Document OJzxYrV4QjxQK1Km2L7b8zn7M

Non-Hodgkin's Lymphomas in Leukemic Phase: Clinicopathologic Correlations (A/KT) From Ihe Medicine Branch. nCT. the 1.ab)ratory of Pathology. IJCBD. National Cancer Institute. and the Clinical Pathology Department, Clinical ( h t e r , National Institutes of Health. Bethesda. M:iryland. Rcqiicsts for reprints shoirld he adtlressctl 10Dr. Elaine S.laffe. Laboratory of PaIliolii~h. iltling 10,Room 2N110. National Inslilutr~sof I Icalth. Bnthesda. Maryland 20205. M;inriscript acccptctl lime 2, 19130. * I'rcscnl iddress: Ik!malology-Oncology IJnit. Tlcfli Israel 1lospital. Boston. Massachrrsetts 02215. Prcsciit address: Departmen1 of Medicine. h l k c University School of Medicine. Durham. North Carolina 27710. Prescnt address: Department of Pathology. The lnhns I Iopkins 1lospilal. Baltimore. Meiyland 21205. A leukemic phase occurred in 30 (14 percent) of 214 patients with non-Hodgkin's lymphoma. To determine the significance of peripheral blood involvement in each type of NHL, patients were subdivided according to a modified Rappaport classification. Each histologic subtype presented a homogeneous clinical picture which differed from that seen in other histologic subtypes. Of particular iruie was the recognition of two distinctive cytologic and clinical subtypes within the category of nodular lymphoma, poorly differentiated lymphocytic lymphoma (NPDL). In one subtype, the predominant cells had cytologic features akin to those of lymphoblasts. In these cases, although the interval to peripheral blood involvement was variable, the median leukemic survival was only two months. In contrast in conventional NPDL the median leukemic survival was 43+ months, and peripheral blood involvement did not appear to exert an independent effect on prognosis. In diffuse large cell lymphomas the median leukemic survival was 0.5 months, with peripheral blood involvement appearing as a terminal evefit assodated with unresponsive disease in multiple sites. The recognition ofadult lymphoblastic lymphoma as a clinicopathologic entity with a high risk of leukemic conversion, 100 percent in this study, is also confirmed. The presence of malignant lymphoid cells in the peripheral blood is an infrequent but well appreciated manifestation of non-Hodgkin's lymphoma [I -31. Early observers considered this phenomenon a lymphatic leukemia developing late in the course of a lymphosarcoma 141. In 1937. Issacs [5] described the morphologic similarity between the circulating malignaht cell and that in the lymphosarcomatousnode, suggesting that the leukemic phase was a manifestation of the underlying disease. Although all but one of Issacs' cases became leukemic preterminally. Schwartz [e],and later Zacharski [7], reported that leukemia with these distinctive cells-lymphosarcoma cell leukemi [LSCL)-could be the presenting manifestation of lymphosarcomi Schnitzer et al. [8],applying the classification of Gall and Rappapor concluded that LSCL was the leukemic phase of poorly differentiate lymphocytic 1yMilphbina. Subsequent reports have established that leukemic phase may occur in other subtypes of non-Hodgkin's lyn phoma [9-121. Most previous descriptions of such a leukemic picture were writte prior to the routine application of Rappaport's classification and prit to recent advances in the teatment of lymphoma which have producc gains in disease-free survival 1131.Our population with lymphoma hi been carefully classified, staged and prospectively followed wit November 1980 The American Journal of Medicine Volume 89 867 -__ Cases (no I 5 Age (yr) Median Range 49 40-57 Marrow involvement at diagnosis 4 Leukemlc at dlagnosls 3 Subsequent leukemlc progression 2 Maximum whlte blood cell count (X 103/mm3) durlng leukemlc phase Median 21 Range 7-31 Ahnormat white blood cell count (%)during leukemic phase' Median 64 Range 46-87 Cerebrosplnal fluid involvement Total survival (mo) 0 Median Range 54 43-176 Survival after onset of leukemic phase (mo) Median 43 Range 20=54 5 47 39-55 1 0 5 15 7-208 90 61-95 2 54 12=115 2 1-3 1 59 59 1 1 0 39 90 ... 0 21 ... 21 ... 6 42 29-65 6 5 1 49 36 12-65 1 0 4 36 2-65 4 4 5 23 1 1 31 11-92 7-17 8-120 85 16 85 50-100 13-32 14-95 2 47 42 10 13-110 4- 13 3-45 37 0 5 4 13-110 02-1 2-35 NOTE: Groups I 4- II = see text. NPM. = nodular poorly dlfferentlatedlymphocytlc; NPDL-B1 = nodular poorly dlfferentlated lymphocytic with blastb cytologic features; NLCL = nodular large cell, DWDL = diffuse well dlfferentlated lymphocytlc: DCCL = diffuse large cell; DLRL = dlffuse Iymphoblastlc ' Hlghesl percentageof peripheral white blood cell dlfferentlalcomposedof malignantcells. ciirrent treatment protocols. Accordingly. w c scl otit to ASSCSS a n d compare tho pathologic and clinical fentiires of leuknmic involvement i n tho various types of nnn- I Ioclgkin's lymphoma. MATERIALS AND METHODS 'l'he clinical records of 214 consccitlive. Iirevioiisly iinlroatctl Iiiiticnls with non-ttodgkin's lymphoma followed at tho hfctlicine Branch of the National Cancer Institute Iictween 1965 nntl 1977werecvnliiatcd. Thirty pntients(l4 pcment] who. at some ptiint in their cniirse. had repeated documentation of a~ leiist 10 Iwrcent miiliflant calls in the white blotd cell count tliffrrrntial. form the tiasis of this report. All 30 patients had crirnpletc I i l o r i d crirints. pcriplieral sinears. ant1\)onemarrow aspiratesand hiopsicsns part nf thr:ir inilia1 cvaliiatinn. Complete hlnotl counts nnri peripheral s i n r i m were nhtninetl at each visit for all pntinnts.Repcat hone iiiiirrnw examinations were performed in 27 patients. Thc p~ripheraslmears and the pathologic material from all dingniislic biopsies and recurrences were reviewed withoiit kiiowleclgo of the patient's clinical course. Each pnlicnt was c:l;issifiod by tho Rappaport system 1141wilh minor moclificn- tinns 1151. The mitotic index per high powered field wns determinod on Iwth primary and all subsequentlymph node biopsy speciinms. 'This figure represents the average number of mitoses per high poweroti field in 20 ciinseciitive high powered fields withniit regard lo intra- or extranodular h a t i o n . For nodiilnr lymphomas. the degree of notliilority of the primary biopsy spcirnon was ~ r a d e don a scale of 1 to 4. One plus indicded vngiir. pnorly drmarcntrtl nodiiliirity: 4 inclir.itrtl shiirlilj ' tlrlinrd rlistincl nodulnrity throoghoril the ncxir I n 16 or tho 30 patients. ncopl;tsticcrlls were niitainrtl from cither peripheral blood antI/or lymph notlcs or splrriis IIU immunotyping. Srrspensions of monnniiclcar cclls werr pi' pared by standt~rdlzed.previously doscrihotl mbthotls 1t"1 Cells were charecterizetl for srirfnce ImmunogIoIiuIins niid rosette fdrmation with erythrocytes coated with 1gM antilincli and complemanl (EAC).crythrocytescoatw! with fgc; antilmh 8llzed sheep erythrocytes[E) by prcvloiiql\ described techniques [1R.171. I n eight patlonth. tcrmin.il deoxynucleotidyl transferase (TdT]activity was nssny~clv methods previotdy reported [in] Drtniled resiilts of the sirface marker studies of these lymphomas ore rrpnrted clw- where [iSj. lrnmunotypes were assigned based on thrsi' standard immunologic assoye. B cell (umors had siirfacr lni- munoglobulin and. In most cows, EAC receptors T coil tumors were identified by their ability to form E rosettes. Tiimnr~ lacking B or T c e l l markers were termed "null." All patlnnls were staged pathologically as oritlinecl by Ihr Ann Arbor conference (201. The inltial treatment of the 311 patlents consisted of cycllcd comliination chemotherapy (In pallent$). sin& Llkylaling agents ( l o w patients). lotnl h h lrradiallon (six patlenle] or Involved field radinthrropy (tn'~~ patientaj. The details of these regimens are descrlbed elw- where [21-24]. b m p l o t e MmfdOh indicated tlinappcarance of all evidence of disease ascertained by restaging one month ofter the ces- ornation or thorapy. The duration the remission was calculatrd from the last day of treatment. Partial remission was delinet1 688 November 1880 The Amerlcan journal of Medicine Volume 68 I I I 31 1. 7 17 R 120 ,is 75 pvri:rtiI rixI1ii:tiwi iif H I I rviih-nlilismsr.I'riiyressinn fir cin Ilicriqiy wiis i ~ i i i i s i ~ l r riiidn rrsIninsc. Siirvivnl wiis miwwrrcl from I h n IwRinniiig r i f trrntnicnl. . .KESULTS Iliirtenn pationts wore Ionkcmicat ! h c ! i ; i ~ui~prescntiilinn; in 17. a leukcmic phasc cvolvctl later in the i:linicnl e:eiursc. The extent of peripheral involvcmcnt illiring tho leiikemic phase varied consitlerahly. ranging from 13 iwrccnt malignant cells in a total white call count of 7,nno/mm3 to 95 percent malignant cells i n n I~itawl hitc count of 208,00O/mm:' (Table I]. No patient had malignant cells in the peripheral hlood in the absence of marrow involvcmcnt ly tiimor. Thc histologic siiliclassificatioiiof tlw 30 patients with Iwkcmic rnanifeslationa appears i n 'I'nblc I. Nodular Lymphomas. In 11 crf lhc 30 patients. tho condition was classified as notliilnr lymphonin. On thc hasis of lymph node histologic and cytologic fcaturcs. these patients c:ould. in turn. ha suhclivided into two groups (Flgure 1). The lymph nntlcs from five patients, hcrenfter referrcd to as nodrilar group I. had tho characterislics of nwlul;~rpoorly tlifferentietcd lymphocytic lymphome (NPDI,] 1161. The neoplastic cells within the notlules consistad prctlominately of small clnnved irregulnr lymphoid cclls with clumpcd compact nuclear chromatin. inconspicuous nucleoli and scant indistinct cytoplasm (Flgure 2). Serial node biopsy specimens revealed retention of nndiilarity and little increase in mitotic index (Table 111. No morphologicdifference WAS d)servetl nmnng the nodes of tho three pntients who werc leirkemic at the time of biopsy compared to the two ih whom leukemic progression subsequently developnil. Further, the histologic and immunologic features i n these five patients(Tab1e 111were indistinguishable from tliose In patientswith NPDL in our pnprilntion in whom leukemic involvement never developed. Lymph nodes from the six patients who coniprisntl nodular group II displayed cytologic features suggestive of le&cellulnr diflcrentiation. In five patients. who had no evidence of leukemic progressinn nt Ihc time of node biopsy. the majority of the cells within the nodillas resembled lymphoblasts (Figure 2) [lo]. n lindinR nnt previously descritwcl in nodular lymphomas. in each of these cases. referred to as nodular poorly clilfercntiatcd lymphocytic lymphoma. blastic (NPDL-Bl]. the prcdominant cells were or moderate size, 15 to 26 niii in diameter. with sparse pale-staining cytoplasm. The nuclear contours were round or oval with slight iridentation&Convoluted nuclei were not oliscrved 1251. Nuclear chromatin was finely disperscrl with 0 to 2 small basophilic. inconspicuous nucleoli. A siilqmpii- lation of small cleaved lymphocytes was also present. This histologic picture can he readily distingriishcrlfrnni the Rappaport nodular mixed lymbhomns in which larger "histiocytic"-appearirig cells. rather than the smaller blast cells, are RSIICI~IC~ with the small cleaved lymphocytes [is]. The mitotic index ('rahle 11) in the primary hioi)sy specimens from these five patients was more than twice that obarved in nodular group 1, and a starry sky pattorn WRS frequent within the nodules. Three of these patients underwent repeat node biopsies. all of which showcd November 1986 The Amerlcen Joirrnelof Mediclne Volume 69 969 - - . - - .. - - - - - -- . _ME! Cases (no ) Degree of nodularlty ( m ~ ; , Milolotlc Indexmf P~IWWYbiopsy (nmin\ Recurrences (nmsnn) Irnmwrotyplng No sttdled Results Termlnel tra~sferrso(tin + - -positivelno.stdk~J) NOTE: &OW I II wllh blastlc cytologir Ionhi = dllfuse Iymphobksl~c * Based on three hlopqy + Based on four bioD3i. CI -- A progression to a diffuse growth pattern. an increased proportion of cells with blastic cytology, an increased mitotic index and prominent starry sky patterns. The neoplastic cells seen in the peripheral blood cliiring the leukemic phase (Flgure 3) accurately re- flected the contrast in lymph node histology between thc patients in nodular group 1 and the five patients with hlastic cytology in nodular group II. The sixth patient in nodular group I1 had a primary diagnosis of nodular lymphoma. large cell ["histiocytic") type (NLC1.J 1151. lmrnunotyping revealed the two "HAYstlidled and the nodular large cell tumor to be of B cell origin. TdT activity was undetectable in these three patients. The clinical features of these two subgroups of nod- ular lymphoma are compared in Table I. In the thrrlr, patients in group I who presented with a leukemic h l d t~icture,a complete remission was achieved a h treatment with the combination of cytoxan, vincristine. and prednisone (CVP).the duration ranging from sevcI1 to 24 months. Although all have had a relapse, two nrc still olive. None of the three relapses occurred in tlw peripheral blood, and a leukemic phase subsequenlly recurred In only one patient. Two patients in nodiil~r group I hecame leukemic at 156 and 46 months aftcr their initial presentation. Both are still livlng. one is in a continuing complote remission after treatment with CVP.In group I the median survival will exceed 54 Figure 3. Peripheral hIrv group 1. Cells have spame I in NPDC. blastic Predomicr and smell nucleoli.OCCHW 61sIn Figure 28 (Wrlght's. finely dlstrlbuted nuclear c l ~ with deeply basophlllc cy14 nucleoli. Wrlght's staln: 1ni3 870 November 1880 The Amerlcan Journalof Medlcine Volume 60 TABLE II Cllnlcal Features ..~-.________ LEI JKEMlC PHASE OF I.YMPI{OMA-COME ET AI,. Cases (no.) noqee of noduiarlty (mean) Mitotollc lndexihpl Prlmnry blopsy (mean) Recurrences (mean) lmmunotyplng No. studled Results Termlnal transferase (no. positiveha. studied) 5 2.8 1.2 1.5' 3 All B ... 5 2.0 2.1 12.5' 2 Both E 012 1 1.0 6 .. . 4 .. . 9 ... 6.6 ... 0.8 9.3 5.7' . . . IO.1 ... 1 34 3 B AllE 2 8 1 null 2 T 2 n u l i 0/1 . . . 011 4t4 NOTE: Groups I t II = see text. NPDL = nodular poorly differentiatedlymphocytlc; NPM-B1 = nodular poorly dlfferentlated lymphocytic with blastic cytologic features; NLCL = nodular large cell; DWDL = dlffusewell dlfterentiatdlymphocytic;DLCL-dlffuse largecell;DLBL = dlffuse lymphoblastic. '' Rased on three blopsy speclmens from two patients. Rased on lour biopsy specimens from four patlents 89 In Figure 28. (Wright's. X 1000). E, Lymphbblastlc lymphoma. Leukernlclymphoblastshave extremely sparse cytoplasm, finely distributed nuclear chromatin and inconspicuous nucleoli. f, Diffuse "hlstlocytlc"lymphoma. Ckculatingcells are large with deeply basophlltc cytoplasm. Nuclel are markedly Irregular In configuration with retlculatedchromatin and prominent nucleoli.Wright's staln: magnlflcationX 1,000, reduced by 13 percent. Novemher 1980 The American lournel of Medlclns Volume 88 671 months. and the mcdian leukemic siirvival will exceed 43 months In the patients with hlastic cytology in nodular group I1 (NPDL-Bl), the chiration of the preleukemic phase varied from 11 to 114 months; however. in contrast to thosc in nodirlar group I. the median leukemic siirvival was two months. During the leukemic phase iii none of thcsc patients was even a partinl remission achiww!. The patirnt in noclulargroup I1 with the large cell (histiocytic)histology entered a transient complcte remisdon hut had a relapse in sites of prior involvement including the pcriphcral blood. Diffuse Well Differentiated Lymphocytic Lymphoma (DWDL). Six patients with leukemic manifestations werc rlassifictl as having DWDI. (Tablc 11) [is] Growth centers composed of large cells with vesicular nuclei and prominent central nucleoli were observed in two of thrm I26.27) Although the histology and initial lymphocytosis in livc patirnts (Table I) is compatible with an alternative diagnosis of chronic lymphocytic leukemia. the young ages and dominant nonhematologic disease manifestations--bulky adenopathy in all, gastrointestinal invnlvrment in two. breasl and cutaneous involvcment in one cnch--prompted their inclusion in this scries The sixth patient became lcukrmic 10 months after diag- nosis. Treatment. which varied in intensity from oral al- kylating agents to total body radiation plus CVP.pro- diiced no complete remissions. In three patients the condition subsequently converted to a less differentiated lymph node histology. In one it progressed to a diffuse lymphoma. poorly differentiated lymphocytic type with lymphohlastic features whereas in the other two diffuse large cell tumors developed of the undifferentiated pleomorphic type, compatlble with Richter's syn- drome. Diffuse Large Cell Lymphoma (DLCL). Four patients in whom a leukemic phase evolved were identified as having a diffuse large cell lymphoma based on initial lymph node histology (Table 11) In three It was sub- classificd as "histiocytic" (Figure 2) (151. In one patient it was suhclassificd as undifferentiated, pleomorphic (non-Bnrkitt's) (151. This was distinguished from the "histiocytic" type by a smaller cell size. less abundant cytoplasm and smaller although still distinct nucleoli. In no patient was a complete remission achieved during initial treatment with combination chemotherapy. and a leukemic phase emerged a median of nine months after diagnosis (Table I) [Figure 3).The number of malignant cells in the peripheral blood in these patients was the smallest among our subgroups, and the median survival after the occurrence of leukemia was only two weeks, the shortest in this series. The leukemic phase was associated with progressive disease in other sites. Diffuse Lymphoblastic Lymphoma (DLBL]. Nine pa- tients with peripheral blood Involvement wore classified hy initial lymph node histology as having diffuse lym- phoblastic lymphoma (Table 11) The predominant 1 , ~ 1 in these tumors had the cytologic features of a ~\ill. phohlast (Figure 2 ) [lo].Nuclear diamcter was 12 i f 1 '11 mu. Convoluted nuclei were conspicuous in two I,+ ticnts, and evidence of lymphoid differentiationitltil progressive condensation of nuclear chromatin \ I , ~ ~ ohserved in three (10.251. Immunotypically, the rclllr patients studied had niiil or T cell markers. and '1 1 activity was detected in each of them This histology wasconfined neither to the young lil,l to those with mediastinal Involvement (Taldr 11 [10,16.25] Five of the nine patients were over age 31) 111 these older patients. extranodal sites of involvemrrli skin. gastrointestinal tract. hone, kidney, hcart- MI I,< particularly common. and a mediastinal mass wns I 11. countered only once Four patients were leukemic ,II diagnosis; in the remaining five, ail of whom had 1 1 ~ ~ marrow involvement initially, a leukemic phasr 11,s veloped a median of 14 months after diagnosis ( F i p i 3).In either event, the leukemic phase of the diseasr I' poorly responsive to treatment. with L-asparngin,l,$ proving to be the most active agent. The mrclian survii 11 in leukemic patients was four months whethrr poiilili era1 blood involvement was an initial or late maiiifr tation of disease. Leptomeningeal Involvemenl. Meningeal lymphorw documented by cerebrospinal fluid cytology on p a t h logic examination, developed in 15 of the 30 patients 111 only two patients did meningeal involvement prcddl the onset of the initial leukemic phase, hy two and t ' months, respectively In five patients, meningeal in,volvement occurred during systemic remission COMMENTS In this series, leukemic manifestations were observrtl in 14 percent of the 214 patients with non-Hodgkin's lymplpna. Although this figure is consistent with pricli reports [3.5,28-30]. it is more meaningful to consiih incidence in relation to pathologic subtype. Periphet J I blood involvement occurred in 10 percent of the patients classified 9s having NPDL (nodular group I) and iii 1; percent of those classified as having DLCL In 0111 over-ell population. In both of these histologic groiip. which are the most common subtypes of non-Hodgkin's lymphoma, an analysis of lymph node histology, aw. sex, stage, Initial sites of involvement, including thv marrow, and Initial complete blood count did not distinguish nonleukemic patients in whom a leukemic phaee was to evolve from those who have never evidenced peripheral blood involvement. Further. thv patients wlth NPDI, who were leukemic at the time d diagnosis cnuld not otherwise be differentiated hy thv aforementioned parameters from their nonleukemic counterparts. Seven patients with nodular lymphoma and blastic cytology (NPDL-El)were noted among the 214 patients with non-Hodgkin's lymphoma. Although none was leukemic at the time of diagnosis five, discussed In this manuscript, had Ieukemic progression. Furlher. all nine 872 November 1880 The Amerlcan Journal of Medlcine Volume 88 i9;ilinnls with tliffiisc 1ymphol)l:istic (DLUL) tumors in 1lll! over-;iIl grniip of 214 i i h ~ i t c l ymanifested pet ililinral I)locicl involvcmcnt. Thus. in contrast to thc i~,iri~mctt!rcsited hcarrin that were not prcdictivc of I,,iikcinic conversion. lymphnlilastic or l)liMic lymph l i , ~ cdy~tolol~yin either a difftlsc o r a nodular growth il;ittcrn tlocs srcm to lie a harliingcr of leitkemic pror:rc!ssion i f it is nnt iilrciitly in evidence. Further. once Ilwkcmic c~nv~!rsinnoccitrrcd in piiticnts with ~ J I W ~ , - I I tIh, e (:niirso rcsom1)lccl that of r)I,Bl, iri its 11511 k t m ic: ph;isc. f~owc?vctrh. crc iippc;ir to he ccrtain diffcrcnccs be- I W ~ I 1~11BL itntl our ciisos nf NPDI,-Di. Cytologically, . l l ~ h o ~ ~cgonl ivtilittcd cc?lls :ire commnn in I11,DL. i n .\Jllt~l.t-nthey worc not seen. Furthermore, the cells of rdPI)I.-l31 tlil'fcrcd from those of DLBL in having a ~ ~ i t n e w h aIitirgcr cell size (15 to 25 mu in NPUL-Rl I ITSIIS 10 to 1'5 niit i n 131.1%) antl more nlirinclant cyto- iilasm ilistingiiis1ial)lc in smears as a thin I)asophilic rim iiiiniitnolo~ic:ally,tlicsc ciiscs arc also tlistingiiishalilc. I'hrrc p;itic?ntswith N P 1 ) I A l wcre stridictl in the tcr~ i i i i i i i lIciikrniic phase, and all had 0 cell mnrkcrs in- 8 Iiitling rnciiioclonal surface irntniinoglol~itlin.This is g.onsisIcn1 with the recognition of notlrilnr lymplioma .IS ii 11call t i l m o r . I n contrast, the sitrf;icc markers nf the h i r piiticrits with l)l,nl, stritlicd were in kcqling with 1 1 rn~ iirkcrs 1)rcvioitsly rcportctl for this tliswsc: one . n i i l l . two wcro 7'. and one had an isolatrtl (:3 re- i q i t o r witliciiit other 13 o r '1' r:clll miirkcrs. F t i r t l i v , i n two ji:itimts with NPI)I.-Dl stiitlicd TtlT was altsent in ~.ontriistn Ihc rilkliiitoiis I)rcsc!nc:c of 'TdT i n I X U I , [In]. I'hlis. tliis rwitlcncc intliciilcs that tlicsc tiiinors a r c hi~ h g i c a l l ytlistinct. 111 a prcvioiis review of follicrilar lymphoma. Spiro ~t ill. 1311 tlist:rilicd five patients with a terminal leukc- inic phase who clinically and cytnlogically rcscmlile our lliilicnls with NPDL-B1. ilowevcr. in that stridy no i,orrcl;ttivc lymph notlc histologic features wcre demd)od either prior to or during the leukemic phase. In lmr patic~itswith NPDI,-Rt. cells with lilnstic cytology w r r priwnt in lymph notlcs tliiriiig the nrtcn prolonged I)rclciikcinic pIi;isc. h.lorphologic classificiition also proved important as risk tlctcrminant of cerebrospinal fluid invnlvcmcnt which tlc\wlopctl in 15 of the 30 leukemic patients. This t:nmpIication w;is restricted to pntients with oithcr lynipholilastic or large cell histologies. nodular or diffirsn.atid occrirrctl in 68 percent of these patients in this swic?s.This liguro is in contrast to the 3 1 percent inci- clt!iicc of iicrvnits system involvemcnt in unselccted Iwtionfs with nnn-llodgkin's lymphoma (32.331, but it is compi1r;ihlc to the 65 pcrccrit incidencc! figrirc for patients with DtCL and marrow involvement 1)rovictusly reported from this institution [34]. An cvaluatinn of prophyhctic central nervous system therapy scrnis warranted in these high risk histologies in which CNS disease both contribittcs to systemic failure antl is clilficult to eradicate once rstaldished. Three clinical patterns of peripheral hlootl involvcment in non-Hodgkin's lymphoma emerge from this study. In NPDL (grrwpI]. the relatively intlolcnt ni\ttiiiil history Is reflected in the leukemic phase which nppcars to have liitlo independent effect on the outcome of the illness. Indeed. the rate and degree of rcsponsr. and the total survival ol patients with NPDL who eithcr prcscnt with peripheral blood involvement or in whom i t clr- velops approach these results in nonlertkemic patients with NPDL. Bloomfield et al 1351 have similarly rrportcd that neither hone marrow nor pcriphcral I)lootl involvement et diagnosis correlate with survival in N P D L The total survival in patients with T)WDI, antl lcrikemic manifestations was also long. althori~hthr rcsponsiveness to therapy and the survival of hicsr patients was somewhat less than that in our nonleiikemic patients with this histology [13] and lcss than t h t observed by Rappnport et al in his cases of DW1)I. ; i n d chronic lymphocytic lertkemia. At the opposite end of the spectrum. in diffiisr Iargr cell lymphomas, leukcmic conversion appears as ii terminal event associatrd with iinrcsponsivc tliseasr in multiple sites and again exerts no important intlrprndent effect on prognosis. Between these poles arc norlrilar lymphomas with blastic cytology [NPDL-nl] and tlifhist? lymphoblastic lymphoma When it cmcrgcs i n thrse patients, a leukemic phasr heconics thc domin.int disease manifestation and signals an ahritpt ch;ingc?i n the clinical course with poor response to cnnventional treatment and rapid death. The almost invariahlr prn- gression to a n aggressive leukemic pictiire nllows us to predict. and therefore possibly to avoid. a poor nittcomr in patients with this cytology and ostensibly more lo- calized disease. Recently, extended diseaso-free sur- vival has been achleved in pediatric patients wlth t>l,BI, following intensive treatment with multiple chrmn- therapeutic agents in the preleukemic phase of the disease 138,371.Preliminary studies indicate that this approach m&ybe tisefril in adults with diflrtse lymphoblastic tumors [38]. It is uncertain whether this s~icccs9can be translated to patients with nodrilar lymphoma And blastic cytology as this disease is biologically different from DLRI. despite its clinically similar leukemic course. However, increased rrcognilion of this cytologic type of lymphoma is a newssary first step to devising more effective forms of therapy. November 1980 The American Journal ot Medicine Volume 89 873 -- ~ , . ...., ...... ... . ,,, . gonnl 1926; 15: 4:) 5. Iss;ics R: IJymphosarcomn r:cll Iciikninia. Ann lnlcrn Mod 1937;11: 657. 6. Schwartz [)I,, Pierre RV, Sclionrcr PP. Reed EC Ir. Linman )W: 1,ymplrosarcoma (:ell loiikcnii;i. Am I Mctl 1965: 38: 778. 7. Zacliarski 1,R. 1.inman IW: (;lirr)iiic lyinpliocytic Icrtkcinia vcrsiis diroiiic lymphnsari:oina cell lcuketnin: analysis of 496 ciiscs. Am I Moil 1969: 47: 75. 8 . Sr:hnitzcr B. 1,rrcsel I S , Reed RE: I.ymphosarcoma cell leu- koniia: a clinicopatlioli~gicstitily. Cancer 1970; 26: 1082. 9. 1'atig;tIis (;A, Niilhwani BN. Ritt)paprirt I!: !vLiiignant lym- phoiiia. well diffcrctitintetl iyinlihoc:ylic. (~hnccr1977; 3!): 9!N. IO. Niithwatii ON. Kim 11. Rappaliort I I: Malignant lymphoma, ~ y t n ~ ) ~ i ~ ~ ~C)a~n;ci cs rl il(9:7.G; 38: 964. 1 1 . Si:Itnitmr 11. Kass I,: 1,ciikcmic p l m o rrl rcfic;itlrim cc11 siir- coniii Ihisliocylic lymphoma) Cnnccr 1!)73: 31: 547. I:?. Slmlort Krocsc WP. (%!ton 1'1, Sotncrs R: 1,ctiknonic progri?ssion in lyn~i)hornntaH. r 1 C;inc:nr 1975; 31 (sitppl 111: 102. 13. Atidmson 'T. l h i d c r RA. Fishrir RI. ct ill.: Coinliinitlion chcinotlicriipy in non-I loilgkin's lyrnlilintn;is: rcscilts ni long-lorin fnIkiw-iip. Ciinccr 'Trcat KC^ 1977:61: 1057. 1.1. Ra~ipapnrJtI : Tiimors of the hcm;ilopoictic systcm. Alliis of tumor pathology, scction 3. Iasc:iclc 8. Washington, D.C.. 1l.S.Armcd Forces Inslitrile i t 1 Pathology. 1966. 15. Ocr:trtl CW. Dorfman RF: I listopnlhology o l malignant lymphomas. In: Ri)scnhcrg SA, ctl.(3inics in hcinatology. Philatlnlphia: W.B. Sauntlors. 1974: 3% 16. Iaffr ES. Dr;tylan RC. Frank MM. Grccn 1. Berartl CW: Ilotcrogcneity of immiinologic: niarkcrs L i n d sitrfaco mor- pltolrigy i n childhood 1ytnI)lioIit;tstic: lymphoma. Blond 1976;48: 213. 17. Rraylnn RC. l a r k ES. Triche'r). ct al.: Striictural and Tunc- tional ~)rnlicrticsof the "hairy" cclls of lcrikcrnic icticu- ~oenr~ot~ic~iCoasnicsc. r 1978;41: 210. 111. ncrnlon !A, lalfc ES. Braylan RC: 'Terminal cleoxynuclcoticlyl 1r:insfcrasc activity in malignant lymphomas. N Engl 1Med 1977;297: 461. 19. jaIfc ES, DrayIan RC. Nanbii K, Frank MM. Rcrard CW: Fitnctional markers: a new pcrspcctivc on malignant lyrnphomns. Cancer Treat Rep 1977:61: 953. 20. Carlmne PP. Kaplan I-IS. Mnsshoff K. Smithcrs DW, Tuhiana M: Report of the committee on Ilndgkin'sdisease staging 21. RicghlesysiCfiMca.tiDoenV. CitaanVcerr, Res 1971; 31: 1880. Bcrartl CW. Cancllos CP: Admixed lymphosarcoma: intcnsivc cyclical combination chcino- therapy with cyclophosptiamidc,vincristinc, and prcdni- sone. Ann Intcrn Mod 1972;70: 227. 22. IkVitil V'T Ir, Scrpick AA. Carlmic PP: Combination inollicrapy in tlic trcalmcnt of adviinccd Hodgkin's (lis ~ , 1 . 1 1 1 1 , 1 1 1 . , , I 1.11,<1 I,,, 3 , . ,<,, t,,,,. 23. Schein PS. I)eVita VI' jr. 1hthl)nrtl S , rt :i1 :W o n l y t 1x3 rinniycin. cyclospliamitlc, vincrislinc antl pri~iliiivtw [BACOP] combination chcmolhcrapy in the twaliiii,n! (I: atlvniicctl dilhtsc histiocytic lymphoma. Ann Iiitfwh h ! 1976: 85:417. 24. Iohnsnn RE. O'Conor GT, I.cvin 11:Primary tnan;ij<iwwt atlvancctl lymphrrsarcom;is with ratlintlieralrv. ( ~ I I I , ,I 1970: 25: 787. 25. I1arcos MP, I.iikt?s RJ: Malignant lymplinmn of c r i t i v t i l i r t v ~ t lytiipliocylcs. A new entity o l ptrssil~lcT-c:cIlyl lw I n ! h C . I m t l ( h l t l c n 1 0 , cd. Conllicts i n chilclliontl i m w r r\!,ti\ York: Alan R. Liss. In(:. 1975: 147. 26. R;trising A: 1,ympliocylic Icrtkarnia iintl lyniphcimn in I!V. adult. Acta Mcd Scnnd 1976;595 (siippl): 1. 27. Dick RF. Maca RD: The lymph node in dirtinii: lyiiilili,i!.\ti, Icitkcmia. Cancor 1978;41: 2H.7. 28. Dic:k R. 13lo~1mficIt(l;I], fhrnning RII: 1nc:itloncr. c y t ~ i l i ~. ! ~ , antl histopathology on non-Ilotlgkin's Iyrnpliiirrws iii !Itr* hone marrow. Cancer 1974:33: 1382. 29. McKcnna RW, ~lorrinlielilCI), Rriinning RI): N ~ I i i I . i t lymphoma: bone miirrnw and ~ h r c nl ianilvstatirins f.;iii. ccr 1975: 36: 428. 30. M a t h G, Pottillart P. Schwiirzcnlicrg et al.: I,iwk;wtnir conversion of non-Hodgkin's malignant lympltom;tl;i Iir I Cancer 1975;3 (siippl 111: 96. 31. Spiro S, Galtori UAG. Wiltshaw E. el d:Folliciilnr lynililiiiniri a swvcy of 7'5 cases with spcial rokcrence to Ihr! syiir1rr)rnr. rcsemhling chronic lymphocytic Icrikncmin. Hr 1 1 h 4 ~ r 1975: 31 (siippl Ill: 80. 32. Griffin jW. Thompson RW. Milchinson MI.tlcKic~wid Wclland FH: I,ymphoinatot~sIqitomctiingitis. A m Ih l u l 1971: 51: 200. 33. Young RC, Howscr DM. Aiidmxm T, et a].: (:cnlml nivviiii9 system complications of non-1loclgkin's lym~ilioriiii'.t'tw Iwtential role for prophylactic therapy. Am 1 Met1 19711;fi(i- 435. 34. Brinn PA, Schein PS, Ranks PM,DcVila VT Ir: Centrnl iii'i- vous systcm complications in patients with cliffiisc Iiisiiv- cytic and undiffcrcntiated lymphoma: leiikcniin rcvisibvl Blood 1976;47: 3. 35. Bloomfield CD, McKcnna RW. Rriinning R1.Y: Significaiw of hacmatologic parameters in the non-Hodgkiii's Iyrtr- phomas. Br I Flemalol 1976; 32: 41. 36. Wollner N, Lieberrnan P. Exelhy P. el ai.: Non-Ilotlgkiii's lymphoma in children: results of lreatmont with I,Sh -I. protocol. Br 1 Cancer 1975; 31 (suppl Ill: 337. 37. . Weinstein 11, Buell D, Cassady IR. Vance Z,j:iffc N: t i n proved prognosis for childhood non-HmIgkin's lyml)ltom:i Rlnotl 1977; 50: 212. 38. Rosen 1'1. Feinslein UI, Pattengale PK. 01 al.: Gonvoltitril lymphocytic lymphoma in adiilts. A clinicopalholo~icon- lily. Ann Inlcrn Mctl 1978;89: 319. 674 Novbmher 1980 The American Journalof Medicine Volume 69