Macroglobulinemia of Waldenstrom Diagnosis and Management

Transcription

1 ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 8, No. 4 Copyright 1978, Institute for Clinical Science Macroglobulinemia of Waldenstrom Diagnosis and Management HARRY L. MESSMORE, M.D., JAWED FAREED, Ph.D., SIMONE SILBERMAN, M.D., GERALD M. GAWLIK, M.D., and EDWARD W. BERMES, JR., Ph.D. Departments o f Pathology, Medicine, Biochemistry and Pharmacology, Loyola University Stritch School o f Medicine, Maywood, IL ABSTRACT W aldenstrom s macroglobulinemia is a syndrome characterized by an IgM globulin in the blood, an infiltration of plasmacytoid lymphocytes in the bone marrow, lymph nodes and other tissues and variable symptoms related to serum hyperviscosity, hemostatic defects and tissue replacem ent by tumor cells. It is a disease of the older patient and commonly has a prolonged course. The median age at onset is 64 years and the median survival four years. The circulating IgM paraprotein is usually in excess of 1,000 mg per dl. Increased plasma volume, increased blood viscosity and impairment of the hem ostatic m echanism are all attributed to the abnorm al protein. The hyperviscosity of the blood results in neurologic deficits consisting of impairment of vision and hearing, cerebral dysfunction and peripheral neuropathy. Hemostatic defect resulting in mucous membrane bleeding is also associated w ith the hyperviscosity syndrome. Treatm ent is not necessary in the asymptomatic patient. Plasmapheresis is very useful as initial therapy in severely symptomatic patients, and can be used as long term treatm ent if necessary in cases of failure of chemotherapy with chlorambucil. Potent combination chem otherapeutic regimens are being evaluated for treatm ent of refractory advanced disease. This disorder, along with others having the same histopathology, might be classified as a sub-category of plasmacytic lymphoma, as suggested by Lukes and Collins.12 W aldenstrom s macroglobulinemia is a m alignant lym phom a of B cell origin characterized by a combination of clinical, im m unologic and pathologic features which distinguish it from IgM myeloma and other diseases associated with monoclonal IgM globulin in the blood. Clinical Features The clinical features commonly found are related to infiltration of the bone marrow, lymph nodes, spleen, liver and occasionally other organs (lung, kidney) or more prom inently, to the presence of the /78/ $01.20 Institute for Clinical Science, Inc.

2 MACROGLOBULINEM IA O F W ALDENSTROM hyperviscosity syndrome or a bleeding disorder secondary to the circulating IgM protein. The hyperviscosity syndrom e includes neurologic symptoms and signs such as dizziness, vertigo, headache, im paired hearing, visual defect, ataxia, nystagm us, pyram idal tra c t signs and peripheral neuropathy.6, 19,20,27 The clinical features of a series of 40 patients reported by MacKenzie and F udenberg13 are shown in table I. A study of 45 patients by Krajny and Pruzanski9 was sim ilar, b ut b leed in g manifestations w ere p resent in 44 percent, predominantly mucous membrane bleeding. The average age at onset of symptoms was 64 years (range 29 to 89 years). No sex predom inance was found. The mean survival time was 49.5 months for those who expired and 43 months for those still living.9 Laboratory Features Selected laboratory abnorm alities9 included anemia (Hgb 8.5 to 1.5 g per dl) in 42 percent, thrombocytopenia in 34 percent, elevated sedimentation rate in most (over 80 mm per h in 56 percent), elev ated BUN in 18 p ercent, serum pro tein > 8 g p er dl in 42 p ercen t cryoglobulinem ia in 16 percent and cold agglutinins in 18 percent. The IgM concentration in th e serum was less than 1,000 mg per dl in six of the 45 patients, and over 5,000 per dl in 14 reported. The light chain type was most commonly of the kappa type. IgG levels w ere subnormal in 40 percent; IgA was low in 60 percent. Bence Jones proteins were present in small amounts in 71 percent of their cases; nine patients had more than 200 mg per 24 h. Similar laboratory findings w ere re p o rte d in o th er p u b lish e d series.13,19,26 T he Coom bs test was positive in only two cases. One patient had a high titer of cold agglutinin. The coagulation defect appears to be in the prim ary hem ostatic m echanism involving platelets. The bleeding tim e is T A B L E I Physical Findings in 40 Patients with Macroglobulinemia D i s o r d e r P e r c e n t Hepatomegaly 55 Splenomegaly 35 Adenopathy 45 Retinopathy 35 Neurologic defect 20 Purpura 10 commonly prolonged, and p latelet function defects have been demonstrated by platelet factor 3 assays, prothrom bin consumption and thromboplastin generation tests designed to show a p latelet d e fect.8,10,16,17 Such a defect was not dem onstrated by P erkins,18 b ut abnorm al bleeding tim e, defective platelet adhesion and modest decrease in factors V and V III w ere show n. In te rfe ren c e w ith polym erization of fibrin m onom ers re sulting in prolonged prothrombin time, partial throm boplastin time or throm bin time is uncommonly seen.10,18 Inhibition of a specific clotting factor (factor VIII) has been reported by Castaldi and Perry.7 Increased plasma volume related to the IgM protein is very common, causes an apparent anem ia by dilutional effect and may make transfusion therapy very hazardous.1 The viscosity of the serum, plasma and whole blood are, as a rule, increased. There is a rough correlation betw een the serum viscosity and the level of IgM, and b e tw e en e ith e r of th ese and the symptomatic hyperviscosity syndrome, but many exceptions to this rule have been known to occur.3,4,14,15 Interaction of the IgM molecule with the red blood cell, and the hem atocrit itse lf are variables affecting the in vivo viscosity. In some instances, tem perature has a striking effect.4,6 Q uantitation of the IgM level may be complicated when viscosity is very high, or w hen some 7s IgM is present,9 m aking it difficult to predict

3 3 1 2 MESSMORE, FA REED, SILBERM AN, GAW LIK AND BERMES viscosity from th e IgM lev el an d hematocrit determination. Each patient has his own symptomatic threshold, and will tend to be symptomatic when that level of IgM is reached, or when that particular viscosity occurs. Interaction of fibrinogen w ith the abnormal protein also effects whole blood viscosity and may also be a factor in some cases. This points out the importance of m easuring whole blood viscosity, or calculating it by the m ethod of Mannik.14,15 Although there is correlation betw een serum viscosity, as m easured in the O stw ald viscom eter or red blood cell pipet, with clinical symptoms in most in stances,31 m easurem ent of whole blood viscosity at a low shear rate in a coneplate viscom eter21,30 shows a better correlation with in vivo events. At low shear rates, the effect of hematocrit is im portant because the aggregation of RBC begins to occur at shear rates of 50 sec-1 or less (figure 1). A recent study using Mannik s formula: log t) (viscosity in centipoises) = m h, w here m = macroglobulin concentration in g per dl and h is the h em a to crit,14 show ed th at symptomatic patients usually had whole blood viscosity values greater than eight no 105 Control Whole Blood Control Serum SHEAR RATE (Sec"') F i g u r e 1. Viscometric studies on the whole blood and serum of H.B. (Case #2) at various shear rates. centipoises w hen calculated from the relative serum viscosity. This was a retrospective analysis of seven patients. In Mannik s study,15 there was insufficient data to perm it correlation of viscosity at a given shear rate with clinical symptoms of hyperviscosity. Histopathology Bone lesions are not a part of the syndrom e and w hen found are most commonly accompanied by the histology of multiple myeloma, the predom inant cell being a plasma cell, with varying degrees o f differentiation and pleom orphism. Occasionally, bone lesions may be seen in malignant lymphoma, lymphocytic type, poorly differentiated in association with an IgM paraprotein.13 The usual histology of the bone marrow is that of a pleom orphic infiltrate of lym phocytes, plasm a c e lls and plasmacytoid lymphocytes.9,26 Mast cells are occasionally present, a useful finding in d istin g u ish in g this co n d itio n from myeloma in some cases.13 Intranuclear and intracytoplasmic inclusions of PAS positive material are commonly seen, but are not specific for macroglobulinemia of W aldenstrom.29 Problem s in diagnosis arise w hen (1) the in filtrate is pure plasm a cells or well differentiated lymphocytes, (2) bone lesions are not present as in myeloma or (3) the patient is not leukem ic as in chronic lym phocytic leukem ia28; however, these conditions are of rare occurrence. Other clinicopathologic states that may have associated monoclonal IgM protein in the serum include IgM m yelom a, chronic lymphocytic leukemia, malignant lymphoma lymphocytic type or histiocytic type, and a variety of autoim m une (rheum atoid arth ritis, autoim m une hem olytic anemia, chronic cold agglutinin disease), non-b cell neoplastic disorders, and chronic diseases.19,26 Other than in IgM m yeloma,13 and W aldenstrom s

4 M ACROGLOBULINEM IA O F W ALDENSTROM m acroglobulinem ia, the IgM level in these disorders is seldom greater than 2,500 mg per dl, and more commonly 250 to 1,000 mg per dl (normal IgM 45 to 150 mg per dl).26 There have been reports of a num ber of exam ples of benign monoclonal gam m opathy of the IgM type, which w ere asymptomatic and had no evidence of an infiltrative process of the bone marrow, lymph nodes or spleen.26 Many will remain this way for years, but som e w ill m anifest an u n d e rly in g neoplasm on subsequent examination. A recent report28 describes seven cases of B cell neoplasm having the same histologic features as classical m acroglobulinemia of Waldenstroom, b u t each was secreting monoclonal IgG or IgA instead of IgM. It is conceivable that examples w ill be found w here only heavy chains or no immunoglobulin is secreted. During this period, the same laboratory saw 190 cases of the IgM variety of Waldenstrom s m acroglobulinem ia, for an incidence of about 3 percent having non- IgM paraprotein. Materials and M ethods In v estig atio n of the w hole blood, plasm a and serum viscosities were carried out on a microviscometer* utilizing various shear rates (figure 1). The routine relative serum viscosities were m easured in a red blood cell pipet.31 The whole blood viscosities w ere perform ed on heparin ized w hole blood (heparin 10 units per ml). Serum im m unoglobulin assays w ere perfo rm ed u tiliz in g com m ercial im m unodiffusion plates. Im m u n o electrophoresis on the serum and urine was perform ed w ith com m ercial m aterials and antisera. Plastic double bag plasm apheresis! sets or a blood cell separator and blood * Wells-Brookfield Microviscometer, Brookfield Engineering Laboratories, Inc., Stoughton, MA. t Fenwall. F ig u r e 2. Serial observations of the blood viscosity of A.R. (Case #1) during plasmapheresis. processor^ (figure 2), were used for plasm aphoresis. A C D-adenine was the anticoagulant used. Precisely 1,500 ml of plasma were exchanged every two weeks and replaced by 1,500 ml of fresh frozen plasm a. Standard 51 Cr RBC tagging m ethods w ere em ployed for blood volume measurements. Case Studies CASE 1. Severe Epistaxis Managed by Plasmapheresis A.R. is an 82 year old white man with recurrent severe epistaxis. He had complaints of headaches, dizziness and hearing loss which had been present for several months. Physical examination revealed moderate pallor and bilateral nasal crusting and bleeding. Bilateral one cm cervical lymph nodes were felt. The liver edge was felt three cm below the right costal margin with a total span of 14 cm. The spleen tip was palpated four cm below the left costal margin. There were no skeletal defects or purpura. Deep tendon reflexes were diminished. His eye grounds could not be visualized because of cataracts. Laboratory Data Data included hem oglobin 10.2 g per dl, hematocrit 29.7 percent, MCV 90 y?, MCHC 33.4, reticulocyte count 0.5 percent, WBC 4,900 with 75 percent segmented neutrophiles, 20 percent lymphocytes, 4 percent monocytes and 1 percent basophiles. The platelet count was 132,000 per /id. Sedimentation rate was 152 mm per hr. Marked rouleaux and a rare plasma cell were seen on the blood smear. The direct Coomb s test was negative. I Hemonetic 30, Hemonetics Corp., Natich, MA.

5 314 MESSMORE, FA REED, SILBERM AN, GAWLIK AND BERM ES Prothrombin time was 11.5 sec (control 11.0 sec), partial thromboplastin time 55 sec (N = 24 to 38 sec), fibrinogen 140 mg per dl (normal 200 to 400 mg per dl). Bleeding time 12 minutes (normal 3 to 6 minutes). Platelet aggregations with ADP, epinephrine, collagen and ristocetin were normal. Prothrombin consumption was 11.8 sec, control more than 15 sec. Thrombin time (5 /* thrombin per ml) was 20 sec (N = 16 to 28 sec). Serum Protein3 and Viscosity Initial studies on the serum proteins revealed a total protein 11.2 g per dl, IgM 5,600, IgG 400 and IgA 35 mg per dl. Serum immunoelectrophoresis showed IgM Kappa monoclonal gammapathy. The urine was negative for Bence Jones protein. Urine specific gravity was 1.014, ph 5.0, no RBC or WBC. The relative serum viscosity was 6.6; normal 1.4 to 1.8. The relative whole blood viscosity was 11.7; normal 5 to 7. Biochemical Data The serum calcium, phosphorus, glucose, bilirubin, alkaline phosphatase and SGOT were normal on an SMA-12 analysis. The BUN was slightly elevated at 22 to 26 mg per dl and creatinine was 0.9 mg per dl. The cholesterol was below 150 mg per dl on two occasions. The uric acid was 7.0 mg per dl, Na 136, K4.6, Cl 100 meq per 1, C meq per 1 and ph Radiographic Studies A bone survey was negative for osteoporosis and lytic lesions. Bone Marrow Examination A bone marrow aspiration and biopsy from the iliac crest revealed a moderately hypercellular bone marrow. M egakaryocytes w ere normal. The erythroid-myeloid ratio was 1:3. Erythropoiesis and granulopoiesis were normal. The marrow contained 10 to 15 percent plasma cells; some were atypical and lymphocytoid. The PAS stain was negative, but the MGP (m ethyl-green-pryonine) was positive indicating very active protein synthesis in these cells. The iron stores were normal. Treatment Plasmapheresis of 13 units was performed in two days, with replacement with nine units of fresh frozen plasma. After this the bleeding stopped, the partial thromboplastin time fell from 55 to 34 seconds (N = 24 to 38) and the platelet count decreased from 132,000 to 117,000 per /nl. The bleeding time was shortened from 12 to 9 minutes. The whole blood viscosity changed to 5.2 (relative) and the serum viscosity to 3.0 (relative). The total protein was reduced to 7.6 g per dl, IgM 3,500 mg per dl, IgG 580 mg per dl and IgA 48 mg per dl. The nasal packing was removed on the third day with no further significant bleeding. An attempt was made to treat him with chlorambucil, 4 mg per day, but this was stopped after six weeks because of neutropenia (WBC 1,800 to 60 percent segm ented neutro p h iles) and throm bocytopenia (platelets 70,000 per fil). There was no apparent benefit from chlorambucil and repeated plasmapheresis was necessary to maintain a serum IgM below 5,000 mg per dl and serum viscosity below 5 (relative). The patient tolerated the procedure well and a stable state has been maintained for one year by removing 1,200 ml to 1,500 ml of plasma and replacing it with 900 ml of normal plasma every two weeks. When a three week schedule was tried, bleeding recurred. During this time there have been no detectable complications of a serious nature related to plasmapheresis. The patient s blood count (RBC, WBC, platelets) remain stable without transfusions, the spleen and liver remain about the same size. His coagulation studies have returned to normal or near normal. His serum viscosity has remained in a range of 4.5 to 5.0 relative to water. CASE 2. Pyelonephritis Managed by I.V. Fluids and Antibiotics H.B. was diagnosed as having macroglobulinemia of Waldenstrom in 1973 when he presented with acute pyelonephritis, a rapidly rising BUN and creatinine and a serum protein of 12 g per dl. He responded rapidly to i.v. fluids and antibiotics and has had no urinary problems since that time. Physical Examination (1973) The patient was well oriented and alert. Blood pressure was 130/74. The eye grounds were normal and there was no lymphadenopathy or hepatosplenomegaly. No bruising or petechiae were present. Laboratory Data Data included hem oglobin 10.4 g per dl, hematocrit 32 percent, WBC 4,700 with 64 percent segmented neutrophiles, 26 percent lymphocytes, 5 percent monocytes and 5 percent eosinophiles. Sedimentation rate was 66 mm per hour. The Coomb s test was negative. Coagulation Data Platelet count was 203,000 per fil. Bleeding time was 13 minutes, prothrombin time 12.3 sec, partial thromboplastin time was 37 sec, fibrinogen 170 mg per dl and thrombin time 15.3 sec (5 fj. per ml thrombin N = 16 to 27 sec). Platelet adhesiveness was 12.1 percent (N = 25 percent), prothrombin consumption time 20 sec (N = 15 sec). Platelet aggregations with ADP, epinephrine, collagen and ristocetin were normal. Serum Protein Data and Serum Viscosity Total serum protein was 11 g per dl, IgM 8,000 mg per dl, IgA 23 mg per dl and IgG 270 mg per dl. Albumin was 2.7 g per dl. Serum viscosity (relative)

6 MACROGLOBULINEM IA O F W ALDENSTROM 315 was 5.2. The urine contained 700 mg of protein, IgM Kappa type of Bence Jones protein. The serum IgM was monoclonal, Kappa type. A test for rheumatoid factor was negative. Cryoglobulins were negative. Blood Volume A whole blood volume of 8,345 ml (estimated 6,000) with a red cell mass of 2,382 ml (estimated 2,700) and a plasma volume of 5,963 (estimated 3,300) were found. A rectal biopsy was negative for amyloid. Bone marrow aspiration and biopsy revealed m oderate hy percellularity w ith adequate megakaryocytes and normal erythroid and granulocytic proliferation. There was an infiltration by atypical plasma cells which contained intranuclear inclusions interpreted as Dutcher bodies. The marrow contained 20 to 40 percent plasmacytoid cells. Radiographic Studies A bone survey was negative for osteolytic or osteoporotic lesions. Colon x-rays were normal. Chest x-ray was normal. Clinical Course and Treatment The patient was treated with chlorambucil, 4 mg daily, for two years with no significant change in his symptoms or laboratory data except for a slight decrease in gamma globulin. The bleeding time had reverted to normal when the BUN and creatinine returned to normal. He was then treated with chlorambucil in gradually decreasing dosages to 2 mg every other day with no significant difference in any clinical or laboratory parameter except for decrease in total IgM from 8,000 in 1973 to 5,200 in 1977, and a decrease in serum viscosity from 5.2 to 3.9 (relative). The urine continued to show small amounts of Bence Jones proteins. The BUN was 17 mg per dl and creatinine 1.5 mg per dl. His hemoglobin and hematocrit are unchanged. Summary of Viscosity and Coagulation Studies The results of viscosity studies on Case #2, H.B., are shown in figure 1. The cone plate viscom eter at various shear rates dem onstrated an abrupt change in viscosity at low shear rates and dem onstrated the contribution of the red blood cells to the whole blood viscosity. The profound effect of plasm apheresis on whole blood and serum viscosity of A.R. is shown in figure 2, demonstrating the optim al am ount of plasm apheresis required to achieve a significant lowering of viscosity. The purpose of performing plasm apheresis, however, was to restore the patient s coagulation defect to normal to p re v e n t b leed in g. His sym ptom s, ow ing to hyperviscosity per se, w ere modest. The changes observed in his coagulation param eters pre and post p lasmapheresis are shown in table II. These data show conclusively that his bleeding has b e e n controlled by num erous plasm aphereses over a one year period. Discussion Two patients are p resented and d iscussed as typical examples of the syndrom e of m acroglobulinem ia of W aldenstrom. W hile some clinicians and T A B L E II Coagulation Profile for Case 1 P r e p la s map h e r e s i s P o s tp ia s m a p h e r e s is * Platelet count Bleeding time Prothrombin time Partial thromboplastin time Fibrinogen Thrombin time (5 \i per ml) Reptilase time Platelet factor 3 Antithrombin III Whole blood viscosity 91,000 9 minutes 13.3 seconds 36.8 seconds 135 mg per dl 14.3 seconds 13.3 seconds 1 minute 15 seconds 88 percent 8.3 centipoiset 68,000 6 minutes 12.8 seconds (control 11 seconds) 32.6 seconds (normal seconds) 145 mg per dl (normal ) 12.8 seconds (normal seconds) (normal < 20 seconds) (normal seconds) 108 percent (normal 100 percent) 5.8 centipoiset *1,500 ml of plasma removed and replaced by plasma (FFP). tnormal for W.B. is at a shear rate of 90 seconds- * in the Wells-Brookfield Viscometer.

7 3 1 6 M ESSMORE, FAREED, SILBERM AN, GAWLIK AND BERMES pathologists include all patients having more than 1,000 mg per dl of IgM paraprotein (monoclonal) in their serum within the definition of W aldenstroms macroglobulinemia,13,24 others feel that there is clinical need of restricting the term W aldenstroms m acroglobulinem ia to those patients having monoclonal IgM paraprotein and the histopathology of plasm acytoid lym phocytic infiltrates. The reason in keeping those entities separate at this tim e is primarily to avoid giving a highly toxic chem otherapeutic regim en to patients who may have prolonged survival w ith little or no therapy5,9,26 and to give more aggressive and toxic therapy to those having the clinicopathologic features of a malignant lymphoma, lym phocytic type, poorly differentiated, histiocytic lymphoma or of a m ultiple myeloma. The symptom s of hyperviscosity or bleeding tendency should be treated by plasm apheresis if chem otherapy is not effective enough when used alone.5,22,23,25 The effectiveness of plasm apheresis in IgM paraproteinemia is due to the fact that 80 percent of the 19s globulin is intravascular, related to its high molecular weight; 7s globulins (IgG, IgA) are much more difficult to remove by this method. A recent report28 cites seven cases of a lymphoproliferative disorder having the same clinical course and the same h istopathology of the bone m arrow and other tissues as m acroglobulinem ia of W aldenstrom s, b u t IgA or IgC m onoclonal serum protein instead o f IgM. This suggests that clinicopathologic features, clinical course and response to various forms of treatment may correlate better with the histopathology than with the type of protein secreted by the cells. Further observations may clarify this. T he classification of this type of h istopathology as plasm acytic lym phom a has been p roposed by L ukes.12 Subcategories of IgM, IgG, IgA, etc. might then be found to be essentially the same or perhaps different in their clinical b e havior and response to therapy when sufficient num bers have been studied. Our experience in studying the whole blood and serum viscosity in these patients and in review ing the literature reveals that more meaningful information is likely to be obtained by m easuring the whole blood viscosity at low shear rates, than by m easuring serum viscosity. At the present tim e, too little inform ation is available on the correlatio n of the sym ptom s o f th e hyperviscosity syndrome, with whole blood viscosity at low shear rates. F u rth e r stu d ies w ill be needed to confirm the preliminary data of MacKenzie,14 showing that symptoms are unlikely to be due to hyperviscosity if the whole blood viscosity is less than eight centipoises as calculated utilizing Mannik s formula.15 In general, it can be stated that treatment of the symptomatic patient should be with plasm apheresis if there is a significant hyperviscosity syndrome or overt bleeding disorder, until the symptoms are controlled by chemotherapy.5,22,23,25 If chlorambucil is not effective at a tolerable dosage, then a decision for long term plasm apheresis versus a different chem o th erap eu tic approach m ust be m ade. Some reports13 suggest that chlorambucil is not as effective as earlier reports have indicated. C om bination chem otherapeutic reg i m ens consisting of m elphelan cyclophosphamide, vincristine, prednisone and BCNU have b een tried and found to be effective in a small num ber of cases refractory to other treatm ent.2,11 P lasm ap h eresis on a biw eek ly to monthly basis is expensive and requires several units of plasm a, album in or plasma protein product each month, with a small risk of hepatitis to be considered, at least with the use of plasma. It is not com pletely free of com plications5 or hazards but is reasonably safe. Some of the problem s include depletion of blood

8 M ACROGLOBULINEM IA O F W ALDENSTROM elem ents such as (platelets), protein, electrolytes and citrate intoxification. Acknowledgments The authors are grateful to Drs. Frank Kraft, Walter Wood and Crawford Hawkins for permitting us to study their patients. The expert technical assistance of Ms. L. Gosnell and Ms. Heidi Novak is acknowledged. References 1. A l e x a n ia n, R.: Blood volume in monoclonal gammopathy. Blood 49: , Ar l in, Z., L e e, B., and C l a r k s o n, B.: Combination chemotherapy (M-2 Protocol) of Waldenstroms macroglobulinemia with melphalan, cyclophospham ide, vincristine, and BCNU. Proc. Amer. Soc. Clin. Oncol. 18:337, Abst. C-283, B a r t h, W.: Viscometry of serum in relation to serum globulins. Serum Proteins and the Dysproteinemias. Sunderman, F. W. and Sunderman, F. W., Jr., eds. Philadelphia, J. B. Lippincott and Co., 1964, pp Bl o c k, K. and M a k i, G.: Hyperviscosity syndromes associated with immunoglobulin abnormalities. Seminars Hemat. 10: , B u s k a r d, N., G a l t o n, D., G o l d m a n, K., and L o w e n t h a l, R.: Plasma exchange in the long term management of Waldenstroms macroglobulinemia. Can. Med. Assoc. J. 777: , B u x b a u m, J.: Hyperviscosity syndrome in dysproteinemias. Amer. J. Med. Sci. 264: , C a s t a l d i, P. and P e n n y, H. A.: Macroglobulins with inhibiting activity against coagulation factor VIII. Blood 35: , G o d a l, H. and B o r c h g r e v in k, C.: The effect of plasmapheresis on the hemostatic function in patients with macroglobulinemia of Waldenstrom and multiple myeloma. Scand. J. Clin. Lab. Invest. 77: (Suppl 84), K r a jn y, M. and P r u z a n s k i, W.: Waldenstroms macroglobulinemia: Review of 45 cases. Can. Med. Assoc. J. 114: , L a c k n e r, H.: Hemostatic abnormalities associated w ith dysproteinem ias. Seminars Hemat. 10: , L e e, B., S a h a k ia n, G., C l a r k s o n, B., and K r a KOFF, I.: Combination chemotherapy of multiple myeloma with alkeran, cytoxan, vincristine, prednisone and BCNU. Cancer 33: , L u k e s, R. and COLLINS, R.: Immunologic characterization of human m alignant lymphoma. Cancer 34: , M a c k e n z ie, M. and F u d e n g e r g, H.: Macroglobulinemia. An analysis of forty patients. Blood 39: , M a c k e n z ie, M. and L e e, T.: Blood viscosity in W aldenstroms macroglobulinem ia. Blood 49: , M annik, M.: Blood viscosity in Waldenstroms m acroglobulinem ia Blood 44:87-98, P achter, M., Johnson, S., and Basinksi, D.: The effect of macro globulins and their dissociatio n units on release o f p la te le t factor 3. Thromb. Diath. Haemorrh. 3: , P achte r, M., Johnson, S., N e b le tt, T., and T ru a n t, J.: Bleeding, platelets and macroglobulinemia. Amer. J. Clin. Path. 31: , Perkins, H., M ackenzie, M., and Fuden- BERG, H.: H em ostatic d efects in d y s proteinemias. Blood 35: , Ritzm ann, D. E., D aniels, J. C., and Levin, W. C.: Paralymphomatous disease: the syndrom e of m acroglobulinem ia. L eukem ia- Lymphoma. (A collection of papers presented at th e 14th A nnual C lin ical C o n feren ce on Cancer, 1969, at the University of Texas M.D. Anderson Hospital and Tum or Institute, Houston.) Chicago, Year Book, 1970, pp Perry, M. and H o a g lan d, H.: The hyperviscosity syndrom e. J. A m er. M ed. Assoc. 236: , R e p lo g le, R., M eiselman, H., and M e r r ill, E.: C linical im plication of blood rheology studies. Circulation 36: , R uiz, V., and A le x a n ia n, R.: M u ltip le myeloma and macroglobulinemia. Advances in treatment. Post Grad. Med. 55: , Simson, L., Lien, D., W arner, C., and O verman, H.: The long term effects of repeated plasmapheresis. Amer. J. Clin. Path. 45: , SHEBATA, W.: Report ofa case of Waldenstroms macroglobulinemia treated and controlled by radiotherapy. Blood 49: , Solom an, A. and Fahey, J.: Plasmapheresis therapy in macroglobulinemia. Ann. Int. Med. 58 : , Stein, R., E llm a n, L., and B lo c k, K.: The clinical correlates of IgM-M components: An analysis of thirty-four patients. Amer. J. Med. Sci. 269: , SOMER, T.: Hyperviscosity syndrome in plasma cell dyscrasias. Adv. Microcirc. 6:1-55, Turz, T., B rouet, J., F la n d rin, G. L., Da- NON, F., C h a u ve l, J., and SELIGMANN, M.: C linical and pathologic features o f W aldenstrom s macroglobulinem ia in seven patients w ith serum monoclonal IgG or IgA. Amer. J. Med. 63: , W alde nstrom, J.: Diagnosis and treatm ent of multiple myeloma. New York, Grune and Stratton, W e lla, R.: Syndromes of hyperviscosity. New Eng. J. Med. 283: , W r ig h t, D. J. and Jenkins, D. E. Jr.: Simplified method for estimation of serum and plasma viscosity in multiple myeloma and related disorders. Blood 36: , 1970.