Access to all articles, new health classes, discounts in our store, and more!
Carbohydrate Metabolism and Cervical (Uterine) Carcinoma
Published in the American Journal of Obstetrics and Gynecology, Vol. 97, No. 6, pp. 817-820, March 15, 1967.
* * *
The classical glucose tolerance test was performed on 26 ambulatory, nonhospitalized, biopsy-proved squamous cell cervical carcinoma patients and compared with similar testing of 26 age- and sex-paired individuals without cancer. Three items deserve special mention . First, the frequency of glycosuria in the cancer group is approximately fourfold that observed in the noncancer persons. Second, the frequency of elevated blood glucose levels in the cancer group is approximately twofold that in the noncancer group . Third, the mean blood glucose values at every temporal point are statistically higher in the oncologic group versus those without cancer.
The past 80 years have yielded well over one hundred reports with regard to the relationship of carcinomatosis and carbohydrate metabolism. The consensus appears to be that a significant correlation does prevail.1 However, the frequency of this relationship remains controversial. We submit that, in part, the discrepancy stems from a lack of recognition of two important facts: (1) a disturbance in carbohydrate metabolism and diabetes mellitus are not synonymous; and (2) the biochemical criteria for diabetes mellitus may be questioned.
Accordingly, this report will concern itself with a consideration of the relationship of blood glucose (rather than diabetes mellitus) and cervical (uterine) carcinoma.
Method of Investigation
Twenty-six ambulatory, nonhospitalized, biopsy-proved, squamous cell carcinoma patients were subjected to the classical glucose tolerance test.2 Venous blood samples fasting, 30 minutes, 1, 2, and 3 hours (Somogyi-Nelson method3-5) were performed with parallel urinalyses for glucose and acetone. Twenty-six sex- and age-paired dental patients served as the control group.
Results
Table I provides the original data showing the case numbers, age, and glucose tolerance patterns for the 52 individuals. Glycosuria is indicated by dagger. An analysis of the means and standard deviations for the ages (Table IV) shows no statistically significant difference of the means (P > 0.500) or variance (P > 0.250).
Table I. Classical glucose tolerance patterns in cancer and noncancer patients
Cancer group* | ||||||
Case No. | Age | Fasting | 30 min. | 1 hr. | 2 hr. | 3 hr. |
12
16 24 08 23 11 07 17 10 14 15 05 03 09 13 22 04 19 20 21 01 06 18 02 25 26 |
33
34 38 40 41 43 45 47 48 48 52 53 55 56 57 57 57 58 58 59 61 62 63 65 60 72 |
102
97† 97 147 92 95 107 102 110 77 90 85 77 82 97 59 82 117 130 85 270† 95 72 93 95 95 |
224
165† 150 175 160 130 160 207 140 115 180 150 120 115 150 100 195 160 242 —- 330† 150 118 110 137 145 |
205†
210† 147 223 185 95 130 207 204 120 210† 184 110 110 185 102 210† 165 320 170 406† 184 140 166 155 210 |
115†
102† 132 210 170† 102 103 155 210† 120 185† 160 77 72 110 103 200† 200 330† 210 364† 155 110 240† 142 175† |
53
90† 127 160 138† 102 80 80 127 110 100† 123 63 82 120 80 127 170 260 155 360† 120 63 170† 92 155 |
*Italics indicate elevated blood glucose levels.
†Glycosuria.
Noncancer group* | ||||||
Case No. | Age | Fasting | 30 min. | 1 hr. | 2 hr. | 3 hr. |
504
528 529 511 586 519 559 590 517 527 523 553 503 560 607 608 598 606 565 571 563 513 494 591 497 537 |
33
34 38 40 41 43 45 48 49 49 52 53 55 56 50 64 55 70 58 59 61 62 60 65 54 52 |
110
74 65 99 91 75 70 180 68 80 75 90 80 85 74 85 74 65 75 70 75 85 85 81 86 82 |
150
120 143 150 133 120 95 234 95 178 155 157 135 135 115 135 104 95 107 125 140 157 157 125 153 133 |
175
55 97 134 140 120 79 306† 57 155 175 155 150 138 165 153 104 68 135 107 133 155 125 81 125 130 |
104†
51 63 132 97 74 88 340† 90 139 117† 117 133 100 107 175† 95 56 132 76 90 155 155 79 80 130 |
90
48 67 82 81 74 88 306† 55 51 57 82 107 82 85 153† 81 50 100 44 95 110 115 83 90 104 |
Table II summarizes the frequency of glycosuria in the two groups. For example, 2 of the 26 individuals (8 per cent) demonstrated glycosuria under fasting conditions; in contrast, none of the noncancer group had sugar in the urine. Several points warrant special mention. First, at every temporal point, the frequency of glycosuria in the cancer group is much higher than noted among the noncancer individuals. Second, the greatest frequency of glycosuria in both groups is at 2 hours. Third, even at this point, glycosuria occurs approximately two and one-half times more frequently in the cancer versus the noncancer group. Last, over-all, the incidence of glycosuria in the cancer group is almost fourfold (19 per cent) versus that observed in the noncancer group (5 per cent). Chi square analysis for the entire group is statistically significant (10.7399 and P < 0.001).
Table II. Frequency of glycosuria
Time | Cancer group | Noncancer group | ||
No. | % | No. | % | |
Fasting
Thirty minutes One hour Two hours Three hours |
2/26
2/25 5/26 10/26 5/26 |
8
8 19 38 19 |
0/26
0/26 1/26 4/26 2/26 |
0
0 4 15 8 |
Total | 24/129 | 19 | 7/130 | 5 |
In Table I elevated blood glucose levels are shown by underline. For example, it is generally recognized that a fasting blood glucose level according to the Somogyi-Nelson technique should not exceed 100 mg. per cent. In Case No. 12, the fasting determination is 102 mg. per cent. This value is accordingly underscored. The consensus is that the peak should not exceed 150 mg. per cent. Case No. 12 shows a 30 minute determination of 224 mg. per cent. This score is therefore underlined. Finally, it is generally recognized that the 2 and 3 hour samples should not be above 100 mg. per cent. Case No. 12 has a 2 hour determination of 115 mg. per cent. These values are also emphasized.
Table III summarizes the frequency of elevated blood glucose levels. For example, under fasting conditions, 8 of the 26 patients (31 per cent) of the cancer group had blood glucose levels above 100 mg. per cent. In contrast, only 2 of those in the noncancer group exceeded 100 mg. per cent (8 per cent). It will be observed that, at every temporal point, the frequency of elevated levels is distinctly higher in the cancer group. It is noteworthy that, at 2 hours, 92 per cent of the individuals in the cancer category showed blood glucose levels in excess of 100 mg. per cent in contrast to a frequency of 50 per cent in the noncancer group. The frequency of elevated blood glucose levels over-all is more than double in the cancer versus the noncancer patients. The statistical significance is emphasized by a chi square of 26.2180 (P< 0.001).
Table III. Frequency of elevated blood glucose levels
Time | Cancer group | Noncancer group | ||
No. | % | No. | % | |
Fasting
Thirty minutes One hour Two hours Three hours |
8/26
11/25 18/26 24/26 16/26 |
31
44 69 92 62 |
2/26
7/26 7/26 13/26 6/26 |
8
27 27 50 23 |
Total | 77/129 | 60 | 35/130 | 27 |
The literature is replete with reports indicating that the upper physiologic limits for blood glucose during the glucose tolerance test are still open to question. Even less attention has been given to the physiologic lower limits. In order to obviate these objections, Table IV and Fig. 1 are a summary of the means, standard deviations, significance of the differences of the means, and significance of the difference of the variances for the two groups. It is noteworthy that the means for all of the blood glucose determinations are significantly different at the 5 per cent level or less. Also, mention should be made of the fact that there are also statistically significant differences of the variances at the fasting and 30 minute levels.
Table IV. Statistical analyses of blood glucose values in cancer versus noncancer groups
Cancer group | Noncancer group | Significance of the differences | ||
Means | Variances | |||
Sample size
Age (mean and S.D.) Fasting blood glucose (mean and S.D.) 30 Minute blood glucoset (mean and S.D.) 1 Hour blood glucose (mean and S.D.) 2 Hour blood glucose (mean and S.D.) 3 Hour blood glucose (mean and S.D.) |
26
52.4 ± 10.0 101.9 ± 38.6 161.1 ± 49.9 182.8 ± 66.6 163.5 ± 71.0 127.2 ± 64.8 |
26
51.8 ± 9.6 83.8 ± 22.1 136.4 ± 29.6 131.4 ± 49.6 114.4 ± 55.9 91.5 ± 50.1 |
P > 0.500P
< 0.050* P < 0.050* P < 0.005* P < 0.010* P < 0.050* |
P > 0.250P
< 0.005* P < 0.010* P > 0.050 P > 0.100 P > 0.100 |
*Statistically significant.
†The sample size for the cancer group is 25 (see Table I).
Fig. 1. A comparison of the classical glucose tolerance test patterns in a group of 26 patients with biopsy-proved squamous cell carcinoma of the cervix uteri versus the pattern for an age- and sex-paired noncancer group. At every temporal point, the mean blood glucose values for the cancer group are statistically significantly higher.
Comment
The data from this study add some validity to the observation of a significant correlation between carbohydrate metabolism and carcinomatosis. This is evident when one examines the relationship by traditional means such as glycosuria and the frequency of elevated blood glucose levels. The correlation appears to be even more striking when one ignores standard parameters and simply equates the blood glucose levels of an age-paired cancer versus noncancer group.
Since the prevalence of cancer in various populations is well-established, it would be interesting to observe the rate of development and frequency in a select group of individuals with a very efficient carbohydrate metabolism versus a hyperglycemic group treated to maintain very stable blood sugar levels.
References Cited:
- Cheraskin, E., and Ringsdorf, W. M., Jr.: Cancer 17: 159, 1964.
- Mosenthal, H. O., and Barry, E.: Ann. Int. Med. 33: 1175, 1950.
- Nelson, N. A.: J. Biol. Chem. 153: 375, 1944.
- Somogyi, M.: J. Biol. Chem. 160: 61, 1945.
- Somogyi, M.: J. Biol. Chem. 160: 69, 1945.