From www.bloodjournal.org by guest on December 29, 2014. For personal use only. Prognostic Value of Cellular Proliferation and Histologic Grade in Follicular Lymphoma By Anita R. Martin, Dennis D. Weisenburger, Wing C. Chan, Elizabeth I. Ruby, James R. Anderson, Julie M. Vose, Philip J. Bierman, Martin A. Bast, Donald T. Daley, and James 0.Armitage The clinical usefulness ofhistologic grading in follicular lymphoma (FL) is controversial and is furthercompromised by the subjective nature and poor reproducibility of most systems in current use. Therefore, we decided t o objectively evaluate the importance of cellular proliferation in FL, along with the current grading systems. We studied 106 patients with FL whowereuniformly staged and aggressively treated. A proliferative index (PI) was determined quantitatively using an automatedimage analyzer and a new Ki-67 antibody thatstains archival paraffin tissues. The cases were also subclassified according t o t h eBerard, Rappaport, LukeCollins, and Jaffe methods, and survival analysis was performed. Patients with a low PI (C40461 had a significantly longer overall survival (OS)than thosewith a high PI (~40%). but the PI did not predict failure-free survival (FFS). The mean PI correlated well with the subgroups in each of the various classifications. All four of the classification methods were predictive ofOS, but only theBerard method appeared t o predict FFS and suggest that a proportion of patients with FL may be curable. In multivariate analysis, histologic classification wasthe only independentpredictor ofOS (Berard method: relative risk, 3.1) and the International Prognostic Indexwas the only independent predictor ofFFS (relative risk, 2.3).We conclude that the Berard method for grading of FL is clinically useful and, along with theInternational Prognostic Index, should beincluded in future clinical studies of FL. The measurement of cellular proliferation does not appear t o add additional useful information in FL. 0 1995 by The American Society of Hematology. F the time of initial diagnosis by members of the Nebraska Lymphoma Study Group (NLSG) between July 30, 1981 and December 20, 1991. Only patients whose initial diagnostic biopsy displayed at least 70% follicularity were included in the study. Patients were evaluated before treatment with a complete history and physical examination, conlplete blood count, chemistry profile, chest radiograph, computed tomography of the abdomen and pelvis, and a bone marrow biopsy. The Ann Arbor system was used to stage the patients. During the early accrual period of this study, patients with localized disease received involved-field irradiation as the primary treatment. Subsequently, all patients received as primary therapy one of four closely related six-drug combination chemotherapy regimens including cyclophosphamide, doxorubicin or mitoxantrone, and procarbazine, plus bleomycin, vincristine, and prednisone or dexamethasone (CAP/ BOP). The details of the four treatment plans have been previously reported.' Of 130 eligible patients enrolled on this protocol, clinical dataand adequate tissue for Ki-67 stains were available for 106 patients; of these 106 patients, 11 (10%) received involved-field irradiation only asthe primary treatment for localized disease. A complete response was defined as the absence of demonstrable disease after the completion of therapy and restaging. Pathologic features. All 106 cases hadbeen previously subclassified using the Berard method" as follicular small cleaved cell (FSC), mixed cell (FM), or large cell ( K C ) type for theNLSG diagnosis of record. For the study, the cases were also subclassified by two pathologists (A.R.M. and W.C.C.) using the Rappaport and OLLICULAR LYMPHOMA (FLJ is generally considered to be a low-grade malignancy characterized by an indolent clinical course and a prolonged survival. Although long complete remissions can be induced in some cases, only a small proportion of patients with FL are thought to be cured with conventional the rap^."^ A number of clinicalfeatureshave been associated with an adverse prognosis inFL, but the value of pathologic subclassification of FL with regard to therapy or prognosis is controversial.1-8 The usefulness of pathologic subclassification of FL is further compromised by the lack of uniform criteria for grading FL and poor diagnostic reproducibility due to the subjective nature of most grading system^.^"^ The level of proliferative activity, as determined by flow cytometric or cell kinetic studies, has been shown to correlate with tumor grade and have important prognostic implications in non-Hodgkin's lymphoma (NHL).12"7The antibody Ki67, which defines a proliferation-associated nuclear antigen,l8.l9has been used in tissue sections of NHL to determine proliferative activity2' and has recently been found to also have prognostic v a l ~ e . Although ~ " ~ ~ a number of small series have shown that proliferation increases with the number of large cells in FL,25-28 only one study has shown the potential utility of proliferation as an independent predictor of survival in low-grade However, the latter studyz9excluded cases of follicular large cell lymphoma, and the patients were not uniformly staged and aggressively treated. Therefore, we decided to objectively evaluate the importance of proliferative activity in a large group of patients with FL who were uniformly staged and aggressively treated. We also evaluated the importance of histologic classification of FL using four accepted methods (Rappaport, Lukes-Collins, Berard, and Jaffe). Finally, we evaluated the various clinical factors of potential prognostic importance in FL, including the recently developed International Prognostic Index for patients with diffuse aggressive NHL,30 and performed a multivariate analysis to determine which factors were independent predictors of survival in F L , MATERIALS AND METHODS Patients. The patients included in this study are all adults who were diagnosed with FL and were uniformly staged and treated at Blood, Vol 85, No 12 (June 15). 1995: pp 3671-3678 From theDepartments of Pathology and Microbiology, Preventive and Societal Medicine, and Internal Medicine, University of Nebraska Medical Center, Omaha, NE. Submitted September 12, 1994; accepted January 31, 1995. Supported in part by US Public Health Service Grant No. CA36727 awarded by the National Cancer Institute, Department of Health and Human Services. Address reprint requests to Dennis D. Weisenburger.MD, Department of Pathology and Microbiology, University of Nebraska Medical Center, 600 S 42nd St, Omaha, NE 68198-3135. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. section 1734 solely to indicate this fact. 0 1995 by The American Society of Hematology. 0006-4971/95/8S12-0$3.00/0 367 1 From www.bloodjournal.org by guest on December 29, 2014. For personal use only. MARTINET 3672 Table 1. Histologic Methods for Grading FL Berard Method (used by the NLSG) FSC 1 5 large noncleaved cells/follicular hpf FM r 5 but 115 large noncleaved cells/follicularhpf FLC 215 large noncleaved cells/follicular hpf Rappaport method 125% large cells/follicle FSC (PDL) 24-49% large cells/follicle FM (Mixed) FLC (Histiocytic) 250% large cells/follicle Lukes-Collins method FSC 1 2 5 % large noncleaved cells/follicle FLC 225% large noncleaved cells/follicle Jaffe modification of the Berard method FSC 1 5 large noncleaved cells/follicular hpf FM 2 5 large noncleaved cells/follicular hpf, but 150% large cells FLC >50% large cells/follicular hpf (predominance of large noncleaved and/or small noncleaved cells) Abbreviations: hpf. high-power field (0.158mm'); PDL, poorlydifferentiated lymphocytic. Lukes-Collins methods" and using a modified version of the Berard method recently described by Jaffe et al" (Table I). The tissue section withthe greatest number of large cells present within the neoplastic follicles was used for this purpose and for immunostaining as described below. In the early phase of the study. more patients with the FLC type were enrolled than those with low-grade disease (FSC or FM) hecause of the existence of a competing noncurative treatment protocol for the latter group. However, since June of 1986. all patients with FL who could tolerate aggressive therapy were treated with CAP/BOP. Because of the small number of cases of FSC type (S cases) identified using the Berard or Jaffe methods, the FSC and FM cases were combined into onegroup for statistical analysis. Ir,l,,lrrnosroini,l~ with Ki-67cmtibrxl~. Five-micron thick paraffin sections of formalin-fixed tissue were deparaffinized and rehydrated. To enhance antigen retrieval, sections were subjected to microwave incubation in citrate buffer at pH 6.0 twice for S minutes at full power and were then allowed to incubate in hot buffer for 10 minutes. B AL The slides were incubated with polyclonal rabbit antihuman Ki-67 antibody (A 047 diluted 150: Dako. Carpinteria, CA)," stained with the avidin-biotin peroxidase complex technique, and developed with H20Zand diaminobenzidine (Sigma Chemical. St Louis. MO). and thenuclei were counterstained with 0.2% ethyl green in sodium acetate buffer. pH 4.0. Qrtnnrirrrriw irncrgr crm/y.si.s. Quantitation of Ki-67 antibody staining was performed with the Cell Analysis Systems (CAS) 200 image analyzer7' using the CAS Quantitative Proliferation Software (Becton Dickinson. Elmhurst, IL) to obtain the percent nuclear area positive (PNA) and the cellular proliferation index (PI). These methods have been previously detailed and ~alidated.'~."."."' Fifteen neoplastic follicular fields were analyzed in each case and the fields were selected from follicles withthe greatest number of stained cells. Positive staining was identified as granular nuclear stain (Fig I). The PNA value is calculated as the immunostained nuclear area divided by the total nuclear area which is ethyl green stained and is expressed as a percentage. The P1 value is similar to the PNA except that the immunostained nuclear area is first divided by the individual immunostained nuclear size estimate and the total nuclear area is divided by the overall nuclear size estimate, thus giving the percentage of positive cells. Srcrristiccr/ cmcr/y.si.s. Curves of overall survival (OS) and failurefree survival (FFS) were drawn for the various histologic subgroups. by PI and PNA. and according to various clinical features of disease using the method of Kaplan and Meier." The cutpoints for the PI andPNA were determined by dividing the cases roughly into quartiles around the median PI (39%) and median PNA (32%). respectively. and evaluation of the curves of OS and FFS. Other cutpoints considering the PI and PNA as continuous variables were also evaluated. OS was defined as the time from initial therapy to death from any cause or to last follow-up. FFS was defined a s the time from the onset of initial therapy to the first occurrence of progression, relapse after response, death from any cause, or to the date of last follow-up if none of these events had occurred. Survival curves were compared using the logrank testAxand all P values reported are two-sided. Although the International Prognostic Index"' describes four risk groups. ie, low, low intermediate. high intermediate, and high, we combined and compared the two low groups with the two high groups because the number of patients in the high risk group was small. A multivariate analysis to determine independent predictors of survival was performed using the Cox proportional hazards model for survival analysis."' "" - .".. """* " " Fig 1. (A) Follicular lymphoma showing prominent delineation of the neoplastic follicles with the Ki-67 antibody stain. (B) Positive nuclear staining marks the cells in the G,, S, GP,and M phases of the cell cycle (immunoperoxidasestains; A, x25; B, ~150.1 From www.bloodjournal.org by guest on December 29, 2014. For personal use only. HISTOLOGIC GRADE IN FOLLICULAR LYMPHOMA Table 2. Clinical Characteristics of 106 PatientsWith F1 No. of Patients (%J Characteristic Age ( v ) <60 260 Sex Male Female Stage I II 111 IV Symptoms A 27 (26) 13 (12) 23 (22) 43 (40) 85 (80) 21 (20) B LDH level Normal Elevated Extranodal sites* 0 1-2 Bulk disease 210 cm Karnofsky scoret 290 5 80 International indextS Low risk High risk Response to treatment Complete Partiahone Not evaluable 79 (75) 27 (25) 63 (59) 43 (41) 21 ( 2 0 ) 78 (76) 24 (241 88 (86) 14 (14) Abbreviation: LDH, lactate dehydrogenase. * Includes bone marrow, but not spleen. t Data not available for all patients. Inctudes the following factors: age 2 6 0 years; LDH > normal; number of extranodal sites 2 2 ; stage III/IV; Karnofsky score 5 7 0 . + RESULTS Clinical features. The clinical characteristics of the entire group are summarized in Table 2. The median age of the patients was 61 years (range, 23 to 84 years) and the median follow-up period was 46 months (range, 2 to 136 months). The predicted 5-year OS was 67% and the 5-year FFS was 39% (Fig 2). By univariate analysis, the only clinical features that were predictive for OS were age (P= .098) and the International Prognostic Index ( P = .074). The clinical features that were predictive for FFS included the stage of disease ( P = .01 l), symptom status ( P = .034), number of involved extranodal sites (P = .01l), Kamofsky score (P = .022), and the International Prognostic Index (P = .014). Pathologic features. The effect of the PI on OS is shown in Fig 3. Patients with a low PI (<40%) had a significantly longer OS than those with a high PI (240%; P = .022). However, the PI did not predict for FFS, and the PNA was not predictive of either OS or FFS. The number of cases by histologic subgroup for each of the classifications is shown 3673 in Table 3.When the FSC and FM categories were combined, the cases in this subgroup in our series were identical when using either the Rappaport or Jaffe methods. Each of the methods categorized the cases quite differently, with the Berard methodhaving the most FLCcases and theRappaport or Jaffe methods having the fewest FLC cases. The mean PI correlated well with the subgroups in each of the classifications (Table 3). All of the classification methods were predictive of OS (Fig 4), but only the Berard method appeared to predict FFS (Fig 5). Interestingly, using the Berard method, patients with the FSC/FM type had a better OS, but poorer FFS, than those with the FLC type (Figs 4A and 5A). The apparent plateau in the FLC FFS curve (Fig 5A) suggests that a proportion of patients with this type may be cured. The predicted 8-year FFS of patients with FLC was 47% (95% confidence interval [CI], 33% to 61%) versus only 17% for those with FSC/FM (95% CI, 2% to 32%). The superiority of the Berard method as compared with the other methods for predicting FFS is shown in Fig 6. Cases classified as FM by the Jaffe method but as FLC by theBerard method, ie, those cases with 15 or more large cells but less than 50% large cells per follicle, had the same FFS as those classified as FLC by both methods (ie, cases with 250% large cells per follicle). In contrast, the FFS of cases classified as FM by both methods, ie, cases with 5 to 14 large cells per follicle, fails to plateauand is characterized by a pattern of continuous relapse. Multivariate analysis. A Cox proportional hazards model for survival analysis wasused to determine which prognostic factors were predictive of OS and FFS when controlling for other factors found to be significant in the univariate analysis. For OS, only the histologic classification was an independent predictor. The Berard method was the best predictor of OS with a relative risk of death of 3.1 for those with FLC as compared with those with FSC/FM, whereas the Rappaport, Jaffe, and the Luke-Collins methods gave relative risks of 2.4,2.4, and 2.1, respectively. Only the International Prognostic Index was an independent predictor of FFS, with a relative risk of failure of 2.3 for the high risk group (Fig 7). DISCUSSION The subjective nature and poor reproducibility of most histologic grading systems of FL,9"' as well as ongoing controversy regarding the clinical value of such systems,"' led us to study our cohort of patients with FL who were uniformly staged and aggressively treated. Because a number of studies have indicated that proliferative activity is an important prognostic indicator in NHL,"-17. 2"2429 we decided to evaluate this parameter in an objective and reproducible manner. Image analysis is an ideal method for assessing proliferation in FL because it is rapid, objective, and reprodUCible,25.27.'5.36.40and the analysis can be limited to the neoplastic follicles with the greatest number of large cells. In our study, we used the automated, quantitative CAS image analysis system25~z7~34-36 and a new Ki-67 antibody that works in paraffin tissue,33thus allowing the study of archival tis- From www.bloodjournal.org by guest on December 29, 2014. For personal use only. MARTIN ET AL 3674 9 7060V) 5050 40 40c L 30. 30 L g 3030 2, 50- g -E 3040- Q) i; 20- 8 10 -- A 0 2 4 Years 6 8 10 B 100 4, 0 2 6 4 10 8 Years Fig 2. OS (A) and FFS (B) of all patients with FL. sues. Previous studies of NHL have used a Ki-67 antibody that only works in snap-frozen tissue, thus limiting the number of cases available for study. However, one obstacle to the use of archival tissues is the variation in Ki-67 antibody staining of tissues that are fixed and processed at different institutions, and a number of cases had to be excluded from our study because of inadequate staining. The results that we obtained for FL with the CAS system are similar to those reported by Schwartz et a1': who also used the same system and methods. However, our results are difficult to compare with other reported results obtained by manual methods because only a small number of cases of FL have been so studied and the Ki-67 immunohistochemical, cell counting, and histologic classification methods used were quite heterogeneous. In four such s t ~ d i e s ? ~the , ~ "manual ~ ~ results were similar to our results, whereas in three other studies:0,22.44 two of which also included diffuse counterparts of FL,20,22 the results were somewhat lower. However, our study is the first to evaluate Ki-67 proliferative activity in FL and relate this parameter to survival. Similar to the study of Macartney et al?9 which evaluated the S-phase fraction of low-grade FL by DNA flow cytometry, we found that patients with FL having a low PI had a significantly longer OS than those with a high PI (Fig 3). However, in our study, the PI was not predictive of FFS and was not an independent predictor of OS in the multivariate analysis. As in other smaller st~dies:~-~* we found that proliferation correlated with the histologic subgroups of FL in each of the various classifications (Table 3), thus confirming the association of cellular proliferation with large-cell cytology in NHL.4' In contrast, Weiss et a16 failed to find such a correlation in FL when using a manual, semiquantitative method to estimate Ki-67 proliferation rates. Interestingly, two studies have failed to show a correlation between the mitotic rate and survival in whereas a recent study found that abundant mitoses predicted for better survival in FLC NHL.48However, our study indicates that the Ki-67 antibody can be used to provide important information with regard to the OS of patients with FL. Because of the subjective nature and poor reproducibility of methods for grading FL that require estimation of the percentage of large cells:"' we adopted the cell counting method of Mann and Berard3' some years ago for use by the NLSG. Nathwani et all' found that it was difficult for hematopathologists to reproducibly separate favorable from unfavorable FLs using subjective morphologic criteria alone and suggested that the cell counting method is superior to the estimation of percentages of large cells for grading FL. Although reproducibility may also be a problem with the Berard method: we have found this method to be less subjecTable 3. Association of PI With Histologic Subtypes of FL Patients Mean PI + SE (%l P* 42 64 32.1 2 1.9 43.4 C 1.5 <'oool 85 21 8.6-59.5 t 1.4 48.3 t 2.0 .0°02 34.2-62.3 64 42 35.3 i 1.7 i 1.6 '0005 8.6-59.0 23.2-62.3 No. of -m ...+.*;; .E > 60- Classification p=o.o22 ........ ..........+....** .............. *, PI 240 ,__. i Berard (NLSG) method FSC F M FLC RappaporYJaffe methods 36.6 FSC FM FLC Lukes-Collins method FSC FLC 44.3 + ................ + + 0 2 4 6 Years Fig 3. OS of patients with FL by PI. 8 10 Abbreviation: SE, standard error. * P Value based on the Student's t-test. PI Range (Oh) 8.6-59.5 10.2-62.3 From www.bloodjournal.org by guest on December 29, 2014. For personal use only. HISTOLOGIC GRADE IN FOLLICULAR LYMPHOMA 3675 tive and more reproducible than the other methods in our daily practice. However, when the Berard method is used, the proportion of cases in the various FL subgroups is different from that found by estimating percentages of large cells (Table 3). In our study, which was biased toward the inclusion of FLC cases (see Materials and Methods), the majority of FLs were classified as FLC type (60%) using the Berard method, whereas lesser numbers of cases were classified as such using the Lukes-Collins or RappapodJaffe methods (40% and 20%, respectively). In a large and unselected survey of cases in the population-based registry of the NLSG,"9 we found that 16% of our cases of FL were classified as Bend O0 90]y!.., \ "r..,.+ pso.10 i..... .." ..I.-.-" Fsc "1 + FM 1 L" -(.".-C A -0 2 4 6 8 10 6 8 10 YSam Berard l..).(*I.....+..l+> ......I.. 90- C - ........., 80. 70- FSC+FM .......I .. it +"*g, ....*..*....- H.."" S .= 60- 5 50- = 40 U) f FLC L 301 O 20 A lo 0 0 2 6 4 8 10 0 Yeam 2 4 Ywm RappapoRNaffe Lukes-Cdlins .- . %+l 70 "r I 0 2 FSC + FM *......I"+ i.....,. ____._,_( ...*+..+.*+.. .............. H.....+ p=o.o082 I. 4 6 8 lo 10 0 c 0 2 4 6 8 10 Years Fig 5. FFS of patients with FL according to the various classification systems. 10 0 2 4 6 8 10 YSarO Fig 4. OS of patients with FL according to the various olassification systems. FSC, 37% as F M , and 47% as FLC type using the Berard method. Our more recent data is as follows: FSC 25%, FM 36%, and FLC 39%. These findings are roughly the reverse of what has been reported using more traditional metho d ~In .the~ only ~ study ~ ~ similar ~ to ours, Nathwani et al" categorized FL according to the Berard method and found the following: FSC 36%, FM 42%, and FLC 22%. However, the cases in this latter study" were highly selected and the distribution of cases may not be comparable to ours. The possibility that an excess of FLC NHL occurs in Nebraska, possibly because of heavy pesticide use,51is also a consideration and will be the subject of a future study. However, From www.bloodjournal.org by guest on December 29, 2014. For personal use only. MARTIN ET AL 3676 these differing findings may have particular relevance to the comparisonand interpretation of clinical studiesthat use different criteria for the selection and grading of FL. In our study, we found that all of the classification methods were useful for predicting OS in FL (Fig 4). In fact, histologic classification was the only independent predictor of OS in the multivariate analysis, with the Berard method being the best predictor of OS (relative risk, 3.1). The Berard method also appeared to predict for FFS, whereas the other grading methods were notpredictive of FFS (Fig 5). The apparent plateau in the FLC FFS curve (Fig 5A) suggests that a proportion of patients with this type of FL may be curable, as has been previously suggested by others using the Rappaport method.52"s although longer follow-upof our serieswill be necessary to confirmthisfinding. However, using the Berard method, Bartlett et a148 have recently reported results similar to ours with regard to prolonged FFS in aggressivelytreatedpatientswith FLC NHL and have suggested that such patients should be treated with curative intent. The superiority of the Berardmethod over the othermethods forpredicting FFS is further evidentin Fig 6. It is apparent from these curves that some cases classified as FM type using the Jaffe method and other methods, ie, those cases with 15 or more large cells but less than 50% large cells per follicle, have an FFS that is identical to those with 50% or morelarge cellsper follicle. In otherwords,suchcases should probably be placed for clinical purposes in the FLC category, as with the Berard method. This same finding was also recently reported by Bartlett et aI4* and indicates that some cases classified as FM type using the other methods may be curable. Further studies with longer follow-up will also be necessary to confirm this finding. However, the results of our study suggest tous that theBerard method yields more clinically relevant information than the other grading methods for FL and that it should be used along with other methods in future clinical studies. In our study, we found that the International Prognostic Index," which was developed as a predictive model for diffuse aggressive NHL, predicts for OS and FFS in FL. Others48,56.57 have also recently reported that the International Prognostic Index is predictive of survival in FL, with one -5 ..... 100-1 .-9 5 v) 90. 80706050- Berardlarge caWJaHa mixed cell (215 large cells. but < 50 % large cells) Berard largeCnWJaffe large Cell (D 50% large cells) Berard rnlred dUJaffe mixed cell (> 5 large cells b U c 15 large cells) " 1 ~ p=0.10 l _ c L"- 0 2 )".,4 _ _ _ _ . _ I 4 Years c- 6 \"Cl 8 10 Fig 6. FFS of patients with FL classified by both the Berard and Jaffe methods (see text for explanation). "l. 5 ao L' . 2 70 2 60 50I"I"il p:0014 -7 0 2 6 4 8 10 Years Fig 7. FFS of patients with FL according to theInternational Prognostic Index. exception.58 In our study, multivariate analysis showed that the International Prognostic Index was the onlyindependent predictor of FFS (relative risk, 2.3; Fig 7). The International PrognosticIndex, or modificationsthereof, along with the Berard method of grading, should be useful in the design of future therapeutic trials of patients with €Xand in the selection of appropriate therapeutic approaches for individual patients. REFERENCES 1. Gallagher CJ, Gregory W,Jones AE, Stansfeld AG, Richards MA, Dhaliwal HS, Malpas JS, Lister TA: Follicular lymphoma: Prognostic factors for response and survival. J Clin Oncol 4:1470, 1986 2. McLaughlin P, Fuller LM, Velasquez WS, Butler JJ, Hagemeister FB, Sullivan-Halley JA, Dixon DO: Stage 111 follicular lymphoma: Durable remissions with a combined chemotherapy-radiotherapy regimen. J Clin Oncol 5:867, 1987 3. Ersboll J, Schultz HB, Pedersen-Bjergaard J, Nissen NI: Follicular low-grade non-Hodgkin's lymphoma: Long-term outcome with or without tumor progression. Eur J Haematol 42:155, 1989 4. Romaguera JE, McLaughlin P, North L, Dixon D, Silvermintz KB, Garnsey LA, Velasquez WS, Hagemeister FB, Cabanillas F: Multivariate analysis of prognostic factors in stage IV follicular lowgrade lymphoma: A risk model. J Clin Oncol 9:762, 1991 5. Soubeyran P, Eghbali H, Bonichon F, Trojani M, Richaud P, Hoerni B: Low-grade follicular lymphomas: Analysis of prognosis i n a series of 281 patients. Eur J Cancer 27:1606, 1991 6. Bastion Y , Berger F, Bryon P, Felman P, Ffrench M, Coiffier B: Follicular lymphomas: Assessment of prognostic factors in 127 patients followed for I O years. Ann Oncol 2:123, 1991 (suppl 2) 7. Anderson JR, Vose JM, Bierman PJ, Weisenburger DD, Sanger WG. Pierson J, Bast M, Armitage JO: Clinical features and prognosis of follicular large-cell lymphoma: A report from the Nebraska Lymphoma Study Group. J Clin Oncol 11:218, 1993 8. Coiffier B, Bastion Y, Berger F, Felman P, Bryon PA: Prognostic factors in follicular lymphomas. Semin Oncol 20:89, 1993 (suppl 5) 9. Metter GE, Nathwani BN, Burke JS, Winberg CD, Mann RB, Barcos M, Kjeldsberg CR, Whitcomb CC, Dixon DO, Miller TP, Jones SE: Morphological subclassification of follicular lymphoma: Variability of diagnoses among hematopathologists, a collaborative study between the Repository Center and Pathology Panel for Lymphoma Clinical Studies. J Clin Oncol 3:25, 1985 From www.bloodjournal.org by guest on December 29, 2014. For personal use only. HISTOLOGIC GRADE IN FOLLICULAR LYMPHOMA 10. Dardick I, Caldwell DR: Follicular center cell lymphoma: Morphologic data relating to observer reproducibility. Cancer 58:2477, 1986 1 1 . Nathwani BN, Metter GE, Miller TP, Burke JS, Mann RB, Barcos M, Kjeldsberg CR, Dixon DO, Winberg CD, Whitcomb CC, Jones SE: What should be the morphologic criteria for the subdivision of follicular lymphomas? Blood 68:837, 1986 12. Bauer KD, Merkel DE, Winter JN, Marder RJ, Hauck W, Wallemark CB, Williams TJ, Variakojis D: Prognostic implications of ploidy and proliferative activity in diffuse large cell lymphomas. Cancer Res 46:3173, 1986 13. Wooldridge TN, Grierson HL, Weisenburger DD, Armitage JO, Sanger WC, Collins MM, Pierson JL, Pauza ME, Fordyce R, Purtilo DT: Association of DNA content and proliferative activity with clinical outcome in patients with diffuse mixed cell and large cell non-Hodgkin’s lymphoma. Cancer Res 48:6608, 1988 14. Griffin NR, Howard MR, Quirke P, O’Brien CJ, Child JA, Bird CC: Prognostic indicators in centroblastic-centrocytic lymphoma. J Clin Pathol 41:866, 1988 15. Silvestrini R, Costa A, Giardini R, Boracchi P, Del Bin0 G, Marubini E, Rilke F: Prognostic implications of cell kinetics, histopathology and pathologic stage in non-Hodgkin’s lymphomas. Hematol Oncol 7:411, 1989 16. Rehn S, Glimelius B, Strang P, Sundstriim C, Tribukait B: Prognostic significance offlow cytometry studies inB-cell nonHodgkin lymphoma. Hematol Oncol 8: 1, 1990 17. Lindh J, Lenner P, Osterman B, Roos G: Prognostic significance of serum lactic dehydrogenase levels and fraction of S-phase cells in non-Hodgkin lymphomas. Eur J Haematol 50:258, 1993 18. Gerdes J, Lemke H, Biasch H, Wacker HH, Schwab U, Stein H: Cell cycle analysis of a cell proliferation-associated human nuclear antigen defined by the monoclonal antibody Ki-67. J Immunol 133:1710, 1984 19. Schluter C, Duchrow M, Wohlenberg C, Becker MHG, Key G, Had HD, Gerdes J: The cell proliferation-associated antigen of antibody Ki-67: A very large, ubiquitous nuclear protein with numerous repeated elements, representing a new kind of cell cycle-maintaining proteins. J Cell Biol 123513, 1993 20. Gerdes J, Dallenbach F, Lennert K: Growth fractions in malignant non-Hodgkin’s lymphomas (NHL) as determined in situ with the monoclonal antibody Ki-67. Hematol Oncol 2:365, 1984 21. Gerdes J, Stein H, Pileri S, Rivano MT, Gobbi M, Ralflciaer E, Nielsen KM, Pallesen G, Bartels H, Palestro G, Delsol G: Prognostic relevance of tumour-cell growth fraction in malignant non-Hodgkin‘s lymphomas. Lancet 2:448, 1987 22. Hall PA, Richards MA, Gregory WM. d’Ardenne AJ, Lister TA, Stansfeld AG: The prognostic value of Ki67 immunostaining in non-Hodgkin’s lymphoma. J Pathol 154: 223, 1988 23. Grogan TM, Lippman SM, Spier CM, Slymen DJ, Rybski JA, Range1 CS, Richter LC, Miller T P Independent prognostic significanceof a nuclear proliferation antigen in diffuse large cell lymphomas as determined by the monoclonal antibody Ki-67. Blood 71:1157, 1988 24. Miller TP, Grogan TM, Dahlberg S, Spier CM, Braziel RM, Banks PM, Foucar K, Kjeldsberg CR, LevyN, Nathwani BN, Schnitzer B, Tubbs RR, Gaynor ER, Fisher RI: Prognostic significance of the Ki-67-associated proliferative antigen in aggressive non-Hodgkin’s lymphomas: A prospective Southwest Oncology Group trial. Blood 83:1460, 1994 25. Schwartz BR, Pinkus G , Bacus S, Toder M, Weinberg DS: Cell proliferation in non-Hodgkm’s lymphomas: Digital image analysis of Ki-67 antibody staining. Am J Pathol 134:327, 1989 26. Tominaga K, Yamaguchi Y, Nozawa Y, Abe M, Wakasa H: Proliferation in non-Hodgkin’s lymphomas as determined by immu- 3677 nohistochemical double staining for Ki-67. Hematol Oncol 10:163, 1992 27. Sebo TJ, Roche PC, Witzig TE, Kurtin PJ: Proliferative activity in non-Hodgkin’s lymphomas: A comparison of the bromodeoxyuridine labeling index with PCNA immunostaining and quantitative image analysis. Am J Clin Pathol 99:668, 1993 28. Swerdlow SH, Westermann CD, Pelstring RJ, Saboorian MH, Williams ME: Growth fraction in centrocytic and follicular center cell lymphomas: Assessment in paraffin sections with a proliferating cell nuclear antigen antibody and morphometric correlates. Hum Pathol 24540, 1993 29. Macartney JC, Camplejohn RS, Moms R, Hollowood K, Clarke D, Timothy A: DNA flow cytometry of follicular non-Hodgkin’s lymphoma. J Clin Pathol 44:215, 1991 30. The International Non-Hodgkin’s Lymphoma Prognostic Factors Project: A predictive model for aggressive non-Hodgkin’s lymphoma. N Engl J Med 329:987, 1993 31. Mann RB, Berard CW: Criteria for the cytologic subclassification of follicular lymphomas: A proposed alternative method. Hematol Oncol 1:187, 1983 32. Jaffe ES, Raffeld M, Medeiros U:Histopathologic subtypes of indolent lymphomas: Caricatures of the mature B-cell system. Semin Oncol 203, 1993 (suppl 5 ) 33. Key G, Petersen JL, Becker MHG, Duchrow M, Schliiter C, Askaa J, Gerdes J:New antiserum against Ki-67 antigen suitable for double immunostaining of paraffin wax sections. J Clin Pathol 46:1080, 1993 34. Bacus JW, Grace LJ: Optical microscope system for standardized cell measurements and analyses. Appl Optics 26:3280, 1987 35. Dawson AE, Norton JA, Weinberg DS: Comparative assessment of proliferation and DNAcontent in breast carcinoma by image analysis and flow cytometry. Am J Pathol 136:1115, 1990 36. Kennedy JC, El-Badawy N, DeRose PB, Cohen C: Comparson of cell proliferation in breast carcinoma using image analysis (Ki-67) andflow cytometric systems. Anal Quant Cytol Histol 14:304, 1992 37. Kaplan GL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457, 1958 38. Pet0 R, Pike MC, Armitage P, Breslow NE, Cox DR, Howard SV, Mantel N, McPherson K, Pet0 J, Smith PG: Design and analysis of randomized clinical trials requiring prolonged observation of each patient. 11: Analysis and examples. Br J Cancer 35:1, 1977 39. Cox DR: Regression models and life tables. J R Stat SOC 34187, 1972 40. Bacus SS, Goldschmidt R, Chin D, Moran G, Weinberg D, Bacus JW: Biological grading of breast cancer using antibodies to proliferating cells and other markers. Am J Pathol 135:783, 1989 41. Schrape S , Jones DB, Wright DH: A comparison of three methods for the determination of the growth fraction in non-Hodgkin’s lymphoma. Br J Cancer 55:283, 1987 42. Holte H, deLang Davies C, Beiske K, Stokke T, Morton PF, Smeland EB, Hoie J, Kvaloy S: Ki67 and 4F2 antigen expression as well as DNA synthesis at relapse/tumor progression in low-grade B-cell lymphoma. Int I Cancer 44:975, 1989 43. Houmand A, Abrahamson B, Tinggaard Petersen N: Relevance of Ki-67 expression in the classification of non-Hodgkin‘s lymphomas: A morphometric and double-immunostaining study. Histopathology 20: 13, 1992 44. Cibull ML, Heryet A, Gatter KC, Mason DY: The utility of Ki-67 immunostaining, nuclear organizer region counting, and morphology in the assessment of follicular lymphomas. I Pathol 158:189, 1989 45. Weinberg DS: The role of cell cycle activity in the generation From www.bloodjournal.org by guest on December 29, 2014. For personal use only. 3678 of morphologic heterogeneity in non-Hodgkin’s lymphoma. Am J Pathol 135:759, 1989 46. Weiss LM, Strickler JG, Medeiros LJ, Gerdes J, Stein H, Warnke RA: Proliferative rates of non-Hodgkin’s lymphomas as assessed by Ki-67 antibody. Hum Pathol 18: 1155, 1987 47. Ellison DJ, Nathwani BN, Metter GE, Turner RR, Burke JS, Miller TP, Mann RB, Barcos M, Kjeldsberg CR. Winberg CD, Whitcomb CC: Mitotic counts in follicular lymphomas. Hum Pathol 18:502, 1987 48. Bartlett NL, Rizeq M, Dorfman RF, Halpern J, Horning SJ: Follicular large-cell lymphoma: Intermediate or low grade? J Clin Oncol 12:1349, 1994 49. Weisenburger DD: Pathological classification of non-Hodgkin’s lymphoma for epidemiological studies. Cancer Res 52:5456s, 1992 (suppl) 50. The Non-Hodgkin’s Lymphoma Pathologic Classification Project: National Cancer Institute sponsored study of classifications of non-Hodgkin’s lymphomas: Summary and description of a working formulation for clinical usage. Cancer 49:2112, 1982 51. Weisenburger D, Zahm S, WardM, Babbitt P, Holmes F, Boysen C, Robel R, Saal R, Vaught J, Cantor K, Blair A: NonHodgkin’s lymphoma associated with the agricultural use of herbicides: Analysis by histologic type. Mod Pathol 3:105A, 1990 MARTIN ET AL 52. Osborne CK, Norton L, Young RC, Garvin AJ, Simon RM, Berard CW, Hubbard S, DeVita VT: Nodular histiocytic lymphoma: An aggressive nodular lymphoma with potential for prolonged disease-free survival. Blood 56:98, 1980 53. Glick JH, McFadden E, Costello W, Ezdinli E, Berard CW. Bennett JM: Nodular histiocytic lymphoma: Factors influencing prognosis and implications for aggressive chemotherapy. Cancer 49340, l982 54. Kantarjian HM, McLaughlin P, Fuller LM,Dixon DO, Osborne BM, Cabanillas FF: Follicular large cell lymphoma: Analysis and prognostic factors in 62 patients. J Clin Oncol 2:811, 1984 55. Horning SJ, Weiss LM, Nevitt JB, Wamke RA: Clinical and pathologic features of follicular large cell (nodular histiocytic) lymphoma. Cancer 59: 1470, 1987 56. L6pez-Guillermo A, Montserrat E, Bosch F, Terol MJ, Campo E, Rozman C: Applicability of the International Index for aggressive lymphomas to patients with low-grade lymphoma. J Clin Oncol 12: 1343, 1994 57. Bastion Y, Coiffier B: Is the International Prognostic Index for aggressive lymphoma patients useful for follicular lymphoma patients? J Clin Oncol 12:1340, 1994 58. Aviles A: The International Index is not useful in the classification of low-grade lymphoma. J Clin Oncol 12:2766, 1994 From www.bloodjournal.org by guest on December 29, 2014. For personal use only. 1995 85: 3671-3678 Prognostic value of cellular proliferation and histologic grade in follicular lymphoma AR Martin, DD Weisenburger, WC Chan, EI Ruby, JR Anderson, JM Vose, PJ Bierman, MA Bast, DT Daley and JO Armitage Updated information and services can be found at: http://www.bloodjournal.org/content/85/12/3671.full.html Articles on similar topics can be found in the following Blood collections Information about reproducing this article in parts or in its entirety may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://www.bloodjournal.org/site/subscriptions/index.xhtml Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.
© Copyright 2019