hodgkin’s disease: histologic and immunohistologic analysis of submitted cases reveals 2 types of Hodgkin disease with a nodular growth pattern and abundant lymphocytes. Blood 2000; 96: 1889–1899. 5. Noordijk EM, Carde P, Mandard AM et al. Preliminary results of the EORTC-GPMC controlled clinical trial H7 in early stage Hodgkin’s disease. Ann Oncol 1994; 5(Suppl 2): S107–S112. 6. Ferme´ C, Eghbali H, Hagenbeek A et al. MOPP/ABV hybrid and irradiation in unfavorable supradiaphragmatic clinical stages I–II Hodgkin’s disease: Comparison of three treatment modalities. Preliminary results of the EORTC-GELA H8-U randomized trial in 995 patients. Proceedings of the 42nd Annual Meeting of the American Society of Hematology, San Francisco, CA, 1–5 December 2000. Blood 2000; 96: 576a. 7. Hagenbeek A, Eghbali H, Ferme´ C et al. Three cycles of MOPP/ABV hybrid and involved-field irradiation is more effective than subtotal nodal irradiation in favorable supradiaphragmatic clinical stages I–II Hodgkin’s disease: Preliminary results of the EORTC-GELA H8-F randomized trial in 543 patients. Proceedings
8. 9. 10. 11.
12. 13.
of the 42nd Annual Meeting of the American Society of Hematology, San Francisco, CA, 1–5 December 2000. Blood 2000; 96: 575a. Aleman BM, Raemaekers JM, Tirelli U et al. Involved-field radiotherapy for advanced Hodgkin’s lymphoma. N Engl J Med 2003; 348: 2396–2406. Rothman KJ, Boice JD. Epidemiologic Analysis with a Programmable Calculator. Boston, MA: Epidemiology Resources Inc. 1982. StataCorp. Stata Statistical Software. Release 7.0. College Station, TX: Stata Corporation 2001. Von Wasielewski S, Franklin J, Fischer R et al. Nodular sclerosing Hodgkin disease: new grading predicts prognosis in intermediate and advanced stages. Blood 2003; 101: 4063–4069. Lukes RJ. Criteria for involvement of lymph node, bone marrow spleen and liver in Hodgkin’s disease. Cancer Res 1971; 31: 1755–1767. Brauninger A, Wacker HH, Rajewski K, Kuppers R, Hansman ML. Typing the histiogenetic origin of the tumor cells of lymphocyte-rich classical Hodgkin’s lymphoma in relation to tumor cells of classical and lymphocyte predominance Hodgkin’s lymphoma. Cancer Res 2003; 63: 1644–1651.
Downloaded from http://annonc.oxfordjournals.org/ at Kainan University on April 26, 2015
Volume 17 | No. 1 | January 2006
doi:10.1093/annonc/mdj037 | 145