Metastasis as a beneficial process

Medical

Hypotheses

5:

351-358,

1979

METASTASIS AS P. BENEFICIAL PROCESS. G. Zajicek. The Hubert II. Humphrey Center for Experimental Medicine and Cancer Research. The Hebrew University-Hadassah Medical School, P.O.B. 1172, Jerusalem, Israel.

ABSTRACT The most important nessage to be found in the "Fourth report on end results in cancer" in the U.S., states that with the progression of cancer, its fcrce of mortality declines. This has been reexamined in the present study. The declining force of mortality implies that as the disease advances th.e chances of the average patient to survive, improve. Since in the cancer patient all vital functions gradually deteriorate, and the only process gaining with time is his tumor load, one has to consider the Rcssibility that the improving chances of the cancer patient could be linked with the amount of tumor mass in his body. These ideas are illustrated by the following example. Cancer could result form a gradual loss of a vital tissue product 'A', to be replaced b.y an analogous tissue product '6' which is of embryonal origin. In the adult, 'B' is produced by stem cells which gradually adapt to the loss of the primary product and increase in number. 'B' is less efficient than 'A' to meet the necessary vital functicns. Its deficiency in quality is however augmented b'y quantity. In order to keep up with the increasing demand, the stem cells proliferate and spread throughout the organism where each metastasis continues to secrete 'B'. The penalty inflicted by this compensatory mechanism is relatively high. Some metastases hit vital functions, and the increasinq tumor load depletes the available energy sources. The net effect however is beneficial since without metastasis the organism would have succumbed to the disease in its earliest stage.

Cancer,

metastasis,

epidemiology,

survival curves.

INTRODUCTION "Are all metastases harmful?" - This question which at first sight appears impertinent or even sacrilegious, has never been seriously studied. Obviously some metastases, like those which destroy vital anatomical structures are harmful, but is there enough evidence to indicate that the pheAnd more, what kind of nomenon of metastasis as a whole is detrimental7 experimental evidence supports the fundamental doyma of oncrjlogy, that metastasis is essentially a detrimental process? One could forward several examples of important and b'eneficial biological processes in the organism, which under certain circumstances cause death. Abscess formation, for instance, represents a mechanism to wall off infection. Tn its absence many such as the "space infections would end up in sepsis. yet some abscesses occuping lesions" in the skull, unless treated, cause death. Could it he that like inflammation, metastasis too is essentially beneficial with the exception of its "space occuping lesions"? Such a possibility has never been explored. The dogma states further that ever. in the absence of a demonstrable destruction of a vital anatomical structure, metastasis is detrimental since it leads tc cachexia, bone marrow insufficiency and sepsis. HCW iut of 157 patients strong is the evidence to support this assumption? In studied by Klastersky et al. (l), 32% died of gram negative infection. 14% infection accompanied another fatal prccess. 11% died from haemorrhaqe. Only in 20% could the death be attributed to neoplastic extension into vital organs. The mere existence of metastases in the organism does net assure their being the primary cause for its destruction. Previously it has been suggested by the present author EC regard cancer as a systemic disease which manifests itself b,y local phenomena like carcinoma (2). It has been proposed further to distinguish in cancer between primary systemic manifestations like cachexia and secondary manifestations like metastasis, which is regarded as sequela to cancer's local manifestation. In many patients metastasis and cachexia proceed side by side which does not On the contrary, both could assure however a cause and effect relationship. The primary systemic could indicate cancer's represent opposing processes. detrimental force while the secondary, its beneficiary opponent. In light of the accepted dogma, a search through the literature for evidence of a possible beneficiary role of metastasis is obviously futile. At best, clinical experience indicates that only in a fraction of patients may death be attribute to neoplastic spread into vital organs (1). The detrimental role of metastasis forms the cornerstone of an international symposium on "Fundamental aspects of metastasis" (3) which embraces all the known aspects of the problem. In spite of these difficulties, the assumption raised in the present study could be supported by the "Fourth report on the end results in cancer" (4). In its introduction, the report mentions a startling observation, that with the progression of cancer its force of mortality declines. A declining force of mortality implies further that with the progression of the disease, the chances of the average patient to survive, improve. How is it possible that in a disease in which all known vital functions of the organism 352

gradually deteriorate, the force of mortality declines.,,Cu;; :;llsr;r;;ct \ a statistical bias or indicate a genuine phenomenon? cancer force of mortality is not explicitly stated, the present study prov ide s first its values for two time intervals: O-5 years and 5-10 year%.

COMPUTATIONAL METHOD The force of mortality or failure rate is known also as hazard rate h(t), lefined as:

lim

Pr Ctcxct+Ct I x-,t)

x(t) =

at

CWO

or the probability of a cancer patient who survived up to time t, to die. The "Tnd result report" provides 5 and 10 year absolute survival rates labelled respectively as 65 and Plo. These were utilized for the estimation of th+ respective hazard rates. surviving until time t: where P

Pi is defined as the cumulative rrcportion = l_O,tmi stands for the mid point interval.

0 Tb,e estimated hazard rate at this point is:

P.

2(1----

1+1

P. 1

) (5)

A Rmi)

= sc1+-_

Pi+1

1

'i ithe number "5" stands for the five years interval). Utilizing this formula, 2.5 and 7.5 year mean hazard rates of all cancers in the report which exhibit at least a 10 year survival have been computed. Cancers of the following organs which do not attain a 10 year survival were excluded: Esophagus, liver, bile ducts, pancreas and acute leukemias.

RESULTS The 7.5 Y mean hazard rate of all the cancers in the report is smaller than that of the 2.5 interval. Thus X(2.5) > X(7.5) for all cancers. The 2.5 Y and 7.5 Y observed hazard rates of female breast cancer for instance, are respectively 0.12 and 0.08. The hazard rates of breast cancer labelled as regional at time of diagnosis are respectively C.14 and 0.10, while the hazard rates of localized breast cancer amount respectively to 0.06 and '3.05 (Table 1). The same phenomenon recurs in all cancers, localized as well Even in a systemic disease like chronic lymphatic leukemia the as regional. hazard rates drop from 0.24 to 0.18 (Table 1). This difference has been found t-values of the paired t-test utilized as significance highly significant. measure exceed 6.0 (Table 2). 353

ON HAZARD RATES A cancer patient may die from cancer as well as from other causes which are here lumped together as "aging". Assuming the two causes are noninteractive, the hazard rate x(t) of a patient represents the sum of the hazard rate due to cancer Xl(t), as that due to aging x,(t) (6)* X(t)

= Al

(t) + h2

(t)

The two hazard rates shape the survival curve. X2(t) increases with age for all adults as evident either from the population life table, or from observed median survival times of cancer patients with age, which decline from age 45 years and onward (4). Since X(t) decreases (Table 1) obviously Al(t) has to decrease even more.

THE INTERPRETATION OF THE DECLINING CANCER HAZARD RATE. Several factors could contribute to the declining hazard rate of the cancer patient. One could argue that at time of diagnosis cancer patients represent two subgroups. Those with a localized disease and the rest in which the disease has spread beyond effective therapeutic means. The improving hazard rate could reflect the death of the latter while the first enjoy a full cure. This argument is clearly refuted by the observed hazard rates due to "regional" cancer which also decline (Table l), or by the fate of chronic leukemia patients, a systemic disease, whose prognosis improves with time. The same fate awaits also women who refuse any treatment, like the 250 females with untreated breast cancer followed at the Middlesex Gospital in the yefrs 1805-1933. All exhibited a declining hazard rate X(2.5) = 0.28 and X(7.5) = 0.25 (7). The decline of the hazard rate could result also from the fact that at the begining of the observation,the disease in some patients is far more advanced than in the others-The death of the first could lead to an apparent decline of the hazard rate in the whole group. Such a selection bias has to be taken into account in any interpretation, yet it has to be remembered that the fourth report describes the fate of 200,000 patients, 1 relatively large sample which could safely be taken to represent the true It seems striking further that in a control group of healthy population. individuals whose ages match those of the cancer patients in the report, The hazard rate of the whole group, which ge the opposite may be observed. nerally reflects its aging process, increases. This group also comprises Obviously the older group several subgroups, some older and some younger. members with the highest hazard rates die first and would be expected to This, as may be apparent from the lead to an overall hazard rate decline. study of life tables does not happen, the mean hazard rate continues to increase. One is lead to conclude that the decreasing hazard rate in the cancer patient reflects at least in part a better chance for survival. The longer 354

he lives the better his chances to survive. What could be the mechanism which supports the patient throughout his Clinical experience shows that all his vital functions such as disease? immunological competence,bone marrow functions, efficient metabolic pathways, whi.ch is highlighted by the encroachment of cachegradually deteriorate, xia. There remains however only one process in his organism which increases throughoutthe disease, and this is the metastatic load. Could this process account for his improving chances for survival? Obviously, metastases which hit vital functions in the organism could These however may be an not account for the decline in the hazard rate. exception. The damage inflicted by metastases which occupy non vital sites like the integument, retroperitoneum or vertebrae is much smaller, and these represent the major metastatic burden. Why not then postulate metastasis to be beneficial, as illustrated by the following example.

A SCENARIO FOR CANCER Cancer results from a gradual loss of a vital tissue product 'A'. In order to compensate for the loss the affected organ has to reorganize for the production of an alternative tissue product 'B' which substitutes in part for the first. Product 'B' is of embryonal character i.e. it had played a major role during fetal development, and was replaced by product 'A' during maturation. In the adult, product 'B' continues to be secreted, in small quantities, by primitive tissue stem cells whose major role is to proliferate and maintain the pool of differentiated cells. The loss of 'A' leads to a gradual enlargement of the stem cell pool which thus adapts to produce more 'B'. 'B' however is less efficient in meeting 'A'S tasks. The gradual loss of 'A' leads to a gradual stem cell pool proliferation which spreads through the organ boundary in form of metastases. These too continue to produce 'B' whose deficiency in quality is augmented by quantity. The penalty inflicted by this compensatory mechanism on the organism is relatively high. Some metastases hit vital functions, while the increasing tumor load initiates a negative energy balance. The net effect however is beneficial since without metastasis the organism would have succumbed to the disease in its earliest stage. Prognosis thus depends mainly upon the rate of deterioration of the 'A' producing machinery as well as upon the size of the tumor load. Since it takes time to develop an adequate tumor load, a rapid deterioration of On the the 'A' producing machinery is accompanied by a grave prognosis. other hand in a gradually deteriorating 'A' producing machinery the compensatory mechanism has sufficient time to build up to save the patient. In such a case the increasing tumor load would lead to a decline in the hazard function. This is not the only scenario which could be forwarded to illustrate a It has been constructed so as to possible beneficial role of metastasis. 355

.

.

m

r-

I+

,-I

co

r-

l-i 4

0

.

m m

m N

cn 0

.

.

0

vi

.

z

CJ .-I

ln

u-l

.

.

w

.

TY

w

w

v-l

CJ

ol-4

m m

.

m N

.

In 0

.

0 N

.

.

co

0 0

N

.

.

w .

al

c-4

e

LT C

2

.

.

.

t2. W. N

.

.

w

m .

N 4

0 N

E.

m

m

Nrl .

.

w

I-

N N

0

.

.

r-v * t-

I

I

I

I

4 0

I

I

l-4

m

m

v

e

N

N

.

.

.

I

I

I

I

I

I

I

I

w e

N

.

m

ul r(

2.

”

.

I

In

.

m

.

Tf

g

vm

I

* .

.

I

.

.

0

.

0”

I

.

or-i

m

0

r-l

m

m

. r-4

?I

-3 I-IV-I .

u

. 0

w

.

N .

. N

0 r-Id .

0

.

.

.

N .

=r

.

0

.

P 0

356

0

ri

W

.

N

In s-4 .

N”

Y

~PUO~hd

(S’Z!

LEZ’

860 -

(S’L)

ZTO‘

i

9c)O'

ZTO'

Y ,

600'

(S’L)Y

080.

?azT

II [Pxn

(5-z)

6EI

-

(S-L)

031’

f Y sabP?s

602

*

ETO'

(s-z,

TTV

S60T'

T6SO'

06ET'

800'

OOTO'

Y b.

I

uea’d

(S’L) u-(S’Z)

Y

plPpuP>S

2

laqulnN

S6'OT

30

X33UP3

d

SE

;adL-$

T0000'0 ;>

86Oi)'

87.

S600'

81

_-

r-. Ln l-0

underline an important biological mechanism active in cancer, the mechanism of adaptation. The normal and healthy organism exists in a steady state, any deviation from which initiates a host of complex processes known as adaptation. The diseased organism usually passes through several adaptational steps known also as compensation, in which the disease remains reversible. Cancer is the only disease in which such compensation has never been demonstrated. The organism is assumed to passively yield to a disease which progresses in irreversible steps. The belief that the patient's only chance depends upon a successful1 elimination of the tumor in its earliest stage, highlights this inconsistency, according to which from its earliest stage the organism inflicted by cancer is in decompensation. Metastasis could therefore represent the major adaptive mechanism of the organism in cancer. Considering the above arguments, such a hypothesis can not be easily dismissed. It has however far reaching implications upon cancer treatment and forwards new opportunities for its study.

REFERENCES

1. Klastersky J, Daneau D, Verherst A. Causes of death in patients with cancer. Europ. J. Cancer 8:149, 1972. 2. Zajicek G. Cancer as a systemic disease. Med. Hypotheses 4:193, 1978. 3. Weiss L. Fundamental aspects of metastasis. North Eolland publ. CO. Amsterdam-Oxford, 1976. 4. Axtell ML, Cuttler SJ, ,"leyersIII!.End results in cancer report no. 4 DIIEW publication no. (HIII)73-272, 1972. 5. Gross AJ, Clark VA. Survival distributions: Reliability application in the biomedical sciences. J. Wiley & Sons M.Y. F. 35, 1975. 6. Gross A,:, Clark VA. ibid p. ST)-9C. 7. Bloom I!JG. The natural history of untreated breast cancer. Ann. N.Y. Acad. Sci. 114:747, 1964.