Altered Metabolism and Mortality in Patients With Colon Cancer Receiving Chemotherapy

[American Journal of the Medical Sciences 310:48-55, 1995]

JOHN A. TAYEK, MD, FACP, FACN, LYNDA SUTTER, BS, SAVITA MANGLIK, MBBS, LINDA B. LILLINGTON, RN, PHD, MARY GROSVENOR, RD, MS, ROWAN T. CHLEBOWSKI, MD, PHD

ABSTRACT: To identify the metabolic effects of 5-fluorouracil and hydrazine sulfate therapy, 22 patients with colon cancer were admitted prospectively to a Clinical Research Center for serial measurement of counter-regulatory hormones, fasting hepatic glucose production (HGP), intravenous glucose tolerance test, plasma leucine appearance (LA) and leucine oxidation. Combined therapy was associated with a significant reduction in fasting glucose level (98±2 mg/dL to 94±2, P < 0.025) without a significant fall in fasting HGP (2.09±0.11 mg/kg/min versus 2.03±0.13; P >0.05). The decreased fasting glucose value was associated with a mild but not statistically improved glucose disposal rate in response to the intravenous glucose tolerance test (1.34±0.07 %/min vs 1.47±0.11, P = 0.15). Plasma leucine appearance was significantly reduced after 2 months of therapy (63.3±3.0mmol/kg/hr vs 57.1±3.9mmol/kg/hr; P < 0.025), but leucine oxidation (11.5±1.1 mmol/kg/hr vs 11.2±1.1 mmol/kg/hr) was not altered. Despite the fact that plasma triiodothyronine concentrations significantly increase with therapy, it was not associated with plasma LA. Half of the patients with cancer died 14±4 months after the study, and the other half were alive 58±2 months later. Survival time can be estimated with 59% accuracy using plasma LA, HGP, carcino-embryonic antigen, and insulin concentration. Multiple regression analysis identified that plasma LA was related directly to length of survival time, and baseline HGP, carcino-embryonic antigen, and insulin concentration were related inversely to length of survival. Unlike an elevated HGP seen in cancer cachexia, based on the association of a higher plasma LA and longer survival rate, an elevation in plasma LA and longer survival rate, an elevation in plasma LA may not be an unfavorable response to cancer. Further research is required to validate the predictability of baseline metabolic markers with survival rate in the cancer population. KEY INDEXING TERMS: Cancer cachexia; Hepatic glucose production; Plasma leucine appearance; Leucine oxidation; Hydrazine sulfate; Growth hormone; IGF-1; Cortisol; Glucagon; Colorectal carcinoma; Survival; Urine urea; Triiodothyronine. [Am J Med Sci 1995; 310(2):48-55.]

Elevations in hepatic glucose production (HGP) occurs in approximately 75% of all published studies of patients with cancer. [1] Two of the four studies of patients with colon cancer identified significant elevations in fasting HGP. [1-3] In addition to an elevation in HGP, research demonstrated that glucose disposal during hyperinsulinemic euglycemic clamp is reduced in patients with colorectal cancer. [4] An increased glucose volume of distribution observed in patients with colon cancer [2,3] may increase the insulin requirements to optimize glucose disposal. Other mechanisms, such as the presence of a reduced triiodothyronine concentration or tumor-derived factors, may be responsible. [5] The abnormal glucose metabolism in patients with colon cancer appears to be due to an increased glucose recycling, [2,6] excessive HGP, reduced glucose disposal, or a combination of these processes.

An increased glucose carbon recycling occurs before an increase in HGP. [2] The resynthesis of glucose is energy costly and may account for up to a 300-kcal loss of energy daily. [7] Patients with colon cancer have reductions in plasma gluconeogenic amino acids (alanine, glycine, glutamine, threonine), which may be used as precursors for glucose production. [8-10] Elevations of plasma amino acid appearance occur in many of the patients with colon cancer. [11-13] A reduction in the plasma levels of gluconeogenic amino acids may result from an increased need for glucose carbons, which would suggest that the two processes may be linked.

Hydrazine sulfate, an inhibitor of phosphoenolpyruvate carboxykinase, [14]has demonstrated metabolic activity, reducing both amino acid appearance rates [15] and HGP [16] in malnourished patients with cancer. In the lung cancer population, a suggestion of clinical benefit [17] remains a question. [18] In the current study, the combined activity of hydrazine sulfate and 5-flu-orouracil (5-FU) chemotherapy in a colon cancer population was explored using a serial study design. The effects of 5-FU chemotherapy on glucose and amino acid metabolism are published earlier. [10]

In the current study, we evaluated 22 patients with stage D colon cancer. Our objective in the study was to identify metabolic abnormalities during the first 2 months of 5-FU and hydrazine sulfate therapy. Base-line metabolic markers, hormonal determinations, carcino-embryonic antigen (CEA), HGP, plasma leucine appearance (LA) and leucine oxidation were evaluated before chemotherapy and monthly for 2 months. Baseline measurements (time zero) also were evaluated as a marker of survival. Our hypothesis was that an elevated HGP and an elevated plasma LA would be markers of a poor survival rate because they are thought to reflect an increased metabolic rate and tumor burden.

Methods

Patients with biopsy proven nonresectable colon carcinoma with or without prior malnutrition were en-rolled once a month for 3 months for a 5-day hospital stay. Patients with diabetes mellitus, hypothyroidism, hyperthyroidism, clinical evidence of cirrhotic liver disease, renal disease, or anemia (hematocrit less than 30) were excluded. Patients with cancer were admitted to the Clinical Research Center at Harbor-UCLA Medical Center under an Institutional Review Board-approved protocol. The isolated effects of 5 days of 5- FU chemotherapy on nitrogen, amino acid, and glucose metabolism was reported previously for this group of patients. [10]

Fasting baseline blood measurements included glucose, insulin, growth hormone, insulin-like growth factor- l, glucagon, cortisol, thyroxine (T4), and triiodothyronine (T3) concentrations. Plasma insulin, growth hormone, glucagon, cortisol, T4, T3, and reverse T3 were measured by standard double antibody radioimmunoassay methods. Insulin-like growth factor-l was assayed after ethanol and acid extraction by radioimmunoassy. [19] The hospital laboratory performed CEA and chemistry profile analysis. A Beckman amino acid analyzer was used to determine plasma amino acids and plasma and urine 3-methylhistidine levels. Urine was collected daily during the 5-day treatment period and measured for creatinine, urea nitrogen, 3-methylhistidine, and free cortisol concentrations. After an overnight fast, a 4-hour combined 6-3H-glucose (25 mu Ci, 14.4 mu Ci/hr) and 1-14C-leucine (4 mu Ci, 1.6 mu Ci/ hr) primed, continuous infusion was started at 8 AM. During hours 3 to 4 of the study, blood was drawn every 20 minutes for plasma 1-14C-leucine specific activity; glucose, and 6-3H glucose specific activity, as previously determined. [20,21] Plasma leucine appearance rates were determined according to the following formula: plasma LA (mu mol/hr) = [IR (mL/hr) X SAi (dpm/mL)]/SAs(dpm/mu mol), where IR = infusion rate; SAi = specific activity of infusate; and SAs = specific activity of leucine in protein-free plasma at equilibrium.

Resting carbon dioxide production was measured using an indirect respiratory calorimetry method and analyzed by a gas mass spectrometer. The indirect calorimetry collections were made on the same day as the isotope infusion study before the start of the infusion, at 120 minutes, and at 20-minute intervals during hours 3 to 4. Each 5-minute collection period was preceded by trapping 1/2 mmol of carbon dioxide in hydroxide of hyamine. Standard scintillation techniques were used to determine 14C carbon dioxide content. Leucine oxidation was determined by the product of the carbon dioxide production rate and the 14C carbon dioxide content in the expired breath. [22]

At hour 4, the patients underwent a 25-g intravenous glucose tolerance test (IVGTT), and blood glucose and insulin were drawn at 0, 5, 10, 15, 30, 45, and 60 minutes. Glucose disappearance was calculated as the natural log of glucose from 5 minutes to 60 minutes. Insulin area, as area above baseline, was calculated before the start of chemotherapy and monthly thereafter.

*----5FU---*                    *-----5FU---*                               *-----5FU----*

                    *-------HYDRAZINE SULFATE 60 mg po TID-------------*

DAY 0  DAY 5                  DAY 30   DAY 35                                DAY 60 DAY 65  

The population of patients with stage D colorectal carcinoma were scheduled to be studied after an over-night fast and serially at three time periods: once before initiation of chemotherapy (5-FU and hydrazine sulfate), with the second and third study periods at 30 days, and 60 days later. The 5FU infusional chemotherapy lasted 5 days, starting at day zero, day 30, and day 60. The studies on day 30 and day 60 were 25 days after a 5-day infusion of 5-FU chemotherapy. The patients were given 60 mg oral hydrazine sulfate therapy 3 times daily (2.6 mg/kg) starting on day 6 of this clinical study for a 2-month period (day 5 to day 65). The relations of study determinations, 5-FU infusional chemotherapy, and hydrazine sulfate use are outlined in Figure 1. 5-FU dosage was 15 mg/kg body weight daily as a continuous infusion for 5 consecutive days. Both the baseline measurements in the 22 patients with colon cancer and the acute effects of 5 days of continuous 5-FU infusion on glucose and amino acid metabolism were reported recently. [10]

The Clinical Research Center dietitian obtained a weight history and performed standardized measurement of anthropometry based on established methods.[23] Glucose and leucine kinetic calculations were per-formed using steady state isotopic equations. Changes in IVGTT, HGP, plasma LA, and leucine oxidation were compared by paired t-tests correcting for multiple comparisons and amino acid concentrations by analysis of variance. Standard linear regression analysis and stepwise multiple regression analysis were performed using the statistical package by BMDP. Survival time, CEA, and insulin concentrations were log transformed into a normal distribution.

Results

Twenty-two patients with colon cancer were enrolled in the study (Table 1). The patients with colon cancer as a group maintained weight during the 2-month treatment period (69 +/- 4 kg, 71 +/- 4, 73 +/- 5; mean +/-SEM). Fifty percent of the patients with colon cancer expired 14 +/- 4 months (median 10 months) after the baseline study. The survivors lived 58 +/- 2 months (median 56 months) after the baseline study (at the time of manuscript submission). The time from diagnosis to study date averaged 15 +/- 2 months. The time from diagnosis to study date was shorter in those patients who died (8 +/- 1 vs. 21 +/- 4 months; P < 0.01). T h e shorter time to baseline study (8 +/- 1 month) and shorter survival time (14 +/- 4 months) suggest that the nonsurvivors had a different progression of disease compared with the longer-term survivors (21 +/- 4 months and 58 +/- 2 months, respectively). Even though the patients all had stage D colon cancer, the non-survival group had a greater plasma CEA concentration (737 +/- 446 ng/mL vs. 171 +/- 64 ng/mL; P < 0.025). Liver function tests were normal in all but one patient, who had an elevated bilirubin concentration (6.6 mg/ dL).

Plasma glucose concentrations fell after 60 days of hydrazine therapy (Table 2). There was no change in insulin, IGF-1 glucagon, or growth hormone concentrations during the 2-month period of study. Triiodothyronine (T3) concentration increased after 55 days of hydrazine sulfate therapy. Serum transferrin concentration increased after 1 and 2 months of therapy. There were no changes in thyroxine (T4), reverse triiodothyronine, cortisol, or CEA concentrations.

Table 2. Hormonal Levels in Patients With Colon Cancer Before and During Chemotherapy
	Colon Cancer (Day 0)	Colon Cancer (Day 30)	Colon Cancer (Day 60)
n	22	18	16
Glucose (mg/dL)	90 +/- 2	95 +/- 4	94 +/- 22*
Insulin ( /mL)	11.0 +/-2 2.1	11.7 +/- 2.9	10.2 +/- 1.5
IGF-1 ( g/L)	135 +/- 20	195 +/- 30	184 +/- 26
Glucagon (pg/mL)	121 +/- 25	164 +/- 47	115 +/- 17
Growth hormone (ng/mL)	3.22 +/- 0.62	4.30 +/- 0.95	3.80 +/- 1.91
Thyroxine (T4) ( g/mL)	6.5 +/- 0.5	6.9 +/- 0.6	7.1 +/- 0.6
Triiodothyronine (T3) (ng/dL)	81 +/- 7	97 +/- 8	106 +/- 8*
Reverse T3 (ng/dL)	28 +/- 4	19 +/- 2	17 +/- 2
Cortisol ( g/dL)	12.8 +/- 1.3	12.7 +/- 1.7	11.5 +/- 1.9
CEA (ng/mL)	430 +/- 235	558 +/- 310	177 +/- 79
Transferrin (me/dL)	253 +/- 15	274 +/- 15*	291 +/- 14*
*P < 0.025 vs. Day 0 colon cancer.
ICF-I = insulin-like growth factor 1.
All data are mean +/- standard error of the mean

Twenty-three percent (5 of 22) of the patients with colon cancer had an abnormal baseline glucose tolerance test (glucose disappearance < 1.0%/min). Eighteen of the 22 patients with cancer tolerated the first month of chemotherapy and hydrazine therapy and returned for follow-up after 1 month. There was no significant fall in fasting HGP (Table 3). The decreased fasting glucose value was associated with a slight in-crease in the glucose disposal rate in response to the IVGTT (1.34 +/- 0.07 %/min vs. 1.47 +/- 0.11 %/min; P = 0.15; Table 3). The insulin response to IVGTT was slightly but not significantly increased after 1 month of hydrazine therapy (Table 3). Seventeen percent of the patients with colon cancer had an abnormal IVGTT after 1 month of chemotherapy, and 20% were abnormal after 2 months. This was not different from the 23% of abnormal IVGTT seen at baseline. All insulin concentrations (n = 73) were significantly correlated with ideal body weight (r = 0.745; P < 0.001). All HGP rates were inversely correlated with body weight (r = -0.525, P < 0.01) and ideal body weight (r = -0.506; P < 0.01), but not correlated with fasting glucose, insulin, glucagon, cortisol, T4, or T3 concentrations. Hepatic glucose production was only weakly correlated with fasting growth hormone concentrations (r = 0.294, P < 0.05).

Table 3. Fasting Hepatic Glucose Production Rates and Insulin and Glucose Response to an IV Glucose Tolerance Test
	Colon Cancer (Day 0)	Colon Cancer (Day 30)	Colon Cancer (Day 60)
n	22	18	16
HGP (mg/kg/min)	2.09 +/- 0.11	2.18 +/- 0.10	2.03 +/- 0.13
Insulin area under curve ( U/mL/hr)	2,053 +/- 253	2,500 +/- 403	2,365 +/- 473
Glucose Kd (%/min)	1.34 +/- 0.07	1.40 +/- 0.12	1.47 +/- 0.11
All data are mean +/- standard error of the mean; no significant difference.

Similar to the effect of 2 months of hydrazine therapy on fasting blood glucose levels, the plasma LA rate was reduced in patients with colon cancer (63.3 +/-3.0 mu mol/ kg/hr vs. 57.1 +/- 3.9 mu mol/kg/hr; P < 0.025; Table 4). Hydrazine sulfate had no effect on leucine oxidation (Table 4). The net incorporation of leucine into protein (leucine appearance minus leucine oxidation) was not significantly decreased (Table 4).

While receiving hydrazine therapy, many of the gluconeogenic amino acids and aromatic amino acids concentrations were increased (Table 5). A 54-78% in- crease in tyrosine and a 77-111% increase in ornithine concentrations were noted during hydrazine therapy. Only essential amino acids concentrations (isoleucine and lysine) were decreased after both months of hydrazine and 5-FU therapy (Table 5). Isoleucine was the only amino acid concentration to decrease after 5- FU chemotherapy, [10] so the change in lysine seen in this study was most likely due to hydrazine treatment.

Table 4. Plasma Leucine Appearance, Oxidation, and Incorporation into Protein
	Colon Cancer (Day 0)	Colon Cancer (Day 30)	Colon Cancer (Day 60)
n	22	18	16
Leucine appearance (mol/kg/hr)	63.3 +/- 3.0	56.9 +/- 3.7	57.1 +/- 3.9*
Leucine oxidation ( mol/kg/hr)	11.5 +/- 1.1	11.7 +/- 1.4	11.2 +/- 1.1
Leucine incorporation ( mol/kg/hr)	51.7 +/- 2.7	45.3 + 3.7	45.9 +/- 4.0
* P 0.025 vs. < Day 0. AU data are mean +/- standard error of the mean.

Twenty-four-hour dietary recall demonstrated that the patients with cancer were averaging 1,200 kcals daily and did not significantly change over time. There was no significant change in the urine urea nitrogen or 3-methylhistidine excretion during the 2-month study (Table 6). Urine-free cortisol excretion was normal.

Individual analysis of the baseline nutritional, hormonal, and metabolic measurements demonstrated significant differences between those patients with cancer who did or did not survive (Table 7). Carcino- embryonic antigen measurements were increased in the nonsurvivors. A small but significant reduction in the fasting blood glucose concentration was noted in those patients who did not survive. Although the fall in plasma glucose was only 8%, the nonsurvivors had a significant increase in growth hormone concentrations without a change in the plasma IGF-1 concentration. Plasma LA was not significantly increased compared with previous healthy volunteers. [10] However, the base-Colon line plasma LA or whole body protein breakdown rate was significantly elevated in the long-term survivors compared with the nonsurvivors (Table 7). The survivors also demonstrated a significant 16% increase in the baseline rate of whole body protein synthesis.

Plasma LA correlated directly with length of survival (r = 0.403, P < 0.05). In contrast, HGP, CEA, and insulin concentration were associated inversely with length of survival (r = -0.461, r = -0.459, r = -0.466, P < 0.05). Based on multiple regression analysis, survival time could be estimated with a 59% accuracy using plasma LA, HGP, CEA, and insulin concentrations (plasma LA correlated directly with survival, and HGP, CEA and insulin correlated indirectly with length of survival; r*r = 0.592, P < 0.05). Fasting transferrin, glucose, GH, IGF-1, or thyroid hormone concentration were not associated with survival time by simple or multiple regression analysis in this small group of patients with colon cancer.

Discussion

Weight loss associated with cancer is frequently associated with a decreased survival rate. [24] In our population of patients with colon cancer, it appears that alterations in metabolic parameters (HGP, plasma LA) may also be an index of survival rate in addition to the traditional use of body weight. In earlier research, it was suggested that in addition to abnormal CEA concentrations, elevations in HGP, amino acid appearance (protein breakdown), and amino acid oxidation were considered markers of diffuse disease and a poor prognosis. [7, 15, 16, 25-29] The elevated metabolic activity could be viewed as an energy costly process that contributes to the progressive development of cancer cachexia. This statement would be true if the metabolic process occurred before the weight loss. Shaw et al [28] demonstrated that an elevated plasma LA occurs before the weight loss in a heterogenous group of patients with cancer. However, our homogenous group of patients with an average weight loss of 9% failed to have a significant elevation in plasma LA, [10] which also was seen by others. [27] Based on the favorable association of an elevated plasma LA with survival rate, the elevation in the amino acid metabolism when observed may not be a marker of catabolism but may reflect an appropriate metabolic response to cancer. An elevation in HGP, however, may reflect a maladaptive response.

Table 6. 24-Hr Urine Collection Before and During Hydrazine Therapy
	Baseline	Month 1	Month 2
n	22	18	16
Creatinine (g/d)	0.93 +/- 0.10	0.97 +/- 0.09	0.94 +/- 0.13
Urea nitrogen (g/g creat/d)	9.3 +/- 0.7	10.3 +/- 0.7	9.9 +/- 0.7
3-Methylhistidine ( mol/g creat/d)	190 +/- 9	195 +/- 14	188 +/- 12
Free cortisol ( g/d)	42 +/- 8	38 +/- 5	30 +/- 5
Values are for an average of 5 days, except for the 3-methlhistidine results, which are for day 1 only, All data are mean +/- standard error of the mean.

Abnormalities in glucose metabolism are quite common in patients with cancer. [1] Glucose intolerance was seen in 23% of these patients with colon cancer despite' the fact that insulin secretion was normal. Glucose in-tolerance may be due to an decrease glucose disposal rate or an increase in HGP. A glucose disposal rate of 1.3 4 +/- 0.0 7 %/min is similar to that rate seen in patients with sarcoma (1.6 4 +/- 0.1 1 %/min and is significantly less than previously published values for healthy volunteers (1.9 7 +/- 0.1 4 %/min [30]). Hepatic glucose production rates increased on average by 24% in non-weight losing patients with cancer and by 40% in weight-losing patients with cancer. [1] The HGP was in-creased by only 17% in this study when compared with the similar weight controls published earlier. [10] The glucose intolerance was seen in only 1/4 of our patients, most likely due to a decreased glucose disposal rate.

Table 7. Baseline Metabolic Measurements in Non-survivors With Colon Cancer
	Colon Cancer Survivors	Colon Cancer Non-survivors
Patients (n)	11	11
Age (yrs)	56 +/- 3	55 +/- 3
Body weight (kg)	73 +/- 4	69 +/- 3
Ideal body weight (%)	113+/-5	111 +/- 4
Transferrin (mg/dL)	240 +/- 21	266 +/- 21
CEA (mg/dL)	171 +/- 69	737 +/- 446*
Fasting glucose (mg/dL)	100+/- 2	92 +/- 3**
Growth hormone (ng/mL)	2.34 +/- 0.45	5.12 +/- 1.2*
IGF-1 ( g/L)	144 +/- 33	126 +/- 23
Hepatic glucose production (mg/kg/min)	2.07 +/- 0.10	2.13 +/- 0.09
Leucine appearance ( mold/kg/hr)	69.4 +/- 4.0	57.2 +/- 3.8*
Leucine incorporation ( mol/kg/hr)	56.0 +/- 3.9	47.3 +/- 3.4*
* P < 0.025    ** P < 0.01 vs survivors (All values from baseline measurements)
CEA = carcino-embryonic antigen; IGF-1 = insulin-like growth factor 1.
All data are mean +/- standard error of the mean.

Administration of growth, hormone can increase HGP in healthy volunteers but does not increase HGP in patients with cancer. [31,32] Based on the weak correlation between growth hormone and HGP, an association is suggested, but caution must be exercised because of the pulsatile nature of growth hormone and due to the fact that no association was seen between 24-hour growth hormone patterns and HGP in patients with lung cancer. [33]

A study of the short-term effects of cytotoxic agents on whole body protein breakdown, oxidation, and synthesis rates demonstrated that the combination of Vinblastine, Cisplatinum, and Bleomycin reduces whole body protein synthesis, oxidation, and break-down. [34] We recently demonstrated that 5 days of 5-FU chemotherapy increases HGP and leucine oxidation by approximately 10%. 5-FU chemotherapy also was associated with a 10% fall in whole body protein synthesis. [10] The daily hydrazine sulfate therapy was well tolerated by the patients with cancer and was associated with a decrease in whole body protein breakdown (LA) after 60 days of therapy. There was no effect on leucine oxidation, protein synthesis, or HGP when studied 25 days after 5-FU chemotherapy. The lack of effect on HGP may be due in part to the stimulation of HGP by 5-FU and/or due to the lack of severely malnourished cancer patients in this study.

In addition to the reduction in plasma LA seen after hydrazine sulfate therapy, hydrazine treatment was associated with increases in tyrosine, T3, and transferrin concentrations. A nonsignificant increase in T3 and serum transferrin concentrations was observed in six hydrazine- treated patients with lung cancer when compared with six placebo-treated patients with lung cancer. [15] In the current study, T3 concentrations also failed to increase significantly after 1 month of therapy, but T3 concentrations increased significantly after 2 months of therapy. Transferrin concentrations significantly in-creased at 30 and 60 days, which may have been due to the larger sample size in the current study (n = 22 vs. n = 6). The reduced protein breakdown and/or increase in select amino acids may have been responsible.

Hydrazine sulfate administration to mice is associated with an acute increase in corticosterone and a fall in T3 concentrations during a 5-hour period. [36] Even though we did not measure acute changes, the long-term changes with hydrazine treatment failed to increase fasting cortisol concentrations or daily urine free cortisol excretion. Similarly, the lack of data available on the acute effects of hydrazine therapy on T3 concentrations in our study does not exclude the possibility that hydrazine therapy acutely decreased T3 concentrations. Even though the patients with colon cancer had an increase in T3 concentration over time, our study was not designed to mea-sure the short-term effects.

Hydrazine sulfate therapy acts via inhibition of phosphoenolpyruvate carboxykinase and was associated with a significant increase in the fasting concentrations of most of the gluconeogenic amino acids (Table 5). Hydrazine therapy in patients with lung cancer also was associated with an increase in alanine concentrations, which was associated with a reduction in plasma lysine appearance (protein breakdown). [15] The similar reduction in plasma LA (protein breakdown) observed after 2 months of hydrazine therapy in the current study probably reflects a mechanism by which hydrazine works to conserve protein mass and prevent malnutrition. A 10% reduction in protein breakdown observed in our study could translate into a prolonged protein half-life and pre-vent the depletion of the amino acid pools. The reduced protein breakdown would be expected to translate into a reduction in the average daily urine urea nitrogen and 3-methylhistidine (3MH) content (Table 6). However, a 10% drop in protein breakdown would only account for a 10% reduction in urine urea nitrogen and 3MH excretion, where the accuracy for urine collections are at best able to detect a 10% change. Therefore, a small change in whole body protein metabolism may not have been detected by urine urea nitrogen or 3MH determinations.

Based on the severely altered amino acid concentrations in this study, hydrazine sulfate may be affecting amino acid metabolism to a greater degree than appreciated by measuring leucine kinetics alone. The significant increase in total T3 levels may be due to the increase of tyrosine available or a net increase in thyroid-binding proteins. Likewise, the decrease in arginine and the increase in ornithine may also be due to an increased conversion of arginine to ornithine in the liver. The specific effect of hydrazine on individual amino acids is not known.

In contrast to earlier studies, hydrazine therapy failed to document a fall in HGP. [16] This may be due to patient selection, because in the earlier work, severely malnourished (19% loss of body weight) patients with cancer were studied. The average weight loss in our study was only 9%, and earlier studies demonstrated that as the amount of weight loss increases, so does the HGP. [25] Previous work demonstrated that 5 days of 5-FU chemotherapy can increase HGP so that the reducing effects of hydrazine sulphate may have been minimized because of the 5-FU effect. [10] This latter explanation may be less likely because the 5-FU was only given for 5 days, and the effects of hydrazine were measured 25 days after 5-FU administration.

The current study, in which all patients received hydrazine sulfate, was not designed to assess clinical efficacy. As such, the relation among hydrazine sulfate-associated metabolic changes and nutritional status or clinical outcome must be determined in future studies.

Figure 2. Plasma leucine appearance in survivors and nonsurvivors receiving chemotherapy. Twenty-two patients with stage D colon cancer were followed for up to 6 years after chemotherapy. Baseline measurements (before chemotherapy) and day 60 measurements of plasma leucine appearance were elevated significantly in the long-term survivors alive at the time of manuscript submission (58 +/- 2 months after baseline). The nonsurvivors, who died after an average of 14 +/- 4 months, had a significantly reduced plasma leucine appearance at baseline and after 2 months of chemotherapy.

Koea and Shaw [36] demonstrated that both amino acid appearance rates and HGP rates are directly proportional to the tumor burden in a large group of heterogeneous patients with cancer. In our group of patients with colon cancer, we were not able to quantify the tumor burden. However, the CEA measurement, used by some clinicians as a marker of tumor burden, was not correlated with amino acid appearance rates or HGP. Novel to this study is the fact that elevations in CEA, HGP, and insulin concentration were associated with a shortened survival rate. In contrast, an elevation in plasma LA was associated with a longer survival rate. Individual CEA, insulin, HGP, or plasma LA may not be sensitive enough to be a single marker of survival rate (Figure 2). The combination of the four factors improve the accuracy to predict survival time. In comparison, serum markers of malnutrition (i.e., transferrin or IGF-1 concentration) were not helpful in predicting outcome in this group of homogenous patients with colon cancer.

In summary, we demonstrate in this study that hydrazine sulfate and 5-FU administration to patients with colon cancer increases T3 concentrations and reduces whole body protein breakdown (LA). This therapy also was associated with a small fall in plasma glucose concentration. Plasma amino acid metabolism was not correlated with the degree of malnutrition, so its measurement may be a metabolic marker of disease. In contrast, HGP was correlated inversely with ideal body weight and appears to be related to the severity of malnutrition. A reduced plasma LA was seen in those patients who died almost 3 years sooner than those with a normal rate of leucine appearance (Table 7). Mechanisms that regulate plasma LA need to be evaluated so that factors responsible for these changes may be understood better.

Acknowledgments

The authors thank Stephanie Griffiths, Mario Paredes, and Maria Lajoie for technical assistance; and Connie Soriano, RN, Merlyn Dubria, RN, and all the nurses of the Clinical Research Center for their cooperation.

References

1. Tayek JA. A review of cancer cachexia and abnormal glucose metabolism in humans with cancer. J Am Coll Nutr. 1992; 11:445-56.
2. Long CL, Merrick H, Grecos G, Blakemore WS, Geiger J. Glucose metabolism in colorectal carcinoma. Metabolism. 1990; 39:495 - 501.
3. Lundholm K, Edstrom S, Karlberg I, Ekman L, Schersten T. Glucose turnover, gluconeogenesis from glycerol, and estimation of net glucose cycling in cancer patients. Cancer. 1982; 50:1142 - 50.
4. Copeland GP, Leinster SJ, Davis JC, Hipkin JL. Insulin resistance in patients with colorectal cancer. Br J Surg. 1987; 74:1031 - 5.
5. Tayek JA. Reduced non-oxidative glucose utilization in cancer patients is associated with a low triiodothyronine concentration. J Am Coll Nutr. 1995; 14:341 - 8.
6. Cori CF, Cori GT. The carbohydrate metabolism of tumors. II. Changes in the sugar, lactic acid, and CO2 combining power of blood passing through a tumor. J Biol Chem. 1925; 66:397 - 405
7. Eden E, Edstrom S, Bennegard K, Schersten T, Lundholm K. Glucose flux in relation to energy expenditure in malnourished patients with and without cancer during periods of fasting and feeding. Cancer Res. 1984; 44:1718 - 24.
8. Norton JA, Gorschboth CM, Wesley RA, Burt ME, Brennan MF. Fasting plasma amino acid levels in cancer patients. Cancer. 1985; 56:1181 - 6.
   
9. Bennegard K. Lindmark L, Eden E, Svaninger G, Lundholm K. Flux of amino acids across the leg in weight-losing cancer patients. Cancer Res. 1984; 44:386 - 93.
10. Tayek JA, Chlebowski RT. Metabolic response to chemotherapy in colon cancer patients. J Parenteral and Enteral Nutr. 1992; 16:65S - 71S.
11. Shaw JHF, Wolfe RR. Whole-body protein kinetics in patients with early and advanced gastrointestinal cancer: The response to glucose infusion and total parenteral nutrition. Surgery. 1988; 103:148 - 55.
12. Eden E, Ekman L, Bennegard K, Lindmark L, Lundholm K. Whole-body tyrosine flux in relation to energy expenditure in weight-losing cancer patients. Metabolism. 1984; 33:1020 - 7.
13. Carmichael MJ, Ciague MB, Keir MJ, Johnston IDA. Whole body protein turnover, synthesis and breakdown in patients with colorectal carcinoma. Br J Surg. 1980; 67:736 - 9.
14. Gold J. Proposed treatment of cancer by inhibition of gluconeogenesis. Oncology. 1968; 22:185 - 207.
15. Tayek JA, Heber D, Chlebowski RT. Effect of hydrazine sulphate on whole-body protein breakdown measured by 14C-lysine metabolism in lung cancer patients: A prospective randomized double-blinded placebo controlled trial. Lancet. 1987; 2:241 - 4.
16. Chlebowski RT, Heber D, Richardson B, Block JB. Influence of hydrazine sulfate on abnormal carbohydrate metabolism in cancer patients with weight loss. Cancer Res. 1984; 44:857 - 61
17. Chlebowski RT, Bulcavage L, Grosvenor M. Hydrazine sulfate influence on nutritional status and survival in non-small-cell lung cancer. J Clin Oncol. 1990; 8:9 - 15.
18. Kosty M, Fleishman S, Herndon J, Coughlin K. Cisplatine, vinblastine, and hydrazine sulphate (NSC150014) in advanced non-small lung cancer. A randomized placebo controlled, double blind phase III study. Proc Amer Soc Clin Oncol. 1992; 11:294.
19. Daughaday WA, Mariz IK, Blether SL. Inhibition of access of bound somatomedin to membrane receptor and immunobinding sites: A comparison of radioreceptor and radioimmunoassay of somatomedin in native and acid-extracted serum. J Clin Endocrinol Metab. 1980; 51:781 - 8.
20. Tayek JA, Bergner EA, Lee WP. Correction of glucose carbon recycling for the determination of "true" hepatic glucose production (HGP) rates by 1-13C-glucose. Biological Mass Spectrometry. 1991; 20:186 - 90
21. Tayek JA, Brasel JA. Effect of tumor necrosis factor alpha on skeletal muscle and Walker 256 carcinosarcoma protein metabolism studied in-vivo. Cancer Res. 1990; 50:2765 - 8.
22. Tayek JA, Bistrian BR, Hehir DJ, Martin R, Moldawer LL, Blackburn GL. Improved protein kinetics and albumin synthesis by branched chain amino acid-enriched total parenteral nutrition (TPN) in cancer cachexia: A prospective randomized crossover trial. Cancer. 1986; 58:147 - 57.
23. Lukaski HC. Methods for the assessment of human body composition: Traditional and new. Am J Clin Nutr. 1987; 46:537 - 56.
24. DeWys WD, Begg C, Lavin PT, et al. Prognostic effect of weight loss prior to chemotherapy in cancer patients. Am J Med. 1980; 69:491 - 7.
25. Tayek JA, Bulcavage L, Chlebowski RT. Relationship of hepatic glucose production (HGP) to growth hormone and severity of malnutrition in a population with colorectal carcinoma. Cancer Res. 1990; 50:2119 - 22.
26. Pisters PWT, Cersosimo E, Rogatko A, Brennan MF. Insulin action on glucose and branched-chain amino acid metabolism in cancer cachexia. Differential effects of insulin. Surgery. 1992; 111:301 - 10.
27. Melville S, McNurlan MA, Calder G, Garlick PJ. Increased protein turnover despite normal energy metabolism and responses to feeding in patients with lung cancer. Cancer Res. 1990; 50:1125 - 31.
28. Shaw JHF, Humberstone DA, Douglas RG, Koea J. Leucine kinetics inpatients with benign disease, non-weight-losing cancer, and cancer cachexia: Studies at the whole-body and tissue level and the response to nutritional support. Surgery. 1991; 109:37 - 50.
29. Heber D, Chlebowski RT, Ishibashi DE, Herrold JN. Block JB. Abnormalities in glucose and protein metabolism in non-cachectic lung cancer patients. Cancer Res. 1982; 42: 4815 - 9.
30. Norton JA, Maher M, Wesley R, White D, Brennan MF. Glucose intolerance in sarcoma patients. Cancer. 1984; 54:3022 - 7.
31. Orskov L, Schmitz O, Jorgensen JOL, et al. Influence of growth hormone on glucose-induced glucose uptake in normal men as assessed by the hyperglycemic clamp technique. J Clin Endocrinol Metab. 1989; 68:276 - 82.
32. Tayek JA, Brasel JA. Failure of anabolism in malnourished cancer patients receiving growth hormone. J Clin Endocrinol Metab. 1995; 80:2082 - 7.
33. Tayek JA. Growth hormone physiology and glucose metabolism in lung cancer patients. Endocrine. 1994; 2:1055 - 9.
34. Herrmann VM, Garnick MB, Moore FD, Wilmore DW. Effect of cytoxic agents on protein kinetics in patients with metastatic cancer. Surgery. 1981; 90:381 - 7.
35. Silverstein R, Turley, BR, Christoffersen CA, Johnson DC, Morrison DC. Hydrazine sulfate protects d-glactosamine-sensitized mice against endotoxin and tumor necrosis factor/ cachectin lethality: Evidence of a role for the pituitary. J Exp Med. 1991; 173:357 - 65.
36. Koea JB, Shaw JHF. The effect of tumor bulk on the metabolic response to cancer. Ann Surg. 1992; 215:282 - 8

This page is designed and hosted by Next Generation Computer Systems, and is the property of the Syracuse Cancer Research Institute. © 1996, Syracuse Cancer Research Institute. All rights reserved.
Last modified on 24 September 1998 by webteam@ngen.com.