Notes on the Treatment of Cancer with Low-Dose Phenothiazines, with Special Reference to Promethazine

Robert Jones MA PhD
More indepth information on the Dr. Robert Jones' Phenergan® cancer protocol.

This article is for interest only and should not form the basis of diagnosis or treatment of any medical condition or illness. If you are unwell or have a specific medical condition, please base your healthcare decisions on the advice of a medical doctor or other healthcare practitioner.


Summary: The objective of this novel form of cancer therapy is to eliminate entirely malignant cells from the body through a process of attrition. Tumour necrosis in vivo is achieved by targetting the mitochondria of the malignant cell. An unusually high degree of selectivity is introduced into cancer treatment, as well as a fundamental paradigm shift. The tumour-necrotising action of endotoxin is simulated by a phenothiazine acting as sole anti-cancer principle. Seven anti-neoplastic manifestations of phenothiazines are listed; aspects especially relevant to the clinical treatment of cancer are discussed. The candidate drug appearing most suitable on various grounds, not the least of which is a striking selectivity of action, is promethazine (Phenergan). Its safety contrasts vividly with the disappointing success rates and the serious iatrogenic effects both in the short- as well as in the long-term of almost all current forms of cancer treatment. Careful adherence to a specific administrative protocol is essential in order to maintain continuous destructive pharmacological pressure against the growth(s). Indirect evidence suggests that the principal mode of action is necrosis, as opposed to programmed cell death (apoptosis). Certain forms of cancer not amenable to orthodox treatment, including non-Hodgkin lymphoma and pancreatic carcinoma, respond to the agent. Promethazine has the further advantages of being neither mutagenic nor carcinogenic, and of displaying activity against metastases. Pitfalls in the therapy of cancer with promethazine are described.

Background: On occasion new uses for certain drugs may be found in the wake of discoveries of unsuspected properties. The development of sulphonylureas for the treatment of diabetes from a chance observation of the hypoglycaemic effect of a sulphonamide and the introduction of aspirin as a prophylactic against stroke are but two examples. Phenothiazines as a group, of which sixty are listed in the eleventh edition of the Merck Index (1989), have already been put to a multiplicity of uses; for example, as anti-psychotics, anti-histaminics, anti-emetics, anti-cholinergics, anti-pruritics, and tranquillizers. Less commonly advantage has been taken of anti-depressive, anti-spasmodic, anti-tussive, analgesic, anti-arrhythmic, anti-inflammatory, coronary vasodilative, radioprotective, sedative, and skeletal muscle-relaxing properties. The addition of yet another property capable of exploitation is therefore hardly surprising. Commercially the concept of cancer treatment with phenothiazines is not new; no less than thirteen Japanese patents for phenothiazines as anti-cancer drugs have been listed in Chemical Abstracts since 1973 [30].

Clinical applications of the first phenothiazines date from sixty years ago [17,59]. The discovery of anti-neoplastic properties began a decade later, but have failed to attract mainstream attention. In 1957 prolongation of survival times in tumour-bearing animals treated with chlorpromazine was described [3]; a selective effect of the same agent in disrupting high-energy phosphate production by isolated tumour mitochondria was reported [40] in 1959; the first description of the anti-neoplastic action of chlorpromazine in man [38] appeared in 1960; and the earliest positive epidemiological study [49] dates from 1972.

Further reports, nearly all of which confirmed the initial findings, followed. In 1985 four lines of evidence for an anti-neoplastic role of phenothiazines (clinical findings, epidemiology, tumour inhibition in animals, mechanism of action) were reviewed [25]. A second survey of the literature [42] appeared later. A brief clinical note [30] was published in 1996; a pilot study describing the responses of forty-seven patients has been submitted for publication [31]. In man treatment schedules displayed key similarities; the same pattern of administering a phenothiazine, generally chlorpromazine, every eight hours was common both to the epidemiological studies and to anecdotal clinical reports of tumour regression and disappearance. Unfortunately the suspicion that differences in cancer incidence might be due to a particular medication was raised in only one epidemiological study [15] and was taken no further. The opportunity of refining the epidemiological data by restricting inclusion to patients with specific experiences, namely, (a) treatment with a phenothiazine, (b) a t.i.d. administrative protocol, and (c) continuity of medication for most or all of the period of study, was missed.

One major difference in treatment was that, in contrast both with current practice in the treatment of mental conditions in institutions where high dose therapy (400-800mg daily) was mandatory and with the early work of Luehrs and his associates [38,39], cancers responded to low doses of phenothiazines. The total daily amounts given ranged from 30mg to 75mg [7,10,11,30,31]; less usually, 150mg [10]. The importance of subjecting the tumour to continuous destructive pharmacological pressure was hinted at by the physician with the most experience of using phenothiazines against cancer (Eicke, personal communication). The impressive powers of observation of this perceptive clinician are illustrated by a note [9] on tardive dyskinesia predating that of Schoenecker [53], who is commonly credited with being the first to recognise the condition.

Inhibition of tumour growth and prolongation of survival times in animal models have been reviewed [25,42]. Ten out of twelve studies successfully demonstrated anti-tumour effects [25]. Of striking interest is the huge study by Driscoll and his associates [8]. A total of 298 substances with known psychotropic properties were examined in rodent tumour screens. Only sixteen displayed activity; of these six were phenothiazines and five were phenothiazine isosteres. In contrast only two investigations [5,6] were negative, though how one author [6] arrived at this conclusion is unclear. In a series of experiments with promethazine she demonstrated a 24% inhibition of tumour growth. In addition, chlorpromazine, administered 0.2%w/w in food, produced a 56% inhibition of tumour growth over 28 days with no loss in body weight, though the mortality was 21% [6]. It is not difficult to see why the other [6] failed; administered once daily, the dose of chlorpromazine (2.5mg/kg) was much too small for destructive pharmacological pressure to be maintained against the tumours for any significant period of time. In addition, drugs administered by the intraperitoneal route, as used by the author [5], are less effective in terms of affecting tumour energy metabolism adversely than when given by subcutaneous injection (Jones, unpublished observations).

Despite the impressive selectivity of action of phenothiazines, their anti-cancer properties as reported in 1988 [8] were considered too feeble to be of value in the therapy of malignant disease. Some years later, with the elusiveness of a successful treatment against cancer becoming more apparent, the idea of inhibiting the spread of new blood vessels within tumours (angiogenesis) was adopted as a means of checking cancerous growth. For its success the practice relies on intensifying ischaemia in the tumour mass, but the effects tend to be minor. In contrast with phenothiazine treatment, inhibition of angiogenesis incorporates neither of the known mechanisms for causing cell death. Meanwhile the concept of cancer attrition by phenothiazine treatment remains outside the orthodox frame. It is therefore hardly surprising that the possibility that its refinement might lead to more effective treatment continues to evade attention.

In sharp contrast with current treatments, which for the most part focus on the nucleus, the purpose of using phenothiazines against cancer is to target mitochondria. The critical importance of this radical shift in emphasis is not to be underestimated. Cell death, either through apoptosis [36] or arising from necrosis, involves changes in energy metabolism in the tumour [26]. The biochemistry of necrosis has been reviewed [27-29] and is outlined in Figure 1. All tumour mitochondria so far investigated display only partially-coupled oxidative phosphorylation (see [28] for review). This `damaged respiration' is associated with aerobic glycolysis (the ability of tumour cells to convert glucose to lactate in the presence of oxygen) and is believed to constitute the Achilles heel of the malignant cell [26,28]. The poor coupling of tumour mitochondria finds its physiological counterpart in tumour hyperthermia [37]. A substantial fraction of the chemical energy generated as high-energy phosphate is dissipated as heat. Cancerous growths sited under the skin can feel hot to the touch; hyperthermia has been reported in tumours of the breast, stomach, bowel and rectum [37]. Necrosis arising from ongoing injury to mitochondria is thought to account for the high rates of attrition observed in the division of malignant cells [20,58].

In the same context destruction of mitochondrial function is an early event in haemorrhagic necrosis brought about by endotoxin in a transplanted mouse sarcoma [21,22]. Alterations in energy metabolism which occur in response to chlorpromazine selectively increase the existing state of uncoupling of oxidative phosphorylation in tumour mitochondria [40] and decrease high-energy phosphate levels [34]; these alterations are consistent with early changes preceding necrosis [22,23,27,28]. It is critically important to recognise that the intention of the current approach is to trigger a phenomenon known to occur in nature rather than to impose an artificial form of cell death upon the cancer cell, one to which the intended purpose is seldom obtained. Subject to the provisions discussed below, all cancers regardless of the cell of origin should on theoretical grounds be susceptible to treatment with promethazine. The existence of certain exceptions currently defies explanation.

More recently three additional properties associated with anti-neoplastic activity have been described in phenothiazines. All are shared with tamoxifen [63], and include calmodulin antagonism (see [16,42] for reviews), inhibition of protein kinase C [2], and reversal of multidrug resistance in cancer cells [48]. Despite the accumulation of such an impressive body of evidence, a call [7] for clinical trials of cancer treatment with chlorpromazine made as far back as 1972 continues to be ignored.

Mechanism Of Injurious Action: Precise information here is limited. What is known is that when mice bearing the S180 sarcoma are injected with endotoxin, the classic tumour injurant [14] isolated from Gram-negative bacteria [13,56], total uncoupling of oxidative phosphorylation in tumour mitochondria [21,22] promptly occurs, followed by disintegration of the organelles as the period of exposure in vivo is prolonged [22]. Consistent with these changes, the tumour ATP level falls by 90% at 4hr and by 99% at 24hr [23]. In the same tumour endotoxin produces extensive haemorrhage after eight hours and massive necrosis a day later [14]. Tumour ATP levels also decrease rapidly in response to a variety of drugs not classified as cytotoxics, as well as to certain hormones [32,35; see [27] for review).

Figure 1 provides a theoretical explanation of the early pre-necrotic changes that occur between drug (or hormone) administration and the falls in tumour ATP levels preceding tumour destruction. It is thought that lipid on or near the inner mitochondrial membrane undergoes peroxidation; the products of hydrolysis, lysophosphatides [64] and hydroperoxy fatty acids [61], are both avid uncouplers of oxidative phosphorylation [27-29]. The details are to some extent conjectural. Tumour injury is believed to be initiated by an increase in free radical flux. How this is achieved is uncertain. It is thought that the proportion of oxygen reduced in a stepwise manner is under endocrine control, and that a hormone such as vasopressin stimulates free radical production by this means [cf 27].

Chlorpromazine has been shown to disrupt oxidative phosphorylation directly in isolated tumour mitochondria [40]. Certain metabolites may also participate. The ortho-quinone metabolite of chlorpromazine, 7,8-dioxo-chlorpromazine, prolonged the life spans of mice bearing an ascites tumour [41]. Non-enzymic reduction can occur with reduced glutathione; reaction between the resulting ortho-hydroquinone and molecular oxygen produces superoxide, hydrogen peroxide and hydroxyl radicals [18]. Together, these reactions theoretically constitute a mechanism for the cyclic generation of active oxygen species [27,30].

The concept of phospholipid(s) existing in either peroxidisable or non-peroxidisable forms is thought to be one reason why the same tumour may be either sensitive or resistant to promethazine [28]. Peroxidisable phospholipid in the drug-sensitive state is thought to contain a polyunsaturated 2-substituent which in a drug-resistant state has been replaced by a saturated or monoenoic acyl group. Drug resistance is unstable [34]; the instability may be accounted for by spontaneous in vivo conversion of non-peroxidisable phospholipid back to the peroxidisable form by acyl exchange.

Nutritional provision of phospholipid precursors is recognised as a novel and potentially important feature of cancer treatment. Evidence in favour of a role for omega-3 polyunsaturated fatty acids in the causation of tumour injury has been reviewed [28]. If a measure of ongoing tumour injury consistent with regression is to be inflicted on the tumour, efforts need to be made to ensure that host tissues do not become excessively depleted of polyunsaturated fatty acids, though at present the chemical identities of neither the polyunsaturated acyl residues nor the phospholipids themselves involved in the process are known.

Lysophosphatide metabolism in injured tumours has been discussed in terms of the necessity of promptly removing the cytotoxic threat [28, 64]. An early report described the presence of elevated levels of ethanolamine phosphate in a series of unspecified tumours in cattle [47]; this may represent greater cytotoxic activity against tumours and increased lysophosphatide breakdown, though in a later study [57] more of the ester was reported present in normal tissues of the rat and rabbit than in rat and mouse tumours. The accumulation of ethanolamine phosphate arising from the action of phospholipase B on lysophosphatides suggests in turn that peroxidisable phospholipid may contain ethanolamine. On the other hand anti-neoplastic activity has been reported on the part of inositol and its hexaphosphate derivative phytic acid [54]; dietary inositol and phytic acid have been recommended for cancer patients in combined amounts of 5-8g daily [55]. Further data [54] suggest that phosphoinositides might also contribute to ongoing tumour injury in animal hosts. This could be due to increases in the level of spontaneous peroxidation as a consequence of providing more peroxidisable lipid. Dietary supplements of choline and inositol (a minimum of 250mg of each daily) are recommended with the intention of maintaining adequate levels of lecithin and phosphatidyl inositides in cancerous growths.

Choice Of Phenothiazine: In an atmosphere laden with incredulity, when a novel and unusual discovery is made the same finding may need to be repeated a number of times before notice is taken. Anecdotal reports of regression and disappearance of cancer in man in response to treatment with a phenothiazine were published from three independent sources [7,10,11,39,40] between 1960 and 1975, and again [30] in 1996; this last paper also sought to identify situations in which phenothiazines are ineffective against cancer, and to relate them to the mode of cytotoxic action of phenothiazines. In addition a preliminary note [46] in which a cytostatic action of chlorpromazine against cancers in man was questioned appeared in 1962, but attempts to trace the detailed report promised at the time have not been successful.

Ideally the impact of the selected agent on the host, and especially on the central nervous system, should be minimal. Partly for this reason, partly because of the experience of the late Dr Riad Mahmud in treating successfully three cases of inoperable pancreatic carcinoma (Mahmud, personal communication; see also [30]), partly on account of its widespread availability without prescription (though not in the USA or Canada) and hence its suitability in a self-medication regime, the choice for the moment has fallen on promethazine, a potent H1-histamine receptor antagonist of weak neuroleptic activity and long duration of action. While it is recognised that other phenothiazines might produce more rapid anti-neoplastic effects than promethazine, experimentation in man to find a more suitable drug has not been possible.

Promethazine was originally introduced into medical practice in the late 1940s as an anti-histaminic [17,59]. It is used [1] in the short term for the symptomatic relief of allergic conditions; in sensitisation reactions to drugs or foreign proteins; against anaphylaxis; as a preoperative sedative and anti-muscarinic; as an anti-emetic; to counteract travel sickness; and as a paediatric sedative. It is safe and relatively innocuous; its side effects are well known and are for the most part well tolerated. No reports of death have been traced in the literature since its introduction in 1947. Neither mutagenic nor carcinogenic activity has been reported in promethazine itself or in any of its metabolites.

In other words, the problem of selectivity of anti-cancer action has, in effect, been solved. Partly for this reason, and partly because their impact on the anti-cancer action of promethazine is speculative, uncertain and unpredictable, other alternative treatments should emphatically not be run simultaneously.

Mode Of Treatment: The recommended procedure needs to be followed in detail [30]; readers are referred to Self-Medication; the Treatment of Cancer with Phenergan [promethazine] in this Section. Serviceable, though less up to date, descriptions of the self-treatment of cancer with promethazine can be obtained from the internet by using Google. The search words are cancer promethazine.

Fifty mg of promethazine are taken orally before retiring, and are followed at eight hours with 25mg thereafter. The approach has been adapted for self-treatment from information provided by Mahmud (personal communication), but does not mirror his technique precisely. He gave 50mg promethazine initially by slow intravenous injection, followed by calcium and papaverine by the same route, with 25mg of the phenothiazine t.i.d. thereafter. This strategy is not open to cancer patients adopting the self-treatment procedure. If promethazine increases the permeability of the plasma membrane to calcium by limiting the availability of chemical energy, then, in addition to activating phospholipase A2 [4], the cation itself might augment mitochondrial injury and tumour damage.

As mentioned above, traditionally phenothiazines have been given in high dose (400-800mg daily) for the control of mental disorders, especially schizophrenia, but substantially lower amounts (75-150mg) suffice in cancer treatment [7,10,11,30,31]. Unpublished evidence indicates that a maximal plateau of effect in disrupting energy metabolism in a mouse sarcoma [35] is reached as the amounts of L-isoproterenol or hydralazine given to tumour-bearing mice are increased; a similar situation is considered likely in the instance of promethazine. Indeed, the dose of 75mg promethazine daily may prove to be higher than necessary. Csatary reported regression of a squamous-cell carcinoma of the larynx in response to 10mg chlorpromazine t.i.d. [7; also Csatary, personal communication].

Nutritional supplementation with trace elements involved in enzymically-mediated free radical scavenging, as protection against the slight possibilities of jaundice or blood dyscrasias developing [30], is strongly advised. The additional provision of omega-3 polyunsaturated fatty acids to participate in the process of tumour destruction [28] is considered essential; the details have been discussed [30]. The inclusion of dietary choline and inositol has been mentioned above. Supplementary vitamin E is not recommended, and should be avoided at all costs (see (4) below).

Duration Of Treatment: The critical importance of uninterrupted cancer treatment with promethazine for an appropriate period of time cannot be overemphasized. The basic problem is that viable tumour remaining when promethazine treatment is discontinued has up to now been found resistant when treatment is recommenced. Predictions of the period of treatment necessary are impossible to make, but as a general rule eighteen months or more may be necessary. The rate of attrition is expected to depend on such factors as the size of the viable tumour mass, the extent of associated vascularisation, the magnitude of host reserves of polyunsaturated fatty acids involved in tumour destruction, the efficiency of their transport to the tumour, and the amount of vitamin E available for tumour protection. The duration of therapy tends to vary widely from one patient to the next [31]. Continuation for six months beyond the visible eradication of disease is suggested as an arbitrary end-point, but therapy should be maintained beyond this time if any doubts persist.

Further research is likely to cut the duration of treatment down sharply. One remarkable case has illustrated the possibility. A rodent ulcer which began in a patient fifteen years earlier developed into a basal cell carcinoma behind the left ear. Six operations and radiotherapy failed to eliminate the tumour; further extensive operation was planned after scanning the area. Fortuitously the patient began promethazine therapy sixteen days before the scan. During that time the tumour disappeared but for a tiny area identified as radiation damage. Relapse later revealed the area to be drug-resistant tumour; successful operation followed. Instances of tumour disappearance over similar periods of time in response to attacks of erysipelas have been reviewed [43].

Patients with a history of occupational exposure to carcinogenic influences or who are genetically predisposed to develop cancer may prefer to remain on medication indefinitely. A lady with non-Hodgkin lymphoma kept herself perfectly well on 25mg promethazine t.i.d. for nine years. No experience of treating any instance of multifocal cancer development, as might for example be expected in occupational exposure to a carcinogenic stimulus, with promethazine has been gained. Remaining indefinitely on treatment could be considered as a form of protection, but no firm advice can be tendered. The unsatisfactory nature of the situation is fully recognised.

Eexperience In The Treatment Of Cancer With Promethazine: Supportive evidence for the anti-neoplastic action of phenothiazines has come from animal experiments (see [25] for review), but has been consistently ignored. The comment has not infrequently been heard that if phenothiazines do indeed possess anti-cancer activity in man, then this property should have been spotted long ago. As pointed out above, it was [7,10,11,38,39]. When phenothiazines are used as adjuncts to suppress vomiting or to control pain during conventional cancer treatments, any genuine anti-neoplastic effects can be easily mistaken for the success of orthodox therapies administered simultaneously [25]. What has not been realised is that intermittent phenothiazine therapy is ineffective.

The discoveries of anti-neoplastic activity in phenothiazines was completely unexpected, and have never been taken seriously. Continuity of medication and avoidance of interfering drugs are essential if therapy is to be successful, but in practice have only been met fortuitously. Eicke (personal communication) described not just the difficulties and frustrations that even he, an established head of department in a university medical school, experienced in seeking to persuade colleagues of the validity of his findings [10,11], but also being refused an opportunity of presenting his conclusions to his peers at a meeting of the German Medical Society. The story is not unfamiliar.

Forms of cancer not amenable to orthodox treatment but which may respond to promethazine include cancer of the pancreas [30; also Mahmud, personal communication]. The drug has no problem in traversing the blood-brain barrier. Limited information from two grade III astrocytomas suggests that growth in brain tumours may be slowed by promethazine, though not halted. The treatment of such cancers should therefore not be attended by excessive optimism. Anafranil (clomipramine) [62] is a more appropriate drug; the internet (Google: cancer brain clomipramine) may be helpful. The drug is only available on prescription.

Anecdotal evidence suggests that when the primary tumour is sensitive, secondary growths too respond to promethazine treatment. In Case V [30], a lady suffering from advanced breast cancer, two out of four bone metastases disappeared in the course of twenty weeks of therapy. Results from further cases have been submitted for publication [31].

Promethazine not only protects against radiation sickness but can also be effective against such short-term effects of radiation as nausea and oedema [52]. Tenuous evidence suggests there may be beneficial effects against radiation-induced peripheral neuritis as well [30].

The writer has personally noticed the only side effect of 25mg promethazine t.i.d. to be trivial, namely, an increase in the frequency of dreaming. On the other hand a small number of patients found 25mg t.i.d. insupportable for on account of sedation; for them the recommendation has been altered to 20mg at night and 10mg during the day at eight-hour intervals. Another, with a history of intolerance to medication, found himself unable to continue. A further patient encountered considerable difficulty in adjusting to promethazine after being on nitrazepam (Mogadon) for some while, but her case may be atypical. Patients are strongly advised to follow the recommendation of the manufacturer not to consume alcohol while taking promethazine. Apart from the slight possibility of injury to the liver, drowsiness may be experienced if the advice is disregarded.

Sources Of Interference In The Treatment Of Cancer With Promethazine: Situations arise where patients, often with advanced disease, are unable to derive benefit from conventional cancer treatment. In certain instances promethazine therapy may be initiated to advantage, but the anti-neoplastic action of a phenothiazine may not be expressed for any one of several reasons. In early studies capricious responses to treatment with chlorpromazine arising from discontinuous medication (see (3) below) may have been too baffling to overcome, with the result that therapy with the phenothiazine was abandoned [cf 38,39,46]. Figure 1 provides a theoretical background to the mechanism of necrosis, from which logical reasons for at least some forms of promethazine resistance can be deduced. The available information is discussed below.

(1) Steroids The synthesis of macrocortin [12], an endogenous inhibitor of phospholipase A2, is stimulated by steroids. On theoretical grounds (Figure 1) the development of cellular injury depends on the functioning of this enzyme, which hydrolyses peroxidised lipid to hydroperoxyfatty acids [61] and lysophosphatides [64]. Both classes of compounds are powerful uncouplers of oxidative phosphorylation [28]. Inhibition of phospholipase A2 arising from steroid administration may therefore inhibit the hydrolytic step and protect the tumour against damage by preventing uncoupling of oxidative phosphorylation (Figure 1) [28,30].

(2) Polyunsaturated fatty acids: deficiency and excess Likewise a paucity of polyunsaturated fatty acids, thought more likely to belong to the omega-3 series than to the omega-6 family [28], may limit the extent of phenothiazine-induced cellular destruction. The problem is discussed above. When a patient with inoperable stomach cancer began medication with promethazine, the growth was so far advanced that food could no longer pass into the small intestine; the patient had already lost about 20kg in weight, and the skin developed a scaliness consistent with polyunsaturated fatty acid deficiency. No response to medication with promethazine was seen.

Two essential conditions for necrosis to take place are thought to include the presence of polyunsaturated fatty acyl moieties capable of undergoing peroxidation, and of phospholipase A2 to hydrolyse the resulting hydroperoxy lipid (Figure 1). The failure of promethazine to elicit a response in a patient apparently deficient in polyunsaturated fatty acids would appear to provide further support for the belief that necrosis is the principal manner in which the phenothiazine brings about the death of malignant cells.

Recently a patient complained of prolonged bleeding. The recommendation was to stop taking polyunsaturates and to halve the amount on resumption three days later. In the event of this uncommon response medical supervision should be requested in order to shorten the bleeding time.

(3) Drug resistance arising from interruption of medication Case V [30] displayed a disturbing form of drug resistance. Treatment with promethazine was discontinued after twenty weeks, during which time scans revealed the apparent disappearance of two out of four metastases in bone. Unfortunately the tumour marker increased, and resumption of promethazine ten weeks failed to halt the rise. As soon as the situation became clear, medication with promethazine was stopped; it was not considered advantageous to offer higher doses. The patient has since died. Other instances of premature discontinuation (cancers of the oesophagus and lung, and one instance each of chordoma and non-Hodgkin lymphoma) also resulted in death as a consequence of poor compliance. No explanation of this form of drug resistance has yet been forthcoming, and no method of overcoming such resistance has been discovered. Meanwhile the inadvisability of prematurely interrupting medication with promethazine cannot be over-emphasised.

(4) Vitamin E Vitamin E breaks peroxidative chain reactions (see [28] for references), and would therefore be expected to afford some measure of protection against necrosis in tumours, especially as the vitamin has been reported to accumulate in malignant growths [45,60]. When the hosts of a transgenic murine brain tumour were fed a diet depleted in esters of vitamins A and E, the tumours grew to a significantly smaller size. Although the effect was interpreted in terms of apoptosis mediated by reactive oxygen species [51], an explanation involving necrosis is considered more likely. Reports of a protective action against carcinogenesis by vitamin E (see [28] for references) and coronary heart disease [50] have appeared. A recent prospective study among smokers in Finland indicated that a daily intake of 50mg of vitamin E (about 50iu) lowers the incidence of prostatic cancer by 32% over an average six-year period; on the other hand the time elapsing between a diagnosis of cancer and death was the same for both groups [19].

Instances of patients whose cancers displayed resistance to promethazine are known. Several were consuming large daily supplements of vitamin E; the recommended dietary allowance [44] is 8-10iu. The first patient (Case VI) was a well-nourished lady about 80 years old who had been taking 1200-1400iu vitamin E daily over the previous fifteen years to control her phlebitis. Thirty months previously she had undergone radical mastectomy and radiation therapy. By the time promethazine was given her arm had turned a purplish colour from lymphoedema. Treatment with promethazine was interrupted after about a week and recommenced some ten days later. The colour of her arm returned to normal during both sessions of treatment with promethazine [cf 52]. At the time it was assumed on pragmatic grounds that the morphine administered at the same time was blocking the anti-cancer action of the phenothiazine [30], but there appeared to be no theoretical reasons for the supposition, and the view has since been abandoned.

Failure to respond to promethazine was observed in a second patient; his daily vitamin E intake was 750iu. The third failure was observed in a patient with prostatic cancer who was taking 400iu daily; at the time it was not realised that this form of cancer may not respond to promethazine. Interestingly, vitamin E supplementation did not prevent treatment with an anti-androgen from lowering the prostate specific antigen value to below unity; this would appear to add weight to the conclusion that cancer treatment with hormone analogues acts by initiating programmed cell death [26,36].

Regarding cancer treatment and vitamin E, the recommended course of action would be to stop high dose dietary supplementation for at least a week before promethazine therapy begins and to forego the advantages of cardioprotection [50].

(5) Other medications

(i) NSAIDs In theory agents which block lipid peroxidation might protect malignant cells against phenothiazines; indomethacin pretreatment afforded extensive protection against the disruptive action of both hydralazine and L-isoproterenol on energy metabolism in a mouse sarcoma [24]. Because of the theoretical possibility that NSAIDs taken during a course of treatment with a phenothiazine might dislocate the process of attrition and establish drug resistance, avoidance of these drugs would appear to be a sensible precaution.

(ii) Previous treatment with a phenothiazine isostere or related compound; spontaneous resistance While the hope is expressed that all the possible ways in which phenothiazine action might be blocked have been identified, the existence of further unidentified forms of phenothiazine resistance in tumours cannot be ruled out. An accurate medical history of a cancer patient displaying resistance to promethazine may prove instructive. By way of example, Eicke [10] reported an anti-cancer effect of the thioxanthene chlorprothixene in man, and Wilkie [62] has described anti-neoplastic activity in both the anti-depressant clomipramine and the leprostatic clofazimine. All three drugs are phenothiazine isosteres, though only chlorprothixene displayed anti-neoplastic activity in rodent screens [8]. Although no studies of the effects of these agents on energy metabolism in tumours appear to have been made, and although Wilkie [62] referred to the modes of action of clomipramine and clofazimine as '… inhibitors of the respiratory chain by different mechanisms…', nonetheless they are likely to share a common mode of action. On these grounds the possibility exists that resistance may be encountered in patients in whom malignancies were already established as a consequence of prior intermittent use of phenothiazines or their isosteres.

(6) Mutation Figure 1 shows how the mechanism of necrosis can be blocked by the inhibition, absence or chemical modification of a single participating component. It has been suggested [29] that the deletion of an enzyme such as, for example, phospholipid peroxidase or phospholipase A2, from the genome would confer a new dimension of protection against oxidative injury by preventing respectively lipid peroxidation or the hydrolysis of peroxidised lipid to lysophosphatides [64] and hydroperoxy fatty acids [61] (cf (1) above). At the time the possibility of a chance mutation was considered remote, but anecdotal evidence now indicates that such an event may occur as a consequence of conventional treatment. In a pilot study [31] only 14% of the patients who went into remission received radiotherapy; in contrast, insensitivity to promethazine or subsequent relapse were seen in 55% of patients after radiation.

Resistance to promethazine has been observed in melanoma, cancer of the prostate and in one case each of mesothelioma and Wilm's tumour. Clearly only an approximation of the success rate can be made, but provided the disease is not too advanced and none of the above factors applies, the chances of remission have been put at around 50-50 [31].

General Considerations: Primitive though this early technology [30] might appear, two major advantages of promethazine are that it is a safe drug which has acquired familiarity in acute use for over half a century, and that it belongs to a family of agents of which a number have been given in chronic high dose regimens. Emphasis has been laid particularly on the frequency and continuity of administration; even though it is expected that most problems have been smoothed out, this is perceived as an area where difficulties may arise. If trials are successful, the problem of reversing drug resistance caused by inadvertent interruption of medication is anticipated to become a focus of attention and to be solved relatively quickly. The current procedure needs to be regarded as a prototype; research is likely to produce improvements, such as shortening the period of therapy and extending categories of sensitive tumours.

For the future, the proportion of patients at present beyond the reach of successful therapy should decrease as experience accumulates. The effective dose for an adult may be lower than that advocated at present. New drugs with lesser effects on the central nervous system and whose mode of action can be related even more clearly to mechanisms of cell death may be developed to replace promethazine. In addition, tumour presensitisation may also find a place in cancer therapy. But before any progress can be made the body of available evidence in favour of treating cancer with a phenothiazine should be considered as an integrated whole; the thesis should not be conveniently rejected in its entirety on the superficial basis of an apparent inadequacy within one particular line of evidence, nor on the ostensible inappropriateness of treating cancer with a drug so innocuous as to be classified both as a paediatric sedative and as an anti-emetic.

A shorter version of these Notes, which have undergone revision, was published with the same title in Non Antibiotics: a New Class of Unrecognised Antimicrobics, eds. AN Chakrabarty, J Molnar, SG Dastidar & N Motohashi. Treatment of cancer with low-dose phenothiazines, with special reference to promethazine, pp.79-93. National Institution of Scientific Communication (NISCOM), Delhi (1998).


  1. ABPI Data Sheet Compendium (1996). Datapharm Publications, p.1400.

  2. Aftab DT, Ballas LM, Loomis CR et al. Structure-activity relationships of phenothiazines and related drugs for inhibition of protein kinase C. Mol Pharmacol 1991; 40: 798-805.

  3. Belkin M, Hardy WG. Effect of reserpine and chlorpromazine on sarcoma 37. Science 1957; 125: 233-34.

  4. Bjornstad P. Phospholipase activity in rat liver mitochondria studied by the use of endogenous substrates. J Lipid Res 1966; 7: 612-620.

  5. Chorazy M. Effect of chlorpromazine on Crocker sarcoma and Ehrlich ascites carcinoma. Nature (Lond) 1959; 184: 200-1.

  6. Cranston EM. Effects of some tranquillizers on a mammary adenocarcinoma in mice. Cancer Res 1958; 18: 897-99.

  7. Csatary LK. Chlorpromazine and cancer. Lancet 1972; ii: 338-39.

  8. Driscoll JS, Melnick NR, Quinn FR et al. Psychotropic drugs as potential antitumour agents: a selective screening study. Cancer Trtmt Rep 1978; 62: 45-74.

  9. Eicke W-J. Beitrag zu der Mitteilung von Kulenkampff und Tarnow, "Ein eigentuemliches Syndrom im oralen Bereich bei Megaphenapplikation". Nervenarzt 1956; 27: 418-19.

  10. Eicke W-J. Guenstige Verlaeufe bei Karzinompatienten unter zusaetzlicher Behandlung mit Phenothiazinderivaten. Med Klin 1973; 68: 1015-18.

  11. Eicke W-J. Guenstige Verlaeufe bei Karzinompatienten unter zusaetzlicher Behandlung mit Phenothiazinderivaten. Med Klin 1975; 70: 186-89.

  12. Flower RJ, Blackwell GJ. Anti-inflammatory steroids induce biosynthesis of a phospholipase A2 inhibitor which prevents prostaglandin generation. Nature (Lond) 1979; 278: 456-59.

  13. Fogg LC. Effect of certain bacterial products upon the growth of mouse tumour. Public Hlth Reps, US Public Hlth Service, Washington 1936; 51: 56-64.

  14. Gerber IE, Bernheim AI. Morphologic study of the reactivity of mouse sarcoma 180 to bacterial filtrates. Arch Pathol 1938; 26: 971-83.

  15. Giel R, Dijk S, van Weerden-Dijkstra JR. Mortality in the long-stay population of all Dutch mental hospitals. Acta Psychiatr Scand 1978; 57: 362-68.

  16. Hait WN, Lazo JS. Calmodulin: a potential target for cancer chemotherapeutic agents. J Clin Oncol 1986; 4: 994-1012.

  17. Halpern BN, Hamburger J. A new synthetic anti-histamine substance derived from phenothiazine (Phenergan, 3,277 R.P.). Canad Med Assoc J 1948; 59: 322-26.

  18. Heikkila RE, Cohen G, Manian AA. Reactivity of various phenothiazine derivatives with oxygen and oxygen radicals. Biochem Pharmacol 1975; 24: 363-68.

  19. Heinonen OP, Albanes D, Virtano J et al. Prostate cancer and supplementation with alpha-tocopherol and beta-carotene; incidence and mortality in a controlled trial. J Natl Cancer Inst 1998; 90: 440-46.

  20. Iversen OH. Kinetics of cellular proliferation and cell loss in human carcinomas. Eur J Cancer 1967; 3: 389-94.

  21. Jones GRN. Cancer: energy uncoupling as a primary event in endotoxin-induced necrosis of the S180 sarcoma in situ. Biochem Soc Trans 1977; 5: 214-16.

  22. Jones GRN. Early mitochondrial damage in the induction of haemorrhagic necrosis in the Crocker sarcoma (S180) by endotoxin. J Cancer Res Clin Oncol 1979; 93: 245-54.

  23. Jones GRN. Early cutback in chemical energy production in the Crocker sarcoma (S180) undergoing haemorrhagic necrosis as a result of endotoxin administration. J Cancer Res Clin Oncol 1980; 96: 53-64.

  24. Jones GRN. Prevention of drug-induced upset to energy production in the S180 sarcoma by inhibition of prostaglandin synthetase. Biochem Soc Trans 1980; 8: 106-107.

  25. Jones GRN. Cancer therapy: phenothiazines in an unexpected role. Tumori 1985; 71: 563-69.

  26. Jones GRN. Cancer therapy at the crossroads: which direction for the future? Physiol Chem Phys Med NMR 1991; 23: 217-20.

  27. Jones GRN. Cancer destruction in vivo through disrupted energy metabolism. I. The endogenous mechanism of self-destruction within the malignant cell, and the roles of endotoxin, certain hormones and drugs, and active oxygen in causing cellular injury and death. Physiol Chem Phys Med NMR 1992; 24: 169-79.

  28. Jones GRN. Cancer destruction in vivo through disrupted energy metabolism. II. Lipid peroxidation and cell death: drug resistance as a consequence of reversible cellular injury. Physiol Chem Phys Med NMR 1992; 24: 181-94.

  29. Jones GRN. Cancer destruction in vivo through disrupted energy metabolism. III. Spontaneous drug resistance, selectivity of antineoplastic action, and strategies for intensifying tumour injury. Physiol Chem Phys Med NMR 1992; 24: 195-212.

  30. Jones GRN. Successful cancer therapy with promethazine: the rationale. Med Hypoth 1996; 46: 25-29.

  31. Jones GRN. A pilot study of the anti-neoplastic action of a phenothiazine in low dose. [Submitted for publication.]

  32. Jones GRN, Baum R, Merkel R et al. Disturbance in vivo to energy production caused by hormones and sympathomimetics in the S180 sarcoma, with consequent drug-induced necrosis. Biochem Soc Trans 1980; 8: 105.

  33. Jones GRN, Ellis, R, Frohn MJN. Drug-induced interference with energy metabolism in the S180 sarcoma: a new principle in the production of selective tumour injury. Oncodev Biol Med 1981; 2: 155-64.

  34. Jones GRN, Frohn MJN. An injurious effect of chlorpromazine in the murine S180 sarcoma (var CB) and its relevance to cancer treatment. Biochem Soc Trans 1984; 12: 679-80.

  35. Jones GRN, Frohn MJN, Beckett AH. Liminal therapy: a strategy for the complete and selective destruction of malignant tissue in situ. Med Hypoth 1984; 15: 349-59.

  36. Kerr JFR, Wyllie AH, Currie AR. Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. Br J Cancer 1972; 26: 239-57.

  37. Lawson RN, Gaston JP. Temperature measurements of localized pathological processes. Ann NY Acad Sci 1964; 121 90-98.

  38. Luehrs W. Ueber spezielle Fragen der Chemotherapie boesartiger Geschwuelste. Abh Dtsch Akad Wiss (Berlin) Klasse Med 1960; 3: 332-37.

  39. Luehrs W, Gertich I. Ueber die Kombinationsbehandlung mit entkoppelnden Substanzen bei malignen Tumoren. Krebsarzt 1962; 17: 405-10.

  40. Luehrs W, Bacigalupo G, Kadenbach B et al. Der Einfluss von Chlorpromazin auf die oxydative Phosphorylierung von Tumormitochondrien. Experientia 1959; 15: 376-77.

  41. Molnar J, Mandi Y, Boda K et al. The effect of ring-substituted and quaternary chlorpromazine derivatives on the survival of NK/Ly ascites tumour bearing mice. Neoplasma 1982; 29: 215-22.

  42. Motohashi N. Phenothiazines and calmodulin. Anticancer Res 1991; 11: 1125-64.

  43. Moullin CM (1898). The treatment of sarcoma and carcinoma by injections of mixed toxins. Bale, Sons and Danielsson, London.

  44. National Research Council. Recommended dietary allowances, 10th ed. Nutr Rev 1990; 48: 28-30.

  45. Nikiforova NV, Kirpatovsky VI, Darenkov AF et al. Liposoluble vitamins E and A in human renal cortex and renal cell carcinomas. Nephron 1995; 69: 449-53.

  46. Osterman E. Does chlorpromazine have a cytostatic effect? Some reflections in connection with cases and modern literature [English translation]. Nordisk Psykiat Tidsskrift 1962; 15: 154-59. Psychopharm Abstr 1962; 1: 2758.

  47. Outhouse EL. XXX. Amino-ethyl phosphoric ester from tumours. Biochem J 1936; 30: 197-201.

  48. Ramu A, Ramu N. Reversal of multidrug resistance by phenothiazines and structurally related compounds. Cancer Chemother Pharmacol 1992; 43: 165-73.

  49. Rassidakis NC, Kelepouris M, Goulis K et al. Malignant neoplasms as a cause of death among psychiatric patients. Int Ment Hlth Res Newslett 1972; 14(2): 1-3.

  50. Rimm EB, Stampfer MJ, Aschiero A et al. Vitamin E consumption and the risk of coronary heart disease in men. New Engl J Med 1993; 328: 1450-56.

  51. Salganik RI, Craig AD, Rodgers J et al. Dietary antioxidant depletion: enhancement of tumour apoptosis and inhibition of brain tumour growth in transgenic mice. Carcinogenesis 2000; 21: 909-14.

  52. Salva J-A, Badell M. Traitement de la maladie des rayons par un antihistaminique de synthese. Presse méd 1949; 57: 888.

  53. Schoenecker M. Beitrag zu der Mitteilung von Kulenkampff und Tarnow, "Ein eigentuemliches Syndrom im oralen Bereich bei Megaphenapplikation". Nervenarzt 1957; 28: 35.

  54. Shamsuddin AM (1998). IP6: Nature's revolutionary cancer fighter. Kensington Publishing Corporation, New York.

  55. Shamsuddin, Sakamoto K (1992). Antineoplastic action of inositol compounds. In: Cancer Chemoprevention, eds L Wattenberg, M Lipkin, CW Bone, GJ Keloff. pp.285-308. CRC Press, Boca Raton.

  56. Shear MJ, Andervont HB. Chemical treatment of tumours. III. Separation of haemorrhage-producing fraction of B. coli filtrate. Proc Soc Exp Biol Med N Y 1936; 34: 323-25.

  57. Smith LC, Rossi FM. Quantitative determinations of ethanolamine phosphate and free aminoacids of normal and neoplastic tissue. Proc Soc Exp Biol Med N Y 1954; 87: 643-46.

  58. Steel GG. Cell loss from experimental tumours. Cell Tissue Kinet 1968; 1: 193-207.

  59. Vallery-Radot P, Hamburger J, Halpern B. Un nouvel antihistaminique de synthese. Presse méd 1947; 55: 662-62.

  60. von Euler B, von Euler H. Vitamin E in Sarkomen. Hoppe-Seyler Zeits physiol Chemie 1940; 264: 141-45.

  61. Watanabe T, Nakamura T. The effects of linoleate hydroperoxide on respiration and oxidative phosphorylation of rat liver. J Biochem (Tokyo) 1979; 86: 1041-47.

  62. Wilkie D. Antimitochondrial drugs in cancer chemotherapy: preliminary communication. J R Soc Med 1979; 72: 599-601.

  63. Wiseman H (1994). Tamoxifen. Molecular basis of use in cancer treatment and prevention. Wiley, Chichester.

  64. Witter RF, Morrison A, Shepardson GR. Effect of lysolecithin on oxidative phosphorylation. Biochem Biophys Acta 1957; 26: 120-29.

By G Robert N Jones MA PhD. London: September 1996, Revised May 2007.