copper and zinc metabolism with solid tumor growth

BIOLOGICAL TRACE ELEMENT RESEARCH 1,203-215 (1979)

Copper and Zinc Metabolism with Solid Tumor Growth

ARTHUR FLYNN

Department of Immunology, Research Division, Cleveland Clinic Foundation, Cleveland, Ohio 44106

Received January 5, 1979; Accepted February 1, 1979

ABSTRACT

The variation in copper and zinc metabolism with tumor growth appears to relate directly to progression or regression of the disease. Historically, elevations in serum copper have been used as clinical indicators in hematologi- cal neoplasms since the early 1960s. More recently, we have monitored breast, colo-rectal, and lung cancer patients for a six-month period through courses of cytotoxic chemotherapy to determine copper and zinc changes with tumor growth. Groups were divided into responders and nonresponders blind to their serum copper and zinc levels. Trends in elevated serum copper with active disease have shown similar trends in decreasing values with effective therapy, but normalization was at a slower rate. Serum zinc levels in the same patients were markedly below normal and did not increase in the study period. The clinical significance or elevated serum copper and depressed serum zinc is discussed and the potential relationship between the two elements is explored. A solid tumor-bearing rat model, mammary adenocarcinoma R 3230 AC, has detailed more of the changes in copper and zinc metabolism with solid tumor growth. Serum copper and zinc varied with tumor mass, as in clinical studies. Liver values of the two essential metals did not change significantly, but liver-related copper-containing enzymes showed marked variations. Ceruloplasmin in serum increased with increasing tumor mass, as would be expected with the increased serum copper levels. Cytochrome c oxidase activity in liver homogenates from tumor-bearing animals was significantly depressed.

Key Words: Copper metabolism, and solid tumor growth; zinc metabolism, and solid tumor growth; tumor growth, and copper metabolism; tumor growth, and zinc metabolism; copper/zinc ratio, and tumor growth; ceruloplasmin; cytochrome c oxidase; leukocyte endogenous mediations; siallyl transferase.

203

Copyright �9 1979 by The Humana Press Inc. All rights of any nature whatsoever reserved. 0163-4984/79/0900-0203 $02.60

204 A. FLYNN

INTRODUCTION

Reports began to appear in the biomedical literature about two decades ago of significant changes in serum copper and zinc with the growth of malignant tumors. In 1959, Adink and Frank (1) reported serum and whole blood zinc levels to be depressed in the majority of patients with neoplastic growth. The following year, Pagliardi and Giangrandi (2) showed elevated serum copper levels to have clinical significance in a hematologic neoplasm, Hodgkin's disease. Subsequent to these reports, numerous citations have appeared that describe phenomenologically elevations in serum copper and tumor copper and zinc and depressions in serum zinc. Table 1 provides a survey of serum copper and zinc levels in 279 cancer cases we gathered over a three-year period. Significant elevation of serum copper and significant

TABLE 1 Serum Copper and Zinc Levels in Normals and Cancer Patients

Group (No.) Copper,/Lg/mL Zinc, t~g/mL Cu/Zn

Normals (30) 1.093 + 0.138 1.080 + 0.123 1.01 Cancer patients (279) 1.715 + 0.264 0.783 + 0.183 2.19

depression of serum zinc can be seen in comparison with normal controls. Serum copper and zinc levels have both been purported to be indicators or prognostic tools, alone or in combination, for the status of the malignant tumor.

There is good basis for implicating these two essential elements in the replication of neoplastic cells. Fundamentally, metal ions are shown to inter- act preferentially with nucleoside bases (3). Enzymes related to replication and transcription of nucleic acids are metal mediated (4). And other studies have shown zinc deficiency effects on tumor cell populations and tumor growth (5-7). The significance of variations in whole animal systems with malignant tumors that are neither deficient or optimal in copper and zinc is still an unanswered question.

Volumes of clinical description have been written about the elevated serum copper (and the related metalloprotein ceruloplasmin) and the use of the level of this metal as a marker of disease progression or regression. Hemato- logic neoplasms (8) and solid tumors (9) have been reported to present with abnormally high serum copper. A number of theories have been likewise postulated to explain this increase: (a) Since elevated serum copper is also found in those using exogenous estrogens, Thorling and Thorling (10) hypothesize the "possibility of cross-reacting in some respects with estrogens (or less likely, stimulating estrogen production or interference with its metabolism)"; (b) A cell mediated system is described in infectious processes, identified as leukocyte endogenous mediators (LEM), which are low-

CU AND ZN METABOLISM IN SOLID TUMORS 205

molecular weight proteins exerting a CNS-mediated liver effect on ceruloplasmin synthesis (11); and (c) a defect in sialyl transferase metabolism where the glycoprotein ceruloplasmin is spared catabolism by resialylation owing to increased synthesis of sialyl transferase by the tumor (12). The elevated serum copper has also been related to blast cell percentages, but no mechanism was postulated (13). The recent report by Wilimas et al. (14) continues to describe the difficulties in using serum copper levels owing to the variability in patient response with tumor growth. It is also difficult to separate out responses to non-neoplasm inducers of ceruloplasmin synthesis.

Subsequent to the report of Adink and Frank (1), studies have been reported that relate serum zinc values to severity of disease (15), leukocyte cell zinc content (16), and disease remissions in leukemics (17). Tumor tissue zinc levels varied with the site of tumor from significantly decreased to significantly increased in relation to comparable normal tissue (18-20). Mikac-Devic in 1970 concluded that in clinical studies too many contradictory reports permit the use of zinc levels in defining malignant tumor states in humans (21).

The combination of serum copper and zinc levels in a copper/zinc ratio has been proposed as a way to sensitize the two serum determinations for better prognostic use in cancer. Strain et al. (22) in 1972 first proposed the copper/ zinc ratio in lung cancer as an optimal way of using the serum trace element levels. Delves et al. (23), likewise, used the ratio to discriminate between un- treated and treated-in-remission leukemic children. Fisher et al. (24) found in osteosarcoma a different response in copper/zinc ratios. Primary osteosarcoma prior to therapy had elevated serum copper and zinc giving a near control ratio, whereas metastic osteosarcoma had elevated serum copper, near normal serum zinc, and an elevated copper/zinc ratio. No consensus is apparent and a question as to the use of ratios can be raised. The copper/zinc ratio implies a functional relationship between the two elements, and at this time no evidence supports such an intimate relationship in malignant tumor growth.

COPPER AND ZINC METABOLISM IN PATIENTS WITH SOLID TUMORS

An in-depth analysis of a series of patients for an extended period was thought to provide an opportunity to assess serum copper and zinc changes with solid tumor growth. The design of the study was to objectively define response to therapy blind to changes in serum copper and zinc. The patient groups were separated as having progression or regression of disease at the end of a 6-month study period. The analysis could then define differences within groups prior to treatment and sequentially throughout treatment. Secondarily, variations and relations in individual serum copper and zinc levels would provide an insight into a link between the two essential elements and a copper/zinc ratio.

206 A. FLYNN

A 6-month prospective follow-up study was conducted with breast, colon and lung cancer patients to determine the prognostic value of serum copper and zinc. All patients had metastatic disease and at the start of the study were being evaluated for cytotoxic chemotherapy. Thirty-six patients were sampled prior to therapy and at intervals of one, three, and six months after the start of treatment. After the 6-month period, an evaluation of the patient's status, based on objective measures of the individual disease, was conducted by the treating physician. Serum copper and zinc were analyzed by atomic absorption spectrophotometry according to tumor type and time sequence.

Time Sequence Changes in Serum Copper and Zinc

The time sequence showed significant variations in serum copper, but insignificant changes in all serum zinc levels in all three solid tumor types. Serum copper values appeared to be above the normal range in the nonresponder group regardless of tumor. Serum zinc levels was generally below the normal range for all tumors regardless of time, treatment, or response.

Figure 1 illustrates the variation in serum copper and zinc with metastatic breast cancer. Differences between responder and nonresponder groups were only evident prior to and after six months of therapy in serum copper. At one month and three months no difference was noted and serum zinc varied little from the subnormal values. Colon cancer patients had significant differences in serum copper not only prior to and at six months, but also at the 3-month time interval, as shown in Fig. 2. Serum zinc values again showed no difference in time sequence or between the two response groups. Respond- ers and nonresponders to therapy were most separable in lung cancer. Figure 3 demonstrates the significant differences in serum copper at all sampling times between the two groups. As was evident in the other two tumor sites, serum zinc reflected no between group differences and was depressed below the normal range.

The trends shown in the figures for the three tumor types illustrate the variability of alterations in serum copper and zinc levels. Differences prior to therapy that are no longer evident one month after the start of therapy pre- clude the wise use of serum copper as a prognostic tool in solid tumors. Variations within or very near the normal range of serum values (even though significant) must likewise be troublesome. Wilimas et al. (14) indicated in their report on 29 children with Hodgkin's disease that serum copper levels alone can give a false positive rate of 47.3~ and an ability to detect relapse in only 607o of recurrent disease. Multiple factors effecting serum copper and zinc can not be easily accounted for and therefore make the use of trace element levels in this manner unreliable.

The changes in serum copper and zinc whether variable or stable reflect imbalances of some sort in the patients with solid tumors. Serum copper may reflect changes in either the catabolism of ceruloplasmin as a glycoprotein or the increased synthesis of the protein. The significance of the increased copper protein in acting as both a ferroxidase and copver transport molecule

C u AND Z N METABOLISM IN SOLID TUMORS 207

2 .0 -

1.5-

#g/ml

! .0-

0 . 5 -

COPPER

. w ,

l

i

[ �9 r

!iiiiiii!i!i!i I _ [ '

: - : o : - - : :

~ = = . . . . .

: '1-:4-:-:': ':

To

1.5

,ug/ml

1.0

0 . 5

ZINC

-!!!!!!i!i!i!i!!!

PRE- R x I MONTH 5 MONTHS 6 MONTHS TIME ON THERAPY

Fit3. I. Serum copper and zinc levels in breast cancer patients (N = 12). Responder (empty bars) and nonresponder (laddered bars) groups (n = 6) differ significantly as indicated by , . Normal ranges are shown as the dot-shaded area. Mean + S.D.

(25) is yet to be defined in tumor systems. Depressed serum zinc may reflect a higher metabolic uptake of zinc in the tumor Schwartz et al. (26) reported a 5.7-fold increase in tissue zinc in malignant versus normal tissue. The decreased serum zinc in patients with solid tumors may reflect a flux of zinc to

208 A. FLYNN

2 .0 -

1 . 5 -

,uglrnl

1.0

0.5

COPPER

ii~!i!~iiiiiiii!i iiiilIiii!i :i:!:!:i:i:i:i:!J

~:.~:T~!~i~i~' -

0

Jiiiiiii!!!!i!! !i!i!iiii!!!ill

t i r ~ - _ , ,L .....

~ w

- - = iii!liiiiii

~.%%%%%%~

m

T

" : i : i : i : i : i : i : i : : i : i : i : ! : ! : i : i :

!

1

I

,,uglml 1.0-

1.5 ~ ZINC

iiiiiiiili',!',ii ....... i ........ iiiiiiiiiii',iTi',i?iiiii',iiiii',ii~'::'~i ....... i',iiiii',iii!i ...... ',iii!iil

ii•ii•ii••i!!i!iiiii••ii:iiiiiiiii•!!iiiiii••!i!ii•!iiiiiiii!i!•••!iii•i•iiiiiii!ii!i!ii!ii•••i•!iiiiiii••ii iiiiiiilili!iiiiiiii!iii!i!iiiiiiiiiiiili:iiiiii!iiiiiiiii!iiiii!iiiill /

0 . 5 -

PRE- F~ I MONTH :3 MONTHS 6 MONTHS TIME ON THERAPY

FIG. 2. Serum copper and zinc levels in colon cancer patients (N = 12). Responder (empty bars) and nonresponder (laddered bars) groups (n = 6) differ significantly as indicated by I . Normal ranges are shown as the dot-shaded area. Mean + S.D.

malignant tissue. The time sequence, however, still does not provide us with an indication of the usefulness of copper/zinc ratios in defining disease.


2.0-

1 . 5 -

,,uglml

1.0

05

T COPPER �9

T I

H':':':'H~ i . . . . . . . .

:.:i:i:-:i:_-::::: 1 ~ - x ~ - ~ �9 . x ~ ~.:.:.:.:.:::!::: �9 * - - - - : ~ 1 7 6

T I

I

( . x o : , = ~ 1 4 9

i:!:!:i:i:-:i:

:i:i:i~i~:~:!:

I �9 D

0

~ 1 7 6 1 7 6 1 7 6 1 7 6 1 7 6 1 7 6

: . : .X':

- . . . , . . . .

?i:i:i :.:.1.:- :.:.:.:~ :.:.:~

1.5~ ZINC

~,~,m, iiiililililiiiiiiiiiiill I iiii!21111ilil i i iiiiiiiiiiiil !ili!i!i!ii! !!ii iiiiiiil Lo-i..i~i~i...i~i~i~i..i!i..i..i......iii..i..i..i..i~i..~iii~..ii~i}i!iiiiiiii.:iiiiiiiiiii~i!~iiiiii.~iiiiiiiiiiiiiii..iiiiiiiiiiiiiiiiiii~i~i?ii~iiiii~iiiiiii~i~iiiiiiiiiii~iii~i~i~.iiiiii~iiiiiiii

i!i!i!i!!!i!i!i!i: i!i:! !i!i~i~i!i!i:':i~iiii!i~iiiiii!i!i!i!!!i!.':.!:!:!!!:i:! 1 i:!:~:~:!:!:!:i:i!i:i:i:i'i'i:!:~:!:~:~: !:i:i:i:~:i:?i:i:i:i:i:i:i'i:i:i:~!~:i:i': :~:i:~:~:~:~:i:~

0 . 5

PRE- F~ I MONTH 3 MONTHS 6 MONTHS TIME ON THERAPY

FIO. 3. Serum copper and zinc levels in lung cancer patients (N = 12). Responder (empty bars) and nonresponder (laddered bars) groups (n = 6) differ significantly as indicated by O. Normal ranges are shown as the dot-shaded area. Mean + S.D.

210 A. FLYNN

Copper/Zinc Ratios in Solid Tumor Patients

We also calculated the copper/zinc ratios on the 180 samples from patients representing the three tumor types indicated above. We analyzed the results in three ways: (a) responders versus nonresponders at the four time intervals; (b) trends in the time sequence within each subgroup; and (c) a correlation of all 180 samples to determine whether a significant negative correlation existed between copper and zinc. Table 2 provides the copper/zinc ratios, the statistical results, and the coefficient of correlation on the serum copper and zinc comparisons. The comparison of responders to nonresponders to therapy show a less sensitive indicator than serum copper alone. Lung cancer patient results were the only data that noted any difference, but again not as sensitive as serum copper alone. No variations in the copper/zinc ratio were noted with time within the tumor site groups. A Pearson r coefficient of correlation was calculated to contrast serum copper and zinc changes. As shown in Table I, the normal copper and zinc ratio is 1.01, and the correlation did not show a negative linear relationship between the two elements, but an insignificant, slightly positive correlation. Although this information does not deal with the possibility of a nonlinear relationship between the two elements, the re- suits as a whole do not support the usefulness of copper/zinc ratios as indicators of disease status.

TABLE 2 Copper/Zinc Ratios in Patients with Solid Tumors ~

Treatment Period

Tumor site Pretreatment 1 month 3 months 6 months

Breast Responder 1.75 + 1.17 2.22 + 0.68 1.56 + 0.86 2.01 _+ 0.52 Nonresponder 1.96 + 0.25 2.16 + 0.53 1.70 • 0.36 2.24 + 0.40

Colon Responder 1.84 + 0.17 2.67 + 0.87 1.77 _+ 0.36 1.80 +_ 0.49 Nonresponder 3.05 + 3.08 3.03 + 0.54 4.22 + 1.39 3.60 + 1.84

Lung Responder 1.91 + 0.50 2.19 + 0.71 1.84 _+ 0.73 1.71 + 0.51 Nonresponder b 3.37 _+ 1.72 3.06 _+ 1.05 3.15 + 0.66 3.09 _+ 0.70

a The overall relationship between serum copper and zinc was tested by the Pearson r coefficient of correlation, r = 0.0495, N.S.

b Significantly different from corresponding responder group by the Student t test, p < 0.05.

COPPER AND ZINC IN SOLID TUMOR-BEARING ANIMALS

The data developed from clinical studies left many questions about the significance of serum changes in copper and zinc with tumor burden un-

2.0-

1.5-

~g/ml

1.0-

0.5-

CU AND ZN METABOLISM IN SOLID TUMORS

SERUM TRACE ELEMENTS

T

I

1

"1"

211

CONTROL TUMOR I WEEK GROWTH REMOVED

FIG. 4. Serum copper (empty bars) and zinc (laddered bars) in Fisher rats under control (n = 8), tumor growth (n = 7), and surgically removed (n = 7) conditions. Mean + S.D.

answered. An animal experiment was designed using a transplantable solid tumor to determine systemic and storage changes in the two elements. The study was designed to: (a) obtain a serum copper and zinc profile similar to the profile noted in clinical studies; (b) determine the effect of the amount of tumor burden has on the copper and zinc levels; (c) measure changes in tissue copper and zinc in a storage tissue, liver; and (d) observe variations in metal-dependent enzymes with distal tumor growth.

Animal studies have mainly focused on serum changes in copper (ceruloplasmin) and zinc, with little interest in the systemic metabolism of the elements. We examined a rat mammary adenocarcinoma (R 3230 AC) ob- tained from the Mason Research Institute. The tumor was transplanted sub- cutaneously on the suprascapular region in 100-150 g female Fisher rats. Tumors were palpable at 10 days and allowed to grow for 21 days when the animals were randomly distributed into three groups: (a) sacrifice and sample; (b) surgically remove tumor mass, allow tumor to grow for seven days, then sacrifice and sample; and (c) surgically remove tumor mass, allow tumor to grow for 21 days, then sacrifice and sample (27). Copper and zinc were analyzed in serum, liver, and tumor by atomic absorption spectroscopy, and cytochrome oxidase was measured in liver by the method of Mills and Dalgarno (28).

As noted in clinical studies, serum copper was significantly elevated and serum zinc was significantly depressed with tumor growth. The group

212 A. FLYNN

5.50-

COPPER

:5.00-

,ug/g

2.50-

LIVER TRACE ELEMENTS

-55

ZINC I

-25

CONTROL

t m

II TUMOR I WEEK

GROWTH REMOVED FIG. 5. Liver copper (empty bars) and zinc (laddered bars) in Fisher rats under control (n = 8), tumor growth (n = 7), and surgically removed (n = 7) conditions. Mean +__ S.D.

sampled 7 days after removal of the tumor mass (microscopic disease still present) rebounded with a marked decrease in serum copper and a minimal increase in serum zinc, as shown in Fig. 4. Liver copper and zinc changes were in opposite directions from serum trace element values. Figure 5 shows the decrease in tissue copper and increase in tissue zinc; the possible significance of these changes will be discussed later. Cytochrome c oxidase variations are illustrated in Fig. 6 and were determined as a possible indicator of copper deficiency (28) or a metabolic defect in copper enzymes. The liver cytochrome c oxidase activity was significantly depressed and did not completely recover with removal of the tumor mass. A definite shift in all facets of copper metabolism measured was noted in the animal model.

SIGNIFICANCE OF COPPER AND ZINC CHANGES IN SOLID T U M O R GROWTH

The data presented in our clinical study did not support the use of serum copper or zinc as prognostic tools in following patients with solid tumors. The differences between those patients with objective response and no response were negligible or so variable with time as to negate the significance of the difference. Variations within the normal range are also difficult to inter-

CU AND ZN METABOLISM IN SOLID TUMORS

0.25-

0.20 -

UNITS

0.15-

0.10-

CYTOCHROME OXIDASE

1

213

CONTROL TUMOR I WEEK GROWTH REMOVED

FIG. 6. Liver cytochrome c oxidase activity under control (n = 8), tumor growth (n = 7), and surgically removed (n = 7) conditions. Cytochrome c oxidase activity units reflect the oxidation of/~mol cytochrome c/min/mg protein in the liver homogenates. Mean _+ S.D.

pret and be useful in a clinical setting. The ratio using copper and zinc is also in a questionable position. Differences between groups with response to treatment and time appear to be related to serum copper levels. The amplifica- tion thought to be provided by serum zinc as reported by Delves et al. (23) and Strain et al. (22) was not noted in the patients with solid tumors. Although links between the two cations has been proposed, the relationship appears to be dietary and not disease oriented (29). There is no proof at present that there is an abnormal accumulation of body copper relating to a decrease in body zinc.

The rat studies with a transplantable solid tumor provide some insight into patterns of shifts in copper and zinc metabolism. Serum copper and zinc was similar to changes seen in humans with solid tumors. Liver copper and zinc levels give some interesting clues to mechanisms of change in the two elements. Of the three hypotheses indicated earlier for the increase in serum copper or ceruloplasmin, the liver values appear to support the cell mediated LEM concept. The decrease in tissue copper and increase in tissue zinc go along

214 A. FLYNN

with the mediator hypothesis (11). The idea that the increase in serum copper owes to increased sialyl transferase does not explain these findings, and there is little data to support the hypothesis involving estrogen involvement.

An interesting observation has recently been reported in a rabbit model in the malignant conversion of Shope's papillomas (30). Gel electrophoretic patterns for ceruloplasmin changed with ulceration of the papilloma with the appearance of a second, more highly mobile band with oxidase activity. Transplanted Vx2 carcinomas in rabbits also demonstrate the double band which may implicate an abnormal ceruloplasmin synthesis with tumor growth. Preliminary studies in our lab (data not shown) have not substan- tiated such a double banding for ceruloplasmin in human serum from cancer patients. The idea certainly deserves further consideration.

Both copper and zinc metabolism with tumor growth need to be explored to greater extents since the systems potentially affected are so vital to the host. The impact of the copper protein ceruloplasmin as a ferroxidase and copper transport molecule are basically unexplored. The relationship of zinc deficits to the immune response has only recently been reported (31, 32). Luecke et al. (31) reported a substantial loss ofT-cell helper function and immune respon- siveness. The same group (32) has recently shown the immune response loss to be responsive to zinc supplementation showing a zinc dependency for that immune function.

One final question is raised by the animal data: Does the suppressed cytochrome c oxidase activity in the liver reflect a copper deficiency ? Mills and Dalgarno (28) conclude that liver cytochrome c oxidase is a good indicator of copper deficit. The idea that a copper deficit is possible during tumor growth raises vital questions on the etiology, metabolism and treatment of solid tumors.

REFERENCES

1. N. W. H. Addink and L. J. P. Frank, Cancer 12, 544 (1959). 2. E. Pagliardi and E. Giangrandi, Acta Haematol. 24, 201 (1960). 3. G. L. Eichhorn, J. J. Butzow, P. Clark, and Y. A. Shin, in Effects of Metals on

Cells, Subcellular Elements and Macromolecules, J. Maniloff, J. R. Coleman, and M. W. Miller, eds., Springfield, I1., Thomas, 1970, pp. 70--84.

4. D. D. Ulmer, in Effects of Metals on Cells, Subcellular Elements and Macro- molecules, J. Maniloff, J. R. Coleman, and M. W. Miller, eds., Springfield, Thomas, 1970, pp. 11-26.

5. H. Rubin, Proc. Nat. Acad. Sci., U.S.A. 69, 712 (1972). 6. D. H. Barr and J. W. Harris, Proc. Sco. Exp. Biol. Med. 144, 284 (1972). 7. W. DeWys, W. J. Pories, M. C. Richter, and W. H. Strain, Proc. Soc. Exp.

Biol. Med. 135, 17 (1970). 8. M. Hrgovcic, C. F. Tessmer, T. M. Minckler, B. Mosier, and G. H. Taylor,

Cancer 21,743 (1968). 9. A. Flynn, in Nutrition and Cancer, M. Seelig, B. N. Nichols, and J. Van Eys,

eds., New York, Spectrum, 1979, in press.


10. E. B. Thorling and K. Thorling, Cancer 38, 225 (1976). 11. P. T. Bailey, F. B. Abeles, E. C. Hauer, and C. A. Mapes, Proc. Soc. Exp.

Biol. Med. 153, 419 (1976). 12. G. L. Fisher and M. Shifrine, Oncology 35, 1 (1978). 13. C. F. Tessmer, M. Hrgovcic, B. W. Porovin, J. Wilbur, and F. B. Thomas,

Cancer 29, 173 (1972). 14. J. Wilimas, E. Thompson, and K. L. Smith, Cancer 42, 1929 (1978). 15. W. Weise, D. Wolansky, and G. Agatha, Excerpta Med. 95, 382 (1975). 16. E. Dennes, R. Tupper, and A. Wormall, Nature 187, 302 (1960). 17. S. Szmigielski and J. Litwin, Cancer 17, 1381 (1964). 18. M. Kew and R. Mallett, Brit. J. Cancer 29, 80 (1974). 19. E. B. Wright and T. L. Normandy, Nature 237, 166 (1972). 20. J. M. Janes, J. T. McCall, and L. R. Elveback, Mayo Cin. Proc. 47, 476 (1972). 21. D. Mikac-Devic, Adv. Clin. Chem. 13, 271 (1970). 22. W. H. Strain, E. G. Mansour, A. Flynn, W. J. Pories, A. J. Tomaro, and O. A.

Hill, Jr., Lancet 1, 1021 (1972). 23. H. T. Delves, F. W. Alexander, and H. Lay, Brit. J. Haematol. 24, 525 (1973). 24. G. L. Fisher, V. S. Byers, M. Shifrine, and A. S. Levin, Cancer 37, 356 (1976). 25. E. Frieden and H. S. Hsieh, Adv. Enzymol. 44, 187 (1976). 26. A. E. Schwartz, G. W. Leddicotte, R. W. Fink, and E. W. Friedman, Surgery

76, 325 (1974). 27. A. Flynn, in Trace Substances in Environmental Health, D. D. Hemphill, ed.,

vol. II, Columbia, Univ. Missouri, 1977, pp. 179-182. 28. C. F. Mills and A. C. Dalgarno, in Trace Element Metabolism in Animals,

C. F. Mills, ed., Edinburgh, Livingstone, 1970, pp. 456-461. 29. A. C. Magee and G. Matrone, J. Nutr. 72, 233 (1960). 30. A. Seto, H. Tokuda, and Y. Ito, Proc. Soc. Exp. Biol. Med. 157, 694 (1978). 31. R. W. Luecke, C. E. Simonel, and P. J. Fraker, J. Nutr. 108, 881 (1978). 32. P. J. Fraker, P. Depasquale-Jardieu, C. M. Zeickl, and R. W. Luecke, Proc.

Natl. Acad. Sci. U.S.A. 75, 5660 (1978).

copper and zinc metabolism with solid tumor growth

Documents