Natural Killer Cell Neoplasms: A Distinctive Group ofHighly Aggressive Lymphomas/LeukemiasMichael M.C. Cheung, John K.C. Chan, and Kit-Fai Wong

Natural killer (NK) cell neoplasms, which include extranodal NK/T-cell lymphoma (nasal and extranasal) and aggressiveNK cell leukemia, are generally rare, but they are more common in people of Oriental, Mexican and South Americandescent. These neoplasms are highly aggressive, and show a strong association with Epstein-Barr virus. ExtranodalNK/T-cell lymphoma most commonly affects the nasal cavity and other mucosal sites of the upper aerodigestive tract.Patients present with nasal obstruction or midfacial destruction. Despite the early stage of disease at presentation,overall survival is poor. Patients with the extranasal form of the lymphoma often present with high-stage disease,commonly involving the skin, gastrointestinal tract, testis, and soft tissue, and the prognosis is even worse.Histologically, the lymphoma can show a broad cytologic spectrum, but apoptosis, necrosis, and angioinvasion arecommon. The most common immunophenotype is CD2�, surface CD3�, cytoplasmic CD3�, CD56�. Based on currentlyavailable data, treatment of nasal NK/T-cell lymphoma should consist of radiotherapy, with or without multiagentchemotherapy. More research is required to ascertain the role of high-dose chemotherapy with stem cell rescue andthat of non–multidrug resistance-related chemotherapeutic agents. Aggressive NK cell leukemia affects youngerpatients, who present with poor general condition, fever, and disseminated disease; they often die within a short timefrom systemic disease or complications such as multi-organ failure. The peripheral blood and bone marrow showatypical large granular lymphocytes, which exhibit an immunophenotype similar to that of extranodal NK/T-celllymphoma. Aggressive NK cell leukemia must be distinguished from T-cell large granular lymphocyte leukemia andindolent NK cell lymphoproliferative disorder, both of which are indolent.Semin Hematol 40:221-232. © 2003 Elsevier Inc. All rights reserved.

NATURAL KILLER (NK) cells represent a dis-tinctive lineage of lymphocyte that is closely

related to T lymphocytes. NK cells possess manyimmunophenotypic and functional similarities withcytotoxic T lymphocytes, but differ in the lack ofexpression of surface CD3 molecule and T-cell recep-tor, and the germline configuration of the T-cell re-ceptor genes.3 These cells characteristically expressCD56 (neuronal cell adhesion molecule [N-CAM]),which is also expressed in a subset of cytotoxic Tlymphocytes. NK cells can lyse target cells withoutprior sensitization (spontaneous antibody-indepen-dent major histocompatibility complex [MHC]-unre-stricted cytotoxicity), mostly via the NK receptors.

Neoplasms of NK cells are rare, and have only beenwell characterized during the past 10 to 15 years.Recognition of the NK cell lineage of a lymphoma isimportant, because NK cell neoplasms are generallyhighly aggressive. The new World Health Organiza-tion (WHO) classification of hematolymphoid tu-mors, recognizes three categories of NK cell neo-plasms.10,28,29

The first category, extranodal NK/T-cell lym-phoma,10 replaces “angiocentric lymphoma” in theRevised European-American Lymphoma (REAL)classification. It is called “NK/T” rather than simply“NK” because while most cases are genuine NK cellneoplasms, some appear to be cytotoxic T-cell neo-plasms. The qualifier “nasal” is appended when the

nasal cavity is the primary site of involvement, and“extranasal” or “nasal-type” when other sites are in-volved.

The second type is aggressive NK cell leukemia,12

which will be discussed here and is also briefly ad-dressed by Lamy and Loughran elsewhere in thisissue.

There is little compelling evidence that the finaltype, described as “blastic NK cell lymphoma,”9 trulyrepresents an NK cell neoplasm. In fact, this may bea neoplasm of precursor dendritic cells related toplasmacytoid monocytes (plasmacytoid dendriticcells).31,48 This entity is covered by Bene et al in thisissue.

Extranodal NK/T-Cell Lymphoma

Clinical Features and BehaviorExtranodal NK/T-cell lymphomas show a predilec-tion for Asians, Mexicans, and South Ameri-cans.2,10,53,77 It most commonly involves the nasal

From the Departments of Clinical Oncology and Pathology,Queen Elizabeth Hospital, Kowloon, Hong Kong.

Address correspondence to Dr Michael M.C. Cheung, Departmentof Clinical Oncology, Room 1116, 11/F, Block R, Queen ElizabethHospital, Gascoigne Road, Kowloon, Hong Kong.

© 2003 Elsevier Inc. All rights reserved.0037-1963/03/4003-0007$30.00/0doi:10.1016/S0037-1963(03)00136-7

Seminars in Hematology, Vol 40, No 3 (July), 2003: pp 221-232 221

cavity, which is the prototype of this form of lym-phoma. The median age of patients is 53 years, andthe male to female ratio is about 3:1.3,14

Nasal NK/T-cell lymphoma. Nasal NK/T-celllymphoma is the predominant histologic type of pri-mary nasal lymphoma, at least in the Asian popula-tion.13 The most common presenting symptoms arenasal obstruction, nasal discharge, and epistaxis.13

The full-blown midfacial destructive and ulcerativelesions are much less commonly seen now, probablydue to earlier presentation.10,13,14 Extension to thenasopharynx and paranasal sinuses occurs in 37% to50% of patients with stage I disease.14,49 In a detailedstudy using computed tomographic (CT) scan andmagnetic resonance imaging, the tumor was found tobe locally destructive, with frequent erosion of bones(78% of cases),70 including the medial maxillary wall,orbital floor, lamina papyracea, palatal bone, andalveolar bone. Besides demonstrating subtle bone de-struction, magnetic resonance imaging is superior toCT scan in delineating the extent of soft-tissue infil-tration and in distinguishing tumoral tissue, inflamedtissue, and accumulated fluid.

Although nasal NK/T-cell lymphoma is usuallylocalized at presentation (82% to 92%),13,43,47,49 sys-temic progression often occurs, usually early in thecourse of disease. Common distant sites of involve-ment are the skin, liver, lung, gastrointestinal tract,and testis.13 In our retrospective study of 79 cases,60% of all failures in patients presenting with local-ized disease were due to systemic progression alone,and almost all occurred within 12 months after treat-ment, mainly chemotherapy.14 Similarly systemicfailures have been reported in patients treated withradiotherapy alone at a median time of 7 months aftertreatment.40 Local resistant disease or relapse is alsocommon and has been reported in 50% of patientswith localized disease who had achieved completeremission with radiotherapy.40 In our series, localrelapse occurred in 31% of all patients who hadachieved complete response.14 On the other hand,neck node involvement at presentation or relapse isrelatively rare; regional failure as the sole site offailure occurs in only 4% to 11% of cases.14,40 He-mophagocytic syndrome may develop during thecourse of disease, characterized by fever, precipitousand otherwise unexplained drop in blood counts,presence of hemophagocytic histiocytes in the bonemarrow, coagulopathy, and rapid deterioration inliver function.13,14,40,64

Extranasal NK/T-cell lymphoma. ExtranodalNK/T-cell lymphoma involving sites outside the nasalcavity or upper respiratory tract is rare.1,5,37,39,56 Mostpatients have multiple anatomic sites of disease at pre-sentation, in the absence of lymphadenopathy. The pre-dominant sites of involvement are skin, gastrointestinaltract, testis, and soft tissues, similar to the sites to which

nasal NK/T-cell lymphoma tends to disseminate duringthe course of disease.1,3,5-7,26,37,55,56,86,93,94 The skin le-sions often appear as nodules or plaques with ulcerationand necrosis, and occasionally manifest as diffuse ery-thematous swelling. When the intestines are involved,perforation is the most frequent presenting symptomdue to prominent tissue necrosis, thus differing fromconventional intestinal lymphomas, which usuallymanifest as bowel obstruction. Other sites of involve-ment appear as mass lesions.

Systemic symptoms such as fever, malaise, andweight loss are common. For patients presentingwith apparently localized disease, additional sites ofdisease are often identified on staging, or appear earlyin the course of disease. Hemophagocytic syndromemay complicate the disease at presentation or in theterminal phase.85

The disease pursues a highly aggressive clinicalcourse, with most patients dying within 6 months.According to one series, the 2-year disease-free sur-vival was only 10%.5 The worse prognosis comparedwith nasal NK/T-cell lymphoma may be attributableto the high-stage disease in most patients.3,5,39,60,63

Pathology

The involved tissue commonly shows extensive co-agulative necrosis and ulceration (Fig 1). The diffuselymphomatous infiltrate varies in cytologic composi-tion from case to case, ranging from small, to me-dium, to large cells (Figs 2 and 3).3,10 While the smallcells often exhibit nuclear irregularities and pale cy-toplasm, they can be indistinguishable from normalsmall lymphocytes, causing great difficulties in diag-nosis using morphologic criteria alone.3,16,82 In Gi-emsa-stained cytologic smears, azurophilic granulescan be identified in the cytoplasm of the tumor cells,as expected of the NK or cytotoxic T-cell nature of theneoplatic cells.

Angiocentric-angiodestructive growth is com-mon, although it may not be identified in small biop-sies (Fig 4).3,36 Necrosis and apoptosis are also prom-inent features (Fig 5). Prominent necrosis sometimesmandates repeated and multiple biopsies to procuresufficient viable tissue in order to determine a histo-logic diagnosis (Fig 6).

Immunobiologic Features

The most common immunophenotype of extranodalNK/T-cell lymphoma is CD2�, surface CD3�, cytoplas-mic CD3��, and CD56� (Table 1).3-5,11,19,21,27,57,65,68,73

The discrepancy in expression of surface CD3 (requir-ing fresh or frozen tissue for demonstration) and cyto-plasm CD3 (demonstrable in fresh, frozen, or paraffin-

222 Cheung, Chan, and Wong

embedded tissue) is a distinctive feature, as expected ofNK cells; subunits of the CD3 molecules are present inthe cytoplasm, but the complete molecule is not assem-bled on the cell surface.8,68 The most consistently ex-pressed NK-associated marker is CD56 (Fig 7). Cyto-toxic molecules are, as expected, positive.3,20,24,59,66,87

Nasal CD3�� lymphomas that are CD56� are cur-rently also placed within the category of NK/T-celllymphoma provided that they also express cytotoxicmolecules and harbor Epstein-Barr virus (EBV).10,18

However, if cytotoxic molecules and EBV are nega-tive, the cases should be diagnosed as peripheralT-cell lymphoma unspecified. The clinical and histo-logic features of the CD56� subgroup of NK/T-celllymphoma are indistinguishable from the CD56�

subgroup. While one study reported more favorableoutcome of the CD56� nasal lymphomas comparedwith CD56� nasal lymphomas, cytotoxic markersand EBV status were not applied to delineate thepossible NK/T-cell subgroup from the peripheral T-cell lymphoma unspecified.13 Since information onthe CD56� subgroup of NK/T-cell lymphoma is lim-ited,13,14,24,40,41,43,47,49,51,75,78,80 it would be helpful in

the future to identify this subgroup separately todetermine its prognostic significance.

The T-cell receptor genes are usually not rear-ranged.3,21,24,32,62,68,76,84 For nasal NK/T-cell lym-phoma, EBV is nearly always positive, irrespective ofthe ethnic origin of the patient. For extranasal NK/T-cell lymphoma, EBV association is strong in Orientalpatients, but less consistent in Caucasians.3,54 Notmuch is known about the genetic changes in extra-nodal NK/T-cell lymphoma. So far, the mostcommonly observed changes are del(6)(q21-q25),del(17)(p12-p13), del(13)(q14-q34), and gain of1p32-pter.82,91,92,95

Treatment and OutcomeOwing to the rarity of the disease, all recently re-ported clinical studies are based on retrospectivepatient series (Table 2).13,14,24,40,41,43,47,49,51,75,78,80,97

This discussion will be limited to the nasal form,because information on extranasal disease is ex-tremely limited. Reports in the literature often havenot included complete immunophenotypic data, and

Figure 2. Nasal NK/T-cell lymphoma. The mucosa is diffuselyand densely infiltrated by lymphoma cells, with the mucosal glandshaving been obliterated.

Figure 1. Nasal NK/T-cell lymphoma. The mucosa is ulceratedand covered by fibrinous exudate.

NK Cell Neoplasms 223

thus the results may apply to nasal lymphomas as awhole and not just to NK/T-cell lymphomas.

All recent publications, with the exception of oneWestern study, reported no convincing clinical ben-efit of adding chemotherapy to radiation therapy. Thelatter series reported a 5-year freedom from progres-sion rate of 71% with chemotherapy compared to38% with radiotherapy alone.51 However, immuno-phenotype was not reported, and most of the patientsshowed involvement of the maxillary sinuses; thuslikely most of the lymphomas were of B-cell lineage.

Most series have reported poor results with initialapplication of conventional chemotherapy, while ra-diotherapy alone or radiotherapy combined with che-motherapy produced better or equivalent results.Kwong et al reported a complete remission rate of75% among 20 patients with localized nasal NK/T-cell lymphomas treated with anthracycline-contain-ing chemotherapy followed by radiation therapy, buttwo thirds of the patients relapsed, resulting in a poormedian overall survival of 12 months.43 Sakata et altreated 16 early-stage nasal “T-cell” lymphomas

mainly with chemotherapy and reported a poor5-year cause-specific survival of 22% and distant dis-semination rate of 75%.78 In the Shikama et al seriesof 20 nasal “T-cell” lymphomas treated primarilywith radiotherapy (with 64% patients receiving post-radiotherapy chemotherapy), the 5-year survival ratewas 72%.80 Li et al studied the outcome of 175 pa-tients according to stage and treatment. Stage I dis-ease was treated either with radiotherapy alone (moreso for limited stage I disease confined to the nose)with or without chemotherapy. There was no differ-ence in 5-year survival between those treated withradiotherapy alone and patients undergoing radio-therapy and chemotherapy, indicating lack of advan-tage in adding drugs to radiation.49 Kim et al reportedapparent lack of efficacy of chemotherapy in treat-ment of 17 nasal NK/T-cell lymphomas; frequentprimary resistance to chemotherapy was noted, andonly six of 15 patients given chemotherapy achievedcomplete remission. On the other hand, eight of 10patients who were alive and well had received pri-mary radiation therapy.41 Ribrag et al reported only

Figure 3. Nasal NK/T-cell lymphoma. (A) This example is predominated by small cells with irregular-shaped and angulated nuclei.Distinction from reactive lymphoid infiltrate can be difficult. (B) This example is predominated by medium-sized cells, and thus a diagnosisof lymphoma should be obvious.

224 Cheung, Chan, and Wong

three complete remissions among 12 patients treatedinitially with chemotherapy, while seven of eightpatients treated with initial radiation therapy or alter-nating chemotherapy-radiation therapy remainedfree of disease.75 Yamaguchi et al treated 12 nasalNK/T-cell lymphomas, with all five patients who re-ceived initial radiotherapy achieving complete remis-sion and four of them remaining alive and well, butonly one of seven patients treated with conventionalchemotherapy achieved complete remission.97 In ouranalysis of 79 early-stage nasal NK/T-cell lymphomasaccording to intent to treat, there was no statisticallysignificant survival difference between patientstreated with radiotherapy (5-year disease-free sur-vival, 31%) compared to chemotherapy (5-year dis-ease-free survival, 36%). In this series, the total resis-tance rate to conventional chemotherapy was 67%.14

In summary, anthracycline-containing combinationchemotherapy given before radiotherapy confers nosurvival benefit, at least for stage I disease.

The high rate of resistance to conventional chemo-therapy may be related to frequent expression ofP-glycoprotein, the product of the multidrug resis-

tance (MDR) 1 gene, and poor drug delivery owing tothe prominent tissue necrosis.14,23,96 Attempts toovercome this problem have used MDR-unrelatedanticancer agents or high-dose therapy with periph-eral stem cell rescue. Among three patients treatedwith combination dexamethasone, etoposide, ifosf-amide, and carboplatin (DeVIC) (two concurrentlywith radiotherapy and one after radiotherapy), allwere alive and well at 12 to 96 months.97 In casereports and small series, L-asparaginase as singleagent or combined with vincristine and dexametha-sone has been reported to be effective as salvagetreatment for refractory diseases including failureafter autologous marrow transplant, producing re-sponse rates up to 50%, including complete remis-sions.61,67,98,99 Among three relapsed patients treatedwith marrow transplant in one series, there were twosurvivors at 12 and 44 months.50 In a pilot study of 20consecutively treated early-stage nasal NK/T-celllymphomas, six consolidated with high-dose therapyand peripheral stem cell support, outcome comparedfavorably to a historical cohort of patients treated atthe same institute. The 2-year disease-free survival of

Figure 5. Nasal NK/T-cell lymphoma. Many apoptotic bodies arefound among the lymphoma cells.

Figure 4. Nasal NK/T-cell lymphoma. There is marked infiltrationof the intima and wall of this blood vessel by lymphoma cells.

NK Cell Neoplasms 225

this group was 63% and that of the historical cohort,38% ( P � .083).15

While systemic progression is the usual mode of

failure, local recurrence is also common despite ra-diotherapy. Studies on the relationship between localcontrol and dose suggest that a total radiation dosehigher than that generally recommended for non-Hodgkin’s lymphomas might be needed for diseasecontrol. One group reported complete remission inall five patients given time-dose-factor (TDF) valuesof 80 or over (equivalent to 50 Gy or more given in 2Gy/fraction, 5 fractions/wk), and only in two of 11patients given a lower total dose.78 Another groupreported a significantly higher 5-year local controlrate with total dose greater than 50 Gy compared witha lower dose (100% v 67%, P � .013).80 In our study,the in-field failure rate (12%) in patients given a totaldose greater than 50 Gy and/or concomitant chemo-irradiation with cisplatinum and having imagingscans done to assist treatment was lower comparedwith those who were not (28%).14 Thus meticulousCT conformal planning with the aid of magneticresonance imaging is recommended to delineate theextent of disease and hence the field of radiationtherapy.

Figure 6. Nasal NK/T-cell lymphoma. This biopsy shows onlynecrotic material, and is thus not diagnostic. The diagnosis isconfirmed on a repeat biopsy.

Table 1. Immunobiologic Profile of Extranodal NK/T-CellLymphoma

● CD2: positive● Surface CD3/Leu4: negative● Cytoplasmic CD3 (polyclonal antibody to CD3� or PS1):

positive● CD56: positive (most cases)● T-cell receptor: negative● Other T-cell markers (CD4, CD5, CD7, CD8): usually negative● CD43 and CD45RO: usually positive● Other natural killer cell markers (CD16, CD57): usually

negative● Cytotoxic molecules (perforin, granzyme B, TIA-1): positive● Some cases may express CD25 or CD30● NK cell receptors: variable expression● Fas and Fas ligand: commonly positive● Epstein-Barr virus: positive

Figure 7. Nasal NK/T-cell lymphoma. The lymphoma cells showstrong cell membrane staining for the NK-associated markerCD56.

226 Cheung, Chan, and Wong

Prognostic FactorsThe clinical factors reported to have prognostic sig-nificance in nasal NK/T-cell lymphomas includestage, performance status, B symptoms, age, andbulk.13,14,40,41,43,47,49 Advanced-stage disease (III andIV) was associated with a very poor outcome (2-yearoverall survival, 13%).13 In our study, the 5-yearoverall survival rates for stage I and II disease were42% and 19%, respectively.14 In a series of 92 patientstreated by radiotherapy alone, the 5-year overall sur-vival rates were 64% for stage I disease and 25% for

stage II disease.40 A study of 175 patients with pre-dominantly early-stage disease reported significantlydifferent 5-year overall survival rates for stage I (75%)and stage II (35%) disease.49

The International Prognostic Index (IPI) has notbeen adequately examined for predictive value in thisdisease. We did not find it to be of independentsignificance.14 However, since more than 80% of thepatients have low-risk IPI (0 or 1), there may not havebeen a wide enough spectrum of disease to detect adifference.

Table 2. Major Recent Series of Primary Nasal Lymphomas Reported in Recent 5 Years

AuthorsNo. of Patients

(stage) Subtypes, No. Treatment Outcome Findings

Kwong et al43 20 (I) NK/T All CMT High relapse (67%) withCT, median OS 12 mo

Poor outcome

Shikama et al80 25 (all I) 20 T*, 5 B 9 RT, 16 RT �CT 5-yr-DFS 72% (T* only) RT dose important; CTadds nothing

Sakata et al78 16 (I–10, II–6) All T* 2 RT, 14 CMT 7/16 CR, 12/16systemic progression,5-yr cause-specificsurvival 22%

RT dose important; veryheterogeneouschemotherapy

Li et al49 175 (I–133 II–28III, IV–14)

46 T* 58 RT, 117 CMTor CT

Stage I 5-yr DFS 68%,5-year OS 75%

Stage and extent instage I prognosticallyimportant;chemotherapy addsno benefit in stage I

Lei et al47 44 (I, II–40, III,IV–4)

25 nasal (75% T*)19 NP (33% T*)(6 NK/T)

CMT 64%, RT 14%,CT 23%

5-yr OS 54%:Nasal 33%,NP 82%

Site, age, bulk areprognostic factors;tumors of nasal andNP are biologicallydifferent

Cuadra-Garciaet al18

17 5 NK/T Unknown 73% NED (“NK/T-celltype”)

Outcome of NK/T-celltype similar to diffuselarge B cell

Kim et al40 92 (I–62, II–30) 80% T* RT alone 66.3% CR, 5-yr OS40.1%

Common early localfailure; stageprognosticallyimportant

Ribrag et al75 20 7 NK/T CMT (8 initial RT,12 CT)

10 alive with no relapse Initial RT better; high CTfailure rate

Yamaguchiet al97

12 (I–9, II–3) All NK/T 7 conventional CMT,5 RT (3 �DeVIC,2 concurrent DeVIC)

CMT: 1/7 CR, 5-yr OS13%,

RT: 5/5 CR, 5-yr OS73%

Resistant toconventional CT;DeVIC, RT good

Kim et al41 17 All NK/T CMT (15 initial CT) 3-yr OS 59%, High CT-resistance;B-symptom, stageprognosticallyimportant

Cheung et al14 79 (I–63, II–16) All NK/T 18 RT, 61 CMT (intendto treat)

5-yr DFS 35.5%,5-yr OS 37.9%

CMT no better than RT;stage andperformance statusprognosticallyimportant

Abbreviations: Subtypes, immunophenotypes; NK/T, CD56� tumors; T* � T or NK/T, but CD56 status unknown; CMT, combined modalitytreatment; RT, radiation therapy; CT, chemotherapy; NED, no evidence of disease; OS, overall survival; DFS, disease-free survival; IPI,International Prognostic Index; NP, nasopharynx; CR, complete remission; DeVIC, dexamethasone, etoposide, ifosfamide, carboplatin.

NK Cell Neoplasms 227

Recommended TreatmentFor stage I disease, radiation therapy is the mainstay;the total dose should be over 50 Gy.78,80 Prophylacticradiation of the central nervous system is not neces-sary because relapse in this site is rare. If chemother-apy is to be added in view of the high rate of systemicprogression or to spare essential organs from radia-tion (such as the eyes) by debulking an originallyvoluminous tumor, radiation should be applied early.

Theoretically, chemotherapy should play an im-portant role in the treatment of higher stage disease.Regimens employing MDR-unrelated drugs that havebeen shown to be effective in refractory diseaseshould be tested in prospective protocols. In fact,since outcome has been poor even in stage I diseasetreated with radiotherapy alone, all medically fit pa-tients should be offered the choice of these latterregimens or intensified therapy with concurrent che-moirradiation and high-dose therapy.

To monitor disease activity, the value of plasma orserum circulating EBV DNA as a surrogate marker forminimal residual disease or early relapse requireslarge-scale and prospective evaluation.45,46 P73 genehypermethylation assay in biopsies also appears to bea promising technique for determining the presenceor absence of minimal disease, since this is a commonfeature in NK/T-cell lymphomas (94%).83

Aggressive NK Cell Leukemia

Clinical Features and BehaviorAggressive NK cell leukemia represents the leukemicend in the range of NK cell neoplasms, analogous tothe acute lymphoblastic leukemia–lymphoblasticlymphoma spectrum and chronic lymphocytic leuke-mia–small lymphocytic lymphoma spectrum.12,33

While there are usually easily identified neoplasticcells in the peripheral blood and bone marrow, in-volvement can be subtle, rendering a distinction fromdisseminated extranodal NK/T-cell lymphoma diffi-cult. Hypersensitivity to mosquito bites may repre-sent a precursor lesion of aggressive NK cell leuke-mia.34,35

Similar to extranodal NK/T-cell lymphoma, aggres-sive NK cell leukemia is much more common in Asiansthan in Caucasians44 and there is a strong associationwith EBV.5,25,30,38,42,71,81 In contrast to extranodal NK/T-cell lymphoma, patients are often younger (mean age,36 years).5,17,22,30,33,42-44,57,58,72,81 They are typicallyvery ill, and present with fever, hepatosplenomegaly,lymphadenopathy, and a leukemic blood picture,sometimes accompanied by hemophagocytic syn-drome.5,17,22,30,33,42-44,57,58,69,72,81 Coagulopathy, bleed-ing tendency and multi-organ failure are common. Se-rum levels of lactate dehydrogenase and Fas ligand areoften markedly elevated.79,88 A highly fulminant course

is usual, with patients succumbing in days to weeks.Response to chemotherapy is poor, and the curability ofthe disease with bone marrow transplantation remainsunproven.5,33,43,44,58

PathologyIn the peripheral blood, there are few to numerouscirculating large granular lymphocytes, which rangein appearance from normal-looking to immature oratypical (Fig 8). The nuclei are often round, andshow condensed or loose chromatin, sometimes withdiscernible nucleoli. The lightly basophilic cyto-plasm contains fine or coarse azurophilic granules. Inthe bone marrow aspirate smears, loose aggregates orsheets of similar cells are present (Fig 9).

Immunobiologic FeaturesThe immunophenotype is indistinguishable from ex-tranodal NK/T-cell lymphoma—CD2�, surface CD3/Leu4�, cytoplasm CD3��, CD56�, and cytotoxic

Figure 8. Aggressive NK cell leukemia, buffy coat smear ofperipheral blood. Many immature cells with prominent nucleoli arepresent. Fine azurophilic granules are seen in the cytoplasm. Thereare admixed neutrophils, band forms, and metamyelocytes.

228 Cheung, Chan, and Wong

markers, except that CD16 is expressed in approxi-mately half of the cases.5,17,22,30,33,43,44,57,58,72,81 EBVis almost always demonstrable in the neoplasticcells.5,25,30,38,42,71,81 The T-cell receptor genes are notrearranged. Similar to extranodal NK/T cell lym-phoma, deletion of 6q21-q25 is common, but in ad-dition there is commonly loss of 17p13.82,95

Differential DiagnosisIt is most important not to confuse aggressive NK cellleukemia with T-cell large granular lymphocyte leu-kemia, which is an indolent condition.52,72 The latteroccurs in older patients, who are asymptomatic orpresent with infection, hepatosplenomegaly, purered cell aplasia, or neutropenia.

Not all NK cell lymphoproliferative disorders areaggressive. The indolent form of NK lymphoprolifera-tive disorder is clinically and morphologically similar toT-cell large granular lymphocyte leukemia, but showsan NK phenotype (surface CD3�, CD56�/�, and germ-line T-cell receptor genes).52,71,74,89,90 The patients are

often asymptomatic, there is no hepatosplenomegaly,CD16 and CD57 are commonly expressed, and there isno association with EBV.71,72

References1. Abe Y, Muta K, Ohshima K, et al: Subcutaneous panniculitis

by Epstein-Barr virus-infected natural killer (NK) cell prolif-eration terminating in aggressive subcutaneous NK cell lym-phoma. Am J Hematol 64:221-225, 2000

2. Arber DA, Weiss LM, Albujar PF, et al: Nasal lymphomas inPeru. High incidence of T-cell immunophenotype and Ep-stein-Barr virus infection. Am J Surg Pathol 17:392-399, 1993

3. Chan JK: Natural killer cell neoplasms. Anat Pathol 3:77-145,1998

4. Chan JK, Ng CS, Lau WH, et al: Most nasal/nasopharyngeallymphomas are peripheral T-cell neoplasms. Am J SurgPathol 11:418-429, 1987

5. Chan JK, Sin VC, Wong KF, et al: Nonnasal lymphomaexpressing the natural killer cell marker CD56: A clinicopath-ologic study of 49 cases of an uncommon aggressive neo-plasm. Blood 89:4501-4513, 1997

6. Chan JK, Tsang WY, Hui PK, et al: T- and T/natural killer-celllymphomas of the salivary gland: A clinicopathologic, immu-

Figure 9. Aggressive NK cell leukemia, bone marrow aspirate smear. (A) The marrow is normocellular, and the neoplastic infiltrate is stilladmixed with many normal hematopoietic cells. (B) Mature-looking large granular lymphocytes (leukemic cells) are seen among thehematopoietic cells.

NK Cell Neoplasms 229

nohistochemical and molecular study of six cases. HumPathol 28:238-245, 1997

7. Chan JK, Tsang WY, Lau WH, et al: Aggressive T/naturalkiller cell lymphoma presenting as testicular tumor. Cancer77:1198-1205, 1996

8. Chan JK, Tsang WY, Pau MY: Discordant CD3 expression inlymphomas when studied on frozen and paraffin sections.Hum Pathol 26:1139-1143, 1995

9. Chan JKC, Jaffe ES, Ralfkiaer E: Blastic NK-cell lymphoma, inJaffe ES, Harris NL, Stein H (eds): Pathology and Genetics,Tumours of Haematopoietic and Lymphoid Tissues. WorldHealth Organization Classification of Tumours. Lyon,France, IARC Press, 2001, pp 214-215

10. Chan JKC, Jaffe ES, Ralfkiaer E: Extranodal NK/T-cell lym-phoma, nasal type, in Jaffe ES, Harris NL, Stein H, et al (eds):Pathology and Genetics, Tumours of Haematopoietic andLymphoid Tissues. World Health Organization Classificationof Tumours. Lyon, France, IARC Press, 2001, pp 204-207

11. Chan JKC, Ng CS, Tsang WY: Nasal/nasopharyngeal lympho-mas: An immunohistochemical analysis of 57 cases on frozentissues. Mod Pathol 6:87A, 1993 (abstr)

12. Chan JKC, Wong KF, Jaffe ES, et al: Aggressive NK-cellleukemia, in Jaffe ES, Harris NL, Stein H, et al (eds): Pathol-ogy and Genetics, Tumours of Haematopoietic and LymphoidTissues. World Health Organization Classification of Tu-mours, Lyon, France, IARC Press, 2001, 198-200

13. Cheung MM, Chan JK, Lau WH, et al: Primary non-Hodgkin’slymphoma of the nose and nasopharynx: clinical features,tumor immunophenotype, and treatment outcome in 113patients. J Clin Oncol 16:70-77, 1998

14. Cheung MM, Chan JK, Lau WH, et al: Early stage nasalNK/T-cell lymphoma: clinical outcome, prognostic factors,and the effect of treatment modality. Int J Radiat Oncol BiolPhys 54:182-190, 2002

15. Cheung MM, Chan JK, Wong KH: Early stage nasal NK/T celllymphoma: Preliminary result of intensifying treatment withconcurrent chemo-radiation and high dose chemotherapy.Ann Oncol 12:82, 2002 (suppl 2, abstr)

16. Chinen K, Kaneko Y, Izumo T, et al: Nasal natural killercell/t-cell lymphoma showing cellular morphology mimick-ing normal lymphocytes. Arch Pathol Lab Med 126:602-605,2002

17. Chou WC, Chiang IP, Tang JL, et al: Clonal disease of naturalkiller large granular lymphocytes in Taiwan. Br J Haematol103:1124-1128, 1998

18. Cuadra-Garcia I, Proulx GM, Wu CL, et al: Sinonasal lym-phoma: A clinicopathologic analysis of 58 cases from theMassachusetts General Hospital. Am J Surg Pathol 23:1356-1369, 1999

19. Dukers DF, Vermeer MH, Jaspars LH, et al: Expression ofkiller cell inhibitory receptors is restricted to true NK celllymphomas and a subset of intestinal enteropathy-type T celllymphomas with a cytotoxic phenotype. J Clin Pathol 54:224-228, 2001

20. Elenitoba-Johnson KS, Zarate-Osorno A, Meneses A, et al:Cytotoxic granular protein expression, Epstein-Barr virusstrain type, and latent membrane protein-1 oncogene dele-tions in nasal T- lymphocyte/natural killer cell lymphomasfrom Mexico. Mod Pathol 11:754-761, 1998

21. Emile JF, Boulland ML, Haioun C, et al: CD5-CD56� T-cellreceptor silent peripheral T-cell lymphomas are natural killercell lymphomas. Blood 87:1466-1473, 1996

22. Engellenner W, Golightly M: Large granular lymphocyte leu-kemia. Lab Med 22:454-456, 1991

23. Fukuda J, Yoshihara T, Arai Y, et al: Nasal NK-cell lymphoma.Rinsho Ketsueki 35:588-592, 1994

24. Gaal K, Sun NC, Hernandez AM, et al: Sinonasal NK/T-celllymphomas in the United States. Am J Surg Pathol 24:1511-1517, 2000

25. Gelb AB, van de Rijn M, Regula DP Jr, et al: Epstein-Barrvirus-associated natural killer-large granular lymphocyte leu-kemia. Hum Pathol 25:953-960, 1994

26. Goodlad JR, Fletcher CDM, Chan JKC, et al: Primary softtissue lymphoma: An analysis of 37 cases. J Pathol 179:42A,1996 (suppl, abstr)

27. Haedicke W, Ho FC, Chott A, et al: Expression of CD94/NKG2A and killer immunoglobulin-like receptors in NK cellsand a subset of extranodal cytotoxic T-cell lymphomas. Blood95:3628-3630, 2000

28. Harris NL, Jaffe ES, Diebold J, et al: The World HealthOrganization classification of neoplastic diseases of the hema-topoietic and lymphoid tissues. Report of the Clinical Advi-sory Committee meeting, Airlie House, Virginia, November,1997. Ann Oncol 10:1419-1432, 1999

29. Harris NL, Jaffe ES, Stein H, et al: A revised European-American classification of lymphoid neoplasms: A proposalfrom the International Lymphoma Study Group. Blood 84:1361-1392, 1994

30. Hart DN, Baker BW, Inglis MJ, et al: Epstein-Barr viral DNA inacute large granular lymphocyte (natural killer) leukemiccells. Blood 79:2116-2123, 1992

31. Herling M, Teitell MA, Shen RR, et al: TCL1 Expression inplasmacytoid dendritic cells (DC2) and the related CD4�

CD56� blastic tumors of skin. Blood (in press)32. Ho FC, Srivastava G, Loke SL, et al: Presence of Epstein-Barr

virus DNA in nasal lymphomas of B and ‘T’ cell type. HematolOncol 8:271-281, 1990

33. Imamura N, Kusunoki Y, Kawa-Ha K, et al: Aggressive naturalkiller cell leukaemia/lymphoma: Report of four cases andreview of the literature. Possible existence of a new clinicalentity originating from the third lineage of lymphoid cells.Br J Haematol 75:49-59, 1990

34. Ishihara S, Ohshima K, Tokura Y, et al: Hypersensitivity tomosquito bites conceals clonal lymphoproliferation of Ep-stein-Barr viral DNA-positive natural killer cells. Jpn J CancerRes 88:82-87, 1997

35. Ishihara S, Yabuta R, Tokura Y, et al: Hypersensitivity tomosquito bites is not an allergic disease, but an Epstein-Barrvirus-associated lymphoproliferative disease. Int J Hematol72:223-228, 2000

36. Jaffe ES, Chan JK, Su IJ, et al: Report of the Workshop onnasal and related extranodal angiocentric T/natural killer celllymphomas: Definitions, differential diagnosis, and epidemi-ology. Am J Surg Pathol 20:103-111, 1996

37. Katoh A, Ohshima K, Kanda M, et al: Gastrointestinal T celllymphoma: Predominant cytotoxic phenotypes, including al-pha/beta, gamma/delta T cell and natural killer cells. LeukLymphoma 39:97-111, 2000

38. Kawa-Ha K, Ishihara S, Ninomiya T, et al: CD3-negativelymphoproliferative disease of granular lymphocytes con-taining Epstein-Barr viral DNA. J Clin Invest 84:51-55, 1989

39. Kern WF, Spier CM, Hanneman EH, et al: Neural cell adhe-sion molecule-positive peripheral T-cell lymphoma: A rarevariant with a propensity for unusual sites of involvement.Blood 79:2432-2437, 1992

40. Kim GE, Cho JH, Yang WI, et al: Angiocentric lymphoma ofthe head and neck: Patterns of systemic failure after radiationtreatment. J Clin Oncol 18:54-63, 2000

41. Kim WS, Song SY, Ahn YC, et al: CHOP followed by involvedfield radiation: Is it optimal for localized nasal natural killer/T-cell lymphoma? Ann Oncol 12:349-352, 2001

42. Kim YJ, Park QE, Kim SH, et al: Clinicopathologic findings of

230 Cheung, Chan, and Wong

aggressive NK cell leukemia. Korean J Pathol 36:95, 2002(suppl, abstr)

43. Kwong YL, Chan AC, Liang R, et al: CD56� NK lymphomas:Clinicopathological features and prognosis. Br J Haematol97:821-829, 1997

44. Kwong YL, Wong KF, Chan LC, et al: Large granular lympho-cyte leukemia. A study of nine cases in a Chinese population.Am J Clin Pathol 103:76-81, 1995

45. Lei KI, Chan LY, Chan WY, et al: Quantitative analysis ofcirculating cell-free Epstein-Barr virus (EBV) DNA levels inpatients with EBV-associated lymphoid malignancies. Br JHaematol 111:239-246, 2000

46. Lei KI, Chan LY, Chan WY, et al: Diagnostic and prognosticimplications of circulating cell-free Epstein-Barr virus DNAin natural killer/T-cell lymphoma. Clin Cancer Res 8:29-34,2002

47. Lei KI, Suen JJ, Hui P, et al: Primary nasal and nasopharyngeallymphomas: A comparative study of clinical presentation andtreatment outcome. Clin Oncol (R Coll Radiol) 11:379-387,1999

48. Leroux D, Mugneret F, Callanan M, et al: CD4(�), CD56(�)DC2 acute leukemia is characterized by recurrent clonalchromosomal changes affecting 6 major targets: A study of 21cases by the Groupe Francais de Cytogenetique Hema-tologique. Blood 99:4154-4159, 2002

49. Li YX, Coucke PA, Li JY, et al: Primary non-Hodgkin’s lym-phoma of the nasal cavity: prognostic significance of parana-sal extension and the role of radiotherapy and chemotherapy.Cancer 83:449-456, 1998

50. Liang R, Chen F, Lee CK, et al: Autologous bone marrowtransplantation for primary nasal T/NK cell lymphoma. BoneMarrow Transplant 19:91-93, 1997

51. Logsdon MD, Ha CS, Kavadi VS, et al: Lymphoma of the nasalcavity and paranasal sinuses: Improved outcome and alteredprognostic factors with combined modality therapy. Cancer80:477-488, 1997

52. Loughran TP: Clonal diseases of large granular lymphocytes.Blood 82:1-14, 1993

53. Lymphoma Study Group of Japanese Pathologists: The WorldHealth Organization Classification of malignant lymphomasin Japan: Incidence of recently recognized entities. Pathol Int50:696-702, 2000

54. Martin AR, Chan WC, Perry DA, et al: Aggressive naturalkiller cell lymphoma of the small intestine. Mod Pathol 8:467-472, 1995

55. Misago N, Ohshima K, Aiura S, et al: Primary cutaneousT-cell lymphoma with an angiocentric growth pattern: Asso-ciation with Epstein-Barr virus. Br J Dermatol 135:638-643,1996

56. Miyamoto T, Yoshino T, Takehisa T, et al: Cutaneous presen-tation of nasal/nasal type T/NK cell lymphoma: Clinicopath-ological findings of four cases. Br J Dermatol 139:481-487,1998

57. Mori KL, Egashira M, Oshimi K: Differentiation stage ofnatural killer cell-lineage lymphoproliferative disordersbased on phenotypic analysis. Br J Haematol 115:225-228,2001

58. Mori N, Yamashita Y, Tsuzuki T, et al: Lymphomatous fea-tures of aggressive NK cell leukaemia/lymphoma with mas-sive necrosis, haemophagocytosis and EB virus infection.Histopathology 37:363-371, 2000

59. Mori N, Yatabe Y, Oka K, et al: Expression of perforin in nasallymphoma. Additional evidence of its natural killer cell deri-vation. Am J Pathol 149:699-705, 1996

60. Mraz-Gernhard S, Natkunam Y, Hoppe RT, et al: Naturalkiller/natural killer-like T-cell lymphoma, CD56�, present-

ing in the skin: an increasingly recognized entity with anaggressive course. J Clin Oncol 19:2179-2188, 2001

61. Nagafuji K, Fujisaki T, Arima F, et al: L-asparaginase induceddurable remission of relapsed nasal NK/T-cell lymphomaafter autologous peripheral blood stem cell transplantation.Int J Hematol 74:447-450, 2001

62. Nakamura S, Katoh E, Koshikawa T, et al: Clinicopathologicstudy of nasal T/NK-cell lymphoma among the Japanese.Pathol Int 47:38-53, 1997

63. Nakamura S, Suchi T, Koshikawa T, et al: Clinicopathologicstudy of CD56 (NCAM)-positive angiocentric lymphoma oc-curring in sites other than the upper and lower respiratorytract. Am J Surg Pathol 19:284-296, 1995

64. Ng CS, Chan JK, Cheng PN, et al: Nasal T-cell lymphomaassociated with hemophagocytic syndrome. Cancer 58:67-71, 1986

65. Ng CS, Chan JK, Lo ST: Expression of natural killer cellmarkers in non-Hodgkin’s lymphomas. Hum Pathol 18:1257-1262, 1987

66. Ng CS, Lo ST, Chan JK, et al: CD56� putative natural killercell lymphomas: Production of cytolytic effectors and relatedproteins mediating tumor cell apoptosis? Hum Pathol 28:1276-1282, 1997

67. Obama K, Tara M, Niina K: L-asparaginase induced completeremission in Epstein-Barr virus positive, multidrug resistant,cutaneous T-cell lymphoma. Int J Hematol 69:260-262, 1999

68. Ohno T, Yamaguchi M, Oka K, et al: Frequent expression ofCD3 epsilon in CD3 (Leu 4)-negative nasal T-cell lympho-mas. Leukemia 9:44-52, 1995

69. Okuda T, Sakamoto S, Deguchi T, et al: Hemophagocyticsyndrome associated with aggressive natural killer cell leuke-mia. Am J Hematol 38:321-323, 1991

70. Ooi GC, Chim CS, Liang R, et al: Nasal T-cell/natural killer celllymphoma: CT and MR imaging features of a new clinicopatho-logic entity. AJR Am J Roentgenol 174:1141-1145, 2000

71. Oshimi K: Lymphoproliferative disorders of natural killercells. Int J Hematol 63:279-290, 1996

72. Oshimi K, Yamada O, Kaneko T, et al: Laboratory findingsand clinical courses of 33 patients with granular lymphocyte-proliferative disorders. Leukemia 7:782-788, 1993

73. Petrella T, Delfau-Larue MH, Caillot D, et al: Nasopharyngeallymphomas: further evidence for a natural killer cell origin.Hum Pathol 27:827-833, 1996

74. Rabbani GR, Phyliky RL, Tefferi A: A long-term study ofpatients with chronic natural killer cell lymphocytosis. Br JHaematol 106:960-966, 1999

75. Ribrag V, Ell Hajj M, Janot F, et al: Early locoregional high-dose radiotherapy is associated with long-term disease con-trol in localized primary angiocentric lymphoma of the noseand nasopharynx. Leukemia 15:1123-1126, 2001

76. Rodriguez J, Romaguera JE, Manning J, et al: Nasal-typeT/NK lymphomas: A clinicopathologic study of 13 cases.Leuk Lymphoma 39:139-144, 2000

77. Rudiger T, Weisenburger DD, Anderson JR, et al: PeripheralT-cell lymphoma (excluding anaplastic large-cell lympho-ma): Results from the Non-Hodgkin’s Lymphoma Classifica-tion Project. Ann Oncol 13:140-149, 2002

78. Sakata K, Hareyama M, Ohuchi A, et al: Treatment of lethalmidline granuloma type nasal T-cell lymphoma. Acta Oncol36:307-311, 1997

79. Sato K, Kimura F, Nakamura Y, et al: An aggressive nasallymphoma accompanied by high levels of soluble Fas ligand.Br J Haematol 94:379-382, 1996

80. Shikama N, Izuno I, Oguchi M, et al: Clinical stage IE primarylymphoma of the nasal cavity: Radiation therapy and chemo-therapy. Radiology 204:467-470, 1997

NK Cell Neoplasms 231

81. Shimodaira S, Ishida F, Kobayashi H, et al: The detection ofclonal proliferation in granular lymphocyte-proliferative dis-orders of natural killer cell lineage. Br J Haematol 90:578-584,1995

82. Siu LL, Chan JK, Kwong YL: Natural killer cell malignancies:Clinicopathologic and molecular features. Histol Histopathol17:539-554, 2002

83. Siu LL, Chan JKC, Wong KF, et al: Specific patterns of genemethylation in natural killer cell lymphoma: p73 is consis-tently involved. Am J Pathol 160:59-66, 2002

84. Suzumiya J, Takeshita M, Kimura N, et al: Expression of adultand fetal natural killer cell markers in sinonasal lymphomas.Blood 83:2255-2260, 1994

85. Takahashi N, Miura I, Chubachi A, et al: A clinicopathologi-cal study of 20 patients with T/natural killer (NK)-cell lym-phoma-associated hemophagocytic syndrome with specialreference to nasal and nasal-type NK/T-cell lymphoma. IntJ Hematol 74:303-308, 2001

86. Takeshita M, Kimura N, Suzumiya J, et al: Angiocentriclymphoma with granulomatous panniculitis in the skin ex-pressing natural killer cell and large granular T-cell pheno-types. Virchows Arch 425:499-504, 1994

87. Takeshita M, Yamamoto M, Kikuchi M, et al: Angiodestruc-tion and tissue necrosis of skin-involving CD56� NK/T-celllymphoma are influenced by expression of cell adhesionmolecules and cytotoxic granule and apoptosis-related pro-teins. Am J Clin Pathol 113:201-211, 2000

88. Tanaka M, Suda T, Haze K, et al: Fas ligand in human serum.Nat Med 2:317-322, 1996

89. Tefferi A: Chronic natural killer cell lymphocytosis. LeukLymphoma 20:245-248, 1996

90. Tefferi A, Li CY, Witzig TE, et al: Chronic natural killer celllymphocytosis: A descriptive clinical study. Blood 84:2721-2725, 1994

91. Tien HF, Su IJ, Tang JL, et al: Clonal chromosomal abnormal-ities as direct evidence for clonality in nasal T/natural killercell lymphomas. Br J Haematol 97:621-625, 1997

92. Wong KF: Genetic changes in natural killer cell neoplasms.Leuk Res 26:977-978, 2002

93. Wong KF, Chan JK, Ng CS: CD56 (NCAM)-positive malig-nant lymphoma. Leuk Lymphoma 14:29-36, 1994

94. Wong KF, Chan JK, Ng CS, et al: CD56 (NKH1)-positivehematolymphoid malignancies: An aggressive neoplasm fea-turing frequent cutaneous/mucosal involvement, cytoplas-mic azurophilic granules, and angiocentricity. Hum Pathol23:798-804, 1992

95. Wong KF, Zhang YM, Chan JK: Cytogenetic abnormalities innatural killer cell lymphoma/leukaemia—Is there a consis-tent pattern? Leuk Lymphoma 34:241-250, 1999

96. Yamaguchi M, Kita K, Miwa H, et al: Frequent expression ofP-glycoprotein/MDR1 by nasal T-cell lymphoma cells. Can-cer 76:2351-2356, 1995

97. Yamaguchi M, Shoko O, Yoshihito N: Treatment outcome ofnasal NK-cell lymphoma: A report of 12 consecutively diag-nosed cases and a review of the literature. J Clin Exp Hema-topathol 41:93-99, 2001

98. Yong W, Zhang Y, Zheng W: The efficacy of L-asparaginase inthe treatment of refractory midline peripheral T-cell lym-phoma. Zhonghua Xue Ye Xue Za Zhi 21:577-579, 2000

99. Yong W, Zheng W, Zhang Y, et al: Clinical characteristics andtreatment of midline nasal and nasal type NK/T cell lym-phoma. Zhonghua Yi Xue Za Zhi 81:773-775, 2001

232 Cheung, Chan, and Wong