Influence of age on the metastatic behavior of breast carcinoma

Influence of Age on the Metastatic Behavior of Breast Carcinoma SUZANNE M, DE LA MONTE, MD, MPH, GROVER M, HUTCHINS, MD, AND G, WILLIAM MOORE, MD, PHD T h e effects of aging on the biologic behavior of malignant neo-

which suggests that hormonal and/or metabolic factors may influence the biologic behavior of breast carcinoma. This hypothesis is further supported by the facts that several risk factors for developing breast carcinoma are endocrine related and many breast carcinomas are responsive to endocrine therapy. 26-29 Greater knowledge of the effects of aging on the biologic behavior of breast carcinomas might lead to improved therapeutic approaches to these tumors. To examine this question, we studied a large group of autopsied patients with metastatic breast carcinoma and analyzed the relationship between age and metastatic distributions of tumor.

plasms are poorly understood. To examine the question of a possible age effect on the metastatic behavior of tumors, w e reviewed the clinical histories, autopsy protocols, and histologic slides of 187 patients with metastatic breast c a r c i n o m a w h o w e r e subjected to c o m p l e t e a u t o p s y at T h e J o h n s Hopkins Hospital. Patients w e r e categorized in four g r o u p s according to the age of onset of breast carcinoma: group I had 12 y o u n g patients (<40 years old); group II had 41 patients, 40 to 46 years old, c o n s i d e r e d to be premenopausal; group II had 89 patients, 47 to 60 years old, c o n s i d e r e d to be early p o s t m e n o p a u s a l ; and group IV had 45 patients over 60 years old, c o n s i d e r e d to be late postmenopausal. G r o u p II patients survived about 50% longer than group III or IV

patients (p < 0.05), and groups I, II, and III patients had significantly greater n u m b e r s of metastases than group IV patients (p < 0.05). T h e r e was a p r o g r e s s i v e d e c l i n e in the frequencies of metastases in several locations as a function of age, s u c h that group I a n d II patients h a d significantly m o r e frequent metastases to the central n e r v o u s system (p < 0.05), endocrine organs (p < 0.01), ovary (p < 0.05), pancreas (p < 0.01), and gastrointestinal tract (p < 0.05) than patients in group III or IV. T h e s e d i f f e r e n c e s in n u m b e r and distribution of metastases w e r e not explainable on the basis of survival, therapy, or initial stage of disease. T h e results s u g g e s t that the p r o c e s s of aging may

METHODS

i n f l u e n c e the metastatic behavior of breast carcinoma; h o r m o n a l and metabolic factors may be implicated. HUM PATHOL 19:529-534, 1988.

Epidemologic studies have suggested that a number of major risk factors may be implicated in the etiology of breast carcinoma, including age at first full-term pregnancyl-3; age at menarche ~'5 and natural menopause4; various h o r m o n e s 6, including exogenous estrogens 7 '8 ; diet and nutrition 9- 13 , race or ethnicity14'1~; socioeconomic status16; lf9a20mily history17--19-; and history of previous benign 9 or 9 21 " mah~nant breast disease. Although breast cancer v v 22 occurs more frequently in postmenopausal women, it is probable that the same risk factors are p r e s e n t -9 23 2"4 in both pre- and postmenopausal panents. ' However, age-related differences in survival among patients with breast cancer have been demonstrated, 25 9

From the Autopsy Pathology Division, Department of Pathology, T h e J o h n s Hopkins Medical Institutions, Baltimore, Maryland. Revision accepted for publication 30 July 1987. Supported by grant LM-03651 from the National Library of Medicine. A d d r e s s c o r r e s p o n d e n c e a n d r e p r i n t r e q u e s t s to Dr. Hutchins: Department of Pathology, The Johns Hopkins Hospital, Baltimore, MD 21205. Keywords: age, breast carcinoma, metastases, risk factors. 9 by W.B. Saunders Company. 0046-8177/88 $0.00 + .25

529

The clinical histories, gross autopsy protocols, and .histologic slides of patients with metastatic breast carcinoma who underwent complete autopsy at The Johns Hopkins Hospital between 1965 and 1983 were reviewed 9 Patients in whom metastases were not present at autopsy or on whom only partial autopsies were done were excluded from the data analysis. The patients were classified into four groups: I, less than 40 years old (young); II, 40 to 46 years (premenopausal); III, 47 to 60 years (early postmenopausal); and IV, greater than 60 years (late postmenopausal). This categorization was based on the age at which symptoms related to the primary tumor were first noted or when the diagnosis was established, whichever was earlier. Information obtained from the clinical histories, autopsy protocols, and review of the histologic slides included 1) location of the primary tumor; 2) stage of disease at the time of evaluation; 3) type of therapy administered; 4) survival from the onset of illness or earliest detection of tumor (whichever was earlier); and 5) locations and extent of metastatic lesions. The locations of metastases were recorded with respect to 1) anatomic region of the body, e.g., head or thorax 2) organ or structure involved, including functional subdivisions of organs such as the adrenal and pituitary glands; 3) type of tissue involved, e.g., pleura or myocardium; and 4) broad categories of the embryologic derivation of the affected tissue, e.g., mesoderm or neural crest. The extent of tumor involvement of each organ or structure was graded from 0 to 4 + as follows: 1 + for a single metastatic focus or less than 5% organ replacement by tumor; 2 + for two to five separate gross metastatic lesions or up to 20% organ replacement; 3 + for more than five separate gross metastases or up to 50% organ replacement;

HUMAN PATHOLOGY

Volume 19, No. 5 [May 1988]

All data were expressed in the form of mnemonic codes and were typed and proofread on a Raytheon VT 1303 video display-based word processor with communications software. Data were transmitted in asynchronous ASCII code by dial-up or direct line to a Digital Equipment Corporation PDP 11/84 minicomputer with a MUMPS operating system and programming language in the Department of Laboratory Medicine of The Johns Hopkins Medical Instutitions. Presence~absence data were analyzed statistically by Fisher's exact test. Quantitative and semiquantitative data were analyzed with correlation coefficients (Pearson's r) and Student's t tests.

and 4 + for extensive tumor infiltration or greater than 50% replacement. In addition, tumor burden in the body was given an overall grade (0 to 4 +) based on the degree of metastatic involvement in several major organs. T u m o r metastases were assigned two different overall histologic grades from 1 to 3 + for anaplasia and 0 to 3 + for desmoplasia. Postmorfem tumors were assigned overall grades for anaplasia and desmoplasia (fig. 1). The histologic features of different grades of anaplasia were as follows: grade 1, uniform tumor cells with minimal nuclear pleomorphism and arranged in an acinar pattern; grade 2, moderate nuclear pleomorphism and little or no acinat arrangement of cells; grade 3, marked nuclear pleomorphism with marked hyperchromasia and chaotic arrangement of cells. Grades of desmoplasia were assigned as follows: grade 0, absent; grade 1, scattered foci of collagen deposition within and around tumor cells; grade 9, moderate deposition of collagen without obfuscation of tumor cells; grade 3, marked deposition of collagen with obfuscation of most tumor cells.

RESULTS

General Population Profile One cluded in tions, the histotogic

h u n d r e d eighty-seven patients were inthis study. The age, race, and sex distribulocations of the primary tumors, and the subtypes of the tumors are shown in table 1.

FIGURE 1. Representative examples of anaplasia and desmopiasia. ,It, Grade 2 anapiasia. B, Grade 3 anaptasia. C, Grade I desmoplasia. D. Grade 3 desmoplasia. JAil, hematoxylin-eosin stain. Original magnification, A and B, x750; C and D, x250.]

530

AGE AND BREASTCARCINOMA [de la Monte et al]

TABLE 1. Profile of Autopsied Patients with Metastatic Breast Carcinoma No. (%) Total Age groups I, <40 yrs (young) II, 40~.6 yrs (premenopausal) III, 47-60 yrs (early postmenopausal) IV, >60 yrs (late postmenopausal) Sex Male Female Race White Black T u m o r histology Duct Lobular Medullary Malignant cystosarcoma phylloides Tumor location Right breast Left breast Bilateral

187 (100) 12 (6) 41 (22) 89 (48) 45 (24) 3 (2) 184 (98) 115 (61) 72 (39) 168 (90) 12 (6) 6 (3) 1 (1) 93 (50) 82 (44) 12 (6)

There were no significant differences with respect to histologic type or location of the primary tumors among the four patient groups. Most patients had been treated surgically (147/ 187, 79%), and most of those (111/147, 46%) also received systemic chemotherapy (usually including cyclophosphamide, methotrexate, and 5-fluorouracil) and/or radiation. In addition, a large proportion (91/ 187, 49%) received endocrine therapy, i.e., tamoxifen and castration. Only a few (11/187, 6%) received no treatment. These modes of therapy were uniformly distributed among the groups, and no correlation was seen between age and any particular therapeutic mode. Race, Survival, Stage of Disease, Number of Metastases, and Tumor Histology with Respect to Age The proportion of black patients with metastatic breast carcinoma increased as a function of increasing mean age, such that there were three times as many whites as blacks in groups I and II, approximately 1.4 times as many whites as blacks in group III, and equal numbers of whites and blacks in group IV (p < 0.05) (table 2). In addition, blacks survived significantly longer than whites (p < 0.05). There was no correlation between race and therapy or race and initial stage of disease. In general, group I and II patients survived significantly longer (66.5 + 9.6 months) than either group III (42.8 - 4.8 months) or IV (37.0 + 5.2 months) patients (p < 0.05) (table 2). However, the survival pattern of group I patients was bimodal such that the seven youngest patients had a mean survival of only 22.3 -+ 4.2 months whereas the remaining older patients had a mean survival of 86.8 + 11.2 months. Thus, a subset of very young patients had

survival durations that were shorter than those of all other groups. Especially poor prognosis of breast carcinoma in women less than 40 years of age has been reported previously. 3~ Thirty-seven patients (20%) had stage 1 disease, 76 (41%) had stage 2 disease, 19 (10%) had stage 3 disease, 18 (10%) had stage 4 disease, and in 37 (20%) the initial stage of disease had not been determined (table 2). Patients with stage 1 disease had the longest mean survival (61.6 + 10.0 months), whereas patients with stage 4 disease had the shortest mean survival (34.6 + 9.8 months). Patients with stage 2 or 3 disease had intermediate mean survivals (42.6 -+ 4.6 and 44.4 +- 9.8 months, respectively). Somewhat higher percentages of patients in groups I and II (36% and 55%, respectively) presented with stage 3 or 4 disease compared to patients in groups I l l (15%) or IV (13%); i.e., patients with stage 3 or 4 disease were significantly younger than those with stage 1 or 2 disease (p < 0.05). However, no significant trends were evident with respect to initial stage of disease and postmortem tumor burden, mean number of metastatic sites, or distribution of metastases. In addition, no significant differences were seen with respect to race and the initial stage of disease; blacks represented 38% of patients at each stage. A progressive decline in the n u m b e r of metastases was evident with increasing age. O f 93 possible metastatic sites, group I patients had the greatest mean number of metastatic sites (14.8 - 2.3 sites), and group IV patients had the fewest (9.6 -- 0.9 sites) (p < 0.05). Group I1 and I l l patients had approximately equal numbers of metastatic sites (12.8 -+ 1.0 and 12.0 - 0.7, respectively). In contrast, tumor burden was not correlated with any particular form of therapy, length of survival, or initial stage of disease. None of the postmortem tumors had grade 1 anaplasia. Grades 2 and 3 anaplasia were observed in duct, lobular, and medullary carcinomas, and there were no significant trends with respect to the degree of anaplasia (observed in postmortem tumors) and histologic subtype of tumor. Also, no correlations were seen between degree of anaplasia (postmortem tumors) and initial stage of diseases or age (group). Desmoplasia was observed most frequently in tumors from group III patients (p < 0.01) and least frequently in tumors from group IV patients (p < 0.05). Although chemotherapy was correlated with significantly lower degrees of desmoplasia (p < 0.05), this form of treatment was uniformly distributed among the f o u r patient groups, and t h e r e f o r e the agerelated differences in the frequency of desmoplasia could not be attributed to this factor. Distributions of Metastases with Respect to Age No correlation was seen between survival and the number or distribution of metastases. In fact, there were a number of patients in groups I and I! who survived for several years with documented widespread metastases, whereas patients in groups IIl or 531

HUMAN PATHOLOGY TABLE 2.

Volume 19, No. 5 [May 1988]

Influence of Age on Survival and Number of Metastases in Patients with Breast C a r c i n o m a

Patients

No.

Age* (yrs)

Sex (M/F)

Survival* (mos)

Metastatic Sites*

Group I (<40 yrs) Group II (40-46 yrs) Group III (47-60 yrs) Group IV (>60 yrs)

12

34.9 + 2.2

(3:1)

1/11

51.6 • 11.4

14.8 • 2.3

41

47.5 +- 1.0

31/10 (3:1)

0/41

66.5 -+ 9.6t

12.8 -+ 1.0

89

56.1 +- 0.6

52/37(1.4:1)

1/88

42.8 -+ 4.8

12.0 -+ 0.7

45

71.6 + 1.2

23/22 (1:1)?

1/44

37.0 • 5.2

Race (W/B) 9/3

9.6 - 0.9w

* Mean -+ SE. :~There were proportionately more blacks in Group IV than in the other groups (p < 0.05). ? Group II patients survived longer than group lII or IV patients (p < 0.05). wGroup IV patients had fewer metastatic sites than did groups I, II, or III patients (p < 0.05).

quencies o f metastases to cortical gray m a t t e r than g r o u p III and IV patients (p < 0.005). G r o u p IV patients had the lowest frequencies o f metastases to cortical gray o r white m a t t e r (p < 0.05). G r o u p 1 and II patients had the least extensive metastases in d u r a m a t e r c o m p a r e d to g r o u p s III and IV (p < 0.05), and .group III patients had the most extensive metastases m d u r a m a t e r (p < 0.001). Endocrine Organs. With increasing m e a n age, t h e r e was a progressive decline in the frequencies o f metastases to e n d o c r i n e organs in general, and in particular to the adrenal gland and pancreas; g r o u p I a n d II patients h a d the highest f r e q u e n c i e s a n d g r o u p IV patients had the lowest frequencies o f metastases to these organs (p < 0.05). In addition, the f r e q u e n c y o f metastases to the thyroid gland in g r o u p I patients was twice that observed in each o f the o t h e r three groups. No age-related trends were evident in the f r e q u e n c y o f metastases to the p a r a t h y r o i d or pituitary glands. Reproductive Organs. T h e f r e q u e n c y o f metastasis to the ovary was significantly h i g h e r a m o n g g r o u p I and II patients t h a n g r o u p III and IV patients (p < 0.01) and significantly lower a m o n g g r o u p IV patients than a m o n g the o t h e r g r o u p s (p < 0.01). T h e r e

IV t e n d e d not to survive very long with large t u m o r b u r d e n s . A m o n g the f o u r groups o f patients, several organs and tissues were u n i f o r m l y involved by metastatic disease, including lymph nodes (165/187, 88%), i n t r a p u l m o n a r y lymphatics (109/187, 58%), serosal surfaces including pleura (144/187, 77%), b o n e (147/ 187, 79%), and liver (120/187, 64%). No significant age-related trends in the f r e q u e n c y or extent o f metastases in these locations were observed. T h e age-related differences in the distributions o f metastases were manifested by progressive and significant declines in the f r e q u e n c y o f metastases to the central nervous system, e n d o c r i n e organs, r e p r o d u c tive organs, and portions o f the gastrointestinal tract (fig. 2). T h e greater n u m b e r s o f metastatic sites observed in g r o u p 1 and II patients c o m p a r e d to g r o u p IV patients are accounted for by the significantly increased frequencies o f metastases to these particular locations. Central Nervous System. With increasing age, t h e r e was a progressive decline in the f r e q u e n c y o f metastases to cortical gray m a t t e r o f the c e r e b r u m and cerebellum, gray m a t t e r nuclei, and the leptomeninges, but not to white m a t t e r or d u r a mater. G r o u p I and II patients had significantly h i g h e r fre-

u, 8 0 "~

CENTRAL NERVOUS SYSTEM CNS

ENDOCRINEORGANS ~ l l

REPRODUCTIVE ORGANS LIVER 8~ GI TRACT

Endocrine

~'20 I

II

IK

l~r

I

II

IK

lV

FIGURE 2. Age alters the metastatic behavior of breast carcinoma, With increasing mean age, there were progressive declines in the frequency of metastases to various regions of the central nervous system, several endocrine organs including pancreas, the ovaries, and portions of the gastrointestinal [G0 tract. In contrast, no significant age-related differences were observed with respect to the frequency of metastases to major organs such as liver, lungs, lymph nodes, or kidneys. Group I, <40 years', group iJ,40 to 46 years; group Ill, 47 to 60 years; group iV, >60 years.

532

AGE AND BREASTCARCINOMA [de [a Monte et al]

were no significant age-related trends in the frequencies of metastases to the uterus, fallopian tubes, or contralateral breast. Gastrointestinal Tract. Group I patients had significantly more frequent metastases to the submucosa of the gastrointestinal tract (p < 0.05), particularly the stomach (p < 0.05). Group I and I1 patients had the most extensive gastric metastases compared to other patients (p < 0.01). Age-related differences in the frequencies of metastases to other regions of the gastrointestinal tract were not observed. Other Organs. Group I patients had more extensive, but not more frequent, metastases to the lungs (p < 0.05) and spleen (p < 0.05) and more frequent metastases to the myocardium of the heart (p < 0.05) than the other three groups. The combined patients in groups I and II had significantly more f r e q u e n t metastases to bones of the extremities (p < 0.005) and ribs (p < 0.05) and less frequent metastases to the gallbladder (p < 0.05) than either group III or IV patients. Group III patients had the highest frequencies of metastases to the gallbladder (p < 0.01) and pericardium (p < 0.05) compared to the other age groups. Group IV patients, having the fewest metastatic sites, had the lowest frequencies of metastases to the liver, spleen, serosal surfaces, and thymus and the highest frequencies of metastases to axillary lymph nodes compared to all other groups (all p < 0.05). The increased frequencies of axillary lymph node metastases among group IV patients at autopsy may reflect the tendency of surgeons to perform more conservative tumor resections in older patients. In summary, age-related difference in the numbers and distributions of metastases were observed, such that young and premenopausal patients had greater numbers of metastatic sites than late postmenopausal patients, and as a function of increasing age, there were progressive declines in the frequencies of metastases to several organs and tissues, including the central nervous system, endocrine organs, ovaries, and gastrointestinal tract. These agerelated differences in the distributions of metastases were not correlated with differences in survival, therapy, or initial stage of disease.

Age-related differences in survival of patients with breast carcinoma, as described herein, have been reported by others, z5 Although it has been suggested that the degree of histologic anaplasia may have prognostic value in terms of survival of patients with breast carcinoma, ~1 we did not observe any significant differences in the degree of anaplasia among the four age groups, and therefore it seems improbable that the age-related differences in survival, number of metastatic sites, and distributions of metastases reported here could be explained solely on the basis of the histologic features of the tumors. Similarly, the age-related differences in the numbers and distributions of metastases were not attributable to differences in therapy or length of survival. Because the risk factors for breast carcinoma are essentially the same for premenopausal and postmenopausal patients, 23'24 it seems likely that breast carcinomas are fundamentally the same tumors irrespective of age of onset. However, the observations reported here suggest that the biologic behavior of breast carcinoma may be, to some degree, age dependent. Because metabolic r a t e Y endocrine organ function, 33 and immunologic functions 34 decline dramatically with increasing age, we propose that these types of environmental or trophic factors may be responsible for the age-related differences in the biologic behavior of breast carcinoma as reported here. REFERENCES 1. MacMahon B, Cole P, Lin TM, et al: Age at first birth and breast cancer risk. Bull W H O 43:209, 1970 2. Rosenwaike I: Breast cancer mortality in relation to a measure of early childbearing. Int J Epidemiol 9:317, 1980 3. Bain C, Willett W, Rosner B, et al: Early age at first birth and decreased risk of breast cancer. Am J Epidemiol 114:705, 1981 4. Lyle KC: Female breast cancer: distribution, risk factors, and effect of steroid contraception. Obstet Gynecol Surv 35:413, 1980 5. Gray GE, Pike MC, Henderson BE: Breast-cancer incidence and mortality rates in different countries in relation to known risk factors and dietary practices. Br J Cancer 39:1, 1979 6. Kelsey JL: A review of the epidemiology of h u m a n breast cancer. Epidemiol Rev 1:74, 1979 7. Jick H, Walker AM, Watkins RN, et al: Oral contraceptives and breast cancer. Am J Epidemiol 112:577, 1980 8. Jick H, Walker AM, Watkins RN, et al: Replacement estrogens and breast cancer. Am J Epidemiol 112:586, 1980 9. Wynder EL: Dietary factors related to breast cancer. Cancer 46:899, 1980 10. Kent S: Diet, hormones, and breast cancer. Geriatrics 34:83, 1979 11. Miller AB: An overview of hormone-associated cancers. Cancer Res 38:3985, 1978 12. Hankin JH, Rawlings V: Diet and breast cancer: a review. Am J Clin Nutr 31:2005, 1978 13. Goldin BR, Adlercreutz H, Dwyer JT, et ah Effect of diet on excretion of estrogens in pre- and postrnenopausal women. Cancer Res 41:3771, 1981 14. Mittra NK, Rush BF Jr., Verner E: A comparative study of breast cancer in the black and white populations of two inner-city hospitals. J Surg Oncol 15:11, 1980 15. Melnik Y, Slater PE, Katz L, et al: Breast cancer in Israel,

DISCUSSION In this study we observed that young and premenopausal patients survived significantly longer than late postmenopausal patients and that there was a significant decline in the number of metastatic sites as a function of increasing age. The decreasing number of metastatic sites observed with increasing age was mainly associated with significantly lower frequencies of metastases to the central nervous system, various endocrine organs (including pancreas and ovaries), and portions of the gastrointestinal tract among late postmenopausal patients. In contrast, the frequencies of metastases to most other organs and tissues were relatively uniform among the different age groups. 533

HUMAN PATHOLOGY

16. 17. 18. 19. 20. 21.

22. 23.

24.

25.

Volume 19, No. 5 [May 1988]

1960-1975: II. Effects of age and origin on survival. E u r J Cancer 16:1017, 1980 Devesa SS, Diamond EL: Association of breast cancer and cervical cancer incidences with income and education among whites and blacks. JNCI 65:515, 1980 Anderson DE: A genetic study of h u m a n breast cancer. JNCI 48:1029, 1972 Anderson DE: Genetic study of breast cancer: identification of a high risk group. Cancer 34:1090, 1974 MacMahon B, Cole P, Brown J: Etiology of h u m a n breast cancer: a review. JNCI 50:21, 1973 Lilienfeld AM: T h e epidemiology of breast cancer. Cancer Res 23:1503, 1963 Lewison EF, Neto AS: Bilateral breast cancer at T h e Johns Hopkins Hospital: a discussion of the dilemma of contralateral breast cancer. Cancer 28:1297, 1971 Saracci R, Repetto F: Epidemiology of breast cancer. Semin Oncol 5:342, 1978 Frankl G: Risk factors in breast cancer: are they important, are they the same in pre-and post-menopausal breast cancer patients? Oncology 37:41, 1980 Paffenbarger RS Jr, Kampert JB, Chang H-G: Characteristics that predict risk of breast cancer before and after the menopause. Am J Epidemiol 112:258, 1980 Mueller CB, Ames F, Anderson GD: Breast cancer in 3,558

534

26.

27. 28. 29. 30. 31. 32. 33. 34.

women: age as a significant determinant in the rate of dying and causes of death. Surgery 83:123, 1978 Palshof T, Mouridsen HT, Daehnfeldt JL: Adjuvant endocrine therapy of primary operable breast cancer: report on the Copenhagen Breast Cancer Trials. Eur J Cancer [Suppl] 1:183, 1980 Gapinski PV, Donegan WL: Estrogen receptors and breast cancer: prognostic and therapeutic implications. Surgery 88:386, 1980 Dao TL: Ablation therapy for h o r m o n e - d e p e n d e n t tumors. A n n u Rev Med 23:1, 1972 McGuire WL, Horwitz KB, Pearson OH, et al: Current status of estrogen and progesterone receptors in breast cancer. Cancer 39:2934, I977 Ribeiro GG, Swindell R: T h e prognosis of breast carcinoma in women aged less than 40 years. Clin Radiol 32:231, 1981 Freedman LS, Edwards DN, McConnell EM, et al: Histological grade and other prognostic factors in relation to survival of patients with breast cancer. Br J Cancer 40:44, 1979 Masoro EJ, Adelman RC, Roth GS: CRC Handbook of Physiology in Aging. Boca Raton, FL, CRC Press, 1981 G r e g e r m a n RI, B i e r m a n EL: A g i n g a n d h o r m o n e s . I n Williams RH (ed): Textbook of Endocrinology, ed 6. Philadelphia, Saunders, 1981 Danon ED, Shock NW, Marois M: Aging: A Challenge to Science and Society: vol I, Biology. New York, Oxford University Press, 1981