Preliminary studies on the validity of in vitro measurement of drug toxicity using HeLa cells III. Lethal action to man of 43 drugs related to the HeLa cell toxicity of the lethal drug concentrations

319

Toxicology Letters, 5 (1980) 319-331 o Elsevier/North-Holland Biomedical Press

PRELIMINARY STUDIES ON THE VALIDITY OF IN VITRO MEASUREMENT OF DRUG TOXICITY USING HeLa CELLS III. LETHAL ACTION TO MAN OF 43 DRUGS RELATED TO THE HeLa CELL TOXICITY OF THE LETHAL DRUG CONCENTRATIONS

BJijRN EKWALL Department of Human Anatomy, S-751 23, Uppsala (Sweden)

University

of Uppsala, Biomedical

Centre, Box 571,

(Received July lst, 1979) (Revision received and accepted January 8th, 1980)

SUMMARY

The human lethal plasma concentrations of 46 drugs were divided by their IC50 for HeLa cells in vitro to make up a series of cytotoxic quotients (CQL,). C&L, was then compared with the recorded lethal action to man of 43 of the drugs. While the 7 drugs with the lowest CQL, values produce a non-cytotoxic interference with neuro-transmission, most of the remaining 36 drugs have a known local or systemic cytotoxicity to man. A majority of the 36 drugs induces a non-specific central nervous system (CNS)-depression at lethal dosage, intermingled with function loss from organs outside CNS in proportion to decreasing drug accumulation in CNS cells and increasing CQL,. The remaining drugs which do not penetrate CNS cells and at lethal dosage induce a widespread injury and function loss of tissues outside the CNS, have a C&L, near unity. Non-specific CNS-depression may thus be the primary human reaction to lethal systemic drug cytotoxicity, while widespread drug injury to various tissues outside CNS - conventionally considered to be cytotoxic in origin .- may be the obligatory human reaction to drugs that do not penetrate cells well. The present findings indicate a relevance to human toxicity of the HeLa toxicity for most drugs.

INTRODUCTION

The relevance of the cytotoxicity in vitro of drugs to their human toxicity is relatively unknown, as is that of recently performed tests [ 1, 21 of combined toxicity to HeLa cells in the MIT-24 system of 37 drugs. This relevance [l, 21 has been evaluated in two previous reports [3,4], the first of which [3] compared the MIT-24 toxicity to HeLa cells of 13 combined [l, 21 Abbreviations:

CNS, central nervous system; CQL, cytotoxic

quotients.

320

and 15 other drugs with their in vitro toxicity to various other cells and, indicated a HeLa toxicity representative also of drug toxicity to more specialized cells. The second study [4] compared HeLa toxicity of all 37 combined [l, 21 and the 15 other [3] drugs with mouse LD50 and available human lethal dose (LD). 7 of the 52 drugs had a human LD considerably smaller than the HeLa-toxic dosage, indicating that the human drug action may not be represented in tissue culture. 39 drugs had a grossly (l-10 times) similar dosage in vitro and in vivo, indicating a human lethal drug action caused by an interference with basic cell functions, found also in HeLa cells. These conclusions were supported by available knowledge of the lethal mode of action of the drugs in man. These results [4] may have been invalidated by several factors, among them the very approximate human LD, the liberally defined similarity between compared doses, the accentuation of MIT-24 dosage compared with human LD caused by the very low cell density of the MIT-24 system. The present study was therefore made to extend the previous study [4] to a comparison of toxic concentrations in vitro and in vivo, METHODS

The similarity of lethal (inhibitory) drug action to HeLa cells and in the human body among the 52 randomly chosen common drugs of the previous study [4] was evaluated by judging the congruence between lethal drug concentrations in equilibrium with cells in either system. The 24 h IC50 to HeLa cells [4] was compared with human lethal plasma concentrations calculated from the i.v. human LD with help of the steady state V,, and also with recorded post mortem lethal plasma concentrations. Because of the few cells in the MIT-24 system [4], a concentration administered to the system will equal that finally equilibrated with cells irrespective of the degree of drug penetration of, or accumulation, in cells. Hence MIT-24 concentrations will be comparable with steady state of post mortem concentrations for all drugs. The congruence of compared in vitro and in vivo concentrations was quantified as so-called CQL, which were diverse human lethal concentration (LC) values divided by the same IC50 for HeLa cells in the MIT-24 system. Depending on a variation of the human LC used, three types of CQL were calculated, i.e. CQL, (Table I), CQL, (Table II), and CQLi (Table III). The CQL ought to be a measure of the HeLa toxicity of the human LC. For the 46 drugs with a defined human LD previously studied [4], CQL, (the cytotoxic quotient for the LC calculated with help of V,) values are presented in Table I. The human LC included in CQL, was derived from previously published human p.o. LD [4] by division of the p.o. LD by the p.o./i.v. ratio of mouse LD50 [4] and the subsequent division of the calculated i.v. LD (Table I) by the steady state distribution volume for a therapeutic drug dose, For 16 of the 46 drugs the V, had to be extrapolated from drug acidity and accumulation data (see note g, Table I).

321

For 23 drugs with published [4] plasma concentrations connected with acute lethal poisoning, CQL, (the cytotoxic quotient for the directly measured LC) are presented in Table II’ CQL, is associated with a perorally induced poisoning for 20 drugs (note b, Table II), while CQL, derives from i.v. induced poisoning. To correlate findings from the comparison of in vitro and in vivo concentrations (Tables I and II) with actual human lethal drug action, C&L, is in Table III compared with CQL,, human cytotoxicity of drugs, the symptoms of acute human drug poisoning, and drug penetration of brain cells. CQL, (the cytotoxic quotient for an approximate initial LC) was calculated as the human i.v. LD (Table I) divided by a distribution volume of 0.15 l/kg corresponding to human extracellular water and by the 24 h IC50 to HeLa cells and is thus only an approximate relationship of the HeLa cytotoxicity to an initial, lethal drug concentration. The remaining data of Table III are approximate and incomplete (see note h, Table III). RESULTS

AND

DISCUSSION

The comparison of in vitro and systemic drug toxicity by use of cytotoxic quotients is not entirely new. Nitta [ 181 found a positive correlation between a cytotoxic quotient comparing dosage (the HeLa MSD-mouse LD50 coefficient) and systemic cytotoxicity of antibiotics. This observation may be extended to a positive correlation between CQL and human systemic cytotoxicity for many types of drugs (Tables I-III of the present study). The comparison of concentrations in the form of CQL, (Table I) gives approximately the same results as the previous comparison of dosage [4]. The same 7 drugs had the largest concentration (dose) discrepancy in both studies (Nos. l-7) indicating organ-specific human lethal action. The remaining 39 drugs with CQL, of 0.008 or more seem to have a larger discrepancy of compared concentrations (CQL, 0.003-3) than the previously found [4] discrepancy of dosage. CQL, (Table II) is generally somewhat larger than CQL,, which may depend on its connection with LDlOO values while CQL, relates to average lethal dosage. For 14 drugs (Nos. 9,10,13,14,16,18,21,26,27,28,36,37,40, and 46) the two cytotoxic quotients are, however, similar. Another 9 drugs (Nos. 1, 2, 4, 8, 15, 17, 23, 24, and 35) have a considerably larger CQL,, which could depend on the oral dosage included in CQL, together with the tendency of these drugs to accumulate in tissues (d and e in Table II). A much higher oral than intravenous dose is required to produce the same critical (lethal) plasma concentration, because of the elimination by tissue binding in the liver of a large fraction of an oral dose. This oral overdosage is reflected by the later steady state concentration, which is higher for oral (CQLd) than for i.v. administration (CQL,) of drugs. If CQL, is corrected for peroral overdosing (e.g. by division of CQLd with V,) most CQL, will equal the CQL,.

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

1

No. 24 h IC50 HeLa, mg/lb

400 180 40000 89 4500 120 280 17 45 200 1300 1600 1800 1700 23 1300 9.5 200 18 8.9 3400 60

Dru$’

Nicotine Strychnine Gallamine Methadone Tubocurarine Epinephrine Methoxamine Imipramine Verapamil Phenformin Prilocaine Nikethamide Hexobarbital Lidocaine Amitriptyline Caffeine Chlorpromazine Tripelennamine Benztropine Methotrimeprazine Methyprylon Noscapine 0.25 0.34 2.9 0.36 0.29 0.29 1.4 4.3 2.7 18 71 48 11 43 4.1 24 7.1 10 3.6 11 71 14

i.v. human LD, mg/ke

t3

Q

B:3 B B B B:2 B B A B B B B B:2 B B A B 11.8 7.9 3.5 8.2 7.9 9.4 0.6 9.3 9.0 10.0 9.2 12.0 6.2

12.0 9.2 9.5

-

Q

B

8.0 8.0

B B:2

pK,

Acid-base nature and PK,~

T

TT

P

e t T e TT t

P

: 0 P

P P 0 P

:

P

0

PBf

TT T OL TR

TAe

-__I_

Tissue and protein binding

2E 5E 0.15 5.0 0.07 2E 2E 30 6.5 7.5 4.0 2E 0.40 1.3 8.9 0.80 20 2E 5E 30 0.5E 5E

V& I/kgg

HUMAN PLASMA CONC. OF I.V. LETHAL DOSAGE DIVIDED BY 50% INHIBITORY VITRO (= CQL,) FOR 46 DRUGS

TABLE I

0.13 0.68 19 0.07 2 4.1 0.15 0.70 0.14 0.42 2.4 18 24 28 33 0.46 30 0.36 5.0 0.72 0.37 142 2.8

Calc. human lethal plasma cont., mg/Ih

0.0003 0.0004 0.0005 0.0008 0.0009 0.001 0.003 0.008 0.009 0.01 0.01 0.02 0.02 0.02 0.02 0.02 0.02 0.03 0.04 0.04 0.04 0.05

Human lethal cone. divided by HeLa IC50 = CQL,’

11

7

7,9 799 10 1

7 7 7 k

7 7

8

Ref.j

CONC. TO HeLa CELLS IN

Chloroquine Propranolol Ouabain Phenol Procaine Phenobarbital Lithium Chloramphenicol Theophylline Thiotepa Aiprenolol Na Fluoride Quinidine Chlordiazepoxide Quinine Hydroxyzine Benzylpenicillin Alcohol Prometazine Sulfisoxazole Colistimethate Methotrexate Methampyrone Phenyibutazone

3700

200 3.8 180 400

570 860 210 140 5.7 48

24

4.3

320

0.04 490 350 540 3700 850 440 400 45 180

5.7 0.071 29 86 37 430 60

4.3 4.5

A:2 A

B:2 B A N B A

9.6 8.4 4.6 8.5 2.1 2.8 9.1 4.9

A B N B:2

8.6

N

9.5 9.9 9.0 7.2 -

B:2

e

Oe e

t

t

TT T TT t t e e

P P P 0 P P P P

0 P

OR P 0 P P

0.54 1.0 0.15E 0.055

0.15E 3.1 1.0 2.3 0.39 2.OE 2.OE 0.3 0.59 2.4”

0.9

2E

20E 4.8

260 5.7 320

12

58 43 67 480 110 59 29 6.5 32 8.3 62 12 8.5

1.2

0.2 0.3 0.4 0.6” 0.8 1 2 2

0.2

0.1

0.1 0.1 0.1 0.1 0.1 0.1

7

1

14 7 7

7,9

12 7

739 7,9

aDrug names according to use in previous papers [l-4]. Drugs arranged in order of decreasing CQL,. bA concentration which inhibits about 50% of HeLa cells in the MIT-24 system as judged by microscopy after 24 h. From previous studies [l-4]. ‘From a previous study [ 41. The dose is very approximative and has been extrapolated for most drugs from oral dosage, by use of the p.o./i.v. ratio for a mouse LD50. dFrom Martindale, Florey, or Avery [ 5-71, A, an acid; B, a base; N, a neutral drug; Q, a quaternary ammonium compound. Drugs with several pKa have got the number of pKa values indicated behind A or B, followed by the highest pK, for bases and the lowest pK, for acids.

23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46

eTA, human tissue accumulation of therapeutic drug concentrations. From a previous study [ 41. 0, no cell penetration; e, even distribution between cells and extracellular water; Oe, slight cellular penetration, or absence of cell penetration tombined with some binding of the drug to cell membrane; t, moderate accumulation; T, high ace.; TT, very high act.; B, act. in bone; L, act. in liver. *PB, binding of drugs to human blood proteins at therapeutic dosage. From a previous study [4 ] ; 0, no or slight binding (O-30%); p, moderate binding (30-70%); P, pronounced binding (7+98%); PP, 99% binding, ‘From studies indicated in reference column (note j). For drugs distributing biphasically, V, represents the second, eliminabased on data on tissue binding (note e) and acid-base nature of drugs tion phase i.e. Varea, VP, or V,.. E, estimations (note d) in Table I, and blood/liver ratios of lethal drug concentrations in a previous study 141. Drug No. 23 got a V, of 20 l/kg in analogy with other basic drugs with a very high cell accumulation (Nos. 8, 17,20, and 25). Nos. 2,19, and 22 got a V, of 5 like other basic drugs with a high accumulation (Nos. 4, 9, 11, 15, 24, and 33). Nos. 1, 6, 7, 12, 18, 27, 37, and 38 got a Vd of 2 like other basic drugs with a moderate tissue accumulation (Nos. 14 and 35). Nos. 21 and 26 got a Vd of 0.5 like other acid drugs with moderate cellular penetration (Nos. 13, 28, 31, and 43). Nos. 32 and 45 got a Vd of 0.15 like other acid drugs without cellular penetration (Nos. 3, 5, and 42). For Nos. 7, 12, 19, and 21 the absence of TA-data made Vd estimation a guess-work. nThe i.v. human LD (note c) divided by V,. !CQL,, the cytotoxic quotient of a human lethal cone. calculated from an i.v. LD with use of V,. ‘V, and other data marked by R were found in literature indicated in the column. kReported from Astra Co., SodertZlje. ‘Reference to V, is found in Tables I and II of a previous study [ 41; mEstimated V, for this drug probably is not valid for lethal dosage - compare blood/liver ratio of lethal human dosage in a previous study [ 41. The CQL, therefore probably is too small. “The reported V, for this drug is remarkably low compared with V, for other phenothiazines (Nos. 17 and 20). The CQL, could therefore be too large.

TABLE I (continued) 9

325 TABLE

II

DIRECTLY MEASURED HUMAN LETHAL PLASMA BY IC50 FOR HeLa CELLS IN VITRO (= CQL,) FOR Druga

Methadoned Verapamild Phenformine Hexobarbital Lidocaine Methyprylon Tripelennaminee Nicotinee Chlordiazepoxide Caffeine Phenobarbital Strychnined Phenol Chlorpromazined Quininee Amitriptylined Propranolold Procaine Quinidinee Phenylbutazone Alcohol Imipramined Chloroquined

Human lethal blood cone.,

w/lb

IC50 for HeLa determined in MIT-24 system by microscopy after 24 h,

1.0 0.6 3.0 30 68 150 10 29 26 106 78 26 70 1.6 12 9.0 7.0 150 45 400 5000 30 47

89 45 200 1800 1700 3400 200 400 320 1300 540 180 490 9.5 60 23 18 350 45 400 3700 17 10

CONCENTRATIONS 23 DRUGS

DIVIDED

Human LC divided by HeLa IC50 = CQLdc

mg/la 0.01 0.01 0.02 0.02 0.04 0.04 0.05 0.07 0.08 0.08 0.1 0.1 0.1 0.2 0.2 0.4 0.4 0.4 1 1 1 2 5

aDrug names and IC50 for HeLa cells from Table I. bConc. from the practice of forensic medicine, connected to acute peroral poisoning, with exceptions of iv. administered hexobarbital and procaine, and of phenol, which was absorbed through skin. From a previous study [ 41. ‘CQL,, the cytotoxic quotient for a directly measured human lethal plasma cont. dA drug which accumulates much in tissue including the liver, which is seen in TA-column, Table I, and as a high liver/blood coefficient of human lethal drug cont. in the Tables of a previous study [ 41. eA drug with a moderate tissue accumulation. See above.

In spite of the incompleteness of data in Table III, some correlations are apparent. CQL, of 0.003 or less is also related to documented non-cytotoxicit and a lethal drug action confined to neurotransmission without tissue injury to CNS cells or intermingled function loss from other than nerve tissues. CQL is related to a low but variable CQL, and a variable cell penetration, indicating a lack of association between CQL, and the latter parameters.

2 3 4 5 6 I 8 10 11 12 13 14 15 16 17 18 19 20 21

1

No.

CQLvb

0.0003 0.0004 0.0005 0.0008 0.0009 0.001 0.003x 0.008 0.01 0.01 0.02x 0.02Y 0.02 0.02 0.02Y 0.02 0.03 0.04x 0.04 0.04x

Drug=

Nicotine Strychnine Gallamine Methadone Tubocurarine Epinephrine Methoxamine Imipramine Phenformin Prilocaine Nikethamide Hexobarbital Lidocaine Amitriptyline Caffeine Chorpromazine Tripelennamine Benztropine Methotrimeprazine Methyprylon

0.004 0.01 0.0005 0.03 0.0004 0.02 0.03 2 0.6 0.4 0.2 0.04Y 0.2 1 O.lY 5 0.3 1 8 0.1

CQLiC

r

r

0 +r

r + r

+

0 + r

0 +

ed Ed ed D ed eD D ed ed eD d eD eD d D

D -

e ed

d Ed e ed

d ed

E d Ed

e

d D ed ed D d d D d

D

e e ed

P

1

1

g

g

kgw

1 k

P 1

P

~1

W

W

b m

eD E -

0

0 0 0 0

Symptoms of acute human lethal drug poisoning referred to separate organs (e, excitation; d, depression; p, lung; 1, liver; k, kidney; g, gastrointest. organs; w, bone marrow; b, blood cells; m, mucous membranes; s, skin; f, fever) -_____Heartf CapilOther tissuesh CNSe lariesg

CYTOTOXICITY

Known human drug cytotoxicity*

CYTOTOXIC QUOTIENTS (CQL, and CQLi) OF 43 DRUGS RELATED TO KNOWN HUMAN SYMPTOMS OF ACUTE HUMAN LETHAL POISONING OF THE DRUGS

TABLE III

S

s

f

f

AND

++ +++ +++ ++

+++

+++ +++ + ++ + ++ +++ +++ ++ +++ +++ +++ +++ +++ ++

Drug access to CNS cells’

g 0-J

Chloroquine Propranolol Ouabain Phenol Procaine Phenobarbital Lithium Chloramphenicol Theophylline Thiotepa Na Fluoride Quinidine Chlordiaxepoxide Quinine Hydroxyzine Benzylpenicillin Alcohol Promethazine Sulfisoxazole Colistimethate Methotrexate Methampyrone Phenylbutazone

0.06X 0.07 0.09 0.1 0.1 0.1 0.1 0.1 0.1 O.lY 0.2 0.2Y 0.2 0.2 0.2 0.3Y 0.4 0.6X 0.8 1 2 2 3Y

8 2 12 0.4 2 0.5 0.08 0.5 0.4 O.lY 1 0.9Y 0.5 3 3 0.6Y 2 10 0.9 5 5 2 1Y ++r + r + r + r ++ + r + r

+ + r + r ++ ++r ++r + ++r + r

+ r

++r

+ r

++r

d d

ed ed

d d d d d d d

D

d d d d D d d

d

ed d

z

D

D

d d d

d d

D D Ed

d -

D eD -

eD d eD ed eD -

ed ed ed eD ed D eD d ed kg

g

1

lkgw

1

W

m

S

s

s s s

s

bmsf+

bmsf+

sf+

msf++

b m

b

b m

b m

k w 1 gWbms 1 kgw bm lkgwbmsf+

g

g

k;W kg g

kg k g 1 g W

1 plkgw

Pl

P

~1

~1

~1

1 kg

f

f

f

f f

f

f

+ +

+++

++

++

++

++ + ++ ++ + ++ ++

++ +++ + ++ ++

aDrugs arranged in order of increasing CQL,.The table includes the drugs of Table I with exception of three drugs (Nos. 9, 22, and 33) without precisely described human lethal poisoning symptoms. bFrom Table I. X, CQ4, made extra uncertain by unprecise V d; see notes g, m, and n, Table I. Y, the drug has a 24-h IC50 considerably (5 times or more) lower than its 7-day IC50 [4]. To account for the 7-day IC50 CQL ought to be increased for drugs indicated by Y. ‘CQL,, the cytotoxic quotient of an approximated initial plasma cont. of an i.v. lethal dose, calculated as the coefficient between an i.v. lethal drug dose (Table I) divided by 0.15 l/kg (the volume of distribution corresponding to ECV) and a 24 h HeLa IC50 (Table I). Y, see note b.

23 24 25 26 27 28 29 30 31 32 34 35 36 37 38 39 40 41 42 43 44 45 46

fe, exitatory heart arrythmias ending in ventricular fibrillation; d, myocardial depression ending in asystole. “e, hypertension; d, hypotension ending in vascular collapse (shock). hThose data are incomplete for many drugs, such as Nos. 7, 18, 19, 25, 32, 38, and 43. Much information could be irrelevant to i.v. poisoning, since it often derives from p.o. poisoning. p, pulmonary distress or pneumonia; 1, liver injury or failure; k, nitrogen retention or kidney failure; g, GIT disturbance including ulcers; w, blood marrow depression including agranulocytosis; b, anemia, leukocytosis, or thrombocytopenia; m, eroded m. membranes; s, bullae or petechiae; m, s, and f, Steven-Johnson’s Syndrome or drug fever, for some of the drugs. ‘From handbooks [5, 7, 15-171, references to Table I, and the pharmacokinetic information in Table I. +, no or slight access to CNS cells; ++, a grossly even drug distribution between CNS cells and ECV; +++, rapid drug accumulation in CNS.

“From handbooks [ 5-7, 15-17 ]. o, no cytotoxic drug action to man. +, some local or systemic (side effects like blood marrow or liver damage) cytotoxicity of the drug to human tissues, at therapeutic or toxic dosage. ++, established human cytotoxicity of the drug, sometimes used in therapy (drugs Nos. 32 and 44). r, pain or local tissue reaction to tablets, suppositories or injections. Note that such reactions may depend on other factors than drug cytotoxicity, such as drug alkalinity (No. 13) or drug-induced vasoconstriction (No. 6). e* f* s* hFrom handbooks 15, 7, 15-171. Describes drug-induced function loss or injury of various organs. Capital letters describe the usual death causes. ‘e, CNS-exitation, i.e. convulsions; d, CNS-depression, i.e. paralysis ending in respiratory arrest. A dash denotes absence of CNS function loss.

III (continued)

329

CQL, of 0.008 or more seems to be directly related to all 4 compared parameters. It is positively correlated with documented cytotoxic action and is related to functional loss especially of the CNS, heart and capillaries. An increasing CQL, appears to be of increasing importance for effects other than those in the CNS and circulation. CQL, of 0.008 and more is inversely proportional to brain cell penetration and is finally related to a CQL, of approximately unity. These correlations all support the value of CQL, as an indicator of human drug cytotoxicity. Firstly, the non-cytotoxic organ-specific lethal action for drugs Nos. l-7 predicted by CQL, is confirmed by the above correlations for a CQL, of 0.003 or less. Secondly, the data on known local or systemic human drug cytotoxicity as well as the described functional loss of various tissues compatible with direct cellular injury are both correlated with CQL, between 0.008 and 3, indicating a cytotoxic lethal action in most instances. Thirdly, the step-less successive change in the symptoms of lethal drug poisoning proportional to a CQL, increasing from 0.008 to 3, together with the correlations of a CQL, of 0.008 or less with CQL, and brain cell penetration of drugs, help to explain the lethal effects as well as the large differences in C&L, resulting from differential cellular penetration. Drugs with a poor cell penetration (Nos. 25, 29,32,39, 42-46) are often of large molecular size (Nos. 25 and 43) or strong acidity (Nos. 39, 42,44, and 46) and spread slowly to all tissues, without initial penetration of CNS or centrally positioned heart or capillary cells. These drugs act lethally by a slow action on most cells of the body at their HeLa toxic concentrations (CQL, near unity). Drugs with a grossly even distribution between cells and ECV (Nos. 16, 18, 21, 26, 27, 28, 30, 31, 34-38, and 40) are often of weak acidity or basicity (Nos. 16, 21, 26, 28, 30,31, 34,36, 38, and 40; see Table I) and first penetrate capillary cells when introduced i.v. Other cells are then penetrated, of which the most sensitive are the CNS and heart conduction tissue. The greater sensitivity to drug cytotoxicity shown by CNS cells in vitro has been demonstrated previously [3] and may explain systemic death at about l/10 the HeLa toxic concentration (CQL, around 0.1). Drugs with a rapid accumulation in cells (Nos. 8, 11,14,15, 17, 19, 20, 24, and 41) are often strong bases. The accumulation explains the large difference between their initial and steady state concentrations (Table I). As foregoing drugs these appear to cause systemic death by cytotoxicity to CNS, heart and centrally positioned capillary cells, at concentrations between their peak and their steady state concentrations. A correct CQL for them therefore should have a value between CQLi (around unity) and CQL, (around 0.02), i.e. near 0.1. The small HeLa toxicity of the steady state concentration (l/100) corresponding to CQL, of 0.01-0.02 explains the relative freedom from tissue injury outside the circulation and CNS. These drugs often cause convulsions, exitatory heart arrythmias and high blood pressure; such excitation probably represents drug cytotoxicity. Excitation seems to be as-

330

sociated with a lower CQL, than depression (see the column on capillaries in Table III). For most of the studied drugs with a CQL, of 0.008 or larger, human death is caused by non-specific CNS-depression [15-l 71 which may include an initial CNS-excitation, but always proceeds to unconsciousness, flaccid paralysis, and respiratory arrest. Pulmonary distress, vascular shock, pneumonia, and a hypo- or hyperthermia are included. According to the above interpretation of events, this is the usual reaction of man to systemic drug (chemical) cytotoxicity at high dosage, where extra sensitive and/or immediately vital CNS cells are injured at about l/10 the HeLa toxicity. The widespread injury to parenchymatous organs by some drugs (e.g. Nos. 42-46) often includes blood marrow damage, and is conventionally attributed to cytotoxic drug action [15-171 as well as the mucocutaneous manifestations connected with fever and thrombocytopenia attributed to hypersensitivity reactions (drug Nos. 13, 23, 28, 30, 35, 37, and 39) [15-171, may both represent the obligatory reaction to serious systemic cytotoxicity of drugs not penetrating cells, notably brain cells, sufficiently. ACKNOWLEDGEMENTS

I am indebted to Dr. Olov Borga, Stockholm, and Prof. Sven Ullberg, Uppsala, for discussions of special aspects of this study. REFERENCES 1 B. Ekwall and B. Sandstrom, Combined toxicity to HeLa cells of 30 drug pairs, studied by a two-dimensional microtitre method, Toxicol. Lett., 2 (1978) 285-292. 2 B. Ekwall and B. Sandstrom, Improved use of the metabolic inhibition test to screen combined drug toxicity to HeLa cells -preliminary study of 61 drug pairs, Toxicol. Lett., 2 (1978) 293-298. 3 B. Ekwall and A. Johansson, Preliminary studies on the validity of in vitro measurement of drug toxicity using HeLa cells, I. Comparative in vitro cytotoxicity of 27 drugs, Toxicol. Lett., 5 (1980) 299-307. 4 B. Ekwall, Preliminary studies on the validity of in vitro measurement of drug toxicity using HeLa cells, II. Drug toxicity to HeLa cells in the MIT-24 system compared with mouse and human dosage of 52 drugs, Toxicol. Lett., 5 (1980) 309-317. 5 A. Wade (Ed.), Martindale’s Extra Pharmacopoeia, Pharmaceutical Press, London, 1977. 6 K. Florey (Ed.), Analytical Profiles of Drug Substances, Vol. l+, Academic Press, New York, 1972-1977. 7 G.S. Avery (Ed.), Drug Treatment, Principles and Practice of Clinical Pharmacology and Therapeutics, 2nd Ed., Adis Press, Sydney, 1979. 8 P. Duvaldestin, D. Henzel, M.R. Gilles, C. Roze and J.M. Desmonts, Etude de la pharmacocinetique de la gallamine chez l’homme, Anesth. Anal. Rban., 34 (1977) 235-239. 9 J.M. van Rossum, C.A.M. van Ginneken, P.Th. Henderson, H.C.J. Ketelaars and T.B. Vree, Pharmacokinetics of Biotransformation, in J.M. van Rossum (Ed.), Kinetics of Drug Action, Handbuch der Experimentellen Pharmacologic, Vol. 47, Springer, Berlin, 1977.

331 10 A. Jorgenssen and V. Hansen, Pharmacokinetics of Amitriptyline infused intravenously in man, Eur. J. Clin. Pharmacol., 10 (1976) 337-341. 11 S. Dahl, Pharmacokinetics of methotrimeprazine after single and multiple doses, Clin. Pharmacol. Ther., 19 (1976) 435-442. 12 J. Ekstrand, G. Alvan, L.O. Boreus and A. Norlin, Pharmacokinetics of fluoride in man after single and multiple oral doses, Eur. J. Clin. Pharmacol., 12 (1977) 311-317. 13 H.G. Boxenbaum, K.A. Geitner, M.L. Jack, W.R. Dixon, H.E. Spiegel, J. Symington, R. Christian, J.D. Moore, L. Weissman and S.A. Kaplan, Pharmacokinetics and biopharmaceutic profile of Chlordiazepoxide HCl in healthy subjects: Single-dose studies by the intravenous, intramuscular, and oral routes, J. Pharmacokin. Biopharmaceut., 5 (1977) 3-23. 14 J. Quinn and R. Calvert, The disposition of promethazine in man, J. Pharm. Pharmacol., 82 (1976) 56P. 15 L.S. Goodman and A. Gilman, The Pharmacological Basis of Therapeutics, MacMillan, New York, 1975. 16 R.H. Dreisbach (Ed.), Handbook of Poisoning, Lange, Los Altos, Calif., 1977. 17 P. Cooper, Poisoning by drugs and chemicals, plants and animals - An index of toxic effects and their treatment, Alchemist, London, 1974. 18 K. Nitta, Studies on the effect of actinomycetes products on the culture of human carcinoma cells (strain HeLa), I. The effect of known antibiotics having no or slight tumor inhibitory effect on HeLa cells, Jap. J. Med. Sci. Biol., 10 (1957) 277-286.