Document gby9VKMJzXq2yeGM32gK6npp3

EFFECT OF CHEMOTHERAPY ON T H E LABELLING INDEX OF MYELOMA CELLS BENJAMINDREWINKOM, D, PHD,* BARRYW. BROWNP, HD,' RONALHDUMPHREYP,HD,' AND RAYMONADLEXANIANM, D* The labelling index, namely the percentage of bone marrow plasma cells incorporating tritiated thymidine after short in vitro exposures, was correlated with tumor mass reduction following chemotherapy in 37 patients with multiple myeloma. Tumor mass reduction was assessed from changes in myeloma protein production rate. Patients untreated or not responding to treatment had the lowest labelling index values. The median labelling index increased more than six-fold when tumor mass was reduced by more than 50%. These findings s u p port the concept that an increased fraction of myeloma cells proliferates during remission and justify clinical trials with cell cycle-active drugs in selected patients with a high labelling index. Camer 34:526-531, 1974. T HE GROWTH KINETICS OF DIFFERENT NEOplastic cell populations have been investi- gated in attempts to define superior treatment regimens.8J4.25 T h e present study was con- ducted in order to clarify some aspects of growth kinetics of neoplastic plasma cells and the influence thereupon of chemotherapeutic drugs. Plasma cell tumors are accessible to re- peated bone marrow aspirations. T h e neo- plastic cells composing the tumor produce a specific protein (myeloma protein), the fluctu- ations of which provide an index of tumor mass change.'*28The labelling index (LI) in- dicates the fraction of cells in DNA synthesis and provides an estimate of the growth frac- tion of cell populations. This study correlated changes in the LI of neoplastic plasma cells with different degrees of tumor mass reduction following chemotherapy. Results indicated From the University of Texas at Houston, M. D. Anderson Hospital and Tumor Institute, Houston, T X . Supported by Grants CA 03195 and CA 11430 from the United States Public Health Service. * Associate Professor of Pathology; Chief, Section of Hematology. t Associate Professor of Biomathematics; Chief, Com- pu ter Sciences. t Professor of Physics; Chief, Section of Cellular Studies. 0 Associate Professor of Medicine. Address for reprints: Dr. Benjamin Drewinko, Dept. of Clinical Chemistry and Laboratory Medicine, The University of Texas M. D. Anderson Hospital and Tumor Institute at Houston, 6723 Bertner Ave., Houston, TX 77025. T h e authors thank Mrs. Elvi Fewell, Mrs. Judith Buckingham, and Ms. Cynthia George for assistance in these studies. Received for publication November 2, 1973. that a reduction in myeloma tumor mass is associated with a marked increase in the fraction of proliferating plasma cells. These data justify clinical trials with cell cycle-activedrugs in myeloma patients during remission in attempts to achieve a maximal tumor reduction. MATERIALASND METHODS Patient Population Thirty-seven patients with multiple myeloma were studied. All had bone marrow plasmacytosis and a monoclonal globulin peak on serum or urine electrophoresis. Twentyfour patients had IgG peaks; 5 had IgA peaks; 8 excreted only Bence Jones protein. Twentyfour patients were male; the median age was 58 years. Thirteen patients were studied before any chemotherapy, and the remainder from 1-16 months after the institution of intermittent courses of melphalan-prednisone combination chemotherapy in accordance with specific treatment protocol^.^,^ T h e degree of remission was evaluated in all patients from the ratio between the myeloma protein production rate at the time of study and the pretreatment value considered as 100%.2~22~28 Patients with disappearance of serum or urine peaks were considered to have a reduction in tumor mass to less than 107, of the pretreatment value. At least three simultaneous bone marrow aspirations were obtained from different anatomical sites in 4 patients. I n S other patients, serial samples were obtained at in- 526 No. 3 CHEMOTHERAANPDYKINETICSOF MYELOMACELLS Drewinko et al. 527 tervals of 1 to 16 months to define serial changes in LI as tumor mass was modified by chemotherapy. I n addition, samples from 6 patients with other malignancies and without bone marrow disease were studied to assess the L I of non-malignant plasma cells. Labelling Procedures Marrow aspirates were placed in a centrifuge tube containing 5 ml of McCoy's medium anticoagulated with Versene. Tritiated thymidine (TdR-3H), I @/ml (S.A. 3.0 Ci/mmole) was added, and the marrow particles were mixed by repeated pipette aspirations. Marrow suspensions were incubated in a 50/, C 0 2 atmosphere at 37C for 30 minutes. Smears were prepared from centrifuged concentrates or directly centrifuged onto slides by means of a cytocentrifugeg and processed for radioautography by the liquid emulsion technique using Ilford K5 emulsion. After an exposure time interval of 1 to 2 weeks, the smears were de- veloped, fixed, and stained either with a modified Wright's stain at pH 5.8 or with acid Giemsa. Differential counts were made, and at least 500 plasma cells were examined. Particular attention was given to the differentiation of plasma cells from other hemopoietic elements, especially red blood precursors. With rare exceptions, this was not a major technical problem. Cells were considered labelled when they exhibited at least five grains overlying the nucleus. Multinucleated plasma cells were counted as one cell since the nuclei were either all labelled or unlabelled. T h e peripheral blood of two patients with plasma cell leukemia with white blood cell counts of 17,6001 mm3 and 37,00O/mm3 respectively were processed in an identical fashion. Correlations of Tumor Mass Correlations were made between the percentage of labelled plasma cells and: 1) the degree of remission expressed as a percent of the pretreatment mass level considered as loo%, and 2) the percent of plasma cells in the marrow suspension. Myeloma protein production rate was calculated in each patient from the serum concentration of the myeloma protein, the catabolic rate for the specihc serum concentration, and the estimated plasma volume derived from the hematocrit and body weight.2922 For the eight patients with only Bence Jones protein, changes in the daily excretion of Bence Jones protein provided an index of changes in tumor production rate. RESULTS Reproducibility of LI Comparisons were made of the differential count and the LI of plasma cells obtained simultaneously from different marrow sites. T h e differential of plasma cells varied considerably in a given patient (Table 1). T h e most marked range was 5-14% (a three-fold difference) in patient W.S. However, the LI of plasma cells was virtually identical from site to site in all patients. Thus, the LI of the plasma cells evaluated in each patient was considered representative of the LI of the total tumor. I n one patient (C.C.V.), the LI was determined twice during a 1-month interval when no reduction in tumor mass was observed. T h e LIs were 1.4 and 1.2 respectively. Multinucleated Plasma Cells Some patients had many multinucleated plasma cells, ranging u p to 25y0 of the plasma cell population. All of the nuclei of multinucleated plasma cells were either labelled or unlabelled. T h e proportion of labelled multinucleated cells was compared with the fraction of labelled mononuclear plasma cells i n one patient (N.H.). There were 32% plasma cells in the marrow aspirate with a LI of 11 for the combined population of mono- and multinucleated cells. Twenty-one percent of the plasma cells were multinucleated. A LI of 4 was measured for mononuclear cells and of 36 for multinucleated cells. T h e x2 test for TABLE1. Percentage of Plasma Cells and Labelled Plasma Cells from Different Anatomical Sites Patient Anatomical site o,/", Plasma cells in marrow % Labelled aspirate plasma cells M.M. V.F.C. W.S. F.J. MS LPIC RPIC us LS RPIC MS LPIC RPIC us MS LS LPIC 8 12 11 41 30 35 5 8 14 20 24 21 36 5 8 * 1 1 2 14 * 15 * 11 7 11 US = upper sternum, M S = mid-sterum, LS = lower sternum, LPIC = left posterior iliac crest, RPIC = right posterior iliac crest, * = no data. 528 CANCESRepte mber 1974 Vol. 34 TABLE2. Labelling Index and Bone Marrow Differential of 13 Untreated Patients with Multiple Myeloma Tumor mass Patient Percent L.I. plasma cells High M .L. W.G. D.J. V.C. C.V. W.T. A.T. M.M. 0.1 0.4 06 1.0 1.4 2.0 2.4 4.0 98 36 27 35 38 15 54 10 B.R. 4 5 E.O. 5.2 44 Low G.T. 8 . 5 30 L.C. 11 7 E.B. 18 15 and 2 respectively, while the other four did not demonstrate a single labelled plasma cell among at least 200 cells examined. Plasma Cell Leukemia or Pleural Effusion T h e blood of one patient with plasma cell leukemia (A Bence Jones Protein), with 17,600 WBC/mm3 and 62% plasma cells, exhibited a L I of 0.4. A second patient with plasma cell leukemia (IgGK) with 37,000 WBC/mm3 and 98% plasma cells, had a LI of 0.1. A third patient with 12,600 WBC/mm3 (68y0plasma cells) in a pleural effusion from a large rib plasmacytoma showed a L I of 6.8 in pleural plasma cells and a LI of 18 for bone marrow plasma cells (16% plasma cells). Correlation of LI with Tumor Mass Patients were considered to have either equality of the LI for mono- and multinuclear "high" or "low" tumor mass depending on cells indicated a significant difference between certain laboratory features. T h e patients were these values (p < 0.01). Although not quanti- assigned to the "low" tumor mass group after fied, a higher L I was always observed for multi- all of the following features were confirmed nucleated cells in other patients. before the institution of chemotherapy: 1) Plasma Cell LI of Patients with other Malignancies hemoglobin concentration in excess of 11 g/ 100 ml; 2 ) serum peak of gamma globulin less than 4 g/100 ml; 3) Bence Jones protein T h e LI of bone marrow plasma cells was excretion less than 3 g/day; 4) no evidence of evaluated in six patients with other malig- hypercalcemia; and 5) minimal bone destruc- nancies not involving the bone marrow. tion. When one or more of these disease fea- Plasma cells constituted 0.2-0.8y0 of the dif- tures was present, patients were considered ferential count. Two patients had an L I of 1 to have "high" tumor mass. Patients without previous chemotherapy (13 patients) showed a LI ranging from 0.1 to 18 High Tumor Mass (median of 2.4). Five untreated patients in the n "low" tumor mass category had a median L I 10/ of 8.4 (range 4 to 18), while the median LI of 8 untreated patients in the "high" tumor 20 mass group was 1.2 (range 0.1 to 4).This differ- ence was statistically significant (p < 0.02) (Table 2). T h e LI of bone marrow plasma cells was correlated with changes in tumor mass in treated patients, as evaluated from the mye- loma protein production rate and degree of bone marrow plasmacytosis (Figs. 1 and 2). Eleven patients with no reduction in tumor mass from chemotherapy had a L I ranging from 0.5 to 11 (median of 6.5) (Fig. 1). There was no significant difference between the LIs i :100100 50 of untreated and treated patients who did not 20 10 show a reduction in tumor mass. T h e LI Percent of Pretreatment Tumor Mass usually increased as tumor mass was reduced FIG. 1. T h e labelling index of bone marrow plasma cells from treated patients correlated with tumor mass, by chemotherapy, assessed either in terms of myeloma protein production rate (Fig. l),or of expressed as a percentage of the pretreatment value. the plasma cell differential (Fig. 2). T h e Spear- No. 3 CHEMOTHERAAPNYD KINETICSOF MYELOMACELLS * Drewinko et al. 529 man rank correlation coefficient was higher (0.619) when the LI values were compared with tumor mass change than when compared with the plasma cell differential count (0.495), but this difference was not statistically significant. There was marked variability in L I values after tumor mass reduction. T h e median LI was 18.5 for patients with a reduction in tumor mass of more than 50%. When the reduction in tumor mass exceeded SO%, the median LI was 15 (range 1-28) for nine patients in the "high" tumor mass category (Fig. 1A); such changes produced a median LI of 26 (range 11-50) in five patients in the "low" tumor mass group (Fig. 1B). This difference although suggestive of a higher LI for the "low" tumor mass group after tumor reduction, was not statistically significant. Sequential determinations of the L I in eight individual patients (Table 3) indicated an inverse relationship between the degree of tumor mass reduction and the LI. All five patients with reductions in myeloma protein production rate showed increments in the LI, while two patients with progressive disease demonstrated decrements in the LI. No change occurred in the LI of one patient who had no reduction in tumor mass. DISCUSSION T h e labelling index of plasma cells in patients with myeloma was evaluated in order to clarify the effect of chemotherapy on plasma cell tumor kinetics in man. Plasma cell tumors were considered to provide a model for neoplasms with low growth rate. T h e fraction of proliferating neoplastic plasma cells was evaluated by assessing the proportion of cells undergoing DNA synthesis as indicated by the incorporation of tritiated thymidine after short in vitro exposures. Adequate samples of plasma cells were procured from bone marrow 50 i I 40t 0 I'0 0 00 P "0 3 , "0 t0 0 N 0 0 , 100 90 80 70 60 50 40 30 20 10 0 Percent Plasrno Cells FIG. 2. T h e LI of bone marrow plasma cells from treated patients correlated with the plasma cell differential count. The large asterisk (*) represcnts the plasma cell LI of six patients with a normal bone marrow. aspirates in a large number of patients with different degrees of tumor mass change. T h e reproducibility of this technique was confirmed in individual patients by assessing simultaneous marrow aspirates from different parts of the skeleton. Previous studies had confirmed the superiority of evaluating plasma cell differential counts in marrow suspensions rather than in direct smears. In each patient, the LI was correlated both with the magnitude of tumor mass change assessed from changes in myeloma protein production rate and the plasma cell differential count of bone marrow suspensions. Other studies have demonstrated the utility of evaluating changes in tumor mass from changes in myeloma protein production rate.2vgz T h e nuclei of multinucleated cells were either all labelled or unlabelled, suggesting synchronized transit by all nuclei through the cell cycle. T h e proportion of plasma cells labelled with tritiated thymidine was much higher for multinucleated cells than for mononuclear cells. Apparently, a larger proportion of multinucleated cells was proliferating. In Tumor mass Decreasing Unchanged Increasing TARL3E. Sequential L I in Patients with Multiple Myeloma Initial study Patient LI Time interval between studies Tumor mass (months) Second study LI Tumor mass M.R.M. M ..4.iVI . D.R.R. A.S.T. J.G.S. 4 . 0 100% 1.0 51% 6 . 0 100% 2 . 4 100% 0 . 8 100% 5 6.4 50% 3 7 41% 1 8.4 81% 2 5 . 8 51% 2 15 46% C.C.V. 1.4 100% 1 1 . 2 100% G.P.L. J.T.G. 26 2 18% 8 11 39% 35% 16 1 . 6 49% 530 CANCERSeptember 1974 VOl. 34 vitro studies on immunoglobulin-producing cells suggest that multinucleated cells may arise from the fusion of separate mononuclear cells.10 The higher frequency of proliferating multinuclear cells could be explained by the transfer of information initiating DNA synthesis from a proliferating cell to a non-cycling one at the moment of cell fusion. Alternatively, multinucleated cells arising from endomitosis may have a greater ability to remain in cycle than mononuclear cells. Plasma cells in the blood of two patients with plasma cell leukemia had a LI less than 1%. These findings differ from those reported by Salmon et a1.21 in one patient with IgE plasma cell leukemia who had a peripheral blood plasma cell LI of 40, but are similar to the low values for peripheral blood LIs reported in patients with other leukemias.GJ9 Our data indicate that the plasma cells present in the peripheral blood of our patients with plasma cell leukemia were either terminal elements incapable of further proliferation or resting cells in a Go phase of their cycle. T h e L I of normal bone marrow plasma cells from patients with malignancies other than multiple myeloma was always less than 2. This demonstrates that most normal plasma cells are either resting (Go pool) or in a terminal stage incapable of further proliferation. T h e fraction of proliferating plasma cells was low in untreated patients with multiple myeloma. Our median LI of 2.4 recorded before the institution of chemotherapy was similar to results of Killmanl? and to those reported by Salm0n.~3These findings support the concept that progressive tumor growth is associated with a steadily declining proliferating fraction. Following reductions in tumor mass to less than 50% of the pretreatment level with chemotherapy, the LI increased to a median of 18.5.Our finding of a higher LI following tumor reduction is similar to that reported by Salmon in patients with multiple myel0ma,~3and conforms with similar findings in patients with acute leukemia.15.19 T h e wide variability in measured L I for a specific degree of tumor mass reduction could be attributed to the marked differences among our patients in absolute cell numbers before the institution of chemotherapy. This was supported by our observations that patients with the most extensive disease before treatment had a lower LI, both before and after tumor reduction than other patients with a similar degree of tumor reduction but with less disease involvement. T h e inverse relationship between LI and tumor mass was further supported by serial studies in eight patients. Our observations imply that a reduction in myeloma tumor mass by alkylating agents shifts large numbers of resting plasma cells into a proliferative pool and that the magnitude of this conversion depends both on the degree of tumor reduction and the initial tumor mass. Kinetic studies have shown that, in most tumors, a large fraction of viable neoplastic cells proliferates very slowly.lS~2~~A3s0 tumor growth progresses, the fraction of slowly proliferating cells increases, the proportion of cells initiating DNA synthesis per unit time decreases, and the rate of growth declines. Thus, tumors expand in an exponentially retarded exponential fashion, best described by a Gompertzian curve.13 T h e short-term behavior of untreated tumors is determined by tlie cell cycle time, the length of the phases of the cycle, the growth fraction, and the cell loss factor.27When a single course of treatment is given, cycling cells are particularly sensitive.15~ 2031Following the death of cycling cells, some nonproliferating cells return to the proliferative pool and become vulnerable to the next treatment.4,16.17Thus, the long-term outcome of treatment depends upon the number of tumor cells eliminated by each course of therapy, the fraction of noncycling cells which return to the proliferative pool, and tlie viability and drug sensitivity of these cells. Similar considerations apply to plasma cell myeloma. Griswold et al. have demonstrated,ll in experimental hamster plasmacytomas, that long-term remissions occur when a cell cycleactive agent is given after treatment with alkylating agents, but not before. T h e major factor in accounting for the effectiveness of such a drug (arabinosylcytosine) is the increased growth fraction following the initial course of alkylating agents. Our studies on the growth kinetics of human plasma cell tumors confirm substantial increments in the proportion of proliferating cells after tumor reduction with alkylating agent chemotherapy. These observations justify the systematic evaluation of cell cycle-active drugs in the treatment of patients with plasma cell tumors after remissions have been induced with alkylating agent combinations. Since superior degrees of remission are associated with longer remission and survival times,2 this approach may improve substantially the prognosis for a large fraction of patients with multiple myeloma. -No. 3 CHEMOTHERAPANYD KINETICSOF MYELOMACELLS Drewinko et al. 531 ,REFERENCES 1. Alexanian, R., Bergsagel, D., Migliore, P., Vaughn, W., and Howe, C. D.: Melphalan theiapy for plasma cell myeloma. Blood 31:l, 1968. 2. Alexanian, R., Bonnet, J., Gehan, E., Haut, A,, Hewlett, J., Lane, M., Monto, R., and Wilson H.: Combination chemotherapy for multiple myeloma. Cancer 30:382, 1972. 3. Alexanian, R., Haut, A,, Kahn, A., Lane, M., McKelvey, E., Migliore, P., Stuckey, W., Jr., and Wilson, H.: Treatment for multiple myeloma-Combination chemotherapy with different melphalan dose regimens. J A M A 208:1680,1969. 4. Barranco, S., Luce, J., Romsdahl, M., and Humphrey, R.: Bleamycin as a possible synchronizing agent for human tumor cells in vivo. Cancer Res. 33:882, 1973. 5. Clarkson, B., and Fried, J.: Changing concepts of treatment in acute leukemia. M e d . Clin. North Ani. 55:561, 1971. 6. Clarkson, B., Ohkita, T., Ota, K., and Fried, J.: Studies of cellular proliferation in human leukemiaI. Estimation of growth rates of leukemic and normal hematopoietic cells in two adults with acute leukemia given single injections of tritiated thymidine. J . Clin. Invest. 46:506, 1967. 7. Deschner, I., and Lipkin, M.: Proliferation and differentiation of gastrointestinal cells in relation to therapy. M e d . Clin. N o r t h A m . 55:601, 1971. 8. DeVita, V.: Cell kinetics and the chemotherapy of cancer. Cancer Chemother. Rep. 2 2 3 , 1971. 9. Drewinko, B., Sullivan, M., and Martin, T.: Use of the cytocentrifuge in the diagnosis of meningeal leukemia. Cancer 31:1331, 1973. 10. Drewinko, B., Trujillo, J., and Tessmcr, C.: Functional Sequences modulated by morphological transitions in human lymphoid cells grown in vitro. Science 171:185, 1971. 11. Griswold, D., Simpson-Herren, G., and Schabel, F. M.: Altered sensitivity of a hamster plasmacytoma to cytosine arabinoside. Cancer Chemother. R e p . 54:337, 1970. 12. Killman, S., Cronkite, E., Fliedner, T., and Bond, V.: Cell proliferation in multiple myeloma studied with tritiated thymidine in vivo. Lab. Invest. 11:845, 1962. 13. Laird, A.: Dynamics of tumor growth. Br. J . Cancer 18:490,1964. 14. Lampkin, B. C., McWilliams, N., and Mauer, A,: Cell kinetics and chemotherapy in acute leukemia. Semin. Hematol. 9:211, 1972. 15. Lampkin, B. C., Nagao, T., and htauer, A.: Drug effect in acute leukemia. J . Clin. Invest. 48:1124, 1969. 16. Lampkin, B. C., Nagao, T., and Mauer, A.: Syn- chronization and recruitment in acute leukemia. .I. Clii7. Invest. 50:2204, 1971. 17. Mauer, A., Saunders, E. R., and Lampkin, B. C.: Importance of the non-dividing leukemic ccll. Cnnrer Res. l7:168, 1969. 18. Mendelsohn, M.: Auioratliographic analysis of cell proliferation in spontaneous breast cancer of C3H mouse-111. T h e growth fraction. J . Natl. Cancer I?i.\t. 28:1015, 19G2. 19. Pileri, A., Gabutti, V., Masera, P., and Gavosto, F.: Proliferative activity of the cells of acute leukemia in relapse and in steady state. Acta. Hemntol. 38:193, 1967. 20. Pittillo, R. F., Schabel, F., and Skipper, H.: T h e sensitivity of resting and dividing cells. Cancer Chentother. Rep. 54:137, 1970. 21. Salmon, S. E., McIntyre, O., and Ogawa, M.: IgE myeloma-Total body tumor ccll number and synthesis of IgE and DNA. Blood 37:696, 1971. 22. Salmon, S. E., and Smith, B. A , : Immunoglobulin synthesis and total body tumor cell number in IgC multiple myeloma. J. Clin. I n 7 m t . 49:115, 1950. 23. Salmon, S. E., and Smith, B. A,: Induction of tumor susceptibility to cycle active agents in IgG multiple myeloma. Clin. Res. 20:570, 1972. 24. Skipper, H. E.: T h e cell cycle arid chcmotherapy of cancer. I n T h e Cell Cycle and Cancer, R. Baserga, Ed. New York, Marcel Dekker, 1971; p. 355. 25. Skipper, H. E.: Kinetic behaviour versus response to chemotherapy. Natl. Cancer I m l . hlonogl-. 3 4 2 , 1971. 26. Skipper, H. E.: Kinetics of mammary tumor cell growth aud implications for therapy. Cancer 28: 1479, 1971. 27. Steel, G. G., and Lamerton, L.: Cell population kinetics and chemotherapy in human tumor cell kinetics. Nutl. Cancer Inst. Monogr. 30,:29, 1969. 28. Sullivan, P., and Salmon, S. E.: Kiuetics of tumor growth and regression in IgG multiple myeloma. J . Clin. Invest. 51:1967, 1972. 29. Tannock, I., and Frei, E.: Possible role of cell kinetics in prediction of the response to caucer therapy. Natl. Cancer Inst,. Monogr. 34:19, 1971. 30. Terz, J.. Curutchet, P., and Lawrence, W.: Analysis of the cell kinetics of human solid tumors. Cancel' 28:1100, 1971. 31. Van Putten, L. M., and Lilicveltl, P.: Factors determining cell killing by chemotherapeutic agents in vivo. Eur. J . Cancer i : l l , 1971. 32. Zubrod, C. G.: Chemical control of cancer. Pl-or. Natl. Acad. Sci. U.S.A.69:1042, 1952.