Er. J. Surg. 1991,Vol. 78,April, 483 487
S. D . Heys"?, K. G. M. Park"?, M. A. McNurlan*, E. Milne*, 0. Eremin?, J. Wernermanl, R . A. Keenan? and P. J. Garlick" *Ro wett Research Institute, Bucksburn, Aberdeen, +Department of Surgery, University of Aberdeen, UK and iHuddinge University Hospital, Karolinska Institute, Stockholm. Sweden Correspondence to
Mr S D Heys, Department of Surgery, University Medical Buildings, Foresterhill, Aberdeen AB9 2ZD. UK
Stimulation of protein synthesis in human tumours by parenteral nutrition: evidence for modulation of tumour growth Eighteen patients with localized colorc~mdcarcinoma were randomized to receive intravenous nutrition or to he ,fhsted (luring the 24 h hefi)re surgery. Protein synthesis, an index of tumour growth, was then measured by the incorporation of' 3Cjleucine into tumour protc'in immediately hefbre surgery. The mean(s.e.m.) rate of' tuniour protcin synthesis in patients receiving nutrition (42.7(3.5)per cent per day) MUS 89 per cent higher than the rate in thefasted group (22.6(1.9)pclr cent per day) (P = 0402). A s tumours consist of a variety of difjerent cell types, in vitro rates of' protein synthesis were measured in malignant cells isolated ,from colorectul tumours and cultured with autologous serum obtained,from the patient in either the ,firsred o r the ,fed state. There Mias a mean increase qf81 per cent in protein sj,nthesi.s whc~n,fi.cl rather than ,fusted serum as used (P < O.O.?), iritlicuting that the malignant cells themselves respond to nutrient supplj~.This increcutj in turnour protein synthesis provides the ,first ecidmce in vivo thut thc exogenous supply qf nutrients can modulutr the rcrte of' grow,th of' a human tumour.
['
Nutritional support has been provided to patients with cancer in an attempt to improve nutritional status, but an unsubstantiated concern has been whether stimulation of tumour growth occurs. Studies in experimental animals have addressed this question by assessing the effects of nutrition on tumour volume. weight, mitotic activity and protein synthesis, using these variables as indices of growth. Conflicting results have arisen, with a stimulation of tumour growth' or no effect4 being reported. These studies, however, used rapidly growing transplanted tumours or induced tumours, which constitute a large mass relative to the host animal and may not represent the human situation. Studies to date have failed to detect a stimulation of tumour growth in patients given nutritional support' - I 0 , but reliable clinical assessment of tumour growth rate in humans is dificult and alternative strategies, such as measurement of tumour protein synthesis", need to be used. Measurement of protein synthesis in human tumour tissue has previously been carried out using constant infusions of stable isotopes and radioisotopes'".". Calculation of rates of protein synthesis using this method may lead to errors because of the inability to measure the enrichment of the amino acid at the site of protein synthesis12. An approach widely used to overcome this problem in experimental animals and in cell cultures has been to use the 'flooding dose' technique, where a large amount of unlabelled amino acid is given with the isotope as a bolus injection"^ ". All amino acid pools are rapidly equilibrated to similar isotopic enrichments, thus reducing this potential source of error in precursor pool measurement. We first described this technique to measure protein synthesis in muscle' and colorectal tumour tissue". This has enabled the derivation of an index of tuniour growth rate and a n assessment of the effect of nutrient intake in r i w on t u m o u r growth in patients with localized colorectal carcinoma. Metabolic studies on tumours iri ciuo represent a summation of effects on a number of different cell types which make up the tumour. To determine the specific effects of nutrition on cancer cells we have performed a parallel study in which tumour
'
(M07 I 7 2 7 91 0404x7 (15
(
1991 Butternorth Heinernam Ltd
cells were isolated from colorectal cancers and rates of protein synthesis were measured it2 ciiro.
Patients and methods Pufiet?l.v All patients were admitted to thc Grampian group of hospitals for the investigation and treatment of colorectal carcinoma. Thcy all had localized carcinoma of the rectum with no evidence of metastases at the time of study, and were randomly allocated to either the 'fasted' or the 'fed' group. The study was approved by the Joint Ethical Committee of the University of Aberdeen and the Grampian Health Board, with all patients giving informed written consent.
fr(JlOC(ll A low-residue diet was given to all patients on days 3 and 2 before surgery, together with Picolax" (Nordic Pharmaceuticals Ltd., Feltham, U K ) as a bowel preparation. On the day preceding surgery all oral intake of nutrients was discontinued in both groups. The fasted group was maintained on fluids only, while the fed group was given intravenous nutrition for 20 h. This consisted of amino acids, equivalent to 1.25 g/kg body-weight of protein (Vamin" 9-glucose, Kabi. Bourne End, UK), and energy, 25 kcalikg body-weight, given as 40 per cent dextrose and thc remainder as lipid (Intralipid". Kabi). together with appropriate supplements of vitamins and electrolytes. Measurements of tumour protein synthesis were made at the end ofthis period immediately before surgery. M~wsrrrrrwc~tii (!f proit+ synihcsis r ~ r t cin vivo Measurement of protein synthesis was carried out by injecting a bolus of ~-[I-"C]leucine (4g/70 kg body-weight, 19.4 atom per cent in 200 ml of 0.9 per cent saline, intravenously over 10 min during the period just before surgery". A t 6&90 min after injection of isotope, anaesthesia was induced. A 100--200mg biopsy of tumour was taken endoscopically, immediately after induction of anaesthesia. and frozen in liquid nitrogen. To confirm that the free leucine enrichment in the tumour wab similar to that in plasma over the whole timc-course of the experiment, four fasted patients were injected with leucine after induction of anaesthesia and a tumour biopsy was taken aftcr 10 min, i t . at the termination of the injection. An additional three fasted patients were
483
Protein synthesis in human tumours: S.D. Heys et al.
injected with labelled leucine after anaesthesia had been induced, and a tumour tissue biopsy was taken after 30 min.
The rate of protein synthesis per unit of RNA (kRNA)was calculated from the formula2":
Srrmplc~prrpara/ion Preparation of tumour biopsies was carried out as described previously, and 13C enrichments of free plasma leucine and tumour tissue protein-bound leucine were determined by isotope ratio and gas chromatography mass spectrometry' '.
The RNA and protein contents of the tumour were measured as described previouslyI5.
Cal~~ulo!ion.s Fractional rates of protein synthesis in tumours, expressed as the percentage of protein synthesized per day ( k $ , per cent/day), were calculated from the following formula'5: k,=(P,,,-P,,,)x loo/A
where Pi,, is the enrichment of protein-bound leucine in the tumour biopsy. P,,, is the natural enrichment of leucine, which was obtained by measurement of plasma proteins isolated from the zero time blood sample of each patient'", and A is the area under the curve for plasma free leucine enrichment wrsus time (Fiqure I ).
kRNA= k,,'(RNA/protein)
Measuremenr of protein synthesis in vitro A further six patients with localized colorectal carcinomas were given the same intravenous nutritional regimen for 20 h before surgery. After surgical removal of the tumour, malignant colorectal cancer cells were isolated as previously describedz1.After 7 days of culture in complete medium (RPMI 20 per cent fetal calf serum), the fetal calf serum was replaced by serum from the patient's own blood. This was taken either immediately before intravenous feeding (i.e. after an overnight fast) or after 20 h of intravenous nutrition. After a further 12 h incubation of the malignant cells in uitro, protein synthesis rates were measured by the incorporation of [3H]phenylalanine into cellular protein, as described previouslyz2.
+
Stati,stirs Results are expressed as mean(s.e.m.). The statistical significance between means was assessed by Student's test for the in oiuo study and by paired t test on log-transformed data for the in uitro study.
2or
Results
OL 30
0
60
90
Time (min)
Figure I Plusniu leucine " C enriclmienrs folluwiny labelled leucine. 0 , Fasted; m. /&i
the,
adniinisrrution
of
Table 1
Clinical deails of the patients in whom tumour protein synthesis was measured are listed in Tables I and 2. There were no significant differences between the nutritional state of the two groups as assessed by body mass index and serum concentrations of albumin and retinol binding protein. The time course of the enrichments of free leucine (mean(s.e.m.)) in the plasma of the 18 patients in whom tumour protein synthesis was measured showed maximum levels of enrichment immediately after the leucine infusion had finished, followed by a linear fall in each subject (Figure I ) . The free leucine enrichment in the tumour biopsies, which were taken from each patient between 45 and 90min after the start of
Clinical details of fasted parienrs ~
Dukes' classification
Patient ~
~
80 60 74 76 73 62 68 73 71
1
2 3 4
5 6 7 8 9
~~
Histological differentiation
-
B B B C B C B C C
Serum albumin
(SiU
~
Moderate Moderate Moderate Moderate Moderate Well Moderate Poor Moderate
40 42 40 36 39 40 43 36 34
~
Retinol binding protein (gill
-
~~~
0 06 0 03
~
~
Body mass index (kg/m2) ~
_
_
Weight loss (kg/6 months)
~
21
~
0 06 0 04
23 21 24
~
0 06 0 09 0 07 0 06 0 09
18 21 21 18 22
-
Body mass index (kg/m2)
Weight loss (kg/6 months)
21 25 24 24 21 19 21 21 22
3
8 -
-
Table 2 Clinical details ofparients rewiring inrravenous nutrilion
Patient -~
Age (years)
Dukes' classification
Histological differentiation
Serum albumiii
(sill
Retinol binding protein (g/l)
Moderate Moderate Moderate Moderate Moderate Moderate Moderate Well Poor
33 44 46 30 38 42 29 40 40
0.06 0.06 0.05 0.09 0.07 0.05 0.05 0.06 0.05
~~~
~
10
II 12 13 14 15
16 17 18
484
68 71 70 14 74 80 72 67 74
~
~
~
-
7 ~
~
Br. J. S u r g . , Vol. 78, No. 4,April 1991
Protein synthesis in human tumours: S. D. Heys et al.
a
a a a a
a
a
-
Ta a
a a
a a
0
8
:
-L
%$ a
a
-Fasted
4
3
.-aC
” ?
a
F .
2
Ea
TI
1
0
-Fasted
Fed
Figure 3 Rules ofprotein .synthesis in vitro in colorecru1 turnour cells incubated ,i,i/h uu~okogousserum ,from the patient taken either when /astiny or ufier 24 /I o?fpurenteralnutrition. dpm, disinteyralionsper min.
injection, was compared with that of plasma at the same time. The ratio of tumour to plasma leucine enrichment was 0.90(0.02)in the fasted group and 0.82(0.01)in the fed group (mean(s.e.m.)).The enrichment of free leucine in the tumour and plasma at two earlier time points after administration of isotope to a separate group of patients demonstrated that the enrichment in the tumour was 0.97(0.03) (n=4) of the corresponding plasma value of 10 min and 0.93(0.03)(n=3) 30 min after the start of the injection. Thus flooding, i.e. different amino acid pools reaching similar enrichments, had occurred. The provision of intravenous nutrition resulted in a significant stimulation of protein synthesis in uiuo (Figure 2a), from 22.6(1-9) per cent per day (mean(s.e.m.)) in the fasted state The to 42.7(3.5)per cent per day in the fed group (P=0.002). rate of synthesis expressed per g R N A (kRNA)was also
Br. J. Surg., Vol. 78, No. 4, April 1991
u u
Fed
Fasted
Fed
significantly increased, from 4.8(0.6)in the fasted group to 10.9(1.0)g proteinlday per g R N A in the group receiving intravenous nutrition (Figure 2h). The R N A concentration in the tumours was not, however, significantly altered by feeding (data not shown). Therefore, if the total R N A can be taken as indicative of ribosomal content of the tissue”, the increase in protein synthesis resulted from a change in the activity of ribosomes rather than an alteration in the total number of ribosomes. The measurements documented above were on total tumour protein, but tumours are well recognized for the heterogeneity of their different constituent cell typesz3. To confirm that malignant cells respond to nutrient intake, colonic carcinoma cells isolated from resected tumours were studied in uitro. The results are shown in Fi gure 3 and demonstrate a significant increase in incorporation of labelled phenylalanine into protein when cells from the tumours of fed patients were incubated with the patient’s own fasting serum compared with ‘fed’ serum
(P< 0.02).
Discussion This study has demonstrated that protein synthesis in colorectal tumours was substantially higher in patients receiving parenteral nutrition compared with those who were fasted. This provides evidence that human tumour growth is susceptible to modulation by nutrient intake. In the present investigation the effect of short-term intravenous feeding (20h ) was studied in patients who were nil by mouth for the day preceding surgery. There had been no other dietary restrictions except the provision of a low-residue diet to both groups for 48 h. The groups were matched for nutritional state and weight loss. Several previous studies have attempted to assess the effects of long-term nutritional supplementation in humans, but have failed to detect any effect on clinical or biochemical markers of increased tumour Determination of the rate of protein synthesis provides an alternative measure of tumour growth and has been used as an in uiuo index in experimental animals6. It has been demonstrated previously that, in human malignant and transformed cells in culture, growth of protein mass is effected by an increase in protein synthesis alone, or an increase in synthesis together with either a decrease or an increase in protein d e g r a d a t i ~ . nProtein ~ ~ ~ synthesis ~ ~ ~ ~ ~is therefore the primary determinant of growth. Unfortunately measurement
485
Protein synthesis in human tumours: S. D . Heys e t al.
of tumour protein breakdown in uivo in humans remains impossible. In the present study, breakdown must have been increased, because the increase in tumour growth must have been small in relation to the 20 per cent per day increase in protein synthesis. T h e average doubling time of human solid tumours has been estimated to be approximately 60 days, which is equivalent t o a fractional growth rate of only approximately 1 per cent per day2". Measurement of the rate of protein synthesis in human tumours iti s i t u by reliable a n d practical methods has not previously been possible. The present technique for measuring protein synthesis, using a stable isotopically labelled a m i n o acid given as a large o r flooding dose, minimizes the errors resulting from precursor compartmentation encountered with other techniquesI 3.1 7.18,27 . T h e single-dose procedure is more readily performed during the hour preceding surgery than the lengthy intravenous infusions used previously'".' I . Several studies suggest that leucine might stimulate protein synthesis in isolated muscle of growing animals. but thcre is n o evidence that the large dose of leucine injected in vioo used in this study might alter protein synthesis in either muscle" or t u m 0 u r Z X . The in ritw study demonstrating that serum taken from patients i n the fed state stimulated tumour cell protein synthesis confirmed that thc malignant cells are able to respond to nutrient intake. These results contradict the concept of the tumour a s a growing parasite feeding off the host tissues regardless of the exogenous supply of nutrients. Indeed. it was recently stated that 'malignant tissue is synthesizing protein at a maximal rate and that its rate of growth cannot be affected by the provision of extra nutrients'"'. This statement was based on the study of Mullen ot ul.'". who reported that tumour protein synthesis was not significantly different in patients who either ate normally o r were given intravenous supplementation, but the experimental groups were not matched a n d contained different spectra of malignant tumours. Emery c/ LII." reported no effect of supplementary feeding o n colorectal tumour protein synthesis in rats, but the relevance to humans of this finding is unclear. T h e measurements carried out on tumour cells in culture showed that the mediators of this substantial and rapid increase in tumour protein synthesis a r e present in the plasma from fed patients. but their nature is not clear. In normal tissues, the main mediators of the response of protein synthesis to food intake are insulin a n d amino acids3(),a n d insulin is known to stimulatc growth of experimental tumours3 . A number of other growth factors (e.g. epidermal growth factor'") have been shown t o stimulate the growth of tumour cell lines. a n d further studics arc needed to elucidate the various possible autocrineparacrine mechanisms responsible for stimulation. Although nutritional support is frequently provided t o patients with malignancy to reverse weight loss a n d improve nutritional status, these results suggest that feeding might stimulate tumour growth, which might be detrimental t o patient survival. Tumour stimulation by nutrient intake may be used to advantage, as experimental animal studies have shown that intravenous nutrition increases the sensitivity of experimental tumours to chemotherapy3' ". This study using stable isotopes to measure protein synthesis iti riro offers a n opportunity to investigate the various factors that modulate growth of human t timours.
'
References I.
2. 3. 4. 5.
6.
7. 8.
9.
10.
11. 12.
13. 14. 15.
16. 17.
Steiger E. Oram-Smith J, Miller E. Kuo L. Vons HM. Effects of nutrition on tumor growjth and tolerance to chemotherapy. .I Simq Rrs 1975; 18: 455 61. Cameron JL, Pavlaf WA. Stimulation of growth of B transplantable hepatoma in rats by parentcral nutrition. J Nrril Comer / n . ~ 1976; / 36: 597-601, Popp MB, Wagner SC, Brito OJ. Host and tumor responses to increasing levels of nutritional support. S r r r c p r j . 1983;94: 300--8. Kishi T, Iwasawa Y, Hiroshi I. Chibata I . Nutritional responses of tumor-bearing rats to oral or intravenous feeding. J Prrrcwter Entcrd Nutr 1982; 6 : 295-300. King WWK, Boelhouwcr RLJ, Kingsnorth A N ef a/. Total parenteral nutrition with and without Tat as substrate for growth of rat$ and transplanted hcpatocarcinoma. J Prrroiio. € n / r , r d Nutr 1985; 9: 422-7. Emery PW. Ward MWN, Lcwin MR. Effect or nutritional support on protein synthesis in tumour and host tissues of rats with colonic cancer. BI .I Surg 1989; 76: 790-2. Nixon DW, Moffitt S, Lawson DH P I r r l . Total parenteral nutrition as an adjunct to chemotherapy of metastatic colorectal cancer. Cmicev Trmr R i y 1981; 65: 121-8. Ota D. Nishoka K. Foulkes M, Grossie B. Nutritional parameters affecting erythrocyte polyamine levels in cancer patients. J Clin O n d 15'84; 2: 1157--64. Shamberger RC, Brennan MF. Goodgame JT C I rrl. A prospective, randomized study of adjuvant parentcral nutrition in the treatment of sarcomas: results of metabolic and survival studies. Siirgerj. 1984; 96: 1-12. Mullen JL, Buzhy G , Gertner MH o t ol. Protein synthesis dynamics in human gastrointestinal malignancies. Suryc,rj, 1980: 87: 331-8. Shaw JHF, Humherstone DA. Canccr: ii metabolic parasite. Br J SU,.{/ 1988: 7 5 : 1262. Waterlow JC, Garlick PJ. M111w.ard D. Proroin Twriowr in Munimuliuri Tr.s.tur., mid iri tlw M'liolc Borlv. Amsterdam: North-Holland. 1978. Henshaw EC, Hirsch CA, Morton BE, Hiatt HH. Control of protein synthesis in mammalian tissues through changcs in ribosome activity. J R i d Ch077 1971; 246: 436-46. Scornik OA. I n ritw rate of translation by ribosomes of normal m d regenerating liver. J Biol C h i 1974; 24: 3876--83. McNurlan MA. Tomkins AM. Garlick PJ. The effect of starvation on (he rate of protein synthesis in rat liver and small intestine. Bioc,/icni J 1979: 178: 373-9. Garlick PJ, McNurlan MA. Preedy V R . A rapid and convenient technique for measuring the rate of protein synthesis in tissues by injection of[3H]phenylalanine. Bioc.lioni J 1980; 192: 719-23. Garlick PJ, Wernerman J. McNurlan MA c't a / . Measurement of the rate of protein synthesis in muscle of postabsorptive young men by injection of 'flooding dose' of [I~"C]leucinc. C ' h r S;i 1989: 77: 329- 36. Heys SD, Keenan R A . Wernerman J ('t ul. The determination of tumour protein synthesis rates in patients with colorectal carcinoma using the 'flooding dose' technique. /'roc, Nrr/r SOC 1989; 48: 101A. Hcys SD. McNurlan MA, Park KGM. Milnc E, Garlick PJ. Baseline measurements for stable isotope studies: an alternative to biopsy. Bionrc~dEncir017 Mass S p c ~1990; 19: I76--8. Millward DJ, Garlick PJ, James WPT. Nnanyelugo DO, Ryatt JS. Relationship between protein synthesis and R N A content in skeletal muscle. Nrrtirrr 1973; 241: 2 0 4 ~ 5 . Kuppner M . Wilkinson S, Casson E. Eremin 0 . It7 iiiro generation of tumour-specific reactivity to colonic carcinoma cclls: comparison with normal colonic cells. Crrricw /nin7mi(~/ Imniunotlirr 1987; 25: 209- 14. McNurlan MA. Clemens MJ. Inhibition of cell proliferation by interferons. Bitx$irni J 1986; 237: 871-6. Heppncr G H . Tumor heterogcncity. Crrncw Re\ 1984: 44:
a
18.
19. 20. 21.
22. 23.
2259-65.
Acknowledgements We are grateful to Mr A . G . Caldcr for mass spectrometric analysis. to Mrs V. Buchan for amino acid analyses. to Miss I. Morrison and Dr J . Broom for provision of parentcral nutrition and to the Department o f Clincal Biochemistry for serum albumin and RBP mt'nsurements. We thank the Wellcome Trust. Grampian Health Board. Scottish Hospital Endowment Rcscarch Trust and Nestec Ltd. for their financial support.
486
24. 25.
26.
Baccino FM, Tessitorc L, Bonelli G. Control of protein degradation and growth phase in normal and neoplastic cells. Tosicol Parhol 1984; 12: 281 7. Lockwood TD. Minassian IA. Protein turnover and proliferation: failure of SV-3T3 cells to incrcasc lysosomal proteases. increase protein degradation and cease net protein accumulation. Biochmr J 1982; 206: 251-8. Killion JJ, Fidler IJ. The biology of tumor metastasis. S c r i i i n Omol 1985'; 16: 106-15.
Br. J . S u r g . , Vol. 78, No. 4, April 1991
Protein synthesis in human turnours:
27.
28. 29.
30. 31.
Martin AF. Rabinowitz M, Blough R, Prior G, Zak R . Measurements of half-life of rat cardiac myosin heavy chain with leucyl-tRNA used as precursor pool. J B i d Clicrn 1977; 252: 3422-9. Ballard FJ.Regulation of protein accumulation in cultured cells. B ~ ( J ~ uJw 1982; ? 208: 275-87. Douglas RG, Shaw JHF. Metabolic effects of cancer. Er J Sury 1990; 77: 246-54. McNurlan MA, Garlick PJ. Influences of nutrient intake on protein turnover. Di&re.v Metah Re[: 1989; 5: 165-89. Heuson JC, Coune A, Heimann R. Cell proliferation induced by insulin in organ cultured of rat mammary carcinoma. E . Y ~Cell
Br. J . Surg., Vol. 78,No. 4,April 1991
32.
33.
S.D. Heys et al
Re.7 1967; 45: 351-60. Torosian M H , Mullen JL, Stein T P , Miller EE, Zinsseer KR, Buzby G P . Enhanced tumor response to cycle-specific chemotherapy by pulse total parenteral nutrition. J Sury Rrs 1985;39: 103-13. Torosian M H , Mullen JL, Miller EE, Wagner KM, Stein T P , Buzby G P . Adjuvant, pulse total parenteral nutrition and tumor response to cycle-specific and cycle-nonspecific chemotherapy. Suryert. 1983; 94: 291-8.
Paper accepted 2 October 1990
487
S. D . Heys"?, K. G. M. Park"?, M. A. McNurlan*, E. Milne*, 0. Eremin?, J. Wernermanl, R . A. Keenan? and P. J. Garlick" *Ro wett Research Institute, Bucksburn, Aberdeen, +Department of Surgery, University of Aberdeen, UK and iHuddinge University Hospital, Karolinska Institute, Stockholm. Sweden Correspondence to
Mr S D Heys, Department of Surgery, University Medical Buildings, Foresterhill, Aberdeen AB9 2ZD. UK
Stimulation of protein synthesis in human tumours by parenteral nutrition: evidence for modulation of tumour growth Eighteen patients with localized colorc~mdcarcinoma were randomized to receive intravenous nutrition or to he ,fhsted (luring the 24 h hefi)re surgery. Protein synthesis, an index of tumour growth, was then measured by the incorporation of' 3Cjleucine into tumour protc'in immediately hefbre surgery. The mean(s.e.m.) rate of' tuniour protcin synthesis in patients receiving nutrition (42.7(3.5)per cent per day) MUS 89 per cent higher than the rate in thefasted group (22.6(1.9)pclr cent per day) (P = 0402). A s tumours consist of a variety of difjerent cell types, in vitro rates of' protein synthesis were measured in malignant cells isolated ,from colorectul tumours and cultured with autologous serum obtained,from the patient in either the ,firsred o r the ,fed state. There Mias a mean increase qf81 per cent in protein sj,nthesi.s whc~n,fi.cl rather than ,fusted serum as used (P < O.O.?), iritlicuting that the malignant cells themselves respond to nutrient supplj~.This increcutj in turnour protein synthesis provides the ,first ecidmce in vivo thut thc exogenous supply qf nutrients can modulutr the rcrte of' grow,th of' a human tumour.
['
Nutritional support has been provided to patients with cancer in an attempt to improve nutritional status, but an unsubstantiated concern has been whether stimulation of tumour growth occurs. Studies in experimental animals have addressed this question by assessing the effects of nutrition on tumour volume. weight, mitotic activity and protein synthesis, using these variables as indices of growth. Conflicting results have arisen, with a stimulation of tumour growth' or no effect4 being reported. These studies, however, used rapidly growing transplanted tumours or induced tumours, which constitute a large mass relative to the host animal and may not represent the human situation. Studies to date have failed to detect a stimulation of tumour growth in patients given nutritional support' - I 0 , but reliable clinical assessment of tumour growth rate in humans is dificult and alternative strategies, such as measurement of tumour protein synthesis", need to be used. Measurement of protein synthesis in human tumour tissue has previously been carried out using constant infusions of stable isotopes and radioisotopes'".". Calculation of rates of protein synthesis using this method may lead to errors because of the inability to measure the enrichment of the amino acid at the site of protein synthesis12. An approach widely used to overcome this problem in experimental animals and in cell cultures has been to use the 'flooding dose' technique, where a large amount of unlabelled amino acid is given with the isotope as a bolus injection"^ ". All amino acid pools are rapidly equilibrated to similar isotopic enrichments, thus reducing this potential source of error in precursor pool measurement. We first described this technique to measure protein synthesis in muscle' and colorectal tumour tissue". This has enabled the derivation of an index of tuniour growth rate and a n assessment of the effect of nutrient intake in r i w on t u m o u r growth in patients with localized colorectal carcinoma. Metabolic studies on tumours iri ciuo represent a summation of effects on a number of different cell types which make up the tumour. To determine the specific effects of nutrition on cancer cells we have performed a parallel study in which tumour
'
(M07 I 7 2 7 91 0404x7 (15
(
1991 Butternorth Heinernam Ltd
cells were isolated from colorectal cancers and rates of protein synthesis were measured it2 ciiro.
Patients and methods Pufiet?l.v All patients were admitted to thc Grampian group of hospitals for the investigation and treatment of colorectal carcinoma. Thcy all had localized carcinoma of the rectum with no evidence of metastases at the time of study, and were randomly allocated to either the 'fasted' or the 'fed' group. The study was approved by the Joint Ethical Committee of the University of Aberdeen and the Grampian Health Board, with all patients giving informed written consent.
fr(JlOC(ll A low-residue diet was given to all patients on days 3 and 2 before surgery, together with Picolax" (Nordic Pharmaceuticals Ltd., Feltham, U K ) as a bowel preparation. On the day preceding surgery all oral intake of nutrients was discontinued in both groups. The fasted group was maintained on fluids only, while the fed group was given intravenous nutrition for 20 h. This consisted of amino acids, equivalent to 1.25 g/kg body-weight of protein (Vamin" 9-glucose, Kabi. Bourne End, UK), and energy, 25 kcalikg body-weight, given as 40 per cent dextrose and thc remainder as lipid (Intralipid". Kabi). together with appropriate supplements of vitamins and electrolytes. Measurements of tumour protein synthesis were made at the end ofthis period immediately before surgery. M~wsrrrrrwc~tii (!f proit+ synihcsis r ~ r t cin vivo Measurement of protein synthesis was carried out by injecting a bolus of ~-[I-"C]leucine (4g/70 kg body-weight, 19.4 atom per cent in 200 ml of 0.9 per cent saline, intravenously over 10 min during the period just before surgery". A t 6&90 min after injection of isotope, anaesthesia was induced. A 100--200mg biopsy of tumour was taken endoscopically, immediately after induction of anaesthesia. and frozen in liquid nitrogen. To confirm that the free leucine enrichment in the tumour wab similar to that in plasma over the whole timc-course of the experiment, four fasted patients were injected with leucine after induction of anaesthesia and a tumour biopsy was taken aftcr 10 min, i t . at the termination of the injection. An additional three fasted patients were
483
Protein synthesis in human tumours: S.D. Heys et al.
injected with labelled leucine after anaesthesia had been induced, and a tumour tissue biopsy was taken after 30 min.
The rate of protein synthesis per unit of RNA (kRNA)was calculated from the formula2":
Srrmplc~prrpara/ion Preparation of tumour biopsies was carried out as described previously, and 13C enrichments of free plasma leucine and tumour tissue protein-bound leucine were determined by isotope ratio and gas chromatography mass spectrometry' '.
The RNA and protein contents of the tumour were measured as described previouslyI5.
Cal~~ulo!ion.s Fractional rates of protein synthesis in tumours, expressed as the percentage of protein synthesized per day ( k $ , per cent/day), were calculated from the following formula'5: k,=(P,,,-P,,,)x loo/A
where Pi,, is the enrichment of protein-bound leucine in the tumour biopsy. P,,, is the natural enrichment of leucine, which was obtained by measurement of plasma proteins isolated from the zero time blood sample of each patient'", and A is the area under the curve for plasma free leucine enrichment wrsus time (Fiqure I ).
kRNA= k,,'(RNA/protein)
Measuremenr of protein synthesis in vitro A further six patients with localized colorectal carcinomas were given the same intravenous nutritional regimen for 20 h before surgery. After surgical removal of the tumour, malignant colorectal cancer cells were isolated as previously describedz1.After 7 days of culture in complete medium (RPMI 20 per cent fetal calf serum), the fetal calf serum was replaced by serum from the patient's own blood. This was taken either immediately before intravenous feeding (i.e. after an overnight fast) or after 20 h of intravenous nutrition. After a further 12 h incubation of the malignant cells in uitro, protein synthesis rates were measured by the incorporation of [3H]phenylalanine into cellular protein, as described previouslyz2.
+
Stati,stirs Results are expressed as mean(s.e.m.). The statistical significance between means was assessed by Student's test for the in oiuo study and by paired t test on log-transformed data for the in uitro study.
2or
Results
OL 30
0
60
90
Time (min)
Figure I Plusniu leucine " C enriclmienrs folluwiny labelled leucine. 0 , Fasted; m. /&i
the,
adniinisrrution
of
Table 1
Clinical deails of the patients in whom tumour protein synthesis was measured are listed in Tables I and 2. There were no significant differences between the nutritional state of the two groups as assessed by body mass index and serum concentrations of albumin and retinol binding protein. The time course of the enrichments of free leucine (mean(s.e.m.)) in the plasma of the 18 patients in whom tumour protein synthesis was measured showed maximum levels of enrichment immediately after the leucine infusion had finished, followed by a linear fall in each subject (Figure I ) . The free leucine enrichment in the tumour biopsies, which were taken from each patient between 45 and 90min after the start of
Clinical details of fasted parienrs ~
Dukes' classification
Patient ~
~
80 60 74 76 73 62 68 73 71
1
2 3 4
5 6 7 8 9
~~
Histological differentiation
-
B B B C B C B C C
Serum albumin
(SiU
~
Moderate Moderate Moderate Moderate Moderate Well Moderate Poor Moderate
40 42 40 36 39 40 43 36 34
~
Retinol binding protein (gill
-
~~~
0 06 0 03
~
~
Body mass index (kg/m2) ~
_
_
Weight loss (kg/6 months)
~
21
~
0 06 0 04
23 21 24
~
0 06 0 09 0 07 0 06 0 09
18 21 21 18 22
-
Body mass index (kg/m2)
Weight loss (kg/6 months)
21 25 24 24 21 19 21 21 22
3
8 -
-
Table 2 Clinical details ofparients rewiring inrravenous nutrilion
Patient -~
Age (years)
Dukes' classification
Histological differentiation
Serum albumiii
(sill
Retinol binding protein (g/l)
Moderate Moderate Moderate Moderate Moderate Moderate Moderate Well Poor
33 44 46 30 38 42 29 40 40
0.06 0.06 0.05 0.09 0.07 0.05 0.05 0.06 0.05
~~~
~
10
II 12 13 14 15
16 17 18
484
68 71 70 14 74 80 72 67 74
~
~
~
-
7 ~
~
Br. J. S u r g . , Vol. 78, No. 4,April 1991
Protein synthesis in human tumours: S. D. Heys et al.
a
a a a a
a
a
-
Ta a
a a
a a
0
8
:
-L
%$ a
a
-Fasted
4
3
.-aC
” ?
a
F .
2
Ea
TI
1
0
-Fasted
Fed
Figure 3 Rules ofprotein .synthesis in vitro in colorecru1 turnour cells incubated ,i,i/h uu~okogousserum ,from the patient taken either when /astiny or ufier 24 /I o?fpurenteralnutrition. dpm, disinteyralionsper min.
injection, was compared with that of plasma at the same time. The ratio of tumour to plasma leucine enrichment was 0.90(0.02)in the fasted group and 0.82(0.01)in the fed group (mean(s.e.m.)).The enrichment of free leucine in the tumour and plasma at two earlier time points after administration of isotope to a separate group of patients demonstrated that the enrichment in the tumour was 0.97(0.03) (n=4) of the corresponding plasma value of 10 min and 0.93(0.03)(n=3) 30 min after the start of the injection. Thus flooding, i.e. different amino acid pools reaching similar enrichments, had occurred. The provision of intravenous nutrition resulted in a significant stimulation of protein synthesis in uiuo (Figure 2a), from 22.6(1-9) per cent per day (mean(s.e.m.)) in the fasted state The to 42.7(3.5)per cent per day in the fed group (P=0.002). rate of synthesis expressed per g R N A (kRNA)was also
Br. J. Surg., Vol. 78, No. 4, April 1991
u u
Fed
Fasted
Fed
significantly increased, from 4.8(0.6)in the fasted group to 10.9(1.0)g proteinlday per g R N A in the group receiving intravenous nutrition (Figure 2h). The R N A concentration in the tumours was not, however, significantly altered by feeding (data not shown). Therefore, if the total R N A can be taken as indicative of ribosomal content of the tissue”, the increase in protein synthesis resulted from a change in the activity of ribosomes rather than an alteration in the total number of ribosomes. The measurements documented above were on total tumour protein, but tumours are well recognized for the heterogeneity of their different constituent cell typesz3. To confirm that malignant cells respond to nutrient intake, colonic carcinoma cells isolated from resected tumours were studied in uitro. The results are shown in Fi gure 3 and demonstrate a significant increase in incorporation of labelled phenylalanine into protein when cells from the tumours of fed patients were incubated with the patient’s own fasting serum compared with ‘fed’ serum
(P< 0.02).
Discussion This study has demonstrated that protein synthesis in colorectal tumours was substantially higher in patients receiving parenteral nutrition compared with those who were fasted. This provides evidence that human tumour growth is susceptible to modulation by nutrient intake. In the present investigation the effect of short-term intravenous feeding (20h ) was studied in patients who were nil by mouth for the day preceding surgery. There had been no other dietary restrictions except the provision of a low-residue diet to both groups for 48 h. The groups were matched for nutritional state and weight loss. Several previous studies have attempted to assess the effects of long-term nutritional supplementation in humans, but have failed to detect any effect on clinical or biochemical markers of increased tumour Determination of the rate of protein synthesis provides an alternative measure of tumour growth and has been used as an in uiuo index in experimental animals6. It has been demonstrated previously that, in human malignant and transformed cells in culture, growth of protein mass is effected by an increase in protein synthesis alone, or an increase in synthesis together with either a decrease or an increase in protein d e g r a d a t i ~ . nProtein ~ ~ ~ synthesis ~ ~ ~ ~ ~is therefore the primary determinant of growth. Unfortunately measurement
485
Protein synthesis in human tumours: S. D . Heys e t al.
of tumour protein breakdown in uivo in humans remains impossible. In the present study, breakdown must have been increased, because the increase in tumour growth must have been small in relation to the 20 per cent per day increase in protein synthesis. T h e average doubling time of human solid tumours has been estimated to be approximately 60 days, which is equivalent t o a fractional growth rate of only approximately 1 per cent per day2". Measurement of the rate of protein synthesis in human tumours iti s i t u by reliable a n d practical methods has not previously been possible. The present technique for measuring protein synthesis, using a stable isotopically labelled a m i n o acid given as a large o r flooding dose, minimizes the errors resulting from precursor compartmentation encountered with other techniquesI 3.1 7.18,27 . T h e single-dose procedure is more readily performed during the hour preceding surgery than the lengthy intravenous infusions used previously'".' I . Several studies suggest that leucine might stimulate protein synthesis in isolated muscle of growing animals. but thcre is n o evidence that the large dose of leucine injected in vioo used in this study might alter protein synthesis in either muscle" or t u m 0 u r Z X . The in ritw study demonstrating that serum taken from patients i n the fed state stimulated tumour cell protein synthesis confirmed that thc malignant cells are able to respond to nutrient intake. These results contradict the concept of the tumour a s a growing parasite feeding off the host tissues regardless of the exogenous supply of nutrients. Indeed. it was recently stated that 'malignant tissue is synthesizing protein at a maximal rate and that its rate of growth cannot be affected by the provision of extra nutrients'"'. This statement was based on the study of Mullen ot ul.'". who reported that tumour protein synthesis was not significantly different in patients who either ate normally o r were given intravenous supplementation, but the experimental groups were not matched a n d contained different spectra of malignant tumours. Emery c/ LII." reported no effect of supplementary feeding o n colorectal tumour protein synthesis in rats, but the relevance to humans of this finding is unclear. T h e measurements carried out on tumour cells in culture showed that the mediators of this substantial and rapid increase in tumour protein synthesis a r e present in the plasma from fed patients. but their nature is not clear. In normal tissues, the main mediators of the response of protein synthesis to food intake are insulin a n d amino acids3(),a n d insulin is known to stimulatc growth of experimental tumours3 . A number of other growth factors (e.g. epidermal growth factor'") have been shown t o stimulate the growth of tumour cell lines. a n d further studics arc needed to elucidate the various possible autocrineparacrine mechanisms responsible for stimulation. Although nutritional support is frequently provided t o patients with malignancy to reverse weight loss a n d improve nutritional status, these results suggest that feeding might stimulate tumour growth, which might be detrimental t o patient survival. Tumour stimulation by nutrient intake may be used to advantage, as experimental animal studies have shown that intravenous nutrition increases the sensitivity of experimental tumours to chemotherapy3' ". This study using stable isotopes to measure protein synthesis iti riro offers a n opportunity to investigate the various factors that modulate growth of human t timours.
'
References I.
2. 3. 4. 5.
6.
7. 8.
9.
10.
11. 12.
13. 14. 15.
16. 17.
Steiger E. Oram-Smith J, Miller E. Kuo L. Vons HM. Effects of nutrition on tumor growjth and tolerance to chemotherapy. .I Simq Rrs 1975; 18: 455 61. Cameron JL, Pavlaf WA. Stimulation of growth of B transplantable hepatoma in rats by parentcral nutrition. J Nrril Comer / n . ~ 1976; / 36: 597-601, Popp MB, Wagner SC, Brito OJ. Host and tumor responses to increasing levels of nutritional support. S r r r c p r j . 1983;94: 300--8. Kishi T, Iwasawa Y, Hiroshi I. Chibata I . Nutritional responses of tumor-bearing rats to oral or intravenous feeding. J Prrrcwter Entcrd Nutr 1982; 6 : 295-300. King WWK, Boelhouwcr RLJ, Kingsnorth A N ef a/. Total parenteral nutrition with and without Tat as substrate for growth of rat$ and transplanted hcpatocarcinoma. J Prrroiio. € n / r , r d Nutr 1985; 9: 422-7. Emery PW. Ward MWN, Lcwin MR. Effect or nutritional support on protein synthesis in tumour and host tissues of rats with colonic cancer. BI .I Surg 1989; 76: 790-2. Nixon DW, Moffitt S, Lawson DH P I r r l . Total parenteral nutrition as an adjunct to chemotherapy of metastatic colorectal cancer. Cmicev Trmr R i y 1981; 65: 121-8. Ota D. Nishoka K. Foulkes M, Grossie B. Nutritional parameters affecting erythrocyte polyamine levels in cancer patients. J Clin O n d 15'84; 2: 1157--64. Shamberger RC, Brennan MF. Goodgame JT C I rrl. A prospective, randomized study of adjuvant parentcral nutrition in the treatment of sarcomas: results of metabolic and survival studies. Siirgerj. 1984; 96: 1-12. Mullen JL, Buzhy G , Gertner MH o t ol. Protein synthesis dynamics in human gastrointestinal malignancies. Suryc,rj, 1980: 87: 331-8. Shaw JHF, Humherstone DA. Canccr: ii metabolic parasite. Br J SU,.{/ 1988: 7 5 : 1262. Waterlow JC, Garlick PJ. M111w.ard D. Proroin Twriowr in Munimuliuri Tr.s.tur., mid iri tlw M'liolc Borlv. Amsterdam: North-Holland. 1978. Henshaw EC, Hirsch CA, Morton BE, Hiatt HH. Control of protein synthesis in mammalian tissues through changcs in ribosome activity. J R i d Ch077 1971; 246: 436-46. Scornik OA. I n ritw rate of translation by ribosomes of normal m d regenerating liver. J Biol C h i 1974; 24: 3876--83. McNurlan MA. Tomkins AM. Garlick PJ. The effect of starvation on (he rate of protein synthesis in rat liver and small intestine. Bioc,/icni J 1979: 178: 373-9. Garlick PJ, McNurlan MA. Preedy V R . A rapid and convenient technique for measuring the rate of protein synthesis in tissues by injection of[3H]phenylalanine. Bioc.lioni J 1980; 192: 719-23. Garlick PJ, Wernerman J. McNurlan MA c't a / . Measurement of the rate of protein synthesis in muscle of postabsorptive young men by injection of 'flooding dose' of [I~"C]leucinc. C ' h r S;i 1989: 77: 329- 36. Heys SD, Keenan R A . Wernerman J ('t ul. The determination of tumour protein synthesis rates in patients with colorectal carcinoma using the 'flooding dose' technique. /'roc, Nrr/r SOC 1989; 48: 101A. Hcys SD. McNurlan MA, Park KGM. Milnc E, Garlick PJ. Baseline measurements for stable isotope studies: an alternative to biopsy. Bionrc~dEncir017 Mass S p c ~1990; 19: I76--8. Millward DJ, Garlick PJ, James WPT. Nnanyelugo DO, Ryatt JS. Relationship between protein synthesis and R N A content in skeletal muscle. Nrrtirrr 1973; 241: 2 0 4 ~ 5 . Kuppner M . Wilkinson S, Casson E. Eremin 0 . It7 iiiro generation of tumour-specific reactivity to colonic carcinoma cclls: comparison with normal colonic cells. Crrricw /nin7mi(~/ Imniunotlirr 1987; 25: 209- 14. McNurlan MA. Clemens MJ. Inhibition of cell proliferation by interferons. Bitx$irni J 1986; 237: 871-6. Heppncr G H . Tumor heterogcncity. Crrncw Re\ 1984: 44:
a
18.
19. 20. 21.
22. 23.
2259-65.
Acknowledgements We are grateful to Mr A . G . Caldcr for mass spectrometric analysis. to Mrs V. Buchan for amino acid analyses. to Miss I. Morrison and Dr J . Broom for provision of parentcral nutrition and to the Department o f Clincal Biochemistry for serum albumin and RBP mt'nsurements. We thank the Wellcome Trust. Grampian Health Board. Scottish Hospital Endowment Rcscarch Trust and Nestec Ltd. for their financial support.
486
24. 25.
26.
Baccino FM, Tessitorc L, Bonelli G. Control of protein degradation and growth phase in normal and neoplastic cells. Tosicol Parhol 1984; 12: 281 7. Lockwood TD. Minassian IA. Protein turnover and proliferation: failure of SV-3T3 cells to incrcasc lysosomal proteases. increase protein degradation and cease net protein accumulation. Biochmr J 1982; 206: 251-8. Killion JJ, Fidler IJ. The biology of tumor metastasis. S c r i i i n Omol 1985'; 16: 106-15.
Br. J . S u r g . , Vol. 78, No. 4, April 1991
Protein synthesis in human turnours:
27.
28. 29.
30. 31.
Martin AF. Rabinowitz M, Blough R, Prior G, Zak R . Measurements of half-life of rat cardiac myosin heavy chain with leucyl-tRNA used as precursor pool. J B i d Clicrn 1977; 252: 3422-9. Ballard FJ.Regulation of protein accumulation in cultured cells. B ~ ( J ~ uJw 1982; ? 208: 275-87. Douglas RG, Shaw JHF. Metabolic effects of cancer. Er J Sury 1990; 77: 246-54. McNurlan MA, Garlick PJ. Influences of nutrient intake on protein turnover. Di&re.v Metah Re[: 1989; 5: 165-89. Heuson JC, Coune A, Heimann R. Cell proliferation induced by insulin in organ cultured of rat mammary carcinoma. E . Y ~Cell
Br. J . Surg., Vol. 78,No. 4,April 1991
32.
33.
S.D. Heys et al
Re.7 1967; 45: 351-60. Torosian M H , Mullen JL, Stein T P , Miller EE, Zinsseer KR, Buzby G P . Enhanced tumor response to cycle-specific chemotherapy by pulse total parenteral nutrition. J Sury Rrs 1985;39: 103-13. Torosian M H , Mullen JL, Miller EE, Wagner KM, Stein T P , Buzby G P . Adjuvant, pulse total parenteral nutrition and tumor response to cycle-specific and cycle-nonspecific chemotherapy. Suryert. 1983; 94: 291-8.
Paper accepted 2 October 1990
487
